Milk yield, age at first calving, and the risk of BSE: an analysis at the farm level in France.

The epidemic of bovine spongiform encephalopathy (BSE) in France, as in the UK, has affected dairy cattle much more than beef cattle. However, the intensification of dairy herd management as a risk factor for BSE has not to date been analyzed. For this purpose, two databases were merged: the French Milk Records database, and the French BSE database, which can be considered as being devoid of notification bias since July 2001, when systematic tests were implemented. Only pure Holstein herds were considered, which represent the vast majority of total and BSE-affected dairy herds in France. A case-control study was designed so that 20 control herds were matched to each case herd according to the location of the farm and the year of birth of the index case. Three thousand and forty five farms were included, among which 145 with a BSE case notified between July 2001 and July 2003, and 2900 controls. With respect to the risk of BSE, odds ratios for each class of milk yield and age at first calving were estimated by using conditional logistic regression models with appropriate adjustments to herd size. The two main results were the following: firstly, whereas most Holstein herds, with average production between 7000 and 10,000kg, had nearly the same BSE risk, a small category of very intensive herds, with annual milk yields above 10,000kg, were significantly more at risk than the other herds. Secondly, a very early first calving (under 26 months of age) was found to be at risk for BSE as compared to other categories, independently of the milk yield. These results are discussed in the light of the known age-dependent susceptibility to BSE.

[The bovine spongiform encephalopathy in the UK and France: an epidemiological perspective]. / Point de vue épidémiologique sur l'encéphalopathie spongiforme bovine au Royaume Uni et en France.

Epidemiology is aimed at the study of the health status of human and animal populations. The present review shows how epidemiologists in the UK and France, contributed to a better knowledge of the bovine spongiform encephalopathy (BSE), its probable origin, and its associated risk factors. Following case studies and case-control surveys, the role of meat and bone meal (MBM) as the source of infection for cattle has been well established. Other epidemiological studies have shown the absence of horizontal transmission, but on the other hand a cohort study concluded to a possible vertical transmission. Backcalculation models, which intend to infer the characteristics of infection from the observations, agree on two results : first, the BSE infects young animals, under 2 years. Secondly, the peak of the French BSE epizootic, probably in the nineteen-eighties, was ignored. Successive measures concerningMBMfailed to completely abolish the epizootic. The analysis of born-after-the-ban (BAB) cases shows that their spatial distribution in France is not random, which argues against a sporadic origin. The French BAB cases may result from two nonexclusive events : residual contaminations of cattle by feed intended for pigs and poultry, and mother-to-calf prion transmission, likely at a low rate. If a total eradication of BSE is not obtained, it will be the task of epidemiologists in particular, to figure out the source of the residual infection.

Trophoblastic interferons: do they modulate uterine cellular markers at the time of conceptus attachment in the pig?

Between days 12 and 20 of pregnancy, the trophectoderm of the porcine conceptus secretes two species of interferons (IFN): IFN-gamma (Type II), which is produced in substantial amounts, and IFN-delta (type I), for which secretion peaks at days 15-16 of gestation. The role of these embryonic IFNs is not known. We made the assumption that, in the pig, one possible role of these IFNs may be the remodelling and/or depolarization of the uterine endometrial epithelium, as a prerequisite for implantation and establishment of a functional placenta. A comparative analysis by immunohistochemistry of several cell membrane markers and ECM components of the cyclic and pregnant uterus was performed at day 15 post-oestrus. The markers were those likely to differ between a pregnant and cyclic uterus, or between different stages of pregnancy. A highly specific marker of IFN-gamma activity, namely MHC class II antigens in the uterine mucosa, was also examined. This study provides so far unreported data: in the endometrial epithelium of the pregnant uterus, we observed a partial relocalization of ZO-1, a marker of epithelial tight junctions, thus suggesting significant changes to the endometrial polarity. Heparan-Sulphate Proteoglycan (HSPG) expression did not differ significantly between cyclic and pregnant uteri. In contrast with the accepted rodent model of trophoblast-uterus adhesion, the porcine trophoblast and luminal epithelium were negative for HSPG. Finally, MHC class II antigens were absent from the cyclic uterus, but markedly induced in the day 15 pregnant uterus, particularly in endothelial cells, suggesting that IFN-gamma may indeed cross the maternal epithelium. This hypothesis was supported by the observation of IFN-gamma immunoreactivity associated with clusters of endometrial cells in the pregnant uterus.

Polarized porcine trophoblastic cell lines spontaneously secrete interferon-gamma.

Following the demonstration of high levels of interferon-gamma (IFN-gamma) synthesis by the pig trophectoderm around implantation, an adequate experimental system was established so as to study the regulation of endogenous IFN-gamma gene. Several stable cell lines have been isolated from Day 14 and 15 pig trophoblast, that could withstand indefinite growth when cultured on collagen-coated supports. Since no feeder cells were used for culture, and cell lines could be successfully cloned, these lines represent the first pure porcine trophoblastic (TB) cell lines isolated so far. These cells were shown to exhibit most differentiation markers of epithelial cells and in addition to express the porcine trophectoderm-specific antigen SN1-38. The TB cell line A (TBA) was characterized in more details upon culture on microporous filters, where a high apico-basal polarity could be obtained. In those conditions, a transient and acute interferon-gamma secretion was detected only in the apical medium. Moreover, the two trophoblast-specific mRNA of 1.3 and 1.4 kb that have been described in the blastocyst collected in vivo were shown to be synchronously transcribed. This polarized synthesis and secretion of IFN-gamma was correlated with the acquisition of maximal levels of electric resistance of TBA monolayers on filters, and was not observed on the same cells cultured on plastic. This differentiation was maintained over 30 passages, demonstrating that the induction of IFN-gamma secretion by pig conceptus is not maternally controlled. This cellular model will be of prime importance for studies on the developmental regulation of the IFN-gamma gene, and more generally for studies on relationship between secretion and polarity in transporting epithelia.

Tetracycline-controlled expression of glycosylated porcine interferon-gamma in mammalian cells.

Tetracycline-controlled expression plasmids that allow inducible expression of proteins in mammalian cells (Gossen & Bujard, 1992), have been used to express porcine interferon-gamma in the RK-13 rabbit kidney cell line. Following neomycin selection, stable clones produced recombinant, glycosylated porcine interferon-gamma (rGPoIFN-gamma) only after removal of tetracycline (Tc). Southern blot analysis of one clone showed that approximately 50 copies of IFN-gamma cDNA were present in the cell genome. In the absence of Tc, stable clones secreted large amounts of rGPoIFN-gamma (up to 16 microg/ml) into the medium supplemented with 10% FCS and high glucose concentration. Molecular weight comparison of 35S-Methionine, labelled rGPoIFN-gamma with natural leukocytic IFN-gamma after immunoprecipitation, revealed 4 major glycoforms with apparent Mr of 27,000; 25,000; 20,000 and 18,500, that are almost identical in both IFN-gamma species. In both cases, all 4 glycoforms resolved into 2 polypeptide monomers with apparent Mr of 16,500 and 14,500 upon deglycosylation with N-glycosydase F. The biological activity of rGPoIFN-gamma was in the same range as that of natural leukocytic PoIFN-gamma (2 x 10(6) U/mg). Eventually, this recombinant mammalian IFN-gamma should constitute a very useful substitute for leukocyte PoIFN-gamma in in vitro or in vivo experiments.

Interferon-delta: the first member of a novel type I interferon family.

We have recently described a novel type I interferon (IFN) co-expressed with IFN-gamma by the trophectoderm of the pig conceptus between day 12 and day 18 of gestation, a development stage that corresponds to implantation in the uterus. This IFN, now officially named IFN-delta, is recognized as the first member of a novel type I IFN family. This paper reviews the main published data on IFN-delta, together with some new data, showing that IFN-delta, while being a true type I IFN, has some very specific structural and biological properties. Sequences related to IFN-delta coding sequence were found in the genome of man and other ungulates but the only other potentially functional gene was found, so far, in the horse. The pig IFN-delta mature protein, with 149 amino acids, is the smallest of all known type I IFNs. It is unusually rich in cysteines (seven residues), and has a very basic isoelectric point. Recombinant IFN-delta expressed in insect cells is glycosylated and has a high antiviral activity on porcine cells, but not on human cells. It has high antiproliferative activity, which is significantly enhanced in the presence of IFN-gamma. This new IFN was shown to bind on pig cells to the same type I receptor as IFN-alpha. IFN-delta and IFN-gamma genes are co-regulated in the pig trophectoderm, whose cells on day 14-16 of development simultaneously secrete both IFN proteins. The biological role of porcine IFN-delta in early pregnancy has been found unrelated to the known antiluteolytic effect of trophoblastic IFN-tau in ruminants.

Cloning of the porcine interferon-gamma receptor and its foeto-endometrial expression in early pregnancy.

In early gestation, trophoblastic cells of porcine preimplanting conceptuses transiently and massively secrete two distinct interferons (IFNs), one of which is IFN-gamma. In order to localize possible cellular target(s) for this IFN-gamma, the expression of the porcine IFN-gamma receptor and its developmental regulation have been investigated on the maternal endometrium and on the embryonic tissues. A cDNA encoding the porcine IFN-gamma binding-chain (pIFNGR1) was isolated. When expressed in COS-7 cells, it displayed a specific binding to radiolabelled pIFN-gamma and was shown to be a glycosylated membrane protein with an apparent molecular mass of 92 kDa. Porcine IFNGR1 mRNA was detected by RT-PCR not only in uterine epithelial cells but also in embryonic tissues from at least as early as day 10 of gestation. Moreover, membrane expression of the pIFN-gamma receptor quantified by binding and crosslinking of 32P-pIFN-gamma was demonstrated in uterine epithelium and in the trophoblast. In the trophoblast, expression of the receptor was found to be developmentally regulated: although expression was weak on days 12 and 15 of gestation, it reached a level similar to that found on some IFN-gamma-sensitive cells on day 16. This study shows that maternal endometrium is not the only possible target for trophoblastic IFN-gamma: the induction of pIFN-gamma receptor expression in the trophoblast around day 16 of gestation could suggest the appearance of responsiveness to pIFN-gamma in this implanted tissue and therefore a possible delayed autocrine effect of trophoblastic pIFN-gamma.

Vaccinia virus protein B18R inhibits the activity and cellular binding of the novel type interferon-delta.

The soluble vaccinia virus-encoded protein B18R inhibits the antiviral activity and cellular binding of the type I interferons (IFN)-alpha, -beta and -omega of different mammalian species. Recently, a novel type I IFN was detected in pigs and classified as a member of a distinct IFN family designated IFN-delta. Our study aimed to determine if the structural properties of this shortest (149 residues long) type I IFN allow its interaction with the type I IFN-binding protein B18R. Experiments using bovine (MDBK) cells demonstrated that B18R neutralized the antiviral activity of porcine IFN-delta with high efficiency. Preincubation of B18R with radiolabelled IFN-delta specifically inhibited binding of IFN to bovine cells. These data indicate that the overall conformation of the novel IFN-delta might be similar to that of other type I IFNs.

Intrauterine infusion of high doses of pig trophoblast interferons has no antiluteolytic effect in cyclic gilts.

In the pig species, the preimplanting trophoblast is known to synthesize and secrete high amounts of interferon during early development. Previous experiments in cyclic gilts using total conceptus secretory proteins suggested that porcine trophoblastic interferons, unlike those of ruminants, exert no effect on the luteal cycle. In the present experiment, cyclic Meishan gilts were divided into two groups, cannulated on both uterine horns, and given daily injections of either a placebo or increasing doses of a mixture of recombinant interferon-gamma and interferon-delta, on Days 11-14 of the estrous cycle. In treated gilts, the injected doses were much higher than those previously found in uterine perfusates from pregnant gilts. However, no significant differences could be found between the control (n = 4) and the treated (n = 5) group concerning the days of the estrous cycle for mid-decrease of progesterone (control: Day 14.5+/-0.57 [mean+/-SD]; treated: Day 15+/-1.25), the day of estrus (control: Day 19+/-0.96; treated: Day 19.6+/-0.55), and the subsequent ovulation rate (control: 14+/-2.2 corpora lutea; treated: 13.1+/-1.1 corpora lutea). These data confirm that pig trophoblastic interferons, unlike those of ruminants, do not themselves exert an antiluteolytic effect. A possible synergistic effect of embryonic estrogens on the luteal functions of nonpregnant sows remains to be determined.

Characterization of conceptus-produced goat interferon tau and analysis of its temporal and cellular distribution during early pregnancy.

Two proteins (17 and 22-24 kDa) produced by day 17 goat conceptuses were purified from in vitro culture media. Analysis of their N-terminal amino acid sequences and of their antiviral activity confirmed that both proteins belonged to the interferon tau family characteristic of ruminant conceptuses. The two molecules were glycosylated (22-24 kDa) or nonglycosylated (17 kDa) isoforms of the same protein. The time course of secretion was plotted and immunoblotting of the protein contents of uterine flushings from day 13 to day 21 of pregnancy was performed. The nonglycosylated isoform (17 kDa) was first detected on day 16; both isoforms were present at day 17 and, thereafter during pregnancy, the two proteins were not present in uterine flushings. Immunohistochemistry was used to show that the goat interferon tau was present in the trophoblastic cells as early as day 14 and until day 17. However, immunostaining was not uniform along the conceptus; labelling was greater at the abembryonic pole than at the embryonic pole. By day 18, as implantation proceeded, goat interferon tau was no longer detected. These results confirmed that the goat conceptus secretes interferon tau during the period of maternal recognition of pregnancy but its rapid decrease suggests that other factors need to be present by day 18 to take over its role in the maintenance of luteal function.

In vivo induction of interferon-alpha in pig by non-infectious coronavirus: tissue localization and in situ phenotypic characterization of interferon-alpha-producing cells.

A low frequency peripheral blood mononuclear cell (PBMC) subpopulation, referred to as natural interferon-producing (NIP) cells, is described as producing interferon-alpha (IFN-alpha) following contact with non-infectious viral structures, namely viral glycoproteins. These cells are characterized in vitro as non-T, non-B, MHC class II+ and CD4+ cells. In this study, NIP cells were analysed in vivo after an intravenous injection of UV-inactivated transmissible gastroenteritis virus in newborn piglets, which resulted in strong serum IFN-alpha production. Splenocytes, but not PBMC, were the IFN-alpha producers in vivo. Using double immunohistochemical labelling for both IFN-alpha and leukocyte markers, we established that splenic NIP cells were not T or B cells. The majority were MHC class II+ and only a minority expressed a macrophage marker. NIP cells were localized in contact with MHC class II-expressing cells and T cells, which suggested that NIP cells might modulate the antiviral immune response.

Atypical spI interferon binds on porcine cells to a major component of type I interferon receptor.

The short porcine type I interferon (spI IFN), encoded by a gene physiologically expressed by the pig embryonic trophoblast during implantation, represents the first member of a novel family type I IFN. Binding and cross-linking experiments were carried out to characterize its cellular receptor. On porcine kidney cells, specific binding of 125I-spI IFN could be displaced significantly by spI IFN, rpIFN-alpha 1, and rhIFN-alpha 1, but not by rhIFN-alpha 2a or by rpIFN-gamma. On the other hand, all these type I IFNs but not rpIFN-gamma were capable of displacing bound 32P-hIFN-alpha A-P1 on these cells. Cross-linking data show that the specific 120 kD complex formed with these two radiolabeled ligands was displaceable by an excess of both spI IFN and rpIFN-alpha 1. These results provide primary evidence that spI IFN shares at least the major binding subunit of type I IFN receptor on porcine cells. On human WISH cells, 125I-spI IFN did not form any complex, nor did spI IFN affect cross-linking complexes of 32P-hIFN-alpha A-P1 on these cells, unlike rpIFN-alpha 1. The lack of antiviral and antiproliferative effects of spI IFN on human cells is primarily a result of its inability to recognize human type I IFN receptor.

Atypical porcine type I interferon. Biochemical and biological characterization of the recombinant protein expressed in insect cells.

A recombinant baculovirus was designed to express short porcine type I interferon (spI interferon), a novel and atypical type I interferon that was recently described as the product of a gene transcribed in pig trophoblast at the time of implantation in the uterus [Lefèvre, F. & Boulay, V.C. (1993) J. Biol. Chem. 268, 19,760-19,768]. The recombinant protein, secreted into the culture medium of Sf9 cells at 3 days post infection (60,000 IU/ml), was purified by ion-exchange and reverse-phase HPLC. N-terminal sequencing confirmed the predicted signal peptide cleavage site and therefore the size of the mature protein (149 amino acids), the shortest of all reported type I interferons. Purified spI interferon, with a specific antiviral activity using Madin-Darby bovine kidney cells of 3.7 x 10(7) IU/mg, is an N-glycosylated monomer of 19 kDa that possesses several physicochemical characteristics of interferons: (a) disulfide bonds are necessary for bioactivity; spI interferon is thermolabile, stable at pH 2, and able to renature after complete denaturation (1% 2-mercaptoethanol, 1% SDS, and 5 M urea); (b) the carbohydrate chain is not essential for bioactivity since no loss of antiviral activity is observed following complete deglycosylation. In this study, antiviral and anti-proliferation activities of spI interferon in cell culture were compared with those of other interferons, especially with porcine type 1 interferon-alpha. A major difference with porcine type 1 interferon-alpha was that spI interferon was not active on human cells in either test, and it was relatively more active on pig cells compared to bovine cells than porcine type 1 interferon-alpha. Serological cross-neutralization results obtained with anti-(spI interferon) serum confirmed that several members of interferon families are not antigenically related to spI interferon, in agreement with previous observations; this provides further evidence that spI interferon could represent a new family of type I interferon.

Interferon response in pigs: molecular and biological aspects.

This paper reviews the current state of knowledge on porcine interferons (IFN). Three type I IFN subfamilies (IFN alpha, beta and omega) and one type II IFN (IFN gamma) have been identified in the porcine species. A list of the known porcine IFN genes and already produced recombinant proteins, as compiled from the literature, was included. Two major aspects of porcine IFN were discussed: (1) IFNs as host responses to infections, and (2) IFNs in pregnancy. The first part mainly focusses on the IFN production by virus-infected pigs and the nature of IFN alpha secreting leukocytes. The second part reviews recent data showing the secretion of two different IFNs by the pig embryo at the time of implantation. One striking finding was that a non-lymphoid tissue, namely the porcine trophoblast, was able to secrete high levels of IFN gamma. In addition, the expression of another, newly described, IFN gene was also shown in pig trophectoderm. This novel trophoblastic type I IFN differs from the others in size and sequence. Several hypotheses are discussed on the possible role(s) of porcine embryonic IFNs at the initial stages of gestation.

Paracrine activities of porcine trophoblastic interferons.

Interferon-gamma (IFN-gamma) and a type I IFN (spI IFN) are transiently coexpressed by trophoblastic cells of pig conceptuses at implantation between day 12 and day 20 of gestation. The local effects of these trophoblastic IFNs were examined on endometrial cells and on trophoblast by measuring antiviral activity and the induction of (2',5')-oligoadenylate synthetase activity. Trophoblastic vesicles were shown to be susceptible to infection by vesicular stomatitis virus and transmissible gastroenteritis virus. Vesicular stomatitis virus multiplied by about 1000 times in trophoblastic vesicles, and endogenous trophoblastic IFNs or exogenous recombinant IFN-gamma or spI IFN had no effect on virus production. No (2',5')-oligoadenylate synthetase activity could be measured on the trophoblast, even after treatment with IFN-gamma or spI IFN. These results clearly show that trophoblastic IFNs cannot induce antiviral resistance or (2',5')-oligoadenylate synthetase activity in the trophoblast, suggesting that these IFNs have no autocrine function. Endometrial epithelial and stromal cells in primary cultures displayed distinct sensitivity to the antiviral effect of IFN-gamma and spI IFN. Stromal fibroblasts were highly sensitive to spI IFN but weakly sensitive to IFN-gamma; epithelial cells were sensitive to both IFNs. The same sensitivity pattern was obtained when measuring the (2',5')-oligoadenylate synthetase activity. Flushing fluid, containing IFN-gamma and type I IFN, was a potent inducer of antiviral effect and (2',5')-oligoadenylate synthetase activity. It is therefore postulated that the endometrial epithelium is the most likely target of trophoblastic IFNs. It is possible that these IFNs play a role in the viral protection of conceptuses.

Age-related increase of porcine natural interferon alpha producing cell frequency and of interferon yield per cell.

Porcine blood mononuclear cells (PBMC) were shown to secrete interferon alpha (IFN-alpha) after induction by a coronavirus, the transmissible gastroenteritis virus (TGEV). IFN-alpha producing cells, referred to as natural interferon alpha producing (NIP) cells, were detected by an ELISPOT assay using anti-porcine IFN-alpha monoclonal antibodies. The frequency of NIP cells among blood cells is low, at most 40-110 per 10(5) PBMC and each NIP cell was found to produce several units of IFN. We have shown that NIP cell frequency and IFN yield per cell gradually increased with the age of the donor animals, from the neonatal period to the adult age, with a significant increase around puberty. Our present results also indicate that NIP cells may be influenced by physiological and genetic factors; thus (1) NIP cell frequency and IFN yield per cell were decreased during lactation; (2) Chinese (Meishan) pigs were found to have higher NIP cell frequency and IFN yield per cell than European (Large White) animals.

[Induction of atypical interferon after heat shock]. / Induction d'un interféron atypique après choc thermique.

Hyperthermia (45 degrees C) induced the release of heat-shock induced factor(s) (HSIF) in the culture medium of Madin-Darby bovine kidney cells (MDBK) during their recovery period at 37 degrees C. HSIF is capable of inducing an increase in the 2'5' oligoadenylate synthetase activity in fresh MDBK cells without any detectable antiviral activity (Chousterman et al. J. Biol. Chem. 1987, 262, 4806-4811). We demonstrated here that even though HSIF did not crossreact antigenically with alpha, beta or gamma bovine interferons, it was still also able to induce an antiviral activity which copurified with its capacity of inducing 2'5' oligoadenylate synthetase activity. Therefore we concluded that HSIF is an atypical bovine interferon induced in response to heat shock.

Nature and possible functions of interferons secreted by the preimplantation pig blastocyst.

In several ungulate species, the preimplantation trophoblast, among various secretions, produces large amounts of antiviral activity that was identified as interferon (IFN). IFNs (types I and II) are pleiotropic cytokines, which in addition to a potent antiviral activity, exert multiple effects on cell growth and differentiation, in particular on the cells of the immune system. In ruminants, trophoblastic IFN, or trophoblast protein-1 (TP-1), was found to consist of a multigenic family related to type I IFN-omega. These IFNs exert hormone-like effects through receptors present on the endometrium, leading to the prolongation of luteal life-span and hence to sustained progesterone secretion. In pigs, a species in which the maternal recognition of pregnancy is controlled by conceptus-derived oestrogens, two IFNs have been found in the preimplantation trophoblast. The major species is IFN-gamma (type II), that so far had been found only in activated T lymphocytes and natural killer (NK) cells. Transcription of the IFN-gamma gene in the pig trophoblast differs from that in mature lymphocytes, since two mRNAs are present. The other component with antiviral activity is a novel type I IFN, distant in sequence from IFN-alpha, beta, omega, and containing seven cysteines in its deduced mature protein. These two unrelated IFNs are temporally co-induced, with maximal secretion at day 16 of pregnancy. Specific receptors for both IFNs have been found on endometrial epithelial cells, but not on the preimplantation trophoblast, suggesting a paracrine effect on the uterus. Different hypotheses as to their role(s) in the establishment or maintenance of implantation are discussed. Whereas an indirect anti-infectious (antiviral) protection of the conceptus by IFNs cannot be ruled out, arguments are presented that do not favour a role in the immune tolerance of the conceptus.

A sensitive immunoassay for porcine interferon-alpha.

Two murine monoclonal antibodies (mAbs) directed against different epitopes on recombinant porcine interferon-alpha (IFN-alpha) were selected and used to construct a two-site ELISA. This ELISA, when performed in a one-step version, detected about 0.5 units ml-1 of IFN-alpha and showed similar sensitivity but better precision than a cytopathic effect inhibition bioassay. Estimates of IFN-alpha in tissue culture medium by the two assays correlated well. In contrast, one or several factors in porcine serum reduced the sensitivity of the ELISA. Measurements of IFN-alpha in porcine serum was, however, possible in a two-step version of the ELISA, with a sensitivity of about 1 unit IFN-alpha ml-1. Results of ELISA and bioassay agreed, except that the ELISA possibly produced false positive results in two out of a total of 91 sera negative in the bioassay. In addition, one of 23 sera positive in the bioassay was negative in the ELISA.

Use of ars18 based vectors to increase protein production in Yarrowia lipolytica.

The isolation of ars sequence from the yeast Yarrowia lipolytica has recently been reported (Fournier et al., 1991). Vectors containing ars18 have been used to increase homologous and heterologous protein production. Examples presented are the Yarrowia lipolytica alkaline extracellular protease (AEP), the porcine alpha 1-interferon and the bovine prochymosin. A 2- to 6-fold increase in the corresponding protein production was observed and in several cases it was established that it corresponded to the copy number of plasmid in the cell.