The C-terminal region of the beta subunit of the human chorionic gonadotrophin (hCG) is implied in heterodimer stability (beta26-110 disulphide bridge), in vitro LH bioactivity (region beta102-110) and in in vivo LH bioactivity (beta CTP). Like the hCG beta, the equine eLH and eCG beta subunits, also possess a C-terminal extension (CTP). But, in contrast to hCG, eLH and eCG bind to both LH and FSH receptors in species other than the horse. This allows investigation of the roles of the beta subunit C-terminal region of a eLH/CG recombinant molecule on both LH and FSH activities. To do so, the CTP was deleted and/or the beta26-110 disulphide bond was mutated and the resulting mutated beta subunits were transiently co-expressed with common alpha subunit in COS7 cells. These regions were also deleted in a betaalphaeLH/CG single chain also expressed in COS7 cells. The hormones produced were characterized by different ELISAs and in vitro LH and FSH bioassays. Mutation of the 26-110 disulphide bond and deletion of the betaCTP led to a decrease in eLH/CG heterodimer production. Double mutation promoted an additive effect on production of the heterodimer and of the corresponding tethered eLH/CG. The elimination of the beta26-110 disulphide bond in the betaalpha single chain had no effect on its production. However, neither the 26-110 disulphide bond nor the CTP mutations affected dimer stability and bioactivities of the secreted heterodimers and/or single chain molecules. Therefore, in contrast to hCG, the 26-110 S-S bond of the recombinant eLH/CG beta subunit does not seem to be essential for eLH/CG dimer stability upon secretion and expressing LH and FSH bioactivities.

Horse LH/CG (eLH/CG) and donkey LH/CG (dkLH/CG) are strictly LH-specific in their respective homologous species. However, both bind to the FSH receptors from non-equid species, whereas the zebra hormone (zbLH/CG) does not. The FSH/LH ratio of eLH/CG and of the alphadkbetae hybrid is about tenfold higher than that of dkLH/CG and of the alphaebetadk hybrid, showing that the betae subunit contains the structural features responsible for the high FSH activity of eLH/CG. Only six amino acid positions (51, 94, 95, 102, 103 and 106) are unique to the betae subunit when compared with the betadk and betazb subunits. The Gly-Pro and Val-Phe sequences in positions 102-103 of betadk and betae respectively were swapped by site-directed mutations and the mutated beta-subunits cDNAs were cotransfected in COS cells with either alphae or alphadk subunit cDNA. Other mutations were also introduced in 102-103 dkLH/CG beta-subunit: Ala-Ala, Gly-Ala or Ala-Pro. These mutations with Ala-Ala, Gly-Ala or Ala-Pro in the 102-103 betadkLH/CG subunit did not change the FSH/LH ratio of dkLH/CG but the Gly(102)-Pro(103)-->Val(102)-Phe(103) mutation promoted a marked increase in the FSH/LH activity ratio. This was observed with the two heterodimers containing alphae or alphadk. Conversely, the Val(102)-Phe(103) mutation in betae led to a dramatic drop in FSH/LH activity ratio of eLH/CG, to a level similar to that of dkLH/CG. Since all FSHs possess a Gly residue at position 104, we introduced the Gly(102)-Pro(103)-Arg(104)-->Val(102)-Phe(103)-Gly(104) mutation in betadk with the expectation that the increase in FSH activity observed with the Gly(102)-Pro(103)-->Val(102)-Phe(103) mutation could be potentiated. In fact, the additional Arg(104)-->Gly(104) mutation was found to abolish the increase in FSH activity observed with Gly(102)-Pro(103)-->Val(102)-Phe(103). Mutations Gly(102)-Pro(103)-->Val(102)-Arg(103) or Gly(102)-Pro(103)-Lys(104)--> Val(102)-Arg(103)-Gly(104) were also introduced in human CGbeta (hCGbeta) to compare the impact of these amino acid changes in the well-studied gonadotrophin hCG. The betahCG mutants obtained, co-expressed either with the human or the horse alpha-subunit, did not display any FSH activity. In conclusion, the 102-104 sequence in eLH/CG beta-subunits appears to be of utmost importance for their binding to FSH receptors. However, these results obtained with equid beta-subunits are not transposable to other gonadotrophins as similar mutations in hCGbeta did not lead to any increase in FSH activity.

Horse (Equus caballus) luteinizing hormone (eLH) and chorionic gonadotrophin (eCG), which have the same amino acid sequence, are unusual in that, although they express only LH activity in equids, they express dual LH and FSH activities in all other species tested. Donkey (Equus asinus) LH (dkLH) and CG (dkCG), which also share an identical peptide backbone, have been less well characterized and conflicting results concerning their FSH activity in heterologous species have appeared in the literature. In order to assess and compare the intrinsic LH and FSH activities of the horse and donkey LHs in heterologous species, recombinant eLH (r.eLH/CG) and recombinant dkLH (r.dkLH/CG) were expressed, for the first time, in COS-7 cells. Their LH activities were assessed in a rat Leydig cell bioassay, and their FSH activities were estimated in a bioassay using Y1 cells stably expressing the human FSH receptor. Human CG (hCG) was expressed (r.hCG) and analysed in the same system. The results showed that, whereas r.dkLH/CG was about twice as active as r.eLH/CG in the LH bioassay, it was five times less active than r.eLH/CG in the FSH bioassay; r.hCG was about three times less active than r.eLH/CG in the LH bioassay but was completely inactive in the FSH bioassay. These results confirm that dkLH/CG possesses significant FSH activity in heterologous species that is not attributable to contamination with FSH.

Journal of Endocrinology (1997) 152, 371–377

Horse LH/chorionic gonadotrophin (eLH/CG) exhibits, in addition to its normal LH activity, a high FSH activity in all other species tested. Donkey LH/CG (dkLH/CG) also exhibits FSH activity in other species, but about ten times less than the horse hormone. In order to understand the molecular basis of these dual gonadotrophic activities of eLH/CG and dkLH/CG better, we expressed, in COS-7 cells, hybrids between horse and donkey subunits, between horse or donkey alpha-subunit and human CG beta (hCG beta), and also between the porcine alpha-subunit and horse or donkey LH/CG beta. The resultant recombinant hybrid hormones were measured using specific FSH and LH in vitro bioassays which give an accurate measure of receptor binding specificity and activation. Results showed that it is the beta-subunit that determines the level of FSH activity, in agreement with the belief that it is the beta-subunit which determines the specificity of action of the gonadotrophins. However, donkey LH/CG beta combined with a porcine alpha-subunit exhibited no FSH activity although it showed full LH activity. Moreover, the hybrid between horse or donkey alpha-subunit and hCG beta also exhibited only LH activity. Thus, the low FSH activity of dkLH/CG requires an equine (donkey or horse) alpha-subunit combined with dkLH/CG beta. These results provide the first evidence that an alpha-subunit can influence the specificity of action of a gonadotrophic hormone.