Studies on recombinant DNA expression of follicle stimulating hormone and its receptor

Abstract

Follicle Stimulating Hormone (FSH), a pituitary gonadotropin, belongs to the family of glycoprotein hormones and plays a crucial role in mammalian reproduction. It is a heterodimer consisting of an -subunit noncovalently associated with a hormone-specific -subunit. The -subunit is common to all glycoprotein hormones, while the -subunit confers upon the hormone its unique biological specificity. Although these proteins have been extensively studied, the structure-function relationships of these hormones have not been clearly established, and more information is needed to identify the domains of these hormones involved in assembly of the subunits, binding to its receptor, and signal transduction.

Moreover, glycoprotein hormones, because of their unique and fascinating structural features, provide an excellent model system to study the folding of these molecules during biosynthesis. The final goal of these studies will be to design reagents that can modulate reproductive and thyroid functions. Extensive investigations carried out by Prof. Moudgal, Indian Institute of Science, Bangalore, India, have unequivocally demonstrated that active immunization of male bonnet monkeys with ovine FSH (oFSH) results in generation of antibodies capable of specifically neutralizing endogenous FSH. Blockade of the hormone抯 action leads to disruption of spermatogenesis, resulting in infertility. This forms the basis for the male contraceptive vaccine currently undergoing Phase 1 clinical trials in India. Successful completion of these trials would require large quantities of highly purified and characterized hormone as the antigen. Moreover, it has not been possible to carry out extensive investigations on FSH per se, as only minuscule amounts of hormone are obtained after purification from its natural source.

The aim of the present work is to develop a recombinant DNA expression system for studying the structure-function relationships of FSH while producing large quantities of this hormone to be used as the antigen for contraceptive vaccines and also to produce biologically active hormone that can be used for clinical and veterinary purposes.

This thesis describes studies carried out on FSH and its receptor. Section I deals with studies to develop an ideal recombinant DNA expression system for FSH subunits. Section II describes studies initiated to understand the molecular mechanisms involved in hormone-receptor interaction.

Large-scale production of glycoprotein hormones using recombinant DNA technology has proved to be a difficult task, and invariably the expression levels are very low. The subunits of glycoprotein hormones are synthesized separately from two different genes, and during biosynthesis they assemble to yield the biologically active heterodimer. Formation of correct disulfide bridges and glycosylation of each subunit appears to be absolutely essential for proper folding, which is a prerequisite for assembling with the counter subunit. Hence it is crucial that any expression system chosen should be able to glycosylate the expressed protein and allow formation of correct disulfide bonds. At the same time, the level of expression should be sufficiently high, and for commercial purposes, the host cells should be easy to grow in simple, defined, inexpensive culture medium. In view of these requirements, the yeast expression systems (Saccharomyces cerevisiae and Pichia pastoris) were chosen for expression of FSH subunits. Yeast as an expression host combines the genetic manipulability and growth characteristics of prokaryotes with the subcellular machinery for performing the post-translational modifications of eukaryotes.

a) Saccharomyces cerevisiae Expression System

Chapter 2 describes the feasibility of using yeast as the expression host for - and -subunits of FSH. The coding regions of bovine FSH (bFSH) - and -subunits, without their cognate signal peptides, were cloned into a yeast expression vector in frame with the signal peptide of the killer toxin of Kluyveromyces lactis, under transcriptional regulation of upstream activating sequences from the galactose operon. These expression constructs were used to transform host cells (S1502B, the protease-deficient strain BJ2168, and the mannan mutant strain mnn9). Both subunits were secreted into the medium and were immunologically similar, though not identical, to pituitary-derived oFSH subunits. Although the levels of expression achieved for both subunits of FSH were low, this study demonstrated the feasibility of using yeast as an expression host for glycoprotein hormones.

b) Pichia pastoris Expression System

In recent times, the methylotrophic yeast Pichia pastoris has emerged as a powerful system for expression of complex eukaryotic proteins. Chapter 3 describes the results of using P. pastoris as the expression host for FSH subunits.

The unique features of the Pichia expression vectors are:

Presence of a very strong methanol-inducible alcohol oxidase (AOX1) promoter, resulting in high-level expression of the heterologous proteins, which can be secreted into a simple, defined medium.

The coding portion of bFSH -subunit cDNA, with and without its cognate signal peptide, was cloned into Pichia expression vectors pHIL-D4/pA0815 (with cognate signal peptide) and pPIC9K (without cognate signal peptide), respectively. The expression vector pPIC9K has the -mating factor signal peptide upstream of the cloning site to direct secretion of the heterologous gene into the culture medium. The host cells GS115 (his4) were transformed with the expression cassettes. The Pichia-expressed FSHP (pFSHP) was secreted into the culture medium and was immunologically very similar to pituitary-derived oFSHP. Replacement of the cognate signal peptide by the yeast -mating factor signal peptide increased expression from 230 ng/mL (cognate signal peptide) to 4 礸/mL (-mating factor signal peptide) of the culture supernatant, the highest expression achieved for a properly folded FSH -subunit. This increase was attributed to higher FSHP message levels. The apparent molecular weight of purified pFSHP on reducing SDS-PAGE is ~22,000, compared with ~20,000 for pituitary-derived oFSHP. pFSHP could assemble with the -subunit to yield a heterodimer capable of binding FSH receptors with almost the same affinity as the native hormone and elicit a biological response. Thus, pFSHP is very similar to its pituitary-derived counterpart in immunological identity and biological function. The -subunit of bFSH was also expressed in Pichia using the same strategy and was immunologically very similar to its pituitary-derived counterpart. This system proved ideal for large-scale production and purification of recombinant FSH subunits for immunocontraceptive, clinical, or veterinary use after appropriate characterization.

Chapter 3 also describes expression of - and -subunits of bFSH in Pichia as a genetically linked single-chain molecule, wherein the C-terminus of the -subunit is translationally fused with the N-terminus of the -subunit [pFSH()], converting the heterodimeric hormone into a single-chain hormone. This fusion molecule was immunologically similar to the pituitary-derived heterodimeric oFSH, could bind FSH receptors with almost the same affinity, and elicit biological responses. This provides a good model to study structure-function relationships without the constraint of assembling individual subunits.

Expression of bFSHP in Escherichia coli

Although Pichia proved ideal for FSH expression, the expression levels were still lower than expected. Chapter 4 describes a regulatory mechanism attenuating FSHP expression in E. coli. Initial attempts using pRSET and pET vectors were unsuccessful due to low expression. Substantial expression was achieved only as a fusion protein with glutathione-S-transferase. Polyclonal antiserum raised against bacterially expressed FSHP recognized pituitary-derived oFSHP, oFSH, and hFSH.

A putative hairpin structure in the first 12 codons of the mature FSHP secreted form attenuated expression in E. coli. By deleting these codons or altering bases without changing amino acids, expression increased 10-fold (30� mg/L). The hairpin likely impedes ribosome movement. E. coli-expressed FSHP did not fold properly and formed inclusion bodies. Efforts are underway to refold and characterize it.

Section II: FSH-Receptor Interaction

Chapter 5 describes studies on the molecular mechanisms of FSH-receptor interaction.

Polyclonal antiserum raised against rat FSH receptor peptide (amino acids 5�9 of extracellular domain, RF3) inhibited FSH binding in vitro and in vivo. Administration of RF3 antiserum to immature male rats resulted in apoptotic death of pachytene spermatocytes, confirming blockade of hormone action. Adult rats showed similar but less drastic effects.

Monoclonal antibodies were raised against bacterially expressed FSH receptor fragment (amino acids 201�9, RF2). Four RF2 monoclonals bound FSH receptors on granulosa cells and inhibited labeled FSH binding. These tools are being pursued to probe hormone-receptor interactions.