High-level synthesis of human prolactin in Chinese-hamster ovary cells

Human prolactin (hPRL), a protein hormone primarily secreted by the anterior pituitary, has a remarkably wide spectrum of biological activities [1]. Whereas PRL is best known for its stimulation of lactation, and growth and development of the mammary gland [2], it has also been shown to mediate a variety of physiological processes, including growth and maturation of oocytes [3], immune modulation [4] and osmoregulation [5]. Basically a singlechain protein with a molecular mass of 23000 Da and three intramolecular disulphide bridges, PRL shows significant molecular polymorphism, including a varying percentage of the glycosylated form, which could underlie its diverse properties [1]. Quite a number of disorders in the area of pituitary function and reproduction have been linked to abnormal levels of PRL in the peripheral blood, making it one of the most frequently determined hormones in clinical laboratory assays [6]. Although routine therapeutic applications of hPRL have not been established, recent studies indicate possible applications of this hormone in the treatment of women presenting with poor lactation associated with temporary pituitary ischaemia [7,8] and in the restoration of normal sperm characteristics in hypoprolactinaemic, infertile men [9]. Administration of PRL also appears to be potentially useful for the treatment of infections and immunosuppression. Thus as recently found with experimental mice, PRL has a protective effect against pathogen-induced infections [10], can restore haemorrhagic effects [11] and may be clinically useful in reversing myelosuppression induced by azidothymidine or in other myeloablative therapies [12]. Furthermore, recent in vitro studies have shown that recombinant hPRL (rec-hPRL) can reverse the antiproliferative effects of transforming growth factor b, indicating its potential application as an immunotherapeutic agent in cases of HIV infection [13]. In view of the above, an increasing need for rec-hPRL can be anticipated. In the production of hPRL, recombinant DNA technology has been utilized as an attractive alternative to laborious pituitary gland extractions. Several laboratories have reported cytoplasmic expression of rec-hPRL in Escherichia coli in the form of inclusion bodies [14–17]. However, recombinant proteins in inclusion bodies are present in a reduced and denatured state, and their solubilization and renaturation usually render low yields of biologically active protein and substantial contamination by partly aggregated and immunologically altered proteins [18,19]. To circumvent such problems, our previous research was directed towards the production of rec-hPRL in a ‘ native ’ form, as secreted into the periplasmic space of E. coli. Whereas our first product was linked to an N-terminal peptide tag [20], more recently, secreted bacterially produced rec-hPRL was obtained in the authentic form, although at relatively low expression levels [6]. The production of secreted rec-hPRL by eukaryotic and, in particular, by mammalian cells, is of major interest since such cells have a potential for secreting the hormone at high levels. Importantly, mammalian-cell-produced hPRL would, in contrast to bacterially produced hPRL, more closely resemble native pituitary hPRL, as it exhibits glycosylated and possibly other post-translationally modified forms. Glycosylated forms of hPRL appear to be of fundamental importance for physiological studies and for potential clinical applications. So far, only two research groups have reported expression of glycosylated (G) and non-glycosylated (NG) rec-hPRL in mammalian cells. Hirooka et al. [21] compared the secretion kinetics of G-hPRL and NG-hPRL produced by three transfected cell lines [COS-1, GH3 and Chinese-hamster ovary (CHO) cells] without, however, addressing the levels of expression of the hormone or a purification process. Price et al. [22], utilizing a system based o