CA1303530C - EXPRESSION VECTORS AND THEIR USE FOR THE PREPARATION OF A PROTEIN HAVING HUMAN .alpha. -ANTITRYPSIN ACTIVITY - Google Patents

Abstract

ABSTRACT OF THE DISCLOSURE

A transformed bacterium expressing human .alpha.1-antitrypsin is disclosed, which bacterium is transformed by a vector capable of expressing human .alpha.1-antitrypsin. The vector comprises a sequence coding for human .alpha.1-antitrypsine under the control of a promoter and a translation initiation region with the following sequence:
Translation sequence Shine/Dalgano and also an origin of efficient replication effective in said bacterium.

Description

~3~:P3~3~
W0 84/02918 - ~ - PCT/FR84/00014 Novel expression vectors and the;r use for the prepar~t;on of ~ protein having human ~l-antitrypsin ~ctivitY

The present invent;on part;cularly relates to no-vel clon;ng and express;on vectors, the bacteria trans-formed with the a;d of these vectors and their use, in particular for the preparat;on of a prote;n hav;ng human -antitrypsin activity.
~1-Antitrypsin is a serum glycoprotein accoun-ting for 90X of the ~1-globul;ns in human serum. It is present ~n n~rmal serum ~t a concentrat;on of 130 to 150 mg/dlr It consists of a single polypept;de chain having a molecular ~eight of 47-52,000 daltons, 10 to 15%
by weight being carbohydrate side chains. Numerous human variants of ~1-antitrypsin are known, and most of these polymorphisms seem to result each time from a s;ngle am;no ac;d subst;tuent. ~1-Ant;tryps;n is synthes;zed predom;nantly ;n the liver.
The function of ~1antitrypsin is that of an antiprotease; it combines with and ;nactivates a ~reat var;ety of proteases, wh;ch ;t renders inactive, in parti-cular trypsin~ thrombin, chymotryps;n and plasm;n.
However, ;ts essent;al ~unction is that of ;nh;-biting neutroph;l elastase, a w;de-spectrum protease

2~ capable of degrad;ng most structural proteins, in part;-cular components of connective tissue~
ALthough it is a circulat;ng serum prote;n, ~1- antitryps;n is ofcons;derable ;mportance as an inhibitor of extravascular elastase. This ;s essentially due to ;ts small s;~e, ~h;ch, ;n contrast to the larger se-rum antielastase ~2-macroglobulin, allows it to diffuse through most t;ssues.
Thus, approx;mateLy 44X of the ~1-ant;tryps;n ;s extravascular, and the most important organs protected by ~1 ant;trypsin are the lungs.
A def;ciency of N1 ant;trypsin may cause a pro-tease~ant;protease ;mbaLance and, for the reasons ;ndica-ted above, the most s;gn;ficant cl;nical consequences p~
' ~3~3~3~
W0 84/02918 - 2 - PCT/FR8~/0001~
occur ;n the lun~s. In these orgDnS~ ;n fact, an excess elastase act;v;ty very frequently leads to destructive lung d;seases, such ~s e~phys~r~
The best kno~n case of a~-ant;tryps;n defic;ency ;s due eO a genet;c d;sorder, ;n the course of wh;ch cer-ta;n var;ant phenotypes o~ ant;tryps;n are ~ssociated w;th a marked def;c;ency of serum ~1-ant;trypsin. ~he most studied var;ant, type Z, ;s an ~m;no acidsubstitut;on (lys;ne replaces glutamic ac;d at pos;tion 53, start;ng from the C-~erm;nal end of the protein~. Th;s change affects ~he degree of glycosylation of the 1-anti-trypsin molecule, ~eading to a defic;ency ;n the secre-t;on of the moleculer an accumulat;on ;n the hepatocytes and an ~-ant;tryps;n serum level represent;ng from 10 to 15% of that of controls. Th;s def;ciency ;s assoc;ated w;th development of emphysema 1n ;nd;v;duals aged from 40 to 50 years, and less frequently w;th l;ver d;sease.
The Z allele has a frequency of between ~/3,000 and 1/4,ûO0 ;n the Caucasian popù~ation of the United ~tates of Amer;ca and also ;n Europe. Th;s approx;mately cor-responds to 50,000 ZZ homozygotes in the Un;td States ofAmer;ca. In add;t;on, about one person ;n 800 in the U.S~A. has the SZ phenotype, ~h;ch ts assoc;ated w;th a serum ~1-ant;trypsin level only 35% of the normal level. This level ;s probably the threshold corresponding to develop-ment of destruct;on in the lungs by a protease/anti-protease imbalance, and the number of persons affected ;n the U.S.A. is of the order of Z50,000. The frequency in Europe is approx;mately of the same order (references 1 to 3).
It ;s clear that any factor which reduces the level of funct;onal ul-antitryps;n Leads to a pro~ease~
ant;protease ;~balance and ;n so doing predisposes to-wards pulmonary complaints. I~ is no~ clearly estab-l;shed that such disorders appear ;n c;garette smokers, c;garette smoke cons;stuting one of the major factors ;nvolveô ;n non-hereditary emphysema. It has been demon-strated both that components of c;garette smoke may in h;b;t the funct;on;ng of ~-antitrypsin by oxidation ....
.. . .
., , ;' , ... `, , ~ .

~3~)3~
.
W0 84/02918 - ~ - PCT/~R84/00014 of ~ methionine residue close to the elastase inhib;tion s;t~ (reference 4), ~nd that the funct;onal ~ct;vity of ~1-antitryps;n is r~duced by ~t lerst a fa1tor Df 2 ;n the lungs of smokers (reference 5~.
It is interesting to note that smokers do not ex-h;bit a drop ;n serum a1-antitryps;n~ but that th;s deficiency is local;2ed ;n the lungs and is due to the di-rect ~ct;on of c;garette smoke. It is very probable that c;garette smoke may also induce a h;gh level of neutro-10 phil elastase in the lungs, thus increasing the proteo-lytic load on the lungs, which are less ~ell protected on account of the inactivat;on of the Q1-ant;trypsin.
; Var;ous other patholog;cal condit;ons may be as-sociated with the reduction in the serum ~1-ant;trypsin level, the more s;gn;f;cant be;ng, ;n part;cular~ acute resp;ratory d;stress syndromes in the ne~born and adults, rheumat;sm, arthrit;s, stero;d-dependent asthma and, pro-bably, cyst;c fibroses treference 1).
It ;s clear that ;t must be poss;ble for any 20 protease/ant;protease imbalance, which may be man;fested at the cLin;cal level (reference 6), to be treated by re-placement therapy The f;rst approaches to th;s technique have con-sisted of intravenous ;njection of human ~1-ant;trypsin 25 prepared from normal serum into individuals having a gene-tic deficiency (ZZ homozygotes). The result of these pre-l;minary studies shows that the serum ~1-antitrypsin level can be maintained at a value greater than 70 mg/dl, that is to say at a level sufficient to ensure effective 30 anti-elastase protection in the lungs. It ;s evident that any deficiency of ~1-antitrypsin may be tackled therapeutically in th;s manner~
For th;s type of therapy~ it is necessary to have available large amounts of very pure a1-AT. However, ex-traction of a1-AT start;ng from serum comes up aga;nst the problems of purify;ng the proteins extracted from the natural med;um, and also problems associated with the use of serum as the bas;c material, which are essentially contam;nation problems.

ir ~3~J3~

These var~ous problems can be solved using a product ohtained by bacter;al ferment~tion - ~s will be sho~n below.
In fact, i~ is known that polypeptides, such 35 enzymes, antigens and hormones, can be pur;fled from a bacterial culture incorporating an expression vector, such as a plasm;d, G~nta;ning a cloned gene wh;ch encodes the said polypeptides.
In order to clarify certain elements of termino-logy wh;ch ~ill be used in the context of the description~it ;s appropr;ate to rev;ew br;efly the pr;nc;pal ele-ments govern;ng express;on of a gene.
For the cloned gene to be "expressed", that ;s to say for the desired product encoded by the said gene to be synthesized by the bacterium, it is appropriate for the vector used to have certain character;stics as-sociated with ~xpression of the genes.
Expression of a cloned gene comprises two major stages: f;rstly transcr;pt;on of the information encoded by the vector DNA into messenger RNA (mRNA), then trans-lation of this mRNA ;nto polypeptides at the r;bosome level.
The transcr;pt;on stage starts by binding of an enzyme, RNA polymerase, ~h;ch ensures "read;ng" of the ;nformation, on a part of the DNA of the vector, called the "promoter", which ;nit;ates transcription.
This b;nding is subject to various types of re-gulation; in part;cular, transcr;pt;on by RNA polymerase can be repressed by b;nd;ng of a prote;n called a "repres-sor", wh;ch attaches ;tself onto a sequence called the "operator", ;n the ne;ghborhood of the promoter (negat;ve control).
OnLy ;f the repressor is inactivated can the RNA
polymerase commence transcript;on.
Inactivat;on of the repressor or ;nduct~on ;s ef-; fected e;ther by the action of a substance ~h;ch ;nfluen-ces the repressor or by another means, such as heat.
Once the bNA ;s transcr;bed ;nto mRNA by means of the RNA polymerase, the m~NA can be translated at the .

.

.

~L3~3~3~

ribosome level into a polypeptide chain.
` In order that this translation may be effected, it is necessary that the mRNA contains a region for binding of ribosomes and the translation initiation site; the DNA
5 of the vector should have the corresponding sequences, called below "translation initiation region".
While the abovementioned elements are necessary for expression of a cloned gene, the expression vector also has another advantageous property. In fact, it is capable 10 of replicating and multiplying in certain bacterial cells called "host cells" and, by this means the number of "copies" of the cloned gene is increased. In most cases where the gene is expressed, this leads to increased production of the protein.
Finally, the expression vector, especially if it is a plasmid, generally contains genes which encode resistance to various antibiotics, and this type of resistance is used as a "marker" allowing selection of the transformed bacterial strains.
Other advantageous elements of the expression vectors will be explained below in the course of the description.
The present invention relates to a novel cloning and expression vector which allows improved production of 25 proteins and/or peptides in a Gram-negative bacterial strain, in particular Escherichia coli.
The invention more particularly relates to a transformed bacterium, expressing human ~l-antitrypsine, said bacterium being transformed by a vector, capable of 30 expressing in that bacterium human a l-antitrypsin, said vector comprising a sequence coding for human ~1-antitrypsine under the control of a promoter and a translation initiation region with the following sequence:
~Translation TCGATAACACAGGAACAGATCTATG
sequence Shine/Dalgano and also an origin of efficient replication effective in said bacterium.

,, , ~ , .
' ~3~;~3~3~
- 5a -As has already been indicated above, the presence of an origin of replication for a plasmid is essential to permit repli~ation of the vector in the corresponding bacterial cells, and in the particular case of E. Coli, the 5 origin , ;. ' ' ' , ~3~3~3i~

of replication of plasmid pBR322 will pre~erably be used.
Plasmid pBR322 has the advantage o~

. .

~ 3~ 3 ~

providing a large number of copies and thus of increasing the amount of plasmids producing the desired protein.
It is of course possible to use other oriyins of replication, in particular that of a plasmid which encodes any resistance factor or an episo(sic)-colicinogenic factor (that is to say one which confers on a bacterium the ability to synthesize a colicin).
However, it is appropriate to choose plasmids whose replication is not too strictly controlled and which does not have undesirable restriction sites, outside the cloning region.
Of the promoters of bacteriophage ~ , the principal le~tward promoter, called ~PL/ will preferably be used. P~ i.s a powerful promoter responsible for early transcription in ~ . Previous publications have already noted the expression at a high level of cloned genes under the control of PL (references 8 and 9).
It is also possible to use other promoters of bacteriophage ~, in particular the rightward promoter, PR ' or the second rightward promoter, P' R .
Although it is possible to usa very diverse translation initiation sequences, the use of that of the cII protein of bacteriophage ~ , called cIIrbs below, is preferred.
While use of the ~cIIrbs region is of more particular advantage, for reasons which will be explained below, it is nevertheless possible to use other translation initiation sequences, in particular that of the E gene which is bounded by restriction sites of interest, but also those o~:

QB COAT GAAACTTTGGGTCAATTTGATCATGCAAAATTAGAGACTAGACTGA

PHIX D TTCAACCACTAATAGGTAAGAAATCATGAGTCAAGTTTACTGAACAATC

~7' ~3~3~
~3 Sau96I NarI
*PHIXFTTCGGCCCCTTACTTGAGGATAAATTATGTCTAATATTCAAACTGGCGCCG

RPOBGACTTGTCAGCGAGCTGAGGAACCCTATGGTTTACTCCTATACCGAGAAAAAA
SPC AGTAGTTGACATTAGCGGAGCCTAAAATGATCCAAGAACA&ACTATGCTGAAC
STR CAAAAGCTAAAACCAGGAGCTATTTAATGGCAACAGTTAACCAGCTGGTACGC
_ _ _ *CII ATTGTTATCTAAGGAAATACTTACATATGGTTCGTGCAAACAAACGCAACGAC
TUFB CGATTTACCGTGTCTTAGAGGGACAATCGATGTGTAAAGAAAAGTTTGAACGTACA
OMP A ATACAGTAACTCACAGGGGCTGGATTGATTATGTACACTTCAGGCTATGCACAT
15 LAM-D TGAACACACCAGTGTAAGGGATGTTTATGACGAGCAAAGAAACCTT'rACCCATT
LAM-E
TGTGCGGCTTTTTTTACGGGATTTTTTTATGTCGATGTACACAACCGCCCAACTGCTG
In some cases a synthetic initiation sequence will be used, in particular:

ATCGA TAA CAC AGGAA GAGATCT ATG
* * *
This initiation sequence incorporates a Shine/Dalga o (sic) (S/D) sequence AGGAA and a translation termination codon TAA which stops the reading of N, the usefulness of which will be seen below.
The region which encodes the cIIrbs translation initiation is of more particular advantage because:
1. It has been demonstrated that a large amount of the protein cII is synthesized under the control of promoter P~, 2. The region in question is bounded by restriction sites permitting easy isolation of the said region; and

3. A restriction enæyme (NdeI) overlaps the translation initiation codon.

~3~35~3L9 W0 84/02918 - ~ - PCTtFR84/00014 Under these conditions insertion of fore;gn genes at this site ~;th reconstruction of the in;tiation codon ATG ~llows expression of non-fused proteins, that is to say those wh;ch conta;n no am;no acidsfore~gn to the 5 prote;n wh;ch ;t ;s ;ntended to prepare, with the e~cep-tion of the ;n;e;al methionine.
In contrast, clon;ng of ~ foreign gene ;n one of the un;que restr;ct;on sites do~nstream of ~cIIrbs leads to expression of a fused prote;n, as w;ll be described be-10 lo~.
It ;s advantageous to be able to obtain fusedor non-fused proteins as required.
A fused prote;n conta;ns, in general, besides the prote;~ of interest, a part or;ginating from the expres-sion of genes of the vector or of certain elements jo;n-;ng the var;ous elements of the vector.
Th;s hence necessitates subsequent cleavage of the expressed protein in order to release the protein sect;on of interest. However, in certain cases it is 20 possible to cause the fused prote;n to be excreted.
In contrast, in the case of a non-fused prote;n, it is not necessary to ef~ect cleavage, but the protein is not generally excreted and must thus be isolated from a bacterial lysate.
Finally, in certain cases, a non-fused prote;n will not be stable in the cell, while the fused prote;n wi ll.
It ;s therefore necessary to study the advantages and d;sadvantages ;n each case, and ;t ;s ;mportant to have available a polyvalent vector capable of allo~ing e~pression of the t~o types of protein.
The cloning zone conta;ning the unique restric-t;on s;tes permits ;nsert;on of fore;gn genes for produc-t;on of fused (or non-fused) prote;ns. These un;que restriction sites are situa~ed do~ns~ream of promoter PL
and of the region of initiation of translation of cII.
Amongst the un;que insertion sites ~hich preferably arise in this zone, there may be ment;oned EcoRI, SalI, AccI, 8amHI, H;ndIII, HindII, and PstI, corresponding to ~, .

~3~3~ii3 lo W0 84/02918 ~ - PCT/~R84/00014 restr;ct;on enzymes ~enerslly used~ ~nsert;on of ~enes wh;ch encode a particul~r protein into such sites Loc~ted 40 to 50 bp downstrciM of the transl~t;on ;n;t;ation codon leads to synthesis of foreign prote;ns fused at the 5 N-term;nus of the bacterial am;no ~c~ds der;ved from cII.
The vector ;n quest;on preferably conta;ns~ in add;t;on, a transcr;ption antiter~inat~on function en-coded, for exarple~ by the N gene of ~,caLled ~. In the presence of the N gene transcr;pt;on product, trar,scrip-10 tion starting from PL continues beyond most stop si~naLs.
This eliminates the problems posed by a premature stop ;n transcription which may occur ;f the fore;gn cloned genes contain such stop s;gnals. In addit;on, it has been shown that express;on starting from PL ;s ;mproved ;n an 15 N env;ronment.
If the promoter ;s not PL or PR but ;s P'R, for example, the ant;term;nat;on funct;on may be d;ffe-rent. P'R ;s respons;ble for transcr;pt;on of late genes of bacter;ophage ~ . The express;on of late genes ;s 20 pos;t;vely regulated by the product of the Q gene (refe-rence 10), ~hich has been shown to act as a transcr;ption antiterminator analogous to the N prote;n (reference 11).
Q acts at the 6S RNA terminator of constituent ~in;t;ated at P'R, thus perm;tt;ng transcription of the d;stal late 25 gene region. The express;on of Q is itself controLled by PR ~hich, like PL, ;s regulated by the repressor cI.
Th;s cascade of ;nteract;ons can be reconstructed on an express;on plasm;d and used in the same way for P~.
In order to el;m;nate the problems of tox;c;ty 30 and ;nstab;Lity of the host-vector system ;n the case of continuous product;on of lar~e amounts of a foreign pro-te;n, ;t is necessary to prov;de control of the act;vity of the promoter by assoc;at;ng with it aLl or part of an ;nduc;ble expression system, ;n part;cular a thermo-; 35 ;nducible expression system.
PreferabLy, tempera~ure control of the synthes;s of the fore;gn protein is carr;ed out at the transcr;p-t;on level by means of a thermosens;t;ve repressor enco-ded ;n the host bac~er;um, for example cI857, which ,, " ~

~3~3~
.,-~. .

represses the act;vity of PL at 28 C but ;s inact;vated at 4~C. The repressor acts on the operator OL adjacent to the promoter PL. `Although in the preced;ng case part of the thermoinducible expression system ;s an integral 5 part of the host bacterium, it is possible to arrange for th;s system to be part of the vector itself.
The vector in question can also contain a gene ~ith resistance to an antibiotic, for example ampicillin ;n the case of pPR322r but other resistance genes can be 10 used, such as those ~ith resistance to tetracycline tTetr) or to chloramphenicol (Cmr).
Incorporation of such a marker ;s n~cessary for selection of bacteria containing the carrier transfor-mants of the plasmid according to the invention during 15 the cloning experiments.
The incorporation of a resistance gene enables the stability of the plasmids to be increased ~hile a selective pressure is applied during fermentation, and also facil;tates isolat;on of the transformants.
For clon;ng, ;t ;s useful to have available a system ~hich allows detection of insertion of a foreign DNA into a plasmid.
~ y-way of example, it is possible to provide the cloning zone with the N-terminal fragment of the E. coli ~-galactosidase (lacZ') by fusing the fragment with the translation in;t;at;on reg;on der;ved from ~cII, whirh puts the translation of the ~ fragment under the con-trol of the cII sequences.
The ~ fragment is complemented by the expression 30 Of the host-encoded C-terminal ~ fragment, which leads to ~-galactosidase act;vity in the cells. This ~-galac-tosidase activity results in blue colonies in the presence of the chromophoric substrate 5-bromo-4-chloro-3-;ndolyl-~ - D - galactoside .
At 28~C, the promoter PL is inactivated, the fragnent ;s not synthesized and the colonies remain wh;te. When the temperature is increased to 42C~, the promoter PL is activated, the ~ fragment is synthesized ~,f'''~`. and the colonies turn blue.

,~ ~3~S~
,.,~.. . 1~
WO 84tO2918 ~ PCT~FR8~/00014 The insertion of fore;~n DNA lnto the clon;ng s;tes s;tuated in this detection system prevents synthe^
s;s of g-~alactosidase and thus leads to uh;te colonies both at 28C and at 42C~
It ;s also possible to replace the lacZ' gene by other ~enes ~hich sllou detect;on~
The present ;nvent;on alsu relates tc, processes uhich ~llo~ c~on;ng ancl express;on of genes which encode part;cular prote;ns, ;n part;cular human ~1-antitrypsin prepared in the fused or non-fused state, as well as the corresponding plasm;ds~
The present invent;on thus relates to a process for the cloning and expression of a gene which en~odes a spec;fic prote;n, allowing preparation of the said pro-tein ;n the non-fused form, ~hich involves cloning the gene in a restrict;on s;te overlapp;ng the translation ;nit;a~ion codon of a vector such as that descr;bed above, in wh;ch vector the translation in;t;at;on codon ;s reconst;tuted.
The invention also relates to a process for clo-n;ng a gene which encodes a specif;c prote;n, allo~ing preparat;on of the said protein in the fused form, ~hich process co~pr;ses clon;ny the gene ;n the clon;ng zone of a vector such as that descr;bed above, downstream of the translation init;at;on sequence.
The ;nvent;on also relates to the vectors ob-ta;ned using these processes and to the bacter;a trans-formed by the var;ous vectors described, ;n part;cular Gram-negative bacteria and more particularly Escherich;a coli.
The invent;on lastly relates to the products pre-pared by fermentation of the strains thus transformed.
F;nally, the present invention relates to a vec-tor, in particular a plasmid, which contains all or part of the sequence which encodes human ~-antitryps;n, and in particular a vector such as that descr;bed above.
Among the plasmids containing the sequence wh;ch encodes -antitrypsin there must be mentioned plasm;d pTG922, the preparation o~ uh;ch is described in the examples.
,. ....

~3~13~

However~ the protein which has the human ~1-anti-tryps;n activity and wh;ch ;s prepared ~hen using pTG922 contains a fused sequence of amino acidsat the N-terminus of the 11-antitrypsin protein. The injection of such a product can ;n some cases evoke an ;mmune response in the pa~ient.
In order to avoid this disadvantage, it is useful to be able to express, in the bacter;um, the natural non-fused gene for human Q1-aneitrypsin.
To effect this in the vectors of the invention, there is provided, preferably in the region of the star-ting codon for the translation of the gene encoding human ~-antitrypsin, the foLlowing structure:
_~_ -CATATGGAGGATCCCCAGG--GT~TACCTCCTAGGGGTCC~
so that the starting codon for the translation is in phase for translation of the gene.
Such a plasmid, permitting the preparation of a non-fused a1-antitrypsin, can be obtained, for example, by modification of plasm;d pTG~22, which plasmid leads to the preparation of a fused protein.
As described in Example 4, the nucleotides en-coding the N-terminus of the cII sequence of the fused ~-antitrypsin gene in pTG922 are linked by unique res-triction sites Ndel and BamHI.
The DNA situated between these two sites is eliminated by cleaving the plasmid with Ndel and BamHI, followed by rejoining of the plasmid so that the starting codon for the translation is fused in phase directly at the beginning of the sequence encoding ~1-antitrypsin.
This rejoining is obtained by using a synthetic adaptor oligonucleotide which permits:
a) the fusion of the Ndel and ~amHI sites;
b) the reconstitution of the starting codon and c) the addition of a codon encoding glutamic acid at the N-terminus of the mature protein, which acid was lack;ng in the construction of pTG922.
Other plasmids were used, in particular by using . , ~L3~

W0 84J02918 - ~ - PCT/FR84/00014 a sequence in pTG929 for in;tiation of the transl~tion:
TCGAT;U~Ci~CA ÇGA~ACAGA~CT ~
Plasm;d pTG956 is thus obt~;ned.
Although this plasmid per~;ts ~n increased synthe-s;s of ~-AT, the y;eld remains relat;vely low. The fact that t~o sites o~ ribosomal attachment ~h;ch are cap~ble of leading ~o sign;fic3nt protein synthes;s do not permit ;nitiat;on of translation when they are coupled to the o~AT gene suggests that some character;stics of the ~-AT
gene ;tself l;mit its expression.
For th;s reason, the beginn;ng of the u-AT gene was mod;f;ed in the following manner:
' protein met glu asp pro g ~ gly asp ala or;ginal sequence DNA ATG GAG GAT CCC CAG GGA GAT GCT
protein met glu asp pro glu gly asp ala altered sequence DNA ~ ATG GAA GAT CCT CAA GGC GAT GCT
* ~ 1r t~
There is thus obtained, starting from pTG95b, plasmid pTG983 which produces much more significant quanti-ties of ~~ATr The invention also relates to the bacteria trans-formed and the product obta;ned by fermentation of these bacteria, called "human ~t-antitrypsin of bacter;al origin" belo~.

,,, \

~3~3S~
, It should be understood th~t ~his product c~n have var;ous structur~l d;fferentes from human serum ~1-~ntitryps;n, but nevertheless has the s~me ~ct;v;ty ;n vivo, in part;cul~r the same ant;-elastase astivity or an equivalent act;v;ty in repsect of serum n1-antitryp~
sin.
Human 1-antitryps;n of bacterial or;g;n is also understood as meaning a product of bacterial orig;n wh;ch ;s obta;ned by fermentat;on but has then been modif;ed chemically or b;olog;cally.
F;nally, the present invention relates to the use of hu~an 1-antitryps;n of bacter;al or;gin as 3 medica-ment for the treatment and/or prevention of human d;sea-ses ;nduced by a defic;ency of ~1-antitrypsin. The di-seases wh;ch can be treated or prevented by a1-ant;tr sin according to the invention include hereditary defi-c;enc;es of 1-ant;trypsin wh;ch are due to a genetic disorder and man;fest themselves, in part;cular, ;n emphy-semas. It ;s also possible to treat certain acute re-sp;ratory distress syndromes, in part;cular ;n the new-born and ;n adults, as ~ell as rheumato;d arthritis, stero;d-dependent asthma and cyst;c f;broses. Th;s me-dicament can 3lso be used to cure certain def;c;encies of ~1-ant;tryps;n ;n cigarette smokers.
Human ~1-ant;trypsin of bacter;al origin can be adm;n;stered by any route, but the parenteral route appears most satisfactory. However, in the case of de-ficienc;es due to cigarette smok;ng, direct ;ntroduction into the lungs by inhalation is also very effect;ve.
The da;ly dosages ot o1-ant;tryps;n to be used depend on the nature of the treatment, that is to say whether prevent;ve or curat;ve, as well as the sign;fi-cance of the defic;ency to be treated. It w;ll be pos-s;ble to adapt these dosages ;n order to ma;nta;n a spe-cific level of ~1-antitrypsin, for example in the serum or ;n the lungs.
The corresponding pharmaceutical composit;ons w;ll also be adapted to the adm;nistrat;on route and con-trolled release will be poss;ble, ;n part;cular in the 3~3~

form of ~n inplant which slowly re~esses the active pr;n-ciple ineo the circulatory system.
The present ;nvention of rourse has other ~spects, in particular certain plasm;ds which ~;lL be described in the Examples~ together w;eh the;r mutants and derivatives~ and, generally, the processes of fermen-~ation of the bacter;a transformed as well as the pro-ducts of these fermentations.
Other character;stics and advantages of the ;n-vention hill be better understood by reading the Examples belo~ and study;ng the attached draw;ngs, in ~hich:
Figure 1 represen~s the strategy allo~;ng prepa-ration; of plasmid pTG907, Figure 2 represents the strategy allow;ng prepa-15 ration of phage M13tg910, Figure 3 represents the structure of phage M13tg-910, Figure 4 represents the strategy allowin0 prepara-t;on of plasm;d pTG908, Figure 5 represents the ;mmunoprecipita~es corre-sponding to the translat;on products of total RNA from the human liver, Figure 6 represents an ;solated sample of a clone ~ of total human cDNA which en~odes ~1-ant;trypsin, : 25 Figure 7 represents the partial sequence of a cDNA
; clone of nl~ant;trypsin ;n compar;son with the sequence published, Figure 8 represents the strategy for preparat;on of plasm;d pTG920, F;gure 9 represents gel analysis of the proteins synthesized by E. col; transformed by a plasmid carrying the human gene Gf ~1-antitrypsin as ~ell as the corres-pond;ng ;mmunoprecip;tates, : F;gure 10 represents the b;olog;cal act;vity of 1-ant;trypsin prepared starting from E. col;, : Figure 11 reprecents the strategy allow;ng prepa-rat;on of plasm;d pTG929, Figure 12 represents ~he strategy allow;ng prepa-ration of plasm;d pTG956, ~3q~3~

F;~ure 13 represents the immunoprecipitates of the cultures labeled with ~nti-~1-AT ant;serum, and F;gure 14 represents the ant;-elast3se issay~
1) General_processes ~) ~acterial stra;ns The bac~er;al str~;ns used in the context of ehe present ;nvention nre as follo~s:
. TGE900, wh;ch is an E. coli strain having the follow;ng character;st;cs: su ~ his ilv bio t~cI857 ~am ~I) . N6437, E. coli strain having the follo~ing character;st;cs: F h;s ilv gal ~8proC : tn1û
lac~ml5 (~cI857ABam~HI).
; . Jm103, which ;s an E. coli stra;n hav;ng the follow;ng characteristics: ~(lac-pro~ supE eh; endA
sbc~15 sttA rK nK /F traD36 proAB
laci lacZAm15.
The abovement;oned stra;ns were used because they were ava;lable, but ;t is of course possible to use other stra;ns ;nsofar as they have certa;n essent;al character-;stics; ~hich are recorded ;n the course of the deta;led descr;pt;on.
b) Preparat;on of the DNA
M;ni-preparat;ons o~ DNA of plasm;ds or of ~5 phages M13 are carr;ed out as descr;bed by Ish-Horowitz ~reference 12), the only difference be;ng that the DNA
;s prec;piSated a second t;me w;th ethanol before be;ng used.
The ~ax;-preparations are carr;ed out as des-cr;bed in the above publicat;on, w;th supplementary pur;-fication by means of a density grad;ent of CsCl/ethid;um brom;de.
c) Clonin~ techniq~s Unless indicated other~ise, the treatment of the DNA w;th restr;ction enzymes ;s carr;ed out under the condit;ons ;ndicated by the manufacturer (New England Biolabs~ Hethesda Research Laboratories and Boehr;nger ~annheim).
Where necessary, the phosphates of the 5' ends ~,:

~3~3~

are el;minated using either a bactcr;al alkal;ne phospha-tase or ~ ca~f intest;nal phosphatase, ;n the restrict;on en2yme buffer ~t 37C for 30 minutes.
The repa;r of the cohesive ends us;ng Klenow 5 polymerase (Boehringer Mannhe;m~ ;s carr;ed out at 25 C
for 15 minutes in a mixture 50 mmolar in Tris HCl, pH
7.8, 5 mmolar in MgCl2, 10 mmolar in ~mercaptoethanol ~nd 0~4 mmolar in dNTPs, ~ith the en2yme and 10 to 200 ~g/ml of DNA~
7he nuclease S1 (M;les) is used in an amount of 2 units/yg of DNA at 25 C for 30 minutes ;n a medium 0.3 molar ;n NaCl, 0.03 molar in ~aOAc, pH 4.8, and 0.003 molar in InCl2.
9al31 ;s used acrord;ng to the process of Panayotatos e~ al (reference 13)~ The ligation react;ons are carr;ed out at 15C (unless indicated otherwise) for

4 to 24 hours using DNA ligase T4 (~oehringer Mannheim) with a mixture 100 mmolar in NaCl, 66 mmolar in Tris HCl pH 7.5, 10 mmolar in MgCl2, 0.5 mmolar in spermidine, 20 O.Z mmolar in EDTA, 2 mmolar ;n DTT, 1 mmolar ;n ATP and w;th 0~1 mg/ml of BSA and 5 to 50 pg~ml of DNA.
About 30 units/ml of ligase are used for ligation of cohesive ends. About 100 un;ts/ml of ligase are used for l;gat;on of blunt ends.
Between the various enzymatic reactions, DNA
samples are extracted w;th a m;xture of phenol/chloroform and then precip;tated ~;th ethanol. Where necessary, tRNA of E. col; or of yeasts ;s used as carrier. Mole-cular adaptors (Collaborative Research, ~ethesda Research 30 Laborator;es, New England ~;olabs~ are prehybrid;zed and used in a 10 to 50 fold molar excess for the blunt DNA
ends, under ~he buffer condit;ons descr;bed above, w;th 100 un;ts/ml of l;gase T$ at 4C for 15 hours. If non-phosphorylated adaptors are used, the unreacted adaptors are removed d;rectly after ligat;on by prerip;tat;Qn ~ith spermine ~etrahydrochloride (Hoopes et al - refe-rence 14).
If phosphorylated adaptors are used, the l;ga-tion mixture ;s first extracted with a phenol/chloroform 3~?3~ $q~
W0 84/02918 - 19 - PCT/~R84/00014 m~xture and 1s then prec;pit~ted with ethanol, before spe-cif;c cutt;n~ with the appropr;ate restr;st;on en~ymes fo~lowed by precipitation ~ith sperm;ne tetrahydrochlo-r;de.
S Competent b~cter;al cells ~re prepared and trans-~ormed w;th plasm;ds Dr transfected by M13 DNA accord;ng to the procedures described by Dagert and Ehrlich (refe-rence 15).

The example of the construction of vectors accor-ding to the ~resent ;nvent;on described below is of course in no ~ay l;mitat;ve, and ;t ;s poss;ble to obtain vectors w;thin the context of the present invention by proceeding in a d;fferent manner.
The preparat;~n of a vector according to the in-vent;on f;rst ;nvolves construct;on of a plasmid conta;n-ing the or;g;n of repl;cation of p~R3Z2, the ampicillin resistance gene of th;s same plasmid, ampR, the promoter PLand the ~N gene, and, in addit;on, ;n the course of this 20 construction, it is necessary to endeavor to remove super-fluous restrict;on s;tes in order to be able to obtain unique restr;ction sites later in the course of the con-struction.
1) Suppression of the PstI s;te ;n p~R322 The base plasmid utilized ;s plasmid pBR322;
ho~ever, this has the disadvantage of having a Pstl res-triction site inside the ampR gene, since a s;te of the same nature w;ll be used later in the clon;ng ~one as the unique restr;ction site. It is thus appropriate to re-30 move th;s PstI res~r;ction site us;ng a mutant of plasm;d p8R322, plasmid pUC8, in wh;ch the ampicillin res;stance gene does not have a PstI restriction site (this site has been eliminated by mutation in vitro). p~R322 is marketed, in part;cular, by ~ethesda Research Labora-35 tories and pUC8 is described in the article of refe-rence 16.
To effect this, the 1,669 bp PvuI/PvuII fragment of paR322 is replaced by the analogous PvuI PvuII frag-ment from plasmid pUC8. To carry out this replacement, i:, ~3V3~
.. . .
~0 84/02918 - 20 - PCT/FR8~tO0014 the pl~sm;ds p9R32Z and pUC~ ~re treated success;vely ~;th PvuI ~nd PvuII ~nd are then c;rcularized by section by a iigase5 ~he resulting plasm;d pTG902 no longer h~s ~ PstI

5 restr;ct;on site and has also lost the NdeI restriction site present on p~R322 at the start ~not shown in Fi~ure 1). In addit;on~ plasm;d pTG902 carr;es a 50 bp portion corresponding to the lac;' se~uence, which includes the site PvuII.
10 2~ Insertion_of the promoter PL and the ~N gene, and preparation of plasmid pTG907 ,, The promoter PL and the ~N gene are isolated from plasmid pKC30 for insertion in pTG902, the segment re-moved also containing the operator OL on which the thermo-15 sensitive repressor cI850 ~ill act, as described ;n theexperiments. In addition, this man;pulat;on enables the EcoRI and HindIII s;tes to be el;minated~ while a unique BamHI s;te ;s preserved downstream of the N gene to enable A cIIrbs to be inserted subsequently. Plasm;d pKC3û is Zû described ;n reference 17.
pTG902 ;s cut ~t the un;que EcoRI restriction site and the protruding 5' ends are eliminated by treatment ~ith S1 nuclease. After digestion w;th ~amHI, the larger fragment is gel purified.
The fragment from pKC30 carrying the promoter PL
and the ~N gene is prepared in the same manner, w;th suc-cessive treatment of the plasmid ~ith PvuI, S1 nuclease and aamHI. After gel purification, the fragments are sub-jected to the act;on of the ligase, which leads to fusion 30 of EcoRI and PvuI and reconstitution of the BamHI s;te.
The ligation mixture ;s used to transform compe-tent host cells TGE900 2t 30C. Th;s strain contains a deleted ~ prophage, ~cI857~Bam~HI, ~hich provides the thermosensit;ve ~ repressor, cI857, which is necessary 35 to block transcription starting from PL.
This is important, since the activity of PL ;n an N background is lethal because of the N antieermina-tion function.
According to analysis by restriction enzymes, the ,.~, ~3~3~

clones con~ain pl~smids of the correct structure, snd they ~re then tested for lack of viability ~t 42C.
One of the plas~ids obt~ined, pTG906, is tre~trd to delete the PvuII-S~lI segment so as to elimin~te the restrict;on sttes conta;ned in this segment and ~l~o ln order to remove the t~o extreme restr;ct;on sites. To ef-fect th;s, pTG906 ;s treated success;vely ~;th SalI, S
nuclease, PvuII and the l;gase.
This g;ves plasm;d pTG907, ~h;ch contains all the elements ment;oned at the start of th;s stage and, fur-thermore, ;s Pst Eco H;nd Sal Ava Nde PvuII . the synthes;s of th;s plasmid is represented in F;gure 1.
3~ Cloning of ehe AcIIrb~ re~ion The second ;mportant phase ;n the synthesis con-s;sts ;n ;nsert;ng the ~cIIrbs region, in the form of anAvaI/TaqI fragment, into the start of the lac~' gene (~fragment of B-galactosidase) wh;ch has been cloned ;n the phage M13 des;gnated M13tg110. Th;s strategy perm;ts a s;mple funct;onal test for the rbs, namely product;on of the lacZ' prote;n, and, as a result, makes it poss;ble to obta;n blue plaques in the presence of IPTG and Xsal;
this also permits rap;d sequencing of the construction us;ng the so-called dideoxy method.
Various der;vat;ves of phage M13 will be men-tioned ;n the course of the exper;ments. Phage M13tg110 has been mentioned above, and other phages of the same type, the construction of wh;ch w;ll be reviewed belo~, will be used later ;n the descr;ption.
The method of construction of these vectors is indicated in Table I belo~, ~h;ch shows the references of the starting phage~ the nature of the restr;ct;on enzyme with wh;ch ;t has been cut, the part;cular treatment to wh;ch the fragments thus obta;ned have been subjected and the nature of the ;nsert ~hich has been attached in the s;tes thus revealed Phage M13~p7 ;s marketed, in part;cular, by Messrs. ~ethesda Research Laboratories~
M13mp70~ ;s obta;ned from M13mp7 by replac;ng the PstI/EcoRI fragment to the right (CTG CAG..~ GAA TTC) by - ~L3~53~
W0 84/02918 - 22 - PCT/~R84/90014 the sequence: CTG CAG CAA TCC.
The constrl~ct;on pr~nclp~e ;s described in the ~ol lowing scheme:

A7G ACC JllC All ACG 1At ICC CCC t1T Ctt TCG AtC ICC ~CC AlT tCl tTC cce 1~13-~701 G~ HI
.
AlCACCitGlrTACtAATTCCCCC6ATCCCTCCACCItCACC~AIlCACTGCCC
lACTCClAtlAATGCTtA~GG6GCCtAC GCAGCTtCAt6TGGlTA~tF6ACeG6 . . i 1 ;
~nA polr~

AtGACCAlGAtTACGAAlTCCCCGGAlC GATCCGTCGACClCC~GCAATTCAClGGCC
TACltClACllAlCCTTAAGtGCCClAC CtAGGCACCTGGACCTCGllA~C16ACCGC
. . .
Linktr HindIli 1, . . . ..
ATG~CCATGl.TlACCAAllCCCCGGAlUCAAGCTTG6 CCAAGClTGG61TCC6TCG~CCTGCAGClRIlClrTCGeC
~hC76CtAtTlAT5CTTilGG6tCC7~GG6JTCCdltC GcTtcG~AcctTAG6cl5ETGGAcctEGT-fAAG~ ccGc 1.
lTCACCJ~TC1TTACGAATTCCCC6CA~CCCA AGCTTG66AlCCCTC6ACCTGCAGCAAT7eiC7CGCe tAClCCtACTAAlGCTtlAtGGGCCTAGGGTlCea AC~CTAGGCAGGTGCACtlCGTTAlGTGACC6~

AlG ACC AT5 ATT ACG AAT ICC CCC GAt CCC AAt l;TT GCC Rlt Ctl C5A CCT GCP. tCA ~IT C,SC TGG CC

Etq It~ Hl Hind 111 I!lan HI 5~ tst~

~13~
- 23 _ ~~ ' wl~l E ~

C~ o _ U _ U ,c~ U
. -~ ~, -o,_, u~ C ~ C cr c . a.
.
. ~) .

J~ C ~ - D C

Il~ Ic ~u cr L c~ Z~_ a Q' U~ ~ .
C ~ ~

~ ., ~

~U U~ CU V o ~ -_ o~ V V ~
. ~J ~{~ W .~ o L U~
O O
_ n Q
~o ~ ~ O - O
~ ~ ~ E E ~ o o al cL . ,_~ ~ ~
A 'SE ~5E: L L
. O O
ID 11~

O O ~
n'0 _ _ Ql EE
E ~ _ ,_ . Z SE~ ~D

r . . 13~3~3~
, Thc protocols for use of the enzymes trestr;c-tion, DNA poLymerase~ T4DNA li~ase) ~re those recommended by the suppliers. The H;ndIII l;nker is ~ synthetis oli-gonucleotide w;th the se~uence 5p CCAAGCTTGG 3' (Collabo-rative Research Inc~, W~lth~m, Mass 0215~, U.S.A.).
The cIIrbs re~ion ;s isol~ted from plasm;d pOG11(ref 18~ which cont~ins a ~HaeIIItSau3A fragment ~hich extends from the cent~r of the crs gene to the center of the re~ion encoding cII tcro and cII being involved, in particular, ;n the regulat;on of lysogenesis of bacterio-phage ~).
The 186 bp Aval/TaqI fragment containing the cIIrbs region in pO611 ;s exc;sed and treated with Klenow polymerase. In addition, phage M13tgllO ;s treated ~ith ~amHI and then w;th Kleno~ polymerase, followed by treat-ment w;th calf ;ntest;nal phosphatase. The fragments ob-tained are subjected to the act;on of Tl ligase~
Exam;nation of the sequences ;n the pOG11 frag-ment conta;n;ng the cIlrbs region and ;n phage M13tg11~
allows the predict;on that ;nsert;on of the cIIrbs frag-ment in the ~amHI site of M13tgl10 ~ust lead to the pro-duct;on of blue plaques after fermentation of $he trans-fected bacter;a, tak;ng ;nto account the fast that the lacZ' gene is in phase for translation starting from AUG
of the cII gene, wh;le the plaques corresponding to the parental M13tg110 strain are wh;te.
The blue plaques are collected, the colon;es are then selected by restriction enzyme analys;s of min;-preparat;ons and ;t ;s then ver;f;ed by sequenc;ng that the correct 30 construction is obtained.
The resulting clone M13t~910 is obta;ned, the over-all structure of ~h;ch ;s shown ;n the ~ower part o~ F;-gure 2, and the deta;led structure in F;gure 3.
It ;s found that ;nsertion of the cIIrbs fragment resonstructs the upstream UamHI and AvaI s;tes and the downstream ~amHI s;te.
Translat;on start;ng from AUG of cII leads to fu-sion of ~he 13 terminal am;no ac;ds of cII to the 8 NH2-term;nal amino acids of the protein lacZ'.

.

`

3~
.

WO 8~/02918 - 25 - PCTtFR84/OOOt4 ~) Insertion of the cIIrbs fr~gment 1nto the plflsmid p~G907 __.
The ~hird st~ge of this synthes;s consists in transferrin~ the cIIrbs/l~cZ' fragment of phage M13tg910 to the pT6907 plasmid vYCtor prepared ~s descr;bed above.
To effect this, ;t ;s appropriate first to elimi-nate the EsoRI, B3mHl and AvaI sieeS upstream of cIIrbs and then to insert 3 ~glII site.
Under these cond;tions, cIIrbs can be exc;sed as a aglII-8glII fragment and placed into the ~amHI s;te dounstream from the promoter PL and the ~N gene of pTG907.
Phage M13tg910 ;s digested w;th EcoRI and then treatéd ~ith ~al31, followed by Klenow polymerase. the fragments obta;ned are then sub;ected to the act;on of the l;gase ;n the presence of non-phosphorylated BgllI
adaptors. The l;gation m;xture obtained ;s used to trans-form competent JM103 cells. The blue plaques are then selected. These clones are then analyzed ;n order to verify that they contain a ~glII s;te and that they no longer contain EcoRI or ~amHI upstream sites. Clones such as M13tg912, the structure of ~h;ch ;s shown in Fi-gure 4, are thus obtained.
Treatment with Bal31 produced a deletion of 101 bp, eliminat;ng the EcoRI, ~amHI and AvaI s;tes; as ~ell as the lac ATG and lac Sh;ne/Dal~arno sequences. The 99lII s;te ;ntroduced ;s placed about 100 bp upstream from the cII ATG and 10 bp downstream from PlaC.
As stated above, transfer of the BglII-BglII
fragment conta;ning cIIrbs and lacZ' of M13tg912 into the 3û ~amHI site of pTG907 prepared at the or;g;n was env;saged.
However, th;s construct;on has not been possible, since it proves to be lethal to the host strains, ~hich is why it has been necessary to develop a d;fferent strategy us;ng an Hgal~SphI adaptor which allows insertion of a BglIItHgaI fragment carry;ng cIIrbs and lacZ' between the 9amHI and SphI s;tes of pTG907.
The ~amHItSphI fragment of ptG907, the BglII HpaI
frag~ent carrying cIIrbs and lacZ' and the phosphorylated adaptor were prehybr;dized in a molar rat;o of 1:2:1 and t~. ~

~l3~3~

W0 84/02918 - 26 - PCT/FR84~00014 then tre~ted w~th the T~ se. Al~quots ~re used to transform strain 6150 competent cells at 30C.
The rells of ;nterest ~re identified by screen;ng tr~ns~ormants with a 32p-l~beLed cIIrbstlac~' fr~gment, and the construct;on ;s conf;rmed by an en~ymatic re-strict;on study.
In order to give an initi3l ind;c~t;on that the various elements of the e~press;on system operate DS de-sired, the plasm;d obt~;ned, pTG908~ ;s transferred ;nto a host strain N6437, ~h;ch possesses both c1857 and the ~galactosidase ~ fragment ~h;ch complements the frag-ment encoded by the plasmid.
; When gro~n on a plate conta;n;ng IPTG + Xgal, the transformants obta;ned are white at 28C and turn blue 5 w;th;n about 30 minutes on transfer to 42C.

It ;s poss;ble to use an alternat;ve promoter to PL, ;n the same manner as descr;bed above.
p~ constructed by V. Pirotta carries the cI857 gene, PR and a part of the cro gene cloned in pBR322.
Th;s clone contains a ~glII site in the cro' gene downstream from the promoter PR (r;ghtward promoter), wh;ch is under the control of cI857 encoded by the plas-m;d.
The a gene, P ' R~ the 6S RNA gene and part of the late S gene are conta;ned in a ~clI fragment (reference 19). Th;s fragment can be ;nserted ;nto the BglII s;te of p8~H3, thus putting the Q gene under the control o~ PR
of the plasmid. P'R, in add;tion to all the cascade con-trol system, w;ll thus be conta;ned in an EcoRI fragment, which can be used to repLace PL and N, ~hich are placed in a 3glII/AvaIII fragment in pTG908. This leads to a con-struction analogous to pTG908~ in ~hich however trans-cription is controlled by P'R.
The follow;ng examples relate to cloning of the gene which encodes ~1-antitrypsin.
To effect th;s9 2 cDNA clone ~hich encodes the eomplete sequence of mRNA encoding human ~1-antitrypsin has been isolated. This sequence has been inserted into ' , ~3~3~

a previously prep~red b~cter;al plasm1d express;on vec-tor , Dnd the synthesis has been demonstrated 1n ~ soli of a lsrge quantity of polypepttdes ~hich react ~ith an antibody corresponding to natural human ~1-antitrypsin and which exhib;t, in vitro, the bi,ological ~ctiv;ty of 1-antitryps;n and its capacity for inhibiting the ac-tivity of elastase. This ~1-antitrypsin produced from bacteria const;tutes the basis of a product ~hich can be used in replacement therapy for the treatment of d;s-orders involving an ~1-antitrypsin def;c;ency.

Cloning and isolation of a complete cDNA which encodes human ~ -antitrypsin, and expression thereof in E. coli 1) Cloning of the cDNA of Q -antitrypsin ~ 1 a) Isolation of polyA mRNA from human liver and detect;on of the activ;ty of ~ -antitrypsin - -mRNA
Human livers are obtained post mortem and are ra-p;dly frozen in liquid nitrogen. RNA is prepared from 5 g of liver using the guan;dine hydrochloride method, as described in reference 20. The RNA is chromatographed over a poly-U-Sepharose (Pharmacia) column ~or enrichment in polyA mRNA.
Total RNA and polyA mRNA are translated in a rabbit reticulocyte lysate treated ~ith the nuclease (BRL~, applying the conditions previously described in reference 20. In order to study post-translation modifi-cation, pancreas membranes from dogs (reference 21) are added to the tran~lation system. The immunoprecipitations us;ng a commerc;al anti-n1-ant;trypsin ant;serum (Nordic Immunology) are described in reference 22.
Figure 5 sho~s the immunoprecipitate`s from the translation products of total RNA from human l;ver. These RNA's have been translated in rabbit reticulocyte lysates treated with 300 ~g/ml nuclease in the presence or ab-sence of microsomal membranes from dog pancreas. L35s}
meth;onine was used as a rad;oactive tracer. ~ uots conta;ning about 106 cpm of TCA ;nsoluble radioactivity .;.L .

~'., ~3~S~
WO 84/02912 - 28 - PCT/FR84tO0014 sre subjected to 10X polyacrylam;de/SDS geL-electrophore-sis followed by fluoro~raphy ~nd autoradiography.
Lrne 1 repre5ents radio~ctive protein standards~
the s;~e being given ;n 10 d3ltons;
Lane 2 represents the immunoprecipitates ~;th ~1-antitrypsin antiserum in the presence of membranes;
Lane 3 is identical to lane 2, but the products obtained have been prepared in the absence of membranes;
and Lane 4 is as lane 3, bu~ us;ng a non-;mmune anti-serum.
Immunoprec1p;tat;on of the translat;on product of toCal RNA from human liver w;th ant;-a~-aneitrypsin antiserum ;ndicates ehat a specific polypeptide of mol~-cular we;ght 45,000 daltons is precip1tated when the translation ls carried out ~n the absence of microsvmal membranes. Addit10n of m1crosomal membranes to the tran-slatlon leads to a slightLy larger ;mmunoprec;pitated polypept;de (4B,OOO daltons). This ;ncrease ;n the size of the product o~ the mRNA o~ a1^antitr'yps;n in the presence of m;crosomal membranes is probably due to a post-translat;on mod;f;cat;on, such as glycosylation.
The small and large ;mmunoprec;pitated polypep~;des are allowed to react competitively by addition of,non-labeled natural human ~1-ant;tryps;n, and r,onf;rms (s;c) the ;dentity of the translation products~ Thus, a b;ologi-callyactive messenger RNA of ~1antitryps;n can eas;ly be detected both in total RNA and in the polyA mRNA prepared from human liver; the amount of ~-AT mRNA is of the order of 1 to 5% in the l;ver.
b) Synthesis of cDNA and preparation of a human liver cDNA clone bank , PolyA mRNA ;s used as a te~plate for re~erse trans~ription ut;li~ing an oligomer ~dr)12 18 as a primer.
Complementary DNA ;s synthes;zed ;n a 100 ~l of reaction m;xture conta;n;ng 100 mmolar Tris HCl pH 8.3, 50 mmolar KCl, 8 mmolar MgCl2, 30 mmolar B-mercaptoethanol, 1 ~9 of ol;go ~dT)12 18~ 5 ~9 ot polyA mRNA, 500 ~molar dATP, dCTP, dGTP, dTtP and 80 units of avian myeloblastosis ~3V3~
WO 84/02918 - 2~ - PCT/~R8~/00014 v;rus reverse tr~nscr;pt~se (L;fe Sci~nces Inc.~ St~
Petersberg, Flor;da).
After incubation at 42C for ~5 minutes, the re-0ct;0n ;s stopped ~nd the cDNA conplex ;s denatured by heatin~ ~t 100C for 3 m;nutes and r~p;d transfer to an ice-bath.
For synthesis of the second DNA strand, ~he reac-tion mixture of the first strand is diluted 5-fold and adjusted to a final concentration of 100 mmolar Hepes-KOH, pH 6.9~ 100 mmolar KCl, 200 Pmolar dATP, dGTP, dTTP, Z00 ~umolar ~2P]-dCTP of specific activity 0.5 Ci/mmoleJ and 10 units E. coli DNA polymerase Klenow fragment (Boeh-r;nger Mannhe;m).
Incubation is carried out at 25C for 2 hours.
The yield of double-stranded cDNA (dscDNA) as measured by radioactivity ;s 970 ng. The react;on mixture is extrac-ted with an equal volume of phenol:chloroform (50:50) saturated ~ith 1û mmolar Tris HCl pH 7.5~ and 1 mmolar ED~A and the cDNA ;s precipitated ~;th ethanol.
The dscDNA is rendered blunt-ended by digestion with 5 units of S1 nuclease in a react;on volume of 100 ~l conta;nins 30 mmolar sod;um acetate pH 4.8, 300 mmolar NaCl and 3 mmolar ZnCl2.
After 1 hour at 37C, EDTA and SDS are added to a final concentration of 10 mmolar and 0.1% respectively, and the reaction mixture ;s heated at 65C for 10 minutes.
The dscDNA treated w;th S1 is then applied to a l;near 5-20% sucrose gradient in a buffer containing 100 mmolar Tris HCl pH 7.5, 1nO mmolar NaCl and 5 mmolar EDTA and is centrifuged at 30,000 rpm at 15C for 15 hours using a ~eckman SW60 Ti rotor. 0.5 ~l fractions are collected and the amount of cDNA in each of the fractions is deter-mined by measuring the ~CA-insoluble rad;oact;vity in aliquots of each of the fract;ons. ~he s;~e of the dscDNA
in each of the fract;ons ;s measured by analys;n~ al;quots on neutral agarose gel followed by autoradiography. The fract;ons conta;ning dscDNA of average size 1 kb or more are collected~ 5 yg of E. col; tRNA are added as carrier ___ ~ and the nucle;c acids are recovered by prec;pitation ~ith L3~5~'~
W0 84/02918 ~ 30 - PCTIFR84/0001 eth~nol followed by centrifu~ation.
The dscDNA is dissolved ;n ~ small volume of 1/1Q
x SSC (15 mmolar NaCl 1.5 rmolar sod~um c;træte pH 7.0).
100 ng of double-stranded sDNA ;s tailed at the S 3 end with ~ homopolymeric dC extension by ;ncubation in a 30 ~l resction m;xture conta;n;ng 140 mmolar sodium cacodyl~te pH 6.8~ 1 mmolar cobalt chloride 0.1 mmolar d;thioerythr;tol 100 ~molar [3H3-dCTP of spec;f;c ac-t;vity 6.7 Ci/mmole and 5 un;ts of term;nal deoxynucleo-tide transferase (~RL) for 15 minutes at 37C.
Short homopolymeric dG tails (appro~imately 13residues per 3 end~ are ~dded in the same reaction mix-ture to plasmid p~R322 prev;ously digested with the en-zyme PstI. The plasmid pBR322 thus obtained is then pur;f;ed from the react;on mixture by extraction ~ith phenol follo~ed by preparat;ve electrophoresis and elec-tro-elution on a lX agarose gel.
The dscDNA rbtained directly from the reaction m;xture ;s m;xed ~ith p~R322-dG in a rat;o of 15 ng o~
2~ cDNA to 500 ng of the vettor ;n ~50 ul of 10 mmolar Tr;s HCl pH 7.5 lQ0 mmolar N~Cl and 1 mmolar EDTA heated at 65C for 10 m;nutes and then incubated at 42C for 2 hours. The react;on m;xture ;s then used to transform (reference 23) E. col; stra;n 1106 and the transformants are selected on an L~ agar plate contain;ng 15 ~g/ml of tetracycline. A total of 15 000 recomb;nants are obta;ned from 30 ng of dscDNA approximately 50% of which contain cDNA insert;ons greater than 1 kb. All the transformants are scraped off the plates and suspended in L~ med;um conta;n;ng 15 ~/ml tetracycline and the bank is stored at -20C a~ter add;t;on of glycerol to 50%.
c~ Strategy for ;solat;on of the ~ Y~ 'n clo_es The sequence of 306 bp of an ;ncomplete human ~1-antitrypsin rDNA clone encoding the 69 terminal aminoac;ds and including 66 nucleotides of the 3 non-encod;ng ; sequence (reference 24) is analyzed by computer in order to determine the oligonucleotide sequence ~hich must be used as a probe for an ~1-an~itrypsin cDNA clone. This . .

~ ' . .
:, .

~3~`3~
WO 84/02918 ~ 31 - PC~/FR8~/00014 ;s carried out by d;vid;n~ the 306 b~se p~irs into ~ll poss;ble groupings of 21 bases (21-mers), which are then analy2ed sep~rately for their possibility of forming a double-stranded intramolecular structure (hairp;n). The 21-mer chosen, 5'-TGAAGCTCTCCAAGGCCTGTG-3' (sho~n in Fi-gure 6), conta;ns no hairp;ns longer than 2 bp. The com-plement of th;s sequerlGe, that ;s to say 5'~
GCACGGCCTT6GAGAGCTTCA-3', is synthes;zed on a sil;ca gel support as has been descr;bed previously (reference 25).
After labeling of the 5' end, the s;ze of the oligonucleo-t;de ;s tested by polyacrylam;de gel electrophores;s and the sequence ;s determined by the Maxam and G;lbert me-thod.;
d) Analysis of the human l;ver cDNA bank .
About 15,000 colonies are grown overnight on LB
agar plates containing 50 ~ug/ml of tetracycl;ne. The follow;ng day, the bacterial colonies are transferred on-to Whatman 540 paper (reference 26). The bacteria re-maining on the plates of origin containing tetracycline zo are regrown for several hours at 37C~
In order to prepare the bacterial plasm;c D~A for hybridization, the filters are washed with 0.5 molar NaOH
for 5 ninutes. After drying in air, the filters are neu-tralized by sequential ~ashing ;n a buffer of 0.5 molar Tris HCl pH 7~5 2 x SSC. The filters are then rinsed ~ith alcohol and dried in air. The rema;ning bacterial debris is eliminated by digestion ~ith Z5 ~g/ml of pro-teinase K ;n 50 mmolar Tr;s HCl pH 7.5, 5 mmolar EDTA and 0.5X SDS for 2 hours at 37 C.
The hybrid;zat;on probe ;s prepared by label;ng the 5' end witb polynucleot;de k;nase (PL B;oche~icals) using 80 n~ of 21-mer ~;th 5û uCi of~ -~32P~-ATP. Hybr;-`d;zation is carr;ed out at 37C overnight ;n 40 ml of

6 x SCC, 1 x Denhardt's solution, 10 mmolar sodium phos-phate pH 7.5, and 5D% formam;de. The filters are then ~ashed with 6 x SSC, 0.1X SDS, several times at room tem-perature. The filters are dried and subjected to auto-radiography overn;ght. Very strong pos;t;ve s;gnals can be clearly detected over background hybridization.

. ", WO 84/02918 - 32 - PCT/FR8~/00014 e) Identif;c~tion of recomb;n~nt pl~smids cont~in;ng ~ acter;al colo~ies corresponding to the strong positive signals are ;dentif;ed by orienting the suto-radiograph with the or;ginal plntes containing tetracy-cl;ne. Pl~sm;d DNA s ~re prepared from small cultures of each of the bacterial colonies treference 27), ænd the size of the cDNA insert of e2ch plasmid is determined by digestion with PstI. One plasmid, kno~n as pTG603, is found to contain an insert of about 1.6 kb possessing 2 AvaI sites and one unique Ba~HI site. The s;ze of the fragments produced by double cligestion with AvaI + PstI
t750 base pairs) is in agreement with that pred;cted for a full length ~1-antitrypsin clone ~references 24, 28).
js In order to confirm that this plasmid in fact contains the ~1 ant;trypsin cDNA insert~ the insert is transferred as a PstI fragment to phage M13mp8 after excision ~ith PstI, and is sequenced using dideoxynucleotides as chain terminators. The sequence obtained (Figure 7) confirms 2û the identity of the cDNA as ~hat of ~1-ant;tryps;n. The sequence obtained is compared ~ith the corresponding re-gion known for human DNA ~h;ch encodes ~1-antitrypsin.
The upper part of Figure 7 shows the partial sequence of the cDNA cLone of complete ~1-ant;tryps;n, and the lower part shows the part;al sequence of pTG6û3 as has been de-term;ned. The homology between the two nucleotide se-quences ;s shown by asterisks. X represents nucleotides which have not been ascertained.
2) Express;on of human ~ -ant;tryps;n cDNA ;n E. coli 3û The human ~ I-antitrypsin cDNA clone pTG603 con-ta;ns a un;que ~amHI restriction site immediately after the codon for the f;rst am;no ac;d of the mature protein.
The capacity ~or cod;ng the ~hole mature polypeptide, w;th the exception of the ;nit;al glutam;c acid, ;s thus conta;ned in a ~amHI/Pstl fr2gment, wh;ch has been cloned ;n ~he expression vector pTG920; in th;s construction, transcr;ption starts at the left~ard A promoter PL, and translation is ;n;t;ated at the AcIIrbs AT~ ~h;ch, accom-pan;ed by the ribosome b;nd;ng site and first 39 bp of ' b, ~P3~
WO 84/~2918 ~ 33 ~ PCTtFR84/00014 the ~clI gene, are fused to the st0rt of the locZ gene, as shown.
A bondin~ region cont~;ning un;que restrict;on s;tes is situ3ted at the junction bet~een cII and the lacZ sequence. Plasm;d pTG920 is a derivative of pTG908 prepared in E~mple 3, in wh;ch the original ~amHI/PstI
fragment situated ~0 bp do~nstre~m from ~he cII ATG ;n pTG908 ;s repl~ced by a ~glII/PstI fragmen~ of M13tg115, as descr;bed ;n F;gure 8.
As a result of th;s process, a Bam~I site is ob-tained ~h;ch is placed in the same translational reading frame as the HamHl site of the Q1-antitrypsin gene.
; F;gure 8 sho~s the cloning strategy allowing pre-paration of the plasmid pTG920 from plas~;d pTG908 and phage M13tg115, and indicates the fragment of pTG603 to be cloned. The sol;d l;nes represent pro~e;n-coding se--quences and, ;n the case of pTG603, represent the sequence wh;ch encodes the mature ~1-antitryps;n polypept;de. The cross-hatched areas represent the regions wh;ch encode the 13 N-termina~ am;no acids of cII, ~h;ch w;ll finally be fused to the ~1-antitrypsin.
Insertion of the BamHI/Pstl fragment of pTG603 between ~amHI and the Pstl s;te of pTG920 w;ll thus lead to expression of a fused polypept;de conta;n;ng ~starting z5 from the NH2 terminal) 13 a~ino acids of the cII protein, 4 amino acids der;ved from the adaptor sequence and the mature ~-ant;tryps;n polypeptide, ~ith the exception of the glutam;c ac;d of the NH2 term;nal.
pTG920 treated w;th BamHI~PstI and alkal;ne phos-phatase is l;nked to pTG603, d;gested w;th 9amHI and PstI~ Transformant TGE900 cells carrying the ~1-anti~
trypsin fragment are ;solated after restriction enzyme analysis. One of these clonesO pTG922, is chosen for study ;n greater deta;l.
As has already been ment;oned~ the vectors accor-din~ to the invent;on can allo~ the prepara~;on, by a variant of the above process, of the non-fused prote;n.
To effec~ th;s, pTG920 ;s digested w;th NdeI and pTG60~
w;th BamHI as above~ and the fragmengs obta;ned are then ` ~3~35;3~
W0 84/02918 _ 34 r PCT/FR84~00014 l;yated by means of ~n ol;gomer;c ~d~ptor, cs is ~escr;bed below, or the sare operat;ons ~re c3rr;ed out on the vec-tor pTG922:

Nd~I ~amHI
J, cII
5 ' Ct~T~TG~ ~GÇATCC 3 ' 3 ' ~-GTATAC~ --~CCT.~G6 5 ' t NdeI E~amHI

TATGGAG 7~me r adaptor ~CCqCC~AG 9~mer TATG GAG GAT~
------G'r AT~CCTC CTA---~

Th;s junct;on reconst;tutes the starting codon ATG, to which ;s fused the gene wh;ch encodes ~1-anti-tryps;n.
Since cells of the E. col; stra;n TGE900 conta;n e;ther the express;on plasmid pTG922, wh;ch conta;ns the human q-antitrypsin ~ene, or the plasmid PTG9?0 alone w;thout ;nsert, the prote;ns are labeled w;th L3ss]-meth;onine for 1 hour at the temperatures ;nd;cated above.
Extracts are prepared as described ;n reference 7 and al;quots are analyzed on 10% polyacrylam;de/SDS gel, -fol-lowed by fluorography and autorad;o~raphy.
The results are shown ;n F;gure 9A:
Lane 1: rad;oactive molecular we;ght standard Lane 2: cellular extract with pTG922, growth at 28C (without ;nduct;on) Lane 3: as for lane 2, but growth at 42C (in-duction) ~ane 4: cell extract contain;n~ plasmid pTG920, growth at 28C, Lane 5: as for lane 4, but growth at 42C.
The results found sho~ the ;nduction of a strong , 3~;~3~

band having 3 mo~ecu~ar wæight of about 45,000 daltons ttigure 9A). Dens;tometr;c trac;ng of the autor~d;ogram indic~tes thot the omount ot n1-~ntitrypsin produced rep-resents between 7.5 and 15X of the tot3l proteins of S E coli ~no adjustment has been made for potential d;ffe--rences bet~een ehe methionine cDntent of the l1-antitryp-sin and that of the total E.-col; proteins).
Labeled extracts of E. coli cells carrying the plasmids pT6920 and ptG922 are prepared as ment;oned above. Al;quots are subjected to immunoprecipitation with an anti~ ant;trypsin ant;serum, as is descr;bed for Figure 5, and are analyzed on 10~ polyacrylam;de/SDS
gel, f;ollowed by fluorography. ~he results are sho~n in Figure 9L1:
Lane 1 corresponds to E. coli TGE900 extracts conta;n;ng pTG922 thaving the ~1-antitrypsin sequence~
labeled at 4ZC and immunoprecipitated with 1-antitryp-sin antiserum, 100 ~l of extract, Lane 2 is as for Lane 1, but with 2 ~l of extract, Lane 3 - blank ~Lane 4, as for lane 1, ~;th a cellular growth a~
a non-permitted temperature ~28C), 100 ~l of extract, Lane 5 shows cell extracts containing the parent plasmid pTG920, labeled at the induction temperature o~
42C, 100 ~l of extract~
Finally, F;gure 9B2 shows the results of immuno-competit;on of the ~1-antitrypsin synthes;zed by E. col;.
The cellular extracts of E. co~; conta;n;ng pTG922 are labeled at 42C as has been descr;bed above. Aliquots are used for the following immunoprec;p;tations:
Lane 1: ;mmunoprecip;tation w;th the anti~
ant;trypsin ant;serum, Lane 2: as for lane 1, but in the presence of 10 ~g of non-labeled natural human ~-antitryps;n, Lane 3, as for lane 2, but 20 ~9 of the competi-tor ~1-ant;trypsin, and Lane 4, as for lane 3, but with 50 ~9 of the com-petitor~
These exper;ments show that the product of ,~

" ~3~S;3~
W0 84/02918 - 3~ ~ PCT/FR8~/00014 lmmunoprecip1t3tion h~s the same molec~L~r wel~ht ~s that ~hich is prec;pitated from the tr~nslation product free from cells originatin~ fro~ ehe ~RNA of hu~n l;ver ;n e he a~sence of membr~ne 5 .
3) a;olos;c~l act;vity of~
E. coli Crude extracts of pTG922 ~re prepared and aliquots are tested for the;r inh;bitory activity agalnst porcine pancrea~ic elastase.
To test the b;olog;cal activ;ty of 1-anti~rypsin prepared from E. col_, crude e~trasts o~ culture are pre-pared as described above, in the absence of a protease inh;b;tor, w;th the exception of an ;nhib~tor of the trypsin from soybean. The exper;ments on elastase are carr;ed out accord;ng to a modif;cation o~ the method o~
Werb ~reference 29)~ Crude extracts of samples of ~-ant;tryps;n àre pre;ncubated w;th 0.25 ~9 of por~ne pan-creat;c elastase for 30 m;nutes at room temperature, be-fore add;ng a ~3H~-elastin substrate and incubat;on at 20 37C for 15 hours. 300 ~l samples are th~n centrifuged and the percentage inhibition is determined by radioactive counting of the supernatants.
The samplcs are as fol~o~s:
1 ) e l a s t a s e 2) elastase 4 50 ~L of induced extract of pTG920, 3-4) eLast~se ~ 20 ~l of induced extract of pTG920 ~ 5 ~9 and 10 ~9 of human a1-anti-tryps~n (Sjg~a) 5-~-7) elastas~ ~ S ~l ~ 20 ~L ~ S0 ~L of induced ex-tract of pT~22 and 8~ eLastase ~ 50 ~l sf non-induced extract of p~G922.
The re~ults sho~ tFigure 10~ that co~pLet~ inhi-bi~ion of the ela~tase i~ obta~ned onlr ~ith extract~ in-dueed br pTG922, and not ~n the nGn-induc~d extracts of pTG922 or pTG~20, ~hleh indicate~ the pre~enc~ of bio-Lo~caLL~ active a1-ant~tryps~n ~n ~he induc~d extr~cts.
~, ~"'' ' ' ~ .

~3~3~ii3~
W0 84/02918 ~ 37 - PC~/FR84100014 4) Estimat;on of the amount of ~ ~ ps;n produced by the_b3cteria From polyocrjlDm1de gels of the labeled induced extrac~s ;t is estimated that 1-snt;trypsin ronstitutes about 15~. of the totæl cellul~r protr;ns.
Assuming 1 pg of protein per cell this ~ive5 0.15 pg of u1-2nt;tryps;n per cell.
After ultr~son;c treatment and centrifugat;on~
about 50% of the Q1-ant;tryps;n ;s lost in the res;due:
the percentage of ~1-ane;tryps;n based ~n the total cellular prote;ns ;n the supernatant is thus 7O5%.
The amount of ~1-antitrypsin produced can also be est;mated from measurements of the ant;-elastase ac~;v;ty.
It has been found that 1ûOX inh;bit;on of 0.25 ~9 of pan-creat;c elastase from p;gs ;s obta;ned w;th 10 ~ug of hu-nan 1-antitryps;n prepared from serum ~in the presence of control bacterial extracts). An equ;valent ;nh;b;tion level ;s observed w;th 50 ~l of ;nduced bacter;al extracts conta;ning ~1-ant;tryps;n that is to say:
< ~ S0 ~l ~f extract - 10 ~ of al-~nt~tryps~n 1.5 ~l of culture z 10 ~9 of ~1-antitrypsin >1.5 x 108 cells = 10 ~ of a1-antitrypsin < - -~150 cell~ ~ 10 ~9 of a1-antitrypsin (sic) 1 cell = ~O.OS pg (that is to say 6X
o$ the total cellu-lar proteins)~
This is ~n good agreement ~ith the leve~ of 7n5X
estimated fro0 the gels after sonication and centrifuga-tion. This calculat;on indicates that the ~-anti~rypsin produced by the bacter;a has an anti-elastase act;v;ty level comparable to that of ~1-antitrypsin prepared from serum and that 1-antitryps;n is produced at a level of about 15 mg/l;tre of culture.
5) ~ 1 ~
Hu~an a~-~nt1tryPsin 1s ~ ~ycoprotein ~hich contain~ thr~e carbohydrat~ s~de ch~;ns. The~e eh~ins are of the N-glucos~de typo eonta1ning N-aeetyl-neur2~-nic ~c1d gal~ctose manno~e ~nd N-~cetyl ~lycosamine and are l;nked to the prote;n by means of an asparginyl ,.3~, ~,, ' '" . , , '" ' `"

~3~ 3~
W0 84/02,18 - 38 - PCT/FR84/00014 rad;c~l. The ol;cJos~ccharides exist either in the A form ~or ~;th ~ branch;ng po;nts) or in the ~ form ~1 brnnching po;nt).
GlyCOsylat;on is probably not necessary for the S biolog;cal activity of the protein, since the product r,b-tained from the bacteria is act;Ye.
It may be necessary to glycosylate the product obta;ned by the bacterium im order to ensure better sta-bility. This is possible in Y;tro, s;nce the cell-free translation product obta;ned from ~1-antitrypsin mRNA in the presence of membrane increases in size as a result o~
glycosylat;on.

., l3t,k3~
WO 84/02~18 - 39 - PCT/~R84/00014 E~AVPLE 4 Obt~;n;n non-fused a-antitrypsin (Fi~ure 11) the start;n~ pl~sm;d ts plasmid pTG922 described ~bove~
P~asm;d pTG922 i~ subjected to compLete restric-tion w;th the restrict;on enzymes Ndel ~New England B;OlAbS) and ~ HI (aethesda Research Labs3 by us;n~ the conditions ;nd;cated by the suppl;er.
By known processes, non-phosphorylated complemen-tary adapter oligonucleot;des are synthes;zed having the follo~ing structure:
5'-dTATGGAG-3' and $'-dGATCCTCCA-3', These oligonucleot;des are prehybrdized and then subjected to ligation at 4C wîth plasmid pTG922 which has been subjected to enzymatic restr;ction in a mole rat~o o~ 50:1 using DNA ligase under known cond;t;ons.
The l;gat;on mixture ;s used to transform compe-tent cells ot stra;n TGE900 and the transformants ob-tained are spread on a culture med;um in the presence of ampicill;nl The colon;es are selected on nitrocellulose fil-ters by hybridizat;on w;th a probe labeled with T4 poly-nucleotide kinase 5'-dCCTG~GATCCTCCA-3'. This probe wholly complements the non-fused construction which it is desired to select, but only comple~ents 7 of the nucleo-tides of the parent plasmid pTG922, and this is insuffi-c;ent to ensure hybr;dization.
6 posit;ve cand;dates are thus obta;ned.
These are cultured at 28C and 42C, and the extracts prepared are tested for their anti-elastase ac~ivity as described in the main patent.
The results are collated in the attached table.

; :~ 3~g 3 ~ 3~

W0 84/02918 ~ 40 - PCT/~R84/00014 TA~LE

PLASMID ~ X INIII~ITION OF LASTASE

pTG920 23C
~2 C ~, 2 p~G922 2~C 93,6 42C 98 ,9 pTG529( ) 26C 4 ,9 42aC 60 10pTG929 ( 2 ) 2&o C ~ 1 7 ~ 7 42C 64 ,4 pTG920 ;s the only vector ~;thout the sequence encoding 1-antitrypsin, pTG922 pTG929 (1) and t2) are d;fferent clones prepared by the pro~ess of the present ;nvention.

The above results show that the plasm;ds according to the ;nvention ensure the express;on of a non-fused human ~1-ant;trypsin in Escherichia coli.
The reduction ;n the level of expression of human 1-antitryps;n in plasmid pTG929 relative to plasm;d pTG922 ;s probably due ~o the disappearance of a part of the cII sequence.

Plasm;d pTG929 only produces lo~ quant;ties of ~AT (less than 0.1 X of the total cellular prote;ns).
In order to 1ncrease the level of expressi~n, cIlrbs is replaced by a synthetic rbs site:

t r~
"~
W0 8~/02918 ~1 - PtT/FR84/00014 translation TC(;ATA;~C~CAGGAACAG~TCTSTG
t Sh;ne/DaLgano sequence Th;s exchan~e ;s performed es represented in ~igure 12. The Hp~I and ~gliI s;tes of pTG929 are re-pl~ced usin~ synthetir inserts by CLaI and XhoI site~
5 respectively.
The synthetic rbs (synth rbs) is then inserted between the NdeI site and the ne~ ClaI site created in the gene N. This man;pulat;on eliminates cIlrbs and leads to a truncated N gene immediately followed by 10 synth rbs. A translation termination codon (TAA) in synth rbs bLocks the translation of N. The plasmid obtained, pTG956, produces 1-AT at the level of 0.1 % of the total cellular prote;ns, as shown by the anti-elas-tase assay 1n vitro.

As has been ;ndicated ;n the descript;on, the or;g;nal sequence of ~-AT ;n plasm;d pTG956 is replaced by an altered sequence:
or;g;nal sequence:
15 prote;n: ~et glu asp pro glu 5Iy as? ala DNA ATG G~G GAT CCC C.~G G~ C-~T GCT

protein: ~et glu Z5~ pro slu gly asp ala DNA : ATG GAA GAT CCT CA~ GGC G~T GCT

Each change ~ind;cated by an aster;sk);s carried out at position 3 of the codons and does not alter the am;no ac;d coded. The ~-AT gene wa~ mod;fied using the technique of ~utagenes;s d;rected at s;tes us;ng a synthe-ic oligonucleot;de which def;nes the part;cuLar changes ; 25 ;n base (reference 30). The chosen changes ;n sequence cone from a statist;cal s~udy ~hich shows a preference for certairl bases in severaL positions ;n the neighbor-hood of the beginning of the E~ coli cJenes ~references 31 and 32). ~hese changes ~;ll also destabil;ze the regions correspond;ng to poss;ble secondary structures ,, , ~

.: :

~ , , ... , ,:
: , ..;
' .
~3~3~
WO 84/02918 - 4~ ~ PCT/FR84/~001~
~hich c~n ~ppear in the mRNA sequence of ~-AT. The oligonucleotide for the mutagenesis is:
.GC.~CGCCT~GG~TCT~CCAT-3' Th;s sequence contains 4 d;fferences with respect to the orig;nal sequence and leads to the modifications indicated above in the~ -AT gene.
The plasmid obta;ned, pTG983, ;s ident`ical to pTG956 ~ith the exception of the changes mentioned above.

RESULTS OBSERVED ~r - IMMUNOPRECIPITATION AND
~ Y DEGRADATION OF ELASTASE FROM HUMAN
LEUCOCYtES
Two series of assays were carried out:
- ;mmunoprec;p;tation of cultures of stra;ns labeled w;th [ S~-methionine, with an ant;-1AT ant;serum:
- and ant;-elastase (from human leucocytes) act;vity.
The results of these assays are illustrated in F;gures 2 (sic) and 3 ~s;c).
In F;gure 2 (s;c) there appear the results of 5 ;mmunoprecipitat;on assays:
- cylinder 1 : control culture (molecular weights ;n kilodaltons), - cylinder 2 : induced pTG9ZO tplasm;d vector w;thout ;nsert) culture, - cyl;nder 3 : ;nduced pTG922 (fused ~AT) culture, - cyl;nder 4 : non-induced pTG922 tfused ~1AT) culture, - cyl;nder 5 : induced pTG983 (non-fused ~1AT) culture.
The arrows ;nd;cate the precipitation zones corresponding to fused ~AT (upper band) and non-fused 3G ~1AT (lower band). The lower band appearing on cylinder 5 probably results from a translation initiation ;nternal to the ~AT sequence.
Th;s first series of tests enables it to be demon-strated that immunoprecip;tat;on, with an anti-~1AT anti-serum, of labele~ proteins from cultures of trains con-taining pTG983 (host strain TGE 900j permits a prote;n of moLecular weight 42 kilodaltons to be obtained~ which corresponds tv the si7e of non-fus~d ~1AT tcf- cylinder 5)~

~`2 '` `' ~

~3~3~3~
W0 84/02918 ~ 43 ~ PCTI~R84/00014 by comparison with th~ fused prote;n of 4~ k;lodaltons obta;ned starting from cultures of str~;ns of pTG92 (cyl;nder 3).
In Figure 3 (sic) there are ind;cated schemati-c~lly the results of the assays for anti-elastase atti-vity. These were carried out using ~ human leucocyte elastase (50 ny) (Elast;n Products tO. Inc.) and methoxy-succinyl ala-ala-pro-val-n;troan;l;de tCAL~IOCHEM) as substrate. The elastase cleavage libera~es the n1tro-anilide which ;5 recovered at 410 nm.
In the graph of Figure 3 (sic), the legend forthe d;fferent traces is as fo~lows-- ~ - ;nduced pTG922 extract - o - induced pTG983 extract ~ ~ - non-induced pTG983 extract - ~ - ;nduced control pTG920 extract.
These ant;-elastase tests performed in vivo show that pTG983 permits the express;on of s;gn;ficant propor-tions of ~1AT~ Figure 3 (s;c) shows that a 50 % inhibi-t;on of elastase ~50 ng) is obtained w;th approximately 4 ~9 of a pTG922 extract and with approximately 14 ~9 of a pTG983 extract. This corresponds, respectively, to approximately 3 % and approximately 1 X of totaL cellular prote;n produced. 50 ng of elastase are ;nh;b;ted to 50 X by 100 to 120 ng of natural human a1AT tsigma).
The difference between the percentages of ~1AT obtained w;th one and other of the assays performed with pTG922 results from the var;at;on in the quantity of ~AT lost, this loss correspond;ng to an ;ntracellular prec;pita-tion. None of these intracelluLar precipitations were demonstrable with cultures of pTG983.
References 1. Gadek J.E. and Crystal R.G. ~1982) " ~-Antitrypsin in the Metabolic Basis of Inherited Disease", ~, .

3~3~
W0 84/02918 - 44 - PCT/FR8~tOOn1~
5th ~d~t1On, ~d3 StDnbury ~yns2arden~ Fr~dr~ckson, 6cl ld3t ~1 n ~nd Qro~n, Mc~ r~
21~ Fdgerhol M.K. ~nd Cox D.W. t1~81), rhe ~t Polymorph~
Genet~c9 ~lochesn~c~l snd Cl~n~t~l A~pect~ of Hu~an a1-Ant1tryps~n ~n Adv~ne~s 1n Hum~n Genetlc3 11, 1-6~. .
3. ~uppers F. ~1973), ~m. J. Hum. Genet~ 25, 6~7-636.
4. Gadek J.E.~ ~ells G.A. and Cry3tal R.6. ~1979), Ç~arette s~cklng induce~ funct~on~l ant1pr~te~se de-f~c~ency ~n the lo~er resp;ratory tr~ct o~ h~m~ns;
Science 206, 1315-1316.
5. Carp Hl and Janoft A. (1977~, Possible ~echanisms of .e~physem~ in smoker~; C;gar~tt~ smske condensate sup-p~esses prote~se inhibit10n ~n vitro; Am. Rev. Resp.
D~s. 11~, 65.
6. Gadek J.E., Klein H.G., Holland P.V. and Crystal R.G.
~1981), Replacement Therapy o~ ~lpha 1-Ant~trypsin De-f i ci ency. Reversal of Protease-Antiprotease I~balance ~ith~n the Aiucolar Structures of Pi Z Sub-jects. J. Clin. Invest. ~8, 1158-1165.

7. Groopman J.E. and 60ttlieb M.S. (1982)~ Karposi's Sar-coma; an oncological looking glass. Nature 299, 103-104 .
: a. Remant E., Staussens P. and Fiers W., Gene 15, 81 93 (19~1) 9~ Yoakum ~.H., Young A.T., Maetes ~.B. and-Gross~an L., Proc. Natl. Acad. Sci. U~A, 79~ 1i66-1770 (1982).
10.Hersko~itz ~. and Signer E., J. Mol. 8iol. 47, 545-556 ~1970)~
11.Grayhack E.U. and Roberts J.W~, Cell, 30, 637-648 ~ 1 98Z) .
12.Ish-Horoll;;tz D. and Burke ~I.F." Nucl. Acid~ Re~. 9, Z~8~-2998 (19~1).
13.P3nayotatos ~. and Tron~ K.~ Nucl. ~c~ds Res. 9, 5679-5~88 ~1981).
14.Hoopes ~.C. and McClure R.R~, Nucl. Acids Res. 9, 5~3-5504 ~1~8t ) .
15.D~g~rt M~ and Ehr~ch S.,D., ~ene, 23-28 ~1979).
.V~elra J. ~nd Messing J., Gene 19j 25~-Z~8.

.~, .

.. ~ 3~3~
Wo 84/OZ918 - 45 - PCT/FR84/OQ014 170 Shim~takt H. nd Ro~nber~ ~." N~tulre, 292, 128-132, t1981 ) .
18. Oppenhe~rer h.~., GDl:tesm~n S~ ~nd 60ttes~m~n M.~ J.
Mo~. 3tol,, 15B, 327-36s6 (19~2i .
19. Daniel~ D.L. and 8lattner F.R., Virolog~ 117, 81-92 (1~82).
20. TOlseoshev P., ~ro R.A., Rennard S.I., ~r~d~r K.H., Tr~pneLl ~.C. ~)nd Cry~t~l R.G D ~1981)o Procollagen P roduct i on ~nd proco l l~gen messeng~ r RN~ l e~e l3 and activity in humDn lun~ fibrobL~sts during periods of rap;d ~n~ st~tionarr ~ro~th, J. ~iol. Chem. 25~, 3135-31 4n.
21_ ~lobel G. ~nd Dobber~tein ~. (1975), J. CellO ~iol.
67, 852-116~c, 15 22. LA~he R., Hirth P., D~ Wilde M., Hartford N. and Lecocq J.~. (1980), Nature 284, 473-474.
23. Dagert M. and Ehrlich S.D. l1979}, Gene 6, 23-28.
Z4. Kurach~ K~o Chandra T., Degen S.~l.F., ~hite T.T., Marchioro T.L."~lao S.L.C. and Davie E.~ R1), Proc. Natl. Acad,. Sci. USA, 78, 68Z6-6830~, 25. Kohli V., ~alland A. and L~cocq J.P. (1982) Patent No. 82 12100.
26. Suggs S.V." l~tallace R~B., H;rose T.~, Ka~ashima E .H. and Itakur~ IC" t1981 ) Proc. Nat l. Acad. Sci, USA, 78, 6613-6617.
27. ~irnboirn H.C. and Doly J. ~1979) Nuc~eic P~cids Re~. 7, 15t3-1523.
28. Leicht ~., Long 6,.L., Chandra T. Kurachi K., Kidd V.J.~ M,~ce M. Jr., Davi~ E.W. and ~too S.L. t198Z~
Nature 297, ~55-659.
29. ~erb Z. ~1981) ~iochem~ J. 1~3, 589~605.
30. ~oller, M.J. and Smith, M. (1983) In Methodo (sic) of Enzymology, Vol. 1.,0 pp 4~-500.
35 31. Sche-er, GoF~~ Walkinsha~, ~l.D., Arnott, S. ar.~
Mol~, D.J. (1980~ N~cl. Ac~d~. ~e~. 8, 3895-3907.
32. 60ld~, L., Pribno~, D., Schneider, T., Shinedlin~, S., Singer, ~.S. a.nd Stormo, G. (1981) Ann. Rev. M;cro-b;ol. }5, 365-6~03.

Claims (13)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. A transformed bacterium, expressing human .alpha.1-antitrypsine, said bacterium being transformed by a vector, capable of expressing in that bacterium human .alpha.1-antitrypsin, said vector comprising a sequence coding for human .alpha.1-antitrypsine under the control of a promoter and a translation initiation region with the following sequence:
Translation TCGATAACACAGGAACAGATCTATG
sequence Shine/Dalgano and also an origin of efficient replication effective in said bacterium.

2. Bacterium according to claim 1, wherein in the region of the starting codon of translation of the gene coding for human .alpha.1-antitrypsin, the vector has the following sequence:
ATG GAA GAT CCT CAA GGCGATGCT
TAC CTT CTA GGA CTT CCGCTACGA

3. A transformed bacterium, expressing human .alpha.1-antitrypsine, said bacterium being transformed by a vector capable of expressing human .alpha.1-antitrypsin, said vector having a sequence coding for human .alpha.1-antitrypsin under the control of a promoter, and an initiation translation region, in which the sequence in the region of the ATG
codon is as follows:

TCGATAACACAGGAACAGATCTATG GAA GAT CCT CAA GGC GAT GCT

4. A bacterium according to claim 1, 2 or 3, in which the promoter used is the promoter PL.

5. A bacterium according to claim 1, wherein the vector contains, in addition, a transcription antitermination function.

6. A bacterium according to claim 5, in which the transcription antitermination function is the .lambda.N gene for the promoter PL or PR, or the .lambda.gene for the promoter P1R.

7. A bacterium according to claim 1, 2 or 3, the expression vector of which contains the origin of replication of pBR322.

3. A bacterium according to claim 1, the expression vector of which contains a gene which encodes resistance to an antibiotic.

9. A bacterium according to claim 8, the expression vector of which contains an ampicillin resistant gene.

10. A bacterium according to claim 1, in which the promoter is controlled by a repressor.

11. A bacterium according to the claim 10, in which the repressor is thermosensitive.

12. A bacterium according to claim 1, 2 or 3, which is a strain of E. coli.

13. A process for preparing human .alpha. 1-antitrypsin of bacterial origin, which comprises cultivation thereof in a culture medium of a transformed bacterium as claimed in claim 1, 2 or 3.