- Email: [email protected]
The steroidogenic effect of single-chain bovine LH analogs in cultured bovine follicular cells
Molecular and Cellular Endocrinology 252 (2006) 136–141
The steroidogenic effect of single-chain bovine LH analogs in cultured bovine follicular cells Ruth Braw-Tal a , Svetlana Pen a , Moran Grinberg b , Sigal Nakav b , David Ben-Menahem b,∗ a
b
Institute of Animal Science, Agricultural Research Organization, The Volcani Center, Beit Dagan 50250, Israel Department of Clinical Pharmacology, Faculty of Health Sciences, Ben–Gurion University of the Negev, Beer-Sheva 84105, Israel
Abstract Single-chain gonadotropin analogs had been constructed for the purpose of structure–function studies and analog design. Incorporation of a spacer derived from the carboxyl terminal peptide (CTP) of the choriogonadotropin (CG)  subunit between the tethered subunit domains of the human gonadotropins is beneficial for the secretion of the single-chain variants without compromising biocactivity. Although the CG subunit containing the CTP domain is expressed only in primates and equids, a CTP-like sequence exists in the untranslated region of the LH gene of several mammalian species, including the bovine species. The CTP encrypted in the bovine LH DNA (designated as ‘boCTP’) and the CTP derived from the human CG subunit (denoted as ‘huCTP’) served as a linker sequence in the design of bovine single-chain luteinizing hormone (LH) analogs. The purpose of the present study was to evaluate steroidogenesis in cultured bovine theca cells following stimulation with these singlechain analogs. The concentration of the LHboCTP␣ and LHhuCTP␣ analogs in the conditioned media of the expressing CHO cells was threeto six-fold higher than that of the “linkerless” LH␣ and LH111␣ variants. The four analogs induced androstenedione and progesterone secretion from the primary theca cells in a dose-dependent manner, but differences in the steroidogenic response were observed. The LHboCTP␣ analog (10 ng/ml) effectively induced androstenedione and progesterone secretion over unstimulated levels (4.0- and 4.4-fold increase for androstenedione and progesterone, respectively). The response to the pituitary bovine LH standard (10 ng/ml) was less pronounced for both steroids (two- to three-fold increase over basal levels). The activities of LHhuCTP␣, LH␣ and LH111␣ were comparable and sightly reduced relative to the LHboCTP␣ activity. The data suggested that LHboCTP␣ was ranked as the most potent and this was even more prominent when analogs were used at a lower dose (1 ng/ml). These data suggest that the design, including the huCTP or boCTP linker, is favorable for the production of single-chain bovine LH analogs. Furthermore, spacing of the tethered subunit domains with the cryptic boCTP sequence that originated from the bovine LH gene appears advantageous for the purpose of stimulating steroid production in the species-specific bioassay. Thus, an effective strategy to produce bioactive single-chain LH analogs in non-primate, non-equid species would be the mutatation of the LH genes with the aim of expressing the cryptic CTP sequence as a spacer derived from the DNA of the same organism. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Bovine LH; Single chain analogs; CTP; Cryptic CTP; Theca cells; Species-specific bioassay
1. Introduction The pituitary gonadotropins, luteinizing hormone (LH), follicle stimulating hormones (FSH) and the placental chorionic gonadotropin (CG) that is produced in primates and equids, are regulators of steroid production in the ovary and testis. Each of these glycoproteins is a non-covalent heterodimer, composed of a common ␣ subunit and a unique  subunit that confers the receptor-specificity. The heterodimer, but not the monomeric subunits, efficiently binds and activates the cognate receptor (Pierce and Parsons, 1981). In order to bypass the rate-
∗
Corresponding author. Tel.: +972 8 6477485; fax: +972 8 6477629. E-mail address: [email protected] (D. Ben-Menahem).
0303-7207/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.mce.2006.03.011
limiting assembly step, to expand the arsenal of muteins for structure–function studies and analog design, and to combine various activities in a single polypeptide single chain, analogs of the human glycoprotein hormones were genetically engineered by in tandem linking of the genes encoding the monomeric subunits (Narayan et al., 1995; Sugahara et al., 1995, 1996; Grossman et al., 1997; Kanda et al., 1999; Garcia-Campayo et al., 2005). Recently, the tethering approach was exploited to gonadotropins of additional species including equine, ovine, bovine and tilapia (Galet et al., 2001; Dirnberger et al., 2001; Fidler et al., 2003; Min et al., 2003; Kasuto and Levavi-Sivan, 2005; Nakav et al., in press). We previously engineered genetically the single-chain bovine LH variants, in which the  and ␣ subunit domains were linked directly or via the heterologous huCTP (derived from the human
R. Braw-Tal et al. / Molecular and Cellular Endocrinology 252 (2006) 136–141
CG subunit) or the homologous boCTP (encrypted in the untranslated region of the bovine LH gene (Nakav et al., 2005) sequences and expressed them in CHO cells (Nakav et al., in press). The data suggested that the incorporation of the boCTP or huCTP linker enhanced the rate and efficiency of secretion of the glycoprotein from the transfected cells by about three-fold. The single-chain analogs induced progesterone production in immortalized rat granulosa cells expressing the rat LH receptor with potency and efficacy similar to those observed for a standard preparation of the native hormone purified from bovine pituitaries (Nakav et al., in press). These results were in consistence with various studies that examined the in vitro activity of human and non-human single-chain gonadotropins. In most of the published studies, the in vitro activity of single-chain gonadotropins was determined using cells that (a) are incapable of producing sex steroids (i.e., bioassay aimed to determine receptor-binding/signal-transduction in transfected CHO or HEK-293 cells expressing the cognate receptor) or (b) steroidogenic cells that express the receptor derived from a species (usually from a rodent origin) that differs from that of the examined gonadotropin. However, little information is available on the steroidogenic activity of mammalian single-chain goandotropins in follicular cells derived from the same species. The aim of our study was to assess the effect of single-chain bovine LH analogs on androstenedione and progesterone secretion from cultured bovine theca cells.
137
domain. For comparison, a truncated LH111␣ [-(1-111)-␣-(1-96)] variant was also constructed (Fig. 1). We made the truncation of the decapeptide because previous analyses showed that, in the LH to CG gene evolution in humans and horses, the naturally occurring CTP of the CG subunit substituted the 7 (as in primates) or 10 (in equids) carboxyl–amino acid residues of the LH subunit (Talmadge et al., 1984; Sherman et al., 1992). In the engineering of the bovine LH-CTP subunit variants, we applied the proposed evolution paradigm of the equine CTP-bearing subunit to the bovine LH subunit (Nakav et al., 2005) that does not contain a CTP, because the remainder of the equine LH/CG gene is similar to the bovine (Sherman et al., 1992) and cattle are phyllogenetically closer to horses than humans. These bovine LH subunit variants were used in the construction of the single-chain LH analogs (Nakav et al., in press). In addition, the analogous truncation of the carboxyl heptapeptide in the LH subunit domain was beneficial to the secretion of the CTP-spaced single-chain, human LH analog with no compromise of the receptor binding and signal transduction (GarciaCampayo et al., 1997). The high degree of homology in the LH gene among mammalian species and, particularly, the conservation of the carboxyl region suggested that the combination of bovine LH carboxylterminal deletion and incorporation of a CTP spacer is favorable for the design of tethered bovine LH analogs (Nakav et al., in press). Media samples were collected from expressing cells following 48 h of incubation in serum free F-12 medium. The samples were concentrated (×20) and washed twice with PBS (10 kDa cut-off; VivaSpin 20 ml concentrators, VivaScience, Hannover, Germany). The concentration of recombinant single-chain analogs was determined by enzyme immunoassay (EIA), as previously described (Wolfenson et al., 2004; Nakav et al., in press). We examined the steroidogenic response to the four different recombinant LH analogs that include or are devoid of a CTP linker that spaces the tethered ␣ and LH subunit domains: LHboCTP␣, LHhuCTP␣, LH111␣ and LH␣·
2.2. Tissue collection and cell incubation 2. Materials and methods 2.1. Design, expression and quantification of the tethered bovine LH variants We expressed in CHO cells single-chain bovine LH variants in which the carboxyl region of the LH subunit was linked to the amino end of the ␣ subunit (Nakav et al., in press). The LH␣ analog encodes the wild type subunit domains [-(1-121)-␣-(1-96)] (Fig. 1). The LH subunit domain of the variants that include a spacer between the subunit domains was truncated from 121 to 111 residues, and was linked through the homologous boCTP (Nakav et al., in press), or the heterologous huCTP (derived from the hCG subunit) to the ␣ subunit
Bovine ovaries were collected at the local abattoir. The theca cells were isolated from medium- and large-sized follicles (5–9 mm in diameter), dispersed enzymatically and cultured. The viability of theca cells as determined with 0.1% (w/v) trypan blue, was 85–95%. Cells (150,000/well) were incubated in Dulbecco’s minimum essential medium with Ham’s F12 1:1 (Biological Industries, Beit Ha’emek, Israel) a final volume of 0.5 ml in a 24 well plates (Nunc, Kampstrup, Denmark). Cells were incubated overnight in a medium containing 1% fetal calf serum (Biological Industries). Medium was discharged and fresh medium containing 0.1% BSA and either bovine LH (USDA-bLH B6) or studied LH variants at different doses was added. Medium was collected after 6 h, and in the initial experiments it changed to fresh medium and was collected after an additional 18 h (see below). Four replicate wells were used for each treatment in a single experiment and dose–response analysis was repeated at least four times in separate experiments. Media were collected from each well and stored at −20 ◦ C pending analysis of steroid concentration. Our preliminary experiments showed that both basal and dose-dependent LH-stimulated rate of steroid secretion decreased markedly as culture time progressed from 6 to 24 h (Fig. 2A-progesterone and B-androstenedione). We therefore cultured the theca cells for 6 h in the subsequent experiments and stimulated the cells with LHand single-chain analogs with doses no greater than 10 ng/ml, unless otherwise specified.
2.3. Steroid analyses
Fig. 1. Design of bovine LH single chain variants. The subunit domains were genetically linked in tandem in the absence and presence of linkers. The human CG carboxyl terminal peptide (huCTP; heterologous linker) or the CTP-like sequence decoded from the bovine LH gene (boCTP; homologous linker) was used to space the tethered subunit domains. Numbers at the top of the diagrams refer to amino acids, continuously numbered.
Progesterone levels were determined by a solid-phase radioimmunoassay (RIA) using a Coat-a-Count Kit (DPC, Los Angeles, CA, USA) according to manufacturers’ instructions, except that standards were prepared in PBS solution. Intra- and inter-coefficients of variation were 8% and 13%, respectively. Androstenedione was determined by RIA using specific antibodies developed and characterized by Dr. F. Cohen (Weizmann Institute of Science, Rehovot, Israel). Intra- and inter-coefficients of variation were 5% and 11%. The results were expressed as mean ± S.E.M. from different experiments as specified in the figure legends. Statistical differences between means were calculated using Student’s t-test.
138
R. Braw-Tal et al. / Molecular and Cellular Endocrinology 252 (2006) 136–141
3. Results 3.1. Quantification of the analogs The concentration of two analogs containing the CTP domain in the condition media was 950 ± 195 and 1220 ± 130 ng/ml (mean ± S.E.M.) for LHboCTP␣ (n = 11) and LHhuCTP␣ (n = 12), respectively. The concentration of “linkerless” analogs was 190 ± 35 and 310 ± 125 ng/ml, for LH111␣ (n = 10) and LH␣ (n = 5), respectively. These results suggest that the design that includes the homologous boCTP- or the heterologous huCTP linker is beneficial for the production of recombinant bovine LH analogs in the single-chain format. 3.2. Steroid production following stimulation with the single chain analogs
Fig. 2. Rate of progesterone (A) and androstenedione (B) accumulation in medium. Bovine theca cells were cultured for 6 h, collected and changed to fresh medium and collected after additional 18 h. Rate of progesterone and androstenedione accumulation per hour during first 6 h (filled bars) and the following 18 h (open bars) was calculated. Values are expressed as mean ± S.E.M. of three separate experiments, each run in quadruplicates and expressed as ng/ml/h.
In bovine follicles up to the preovulatory stage, theca interna is the exclusive site for LH receptors, and the main steroids secreted from these cells are androstenedione and progesterone. Androstenedione and progesterone basal secretion by theca cells during 6 h of culture were 0.85 ± 0.18 and 1.69 ± 0.38 ng/ml (mean ± S.E.M., n = 11), respectively (Fig. 3). Bovine LH (10 ng/ml) increased androstenedione and progesterone secretion to 2.13 ± 0.30 and 5.20 ± 0.79 ng/ml, respectively. Thus, LH-stimulated androstenedione secretion increased 2.5-fold and progesterone secretion increased about 3-fold compared to the basal level. The steroidogenic dose response profiles of the four singlechain LH analogs were studied in the same bioassay and some differences were noticed between the analogs (Fig. 3). The LHboCTP␣ analog was the most potent in its effect
Fig. 3. Steroidogenic response to LHboCTP␣ (triangles), LHhuCTP␣ (squares) (panels A and B), and to LH111␣ (diamonds) and LH␣ (filled squares) (panels C and D). Progesterone (A and C) and androstenedione (B and D) accumulation in theca-cell culture medium was determined after 6 h. The effect of bovine LH is also shown (filled circles). Values are mean ± S.E.M. of 4–7 (single-chain variants) or 10 (bovine LH) of separate experiments, each run in quadruplicates. * P < 0.05; ** P < 0.01 vs. unstimulated steroid levels.
R. Braw-Tal et al. / Molecular and Cellular Endocrinology 252 (2006) 136–141
Fig. 4. The net steroidogenic response expressed as fold increase relative to basal level for androstenedione (open columns) and progesterone (filled columns). The levels in the culture media were determined after 6 h of stimulation with 1 ng/ml (A) or 10 ng/ml (B) of bovine LH or the single-chain analogs. Values are mean ± S.E.M. of 4–7 (single-chain variants) or 10 (bovine LH) experiments, run in quadruplicates. * P < 0.05; ** P < 0.01 vs. bovine LH.
on steroid secretion by theca cells. Addition of 1 ng/ml of LHboCTP␣ increased androstenedione and progesterone accumulation (P = 0.01 and P < 0.05, respectively, Fig. 3, panels A and B). LHhuCTP␣ seemed to be less biologically potent, as a significant increase in androstenedione and progesterone secretion was observed when added to the medium at a concentration of 10 ng/ml, but not 1 ng/ml (Fig. 3A and B). The LH␣ analog significantly increased progesterone accumulation in medium at a concentration of 10 ng/ml but not 1 ng/ml (Fig. 3C). In comparison to unstimulated cultures, the effect of LH␣ on androstenedione was not significant (Fig. 3D). LH111␣ significantly increased progesterone secretion at a concentration of 10 ng/ml (P < 0.05) and androstenedione secretion at a concentration of 20 ng/ml, (P < 0.05, data not shown) (Fig. 3C and D). As expected, increasing volumes (1, 10 and 100 l) of the conditioned media derived from empty vector transfected cells (pCINeo) did not enhance steroidogenesis above the basal levels (results not shown). To neutralize the inter-assay variations at steroid levels and to highlight some of the differences we observed between the analogs, the results were summarized as a stimulated/unstimulated (basal) ratio of androstenedione and progesterone production in the theca cells. Fig. 4 shows the hormone- or analog-stimulated steroid concentration increase in the culture media, relative to the basal secretion (fold increase). This net effect was calculated for the stimulant at a dose of 1 ng/ml (panel A) and 10 ng/ml (panel B). These data suggest that LHboCTP␣ was the most potent analog as
139
judged by progesterone and androstenedione secretion. Whereas pituitary bLH (1 ng/ml) increased 1.85 ± 0.45 and 1.25 ± 0.2fold of androstenedione and progesterone respectively, the LHboCTP␣ had a significantly stronger effect. It increased androstenedione 3.15 ± 0.4-fold and progesterone 2.7 ± 0.3fold over the basal levels (Fig. 4, panel A). Similar results were observed at the higher dose (10 ng/ml): pituitary bLH stimulated androstenedione 2.8 ± 0.35-fold and progesterone 2.1 ± 0.30fold over the basal levels. Addition of LHboCTP␣ resulted in 4.0 ± 0.45 (P < 0.05) and 4.4 ± 0.65 (P < 0.01) fold increase for androstenedione and progesterone, respectively (Fig. 4, panel B). The net steroidogenic response of the LHhuCTP␣ and LH111␣ variants was comparable. These single-chain variants stimulated a slightly reduced androstenedione accumulation compared to the LHboCTP␣ effect at 1 and 10 ng/ml doses (Fig. 4, panels A and B). The androstenedione response in the cultures treated with LH␣ was reduced as compared to the cells stimulated with the bLH, particularly at 10 ng/ml (P < 0.01 compared to the bLH). Nevertheless, at a further increased dose (20 ng/ml), LH␣-induced androstenedione production was similar as compared to the other three single-chain variants and LH (data not shown). This implies that the lower extent of the theca cells’ response to LH␣ as compared to other variants (Fig. 4) reflects a reduced sensitivity of the theca cells to this analog, rather than a decrease in the efficacy of the action. Analysis of the variance (ANOVA) between the net steroidogenic effects of the different recombinant analogs showed a higher androstenedione secretion in response to LHboCTP␣ comapred to LH␣ (1 and 10 ng/ml; P < 0.05). In summary, these data suggest that the extent of the response to low doses of the analogs is not identical, and the net steroidogenic induction using the LHboCTP␣ variant was more potent than the LH␣ effect. 4. Discussion Our study indicates that the incorporation of a boCTP or hCG-CTP linker and, to a less extent, the boCTP sequence increased the amount of single-chain bovine LH in the culture media by more than three-fold compared to the variants that lack a spacer. This enhancement may result from the accelerated kinetics and improved efficiency of the LHboCTP␣ and LHhuCTP␣ secretion, relative to the “linkerless” muteins that were previously observed in pulse-chase experiments (Nakav et al., in press). This notion is in consistence with studies on human gonadotropins and TSH single-chain analogs, suggesting that the determinants for enhanced secretion are preserved in the hCG-CTP when used as a linker sequence (for example, see Sugahara et al., 1996, recently reviewed in BenMenahem, 2004). Improved synthesis in the transfected cells may also contribute to the enhanced presence of the huCTP␣ and LHboCTP␣ analogs in the culture media. Currently it is not clear whether the sequence or the length of the CTP linker is the critical factor to allow the bovine LH subunit domains to fold into a structure that is efficiently secreted and activates the receptor. When tethering human FSH, a linker sequence compromised of half of the hCG-CTP was
140
R. Braw-Tal et al. / Molecular and Cellular Endocrinology 252 (2006) 136–141
sufficient to enhance secretion and preserve bioactivity of the analog (Sugahara et al., 1996). These data suggest that the prosecretory determinants are located on the N-terminal region of the CTP or that the spacing that spans 15 residues is sufficient to relieve potential constraints in the structure of the single chain analog. Although alternative linker spacers (repetitive Ser/Gly sequences) had been used in the construction of single-chain human CG analogs (Heikoop et al., 1997), the application of the CTP as a spacer combines various benefits for the design of tethered gonadotorpin analogs: (a) the stretch is flexible and lacks secondary structure; (b) the CTP is encrypted into the LH gene of several mammals and is naturally expressed in primates and equids; (c) the domain contains determinants for secretion without prevention of receptor activation that is specified by the glycoprotein hormone  subunit. In order to estimate the bioactivity of single-chain bovine LH variants, we used cultured bovine theca cells. Our results show that the examined analogs were bioactive and stimulated steroidogenic response at a high affinity and in a dose–response manner. The extent of the stimulated androstenedione and progesterone secretion over basal level (2.5–3-fold increase) is in accordance with previous reports of the effect of native LH on primary bovine theca cells (for example, McNatty et al., 1984; Berndtson et al., 1995; Roth et al., 2001). This window of stimulated/unstimulated steroid ratio (especially for androstenedione) is probably smaller than the measurement of cAMP levels in these cells or the use of cells derived from inbread mice or rats for the estimation of the activity. In addition, to deal with slaughterhouse material, ovaries that come from cows of different ages, day of the cycle and health conditions, are likely to contribute to inter-assay variations and require a large number of repetitions to enable a reliable interpretation of the results. We measured steroid secretion in response to LH rather than cAMP, mainly because the concern of introducing additional variability originated from atretic follicles that may be present in the pools of the collected ovaries, since it has been shown that early atretic bovine follicle theca cells were capable of secretion of cAMP but not the androgen in response to LH (McNatty et al., 1985). Although the steroidogenic system used by us is not appealing, the advantages of measurement of steroids secreted from cells that are targeted by bovine LH in the ovary of the cattle surpass the above-mentioned disadvantages for the purpose of further evaluation of the bioactivity of the recombinant analogs. The general conclusion from these experiments is that, similar to the heterodimer, the bovine LH analogs induced the production of two main steroids (androstenedione and progesterone) in the same-species bioassay. A variation in the potency of LHboCTP␣, LHhuCTP␣, LH111␣ and LH␣ was detected, and the LHboCTP␣ analog was ranked as the most effective. The structural changes associated with the conversion of the native bovine LH into the tethered variants include: (a) fusion of two subunit treminii, (b) deletion of the carboxy-decapeptide of the bovine LH domain in the construction of the LHboCTP␣, LHhuCTP␣ and LH111␣ analogs and (c) in two of the engineered proteins, the addition of either the huCTP or boCTP linker. These modifications are likely to result in variants with differences in
the quaternary structure that, although tolerated, may affect the fruitful interaction with the bovine LH receptor. The reason why the steroidogenic response in the bovine theca cells is not identical for LHboCTP␣, LHhuCTP␣, LH111␣ and LH␣ is currently unknown. In the future, systematic receptor-binding/signal-transduction in the primary theca cultures should reveal the basis for the differences in steroid production observed among these single-chain analogs. The LHboCTP␣ variant showed the highest potency in the species-specific bioassay. This suggests that this design, that the tethered analog is spaced and has the sequence entirely derived from the bovine genome, activates the bovine LH receptor more strongly than the alternative analogs. Hence, the conformation of the LHboCTP␣ analog may be favorable for the activation of the bovine LH receptor or the boCTP contains determinants contributing to the steroidogenic response. Interestingly, the progesterone production profiles that we previously observed in immortalized rat granulosa cells expressing the rat LH receptor were nearly identical for the same four variants (Nakav et al., in press). That the steroidogenic response to the analogs is literally the same in the heterologous bioassay, but not in the homologous bioassay, implies an action which is different in the rat and bovine LH receptors. Possibly, the rat receptor tolerates the mutagenesis of the bovine LH as induced in the single-chain analogs to a greater extent than the bovine LH receptor. This hypothesis is in accordance to previous studies that showed that, unlike the human LH receptor, the rat receptor efficiently binds LHs originated from various species (Jia et al., 1991). The production of recombinant bovine LH in the single-chain format is advantageous, because it simplifies expression and circumvents the subunit assembly which is a delicate process and a limiting factor in the biosynthesis of heterodimeric LH (Corless et al., 1987; Kaetzel and Nilson, 1988). Activation of the immune system is a concern, when protein analogs are designed for potential in vivo application. To minimize possible antigenicity, the primary sequence of the bovine genes is maintained in the single-chain analogs, except for the introduction of a CTP domain; the huCTP as well as the boCTP sequences are not expressed in the bovine genes and thus could be immunogenic. However, the antigenicity of hCG-CTP domain is low (Birken et al., 1980; Berger et al., 1996) and since the boCTP is derived from the bovine DNA and, similar to the huCTP, the amino acid composition of the boCTP predicts a lack of secondary structure, it is also expected to be a weak antigen. On this basis, the cryptic boCTP linker is not anticipated to stimulate a strong immune response in cattle. The bovine LH single-chain analogs were bioactive, although the dose–response curves and effects over unstimulated – control – theca cells, especially at a low analog dose, were not identical for all four. These results provide a basis for the development of LH analogs that exhibit a spectrum of steroidogenic profiles for potential veterinary application in assisted reproduction procedures. Bioactive analogs capable of eliciting various levels of steroidogenesis in vivo could be beneficial for the purpose of fine tuning of the hormonal milieu at different stages of follicular growth in order to reduce inconsistency and undesired effects
R. Braw-Tal et al. / Molecular and Cellular Endocrinology 252 (2006) 136–141
associated with the contemporary superovulation procedures in cattle. Acknowledgements The authors thank Dr. Albert Parlow and the National Hormone & Peptide Program (NHPP) for the generous provision of the bovine LH and Dr. Fortuna Kohen for antibodies essential to the study. This research was supported by a grant from the Israel Science Foundation – The Charles H. Revson Foundation (448/03). We thank Angela Cohen for her excellent editorial assistance. References Ben-Menahem, D., 2004. Single chain variants of the glycoprotein hormones and their receptors as tools to study receptor activation and for analogue design. J. Neuroendocrinol. 16, 171–177. Berger, P., Bidart, J.M., Delves, P.S., Dirnhofer, S., Hoermann, R., Isaacs, N., Jackson, A., Klonisch, T., Lapthorn, A., Lund, T., Mann, K., Roitt, I., Schwarz, S., Wick, G., 1996. Immunochemical mapping of gonadotropins. Mol. Cell. Endocrinol. 12, 33–43. Berndtson, A.K., Vincent, S.E., Fortune, J.E., 1995. Low and high concentrations of gonadotropins differentially regulate hormone production by theca interna and granulosa cells from bovine preovulatory follicles. Biol. Reprod. 52, 1334–1342. Birken, S., Canfield, R., Lauer, R., Agosto, G., Gabel, M., 1980. Immunochemical determinants determinants unique to human chorionic gonadotropin: importance of sialic acid for antisera generated to the human chorionic gonadotropin -subunit COOH terminal peptide. Endocrinology 106, 1659–1664. Corless, C.L., Matzuk, M.M., Ramabhadran, T.V., Krichevsky, A., Boime, I., 1987. Gonadotropin beta subunits determine the rate of assembly and the oligosaccharide processing of hormone dimer in transfected cells. J. Cell. Biol. 104, 1173–1181. Dirnberger, D., Steinkellner, H., Abdennebi, L., Remy, J.J., van de Wiel, D., 2001. Secretion of biologically active glycoforms of bovine follicle stimulating hormone in plants. Eur. J. Biochem. 268, 4570–4579. Fidler, A.E., Lin, J.S., Lun, S., Ng Chie, W., Western, A., Stent, V., McNatty, K.P., 2003. Production of biologically active tethered ovine FSH␣ by the methylotrophic yeast Pichia pastoris. J. Mol. Endocrinol. 30, 213–225. Galet, C., Le Bourhis, C.M., Chopineau, M., Le Griec, G., Perrin, A., Magallon, T., Attal, J., Viglietta, C., Houdebine, L.M., Guillou, F., 2001. Expression of a single ␣ chain protein of equine LH/CG in milk of transgenic rabbits and its biological activity. Mol. Cell. Endocrinol. 174, 31–40. Garcia-Campayo, V., Sato, A., Hirsch, B., Sugahara, T., Muyan, M., Hsueh, A.J., Boime, I., 1997. Design of stable biologically active recombinant lutropin analogs. Nat. Biotechnol. 15, 663–667. Garcia-Campayo, V., Boime, I., Ma, X., Daphna-Iken, D., Kumar, T.R., 2005. A single-chain tetradomain glycoprotein hormone analog elicits multiple hormone activities in vivo. Biol. Reprod. 72, 301–308. Grossman, M., Wong, R., Szkudlinski, M.W., Weintraub, B.D., 1997. Human thyroid-stimulating hormone (hTSH) subunit gene fusion produces hTSH with increased stability and serum half-life and compensates for mutagenesis-induced defects in subunit association. J. Biol. Chem. 272, 21312–21316. Heikoop, J.C., van Beuningen-de Vaan, N.M.M., van den Boogart, P., Grootenhuis, P.D., 1997. Evaluation of subunit truncation and the nature of the spacer for single chain human gonadotropins. Eur. J. Biochem. 245, 656–662.
141
Jia, K.C., Oikawa, M., Bo, M., Tanaka, T., Ny, T., Boime, I., Hsueh, A.J., 1991. Expression of human luteinizing hormone (LH) receptor: interaction with LH and chorionic gonadotropin from human but not equine, rat, and ovine species. Mol. Endocrinol. 5, 759–768. Kaetzel, D.M., Nilson, J.H., 1988. Methotrexate-induced amplification of the bovine lutropin genes in Chinese hamster ovary cells. Relative concentration of the alpha and beta subunits determines the extent of heterodimer assembly. J. Biol. Chem. 263, 6344–6351. Kanda, M., Jablonka-Shariff, A., Sato, A., Pixley, M.R., Hiro’oka, T., BenMenahem, D., Boime, I., 1999. Genetic fusion of an alpha-subunit gene to the follicle-stimulating hormone and chorionic gonadotropin-beta subunit genes: production of a bifunctional protein. Mol. Endocrinol. 13, 1873–1881. Kasuto, H., Levavi-Sivan, B., 2005. Production of biologically active tethered tilapia LH␣ by the methylotrophic yeast Pichia pastoris. Gen. Comp. Endocrinol. 140, 222–232. McNatty, K.P., Heath, D.A., Lun, S., Fannin, J.M., McDiarmid, J.M., Henderson, K.M., 1984. Steroidogenesis by bovine theca interna in an in vitro perifusion system. Biol. Reprod. 30, 159–170. McNatty, K.P., Lun, S., Heath, D.A., Keiboom, L.E., Henderson, K.M., 1985. Influence of follicular atresia on LH-induced cAMP and steroid synthesis by bovine thecae interna. Mol. Cell. Endocrinol. 39, 209–215. Min, K.S., Kang, M.H., Yoon, J.T., Jin, H.J., Seong, H.H., Chang, Y.M., Chung, H.J., Oh, S.J., Yun, S.G., Chang, W.K., 2003. Production of biological active single chain bovine LH and FSH. Asian-Aust. J. Anim. Sci. 16, 498–503. Nakav, S., Jablonka-Shariff, A., Kaner, S., Chadna-Mohanty, P., Grotjan, H.E., Ben-Menahem, D., 2005. The LH gene of several mammals embeds a carboxyl-terminal peptide-like sequence revealing a critical role for mucin oligosaccharides in the evolution of lutropin to chorionic gonadotropin in the animal phyla. J. Biol. Chem. 280, 6676–16684. Nakav, S., Dantes, A., Pen, S., Braw-Tal, R., Chadna-Mohanty, P., Amsterdam, A., Grotjan, H.E., Ben-Menahem, D., Design of bioactive singlechain bovine LH analogs: homologous and heterologous carboxyl terminal peptides (CTP) linker-sequences enhance the secretion. Exp. Clin. Endocrinol. Diab., in press. Narayan, P., Wu, C., Puett, D., 1995. Functional expression of yoked human chorionic gonadotropin in baculovirus-infected insect cells. Mol. Endocrinol. 9, 1720–1726. Pierce, J.G., Parsons, T.F., 1981. Glycoprotein hormones: structure and function. Ann. Rev. Biochem. 50, 495–565. Roth, Z., Meidan, R., Shaham-Albalancy, A., Braw-Tal, R., Wolfenson, D., 2001. Delayed effect of heat stress on steroid production in medium-sized and preovulatory bovine follicles. Reproduction 121, 745–751. Sherman, G.B., Wolfe, M.W., Farmeric, T.A., Clay, C.M., Threadgill, D.S., Sharp, D.C., Nilson, J.H., 1992. A single gene encodes the -subunits of equine luteinizing hormone and chorionic gonadotropin. Mol. Endocrinol. 6, 951–959. Sugahara, T., Pixley, M.R., Minami, S., Perlas, E., Ben-Menahem, D., Hsueh, A.J.W., Boime, I., 1995. Biosynthesis of a biologically active single peptide chain containing the human common ␣ and chorionic gonadotropin  subunits in tandem. Proc. Natl. Acad. Sci. U.S.A. 92, 2041– 2045. Sugahara, T., Sato, A., Kudo, M., Ben-Menahem, D., Pixley, M., Hsueh, A.J.W., Boime, I., 1996. Expression of biologically active fusion genes encoding the common ␣ subunit and the follicle-stimulating hormone  subunit. J. Biol. Chem. 271, 10445–10448. Talmadge, K., Vamvakopoulos, N.C., Fiddes, J.C., 1984. Evolution of the genes for the beta subunits of human chorionic gonadotropin and luteinizing hormone. Nature 307, 37–40. Wolfenson, D., Inbar, G., Roth, Z., Kaim, M., Bloch, A., Braw-Tal, R., 2004. Follicular dynamics and concentrations of steroids and gonadotropins in lactating cows and nulliparous heifers. Theriogenology 62, 1042– 1055.
The steroidogenic effect of single-chain bovine LH analogs in cultured bovine follicular cells Ruth Braw-Tal a , Svetlana Pen a , Moran Grinberg b , Sigal Nakav b , David Ben-Menahem b,∗ a
b
Institute of Animal Science, Agricultural Research Organization, The Volcani Center, Beit Dagan 50250, Israel Department of Clinical Pharmacology, Faculty of Health Sciences, Ben–Gurion University of the Negev, Beer-Sheva 84105, Israel
Abstract Single-chain gonadotropin analogs had been constructed for the purpose of structure–function studies and analog design. Incorporation of a spacer derived from the carboxyl terminal peptide (CTP) of the choriogonadotropin (CG)  subunit between the tethered subunit domains of the human gonadotropins is beneficial for the secretion of the single-chain variants without compromising biocactivity. Although the CG subunit containing the CTP domain is expressed only in primates and equids, a CTP-like sequence exists in the untranslated region of the LH gene of several mammalian species, including the bovine species. The CTP encrypted in the bovine LH DNA (designated as ‘boCTP’) and the CTP derived from the human CG subunit (denoted as ‘huCTP’) served as a linker sequence in the design of bovine single-chain luteinizing hormone (LH) analogs. The purpose of the present study was to evaluate steroidogenesis in cultured bovine theca cells following stimulation with these singlechain analogs. The concentration of the LHboCTP␣ and LHhuCTP␣ analogs in the conditioned media of the expressing CHO cells was threeto six-fold higher than that of the “linkerless” LH␣ and LH111␣ variants. The four analogs induced androstenedione and progesterone secretion from the primary theca cells in a dose-dependent manner, but differences in the steroidogenic response were observed. The LHboCTP␣ analog (10 ng/ml) effectively induced androstenedione and progesterone secretion over unstimulated levels (4.0- and 4.4-fold increase for androstenedione and progesterone, respectively). The response to the pituitary bovine LH standard (10 ng/ml) was less pronounced for both steroids (two- to three-fold increase over basal levels). The activities of LHhuCTP␣, LH␣ and LH111␣ were comparable and sightly reduced relative to the LHboCTP␣ activity. The data suggested that LHboCTP␣ was ranked as the most potent and this was even more prominent when analogs were used at a lower dose (1 ng/ml). These data suggest that the design, including the huCTP or boCTP linker, is favorable for the production of single-chain bovine LH analogs. Furthermore, spacing of the tethered subunit domains with the cryptic boCTP sequence that originated from the bovine LH gene appears advantageous for the purpose of stimulating steroid production in the species-specific bioassay. Thus, an effective strategy to produce bioactive single-chain LH analogs in non-primate, non-equid species would be the mutatation of the LH genes with the aim of expressing the cryptic CTP sequence as a spacer derived from the DNA of the same organism. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Bovine LH; Single chain analogs; CTP; Cryptic CTP; Theca cells; Species-specific bioassay
1. Introduction The pituitary gonadotropins, luteinizing hormone (LH), follicle stimulating hormones (FSH) and the placental chorionic gonadotropin (CG) that is produced in primates and equids, are regulators of steroid production in the ovary and testis. Each of these glycoproteins is a non-covalent heterodimer, composed of a common ␣ subunit and a unique  subunit that confers the receptor-specificity. The heterodimer, but not the monomeric subunits, efficiently binds and activates the cognate receptor (Pierce and Parsons, 1981). In order to bypass the rate-
∗
Corresponding author. Tel.: +972 8 6477485; fax: +972 8 6477629. E-mail address: [email protected] (D. Ben-Menahem).
0303-7207/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.mce.2006.03.011
limiting assembly step, to expand the arsenal of muteins for structure–function studies and analog design, and to combine various activities in a single polypeptide single chain, analogs of the human glycoprotein hormones were genetically engineered by in tandem linking of the genes encoding the monomeric subunits (Narayan et al., 1995; Sugahara et al., 1995, 1996; Grossman et al., 1997; Kanda et al., 1999; Garcia-Campayo et al., 2005). Recently, the tethering approach was exploited to gonadotropins of additional species including equine, ovine, bovine and tilapia (Galet et al., 2001; Dirnberger et al., 2001; Fidler et al., 2003; Min et al., 2003; Kasuto and Levavi-Sivan, 2005; Nakav et al., in press). We previously engineered genetically the single-chain bovine LH variants, in which the  and ␣ subunit domains were linked directly or via the heterologous huCTP (derived from the human
R. Braw-Tal et al. / Molecular and Cellular Endocrinology 252 (2006) 136–141
CG subunit) or the homologous boCTP (encrypted in the untranslated region of the bovine LH gene (Nakav et al., 2005) sequences and expressed them in CHO cells (Nakav et al., in press). The data suggested that the incorporation of the boCTP or huCTP linker enhanced the rate and efficiency of secretion of the glycoprotein from the transfected cells by about three-fold. The single-chain analogs induced progesterone production in immortalized rat granulosa cells expressing the rat LH receptor with potency and efficacy similar to those observed for a standard preparation of the native hormone purified from bovine pituitaries (Nakav et al., in press). These results were in consistence with various studies that examined the in vitro activity of human and non-human single-chain gonadotropins. In most of the published studies, the in vitro activity of single-chain gonadotropins was determined using cells that (a) are incapable of producing sex steroids (i.e., bioassay aimed to determine receptor-binding/signal-transduction in transfected CHO or HEK-293 cells expressing the cognate receptor) or (b) steroidogenic cells that express the receptor derived from a species (usually from a rodent origin) that differs from that of the examined gonadotropin. However, little information is available on the steroidogenic activity of mammalian single-chain goandotropins in follicular cells derived from the same species. The aim of our study was to assess the effect of single-chain bovine LH analogs on androstenedione and progesterone secretion from cultured bovine theca cells.
137
domain. For comparison, a truncated LH111␣ [-(1-111)-␣-(1-96)] variant was also constructed (Fig. 1). We made the truncation of the decapeptide because previous analyses showed that, in the LH to CG gene evolution in humans and horses, the naturally occurring CTP of the CG subunit substituted the 7 (as in primates) or 10 (in equids) carboxyl–amino acid residues of the LH subunit (Talmadge et al., 1984; Sherman et al., 1992). In the engineering of the bovine LH-CTP subunit variants, we applied the proposed evolution paradigm of the equine CTP-bearing subunit to the bovine LH subunit (Nakav et al., 2005) that does not contain a CTP, because the remainder of the equine LH/CG gene is similar to the bovine (Sherman et al., 1992) and cattle are phyllogenetically closer to horses than humans. These bovine LH subunit variants were used in the construction of the single-chain LH analogs (Nakav et al., in press). In addition, the analogous truncation of the carboxyl heptapeptide in the LH subunit domain was beneficial to the secretion of the CTP-spaced single-chain, human LH analog with no compromise of the receptor binding and signal transduction (GarciaCampayo et al., 1997). The high degree of homology in the LH gene among mammalian species and, particularly, the conservation of the carboxyl region suggested that the combination of bovine LH carboxylterminal deletion and incorporation of a CTP spacer is favorable for the design of tethered bovine LH analogs (Nakav et al., in press). Media samples were collected from expressing cells following 48 h of incubation in serum free F-12 medium. The samples were concentrated (×20) and washed twice with PBS (10 kDa cut-off; VivaSpin 20 ml concentrators, VivaScience, Hannover, Germany). The concentration of recombinant single-chain analogs was determined by enzyme immunoassay (EIA), as previously described (Wolfenson et al., 2004; Nakav et al., in press). We examined the steroidogenic response to the four different recombinant LH analogs that include or are devoid of a CTP linker that spaces the tethered ␣ and LH subunit domains: LHboCTP␣, LHhuCTP␣, LH111␣ and LH␣·
2.2. Tissue collection and cell incubation 2. Materials and methods 2.1. Design, expression and quantification of the tethered bovine LH variants We expressed in CHO cells single-chain bovine LH variants in which the carboxyl region of the LH subunit was linked to the amino end of the ␣ subunit (Nakav et al., in press). The LH␣ analog encodes the wild type subunit domains [-(1-121)-␣-(1-96)] (Fig. 1). The LH subunit domain of the variants that include a spacer between the subunit domains was truncated from 121 to 111 residues, and was linked through the homologous boCTP (Nakav et al., in press), or the heterologous huCTP (derived from the hCG subunit) to the ␣ subunit
Bovine ovaries were collected at the local abattoir. The theca cells were isolated from medium- and large-sized follicles (5–9 mm in diameter), dispersed enzymatically and cultured. The viability of theca cells as determined with 0.1% (w/v) trypan blue, was 85–95%. Cells (150,000/well) were incubated in Dulbecco’s minimum essential medium with Ham’s F12 1:1 (Biological Industries, Beit Ha’emek, Israel) a final volume of 0.5 ml in a 24 well plates (Nunc, Kampstrup, Denmark). Cells were incubated overnight in a medium containing 1% fetal calf serum (Biological Industries). Medium was discharged and fresh medium containing 0.1% BSA and either bovine LH (USDA-bLH B6) or studied LH variants at different doses was added. Medium was collected after 6 h, and in the initial experiments it changed to fresh medium and was collected after an additional 18 h (see below). Four replicate wells were used for each treatment in a single experiment and dose–response analysis was repeated at least four times in separate experiments. Media were collected from each well and stored at −20 ◦ C pending analysis of steroid concentration. Our preliminary experiments showed that both basal and dose-dependent LH-stimulated rate of steroid secretion decreased markedly as culture time progressed from 6 to 24 h (Fig. 2A-progesterone and B-androstenedione). We therefore cultured the theca cells for 6 h in the subsequent experiments and stimulated the cells with LHand single-chain analogs with doses no greater than 10 ng/ml, unless otherwise specified.
2.3. Steroid analyses
Fig. 1. Design of bovine LH single chain variants. The subunit domains were genetically linked in tandem in the absence and presence of linkers. The human CG carboxyl terminal peptide (huCTP; heterologous linker) or the CTP-like sequence decoded from the bovine LH gene (boCTP; homologous linker) was used to space the tethered subunit domains. Numbers at the top of the diagrams refer to amino acids, continuously numbered.
Progesterone levels were determined by a solid-phase radioimmunoassay (RIA) using a Coat-a-Count Kit (DPC, Los Angeles, CA, USA) according to manufacturers’ instructions, except that standards were prepared in PBS solution. Intra- and inter-coefficients of variation were 8% and 13%, respectively. Androstenedione was determined by RIA using specific antibodies developed and characterized by Dr. F. Cohen (Weizmann Institute of Science, Rehovot, Israel). Intra- and inter-coefficients of variation were 5% and 11%. The results were expressed as mean ± S.E.M. from different experiments as specified in the figure legends. Statistical differences between means were calculated using Student’s t-test.
138
R. Braw-Tal et al. / Molecular and Cellular Endocrinology 252 (2006) 136–141
3. Results 3.1. Quantification of the analogs The concentration of two analogs containing the CTP domain in the condition media was 950 ± 195 and 1220 ± 130 ng/ml (mean ± S.E.M.) for LHboCTP␣ (n = 11) and LHhuCTP␣ (n = 12), respectively. The concentration of “linkerless” analogs was 190 ± 35 and 310 ± 125 ng/ml, for LH111␣ (n = 10) and LH␣ (n = 5), respectively. These results suggest that the design that includes the homologous boCTP- or the heterologous huCTP linker is beneficial for the production of recombinant bovine LH analogs in the single-chain format. 3.2. Steroid production following stimulation with the single chain analogs
Fig. 2. Rate of progesterone (A) and androstenedione (B) accumulation in medium. Bovine theca cells were cultured for 6 h, collected and changed to fresh medium and collected after additional 18 h. Rate of progesterone and androstenedione accumulation per hour during first 6 h (filled bars) and the following 18 h (open bars) was calculated. Values are expressed as mean ± S.E.M. of three separate experiments, each run in quadruplicates and expressed as ng/ml/h.
In bovine follicles up to the preovulatory stage, theca interna is the exclusive site for LH receptors, and the main steroids secreted from these cells are androstenedione and progesterone. Androstenedione and progesterone basal secretion by theca cells during 6 h of culture were 0.85 ± 0.18 and 1.69 ± 0.38 ng/ml (mean ± S.E.M., n = 11), respectively (Fig. 3). Bovine LH (10 ng/ml) increased androstenedione and progesterone secretion to 2.13 ± 0.30 and 5.20 ± 0.79 ng/ml, respectively. Thus, LH-stimulated androstenedione secretion increased 2.5-fold and progesterone secretion increased about 3-fold compared to the basal level. The steroidogenic dose response profiles of the four singlechain LH analogs were studied in the same bioassay and some differences were noticed between the analogs (Fig. 3). The LHboCTP␣ analog was the most potent in its effect
Fig. 3. Steroidogenic response to LHboCTP␣ (triangles), LHhuCTP␣ (squares) (panels A and B), and to LH111␣ (diamonds) and LH␣ (filled squares) (panels C and D). Progesterone (A and C) and androstenedione (B and D) accumulation in theca-cell culture medium was determined after 6 h. The effect of bovine LH is also shown (filled circles). Values are mean ± S.E.M. of 4–7 (single-chain variants) or 10 (bovine LH) of separate experiments, each run in quadruplicates. * P < 0.05; ** P < 0.01 vs. unstimulated steroid levels.
R. Braw-Tal et al. / Molecular and Cellular Endocrinology 252 (2006) 136–141
Fig. 4. The net steroidogenic response expressed as fold increase relative to basal level for androstenedione (open columns) and progesterone (filled columns). The levels in the culture media were determined after 6 h of stimulation with 1 ng/ml (A) or 10 ng/ml (B) of bovine LH or the single-chain analogs. Values are mean ± S.E.M. of 4–7 (single-chain variants) or 10 (bovine LH) experiments, run in quadruplicates. * P < 0.05; ** P < 0.01 vs. bovine LH.
on steroid secretion by theca cells. Addition of 1 ng/ml of LHboCTP␣ increased androstenedione and progesterone accumulation (P = 0.01 and P < 0.05, respectively, Fig. 3, panels A and B). LHhuCTP␣ seemed to be less biologically potent, as a significant increase in androstenedione and progesterone secretion was observed when added to the medium at a concentration of 10 ng/ml, but not 1 ng/ml (Fig. 3A and B). The LH␣ analog significantly increased progesterone accumulation in medium at a concentration of 10 ng/ml but not 1 ng/ml (Fig. 3C). In comparison to unstimulated cultures, the effect of LH␣ on androstenedione was not significant (Fig. 3D). LH111␣ significantly increased progesterone secretion at a concentration of 10 ng/ml (P < 0.05) and androstenedione secretion at a concentration of 20 ng/ml, (P < 0.05, data not shown) (Fig. 3C and D). As expected, increasing volumes (1, 10 and 100 l) of the conditioned media derived from empty vector transfected cells (pCINeo) did not enhance steroidogenesis above the basal levels (results not shown). To neutralize the inter-assay variations at steroid levels and to highlight some of the differences we observed between the analogs, the results were summarized as a stimulated/unstimulated (basal) ratio of androstenedione and progesterone production in the theca cells. Fig. 4 shows the hormone- or analog-stimulated steroid concentration increase in the culture media, relative to the basal secretion (fold increase). This net effect was calculated for the stimulant at a dose of 1 ng/ml (panel A) and 10 ng/ml (panel B). These data suggest that LHboCTP␣ was the most potent analog as
139
judged by progesterone and androstenedione secretion. Whereas pituitary bLH (1 ng/ml) increased 1.85 ± 0.45 and 1.25 ± 0.2fold of androstenedione and progesterone respectively, the LHboCTP␣ had a significantly stronger effect. It increased androstenedione 3.15 ± 0.4-fold and progesterone 2.7 ± 0.3fold over the basal levels (Fig. 4, panel A). Similar results were observed at the higher dose (10 ng/ml): pituitary bLH stimulated androstenedione 2.8 ± 0.35-fold and progesterone 2.1 ± 0.30fold over the basal levels. Addition of LHboCTP␣ resulted in 4.0 ± 0.45 (P < 0.05) and 4.4 ± 0.65 (P < 0.01) fold increase for androstenedione and progesterone, respectively (Fig. 4, panel B). The net steroidogenic response of the LHhuCTP␣ and LH111␣ variants was comparable. These single-chain variants stimulated a slightly reduced androstenedione accumulation compared to the LHboCTP␣ effect at 1 and 10 ng/ml doses (Fig. 4, panels A and B). The androstenedione response in the cultures treated with LH␣ was reduced as compared to the cells stimulated with the bLH, particularly at 10 ng/ml (P < 0.01 compared to the bLH). Nevertheless, at a further increased dose (20 ng/ml), LH␣-induced androstenedione production was similar as compared to the other three single-chain variants and LH (data not shown). This implies that the lower extent of the theca cells’ response to LH␣ as compared to other variants (Fig. 4) reflects a reduced sensitivity of the theca cells to this analog, rather than a decrease in the efficacy of the action. Analysis of the variance (ANOVA) between the net steroidogenic effects of the different recombinant analogs showed a higher androstenedione secretion in response to LHboCTP␣ comapred to LH␣ (1 and 10 ng/ml; P < 0.05). In summary, these data suggest that the extent of the response to low doses of the analogs is not identical, and the net steroidogenic induction using the LHboCTP␣ variant was more potent than the LH␣ effect. 4. Discussion Our study indicates that the incorporation of a boCTP or hCG-CTP linker and, to a less extent, the boCTP sequence increased the amount of single-chain bovine LH in the culture media by more than three-fold compared to the variants that lack a spacer. This enhancement may result from the accelerated kinetics and improved efficiency of the LHboCTP␣ and LHhuCTP␣ secretion, relative to the “linkerless” muteins that were previously observed in pulse-chase experiments (Nakav et al., in press). This notion is in consistence with studies on human gonadotropins and TSH single-chain analogs, suggesting that the determinants for enhanced secretion are preserved in the hCG-CTP when used as a linker sequence (for example, see Sugahara et al., 1996, recently reviewed in BenMenahem, 2004). Improved synthesis in the transfected cells may also contribute to the enhanced presence of the huCTP␣ and LHboCTP␣ analogs in the culture media. Currently it is not clear whether the sequence or the length of the CTP linker is the critical factor to allow the bovine LH subunit domains to fold into a structure that is efficiently secreted and activates the receptor. When tethering human FSH, a linker sequence compromised of half of the hCG-CTP was
140
R. Braw-Tal et al. / Molecular and Cellular Endocrinology 252 (2006) 136–141
sufficient to enhance secretion and preserve bioactivity of the analog (Sugahara et al., 1996). These data suggest that the prosecretory determinants are located on the N-terminal region of the CTP or that the spacing that spans 15 residues is sufficient to relieve potential constraints in the structure of the single chain analog. Although alternative linker spacers (repetitive Ser/Gly sequences) had been used in the construction of single-chain human CG analogs (Heikoop et al., 1997), the application of the CTP as a spacer combines various benefits for the design of tethered gonadotorpin analogs: (a) the stretch is flexible and lacks secondary structure; (b) the CTP is encrypted into the LH gene of several mammals and is naturally expressed in primates and equids; (c) the domain contains determinants for secretion without prevention of receptor activation that is specified by the glycoprotein hormone  subunit. In order to estimate the bioactivity of single-chain bovine LH variants, we used cultured bovine theca cells. Our results show that the examined analogs were bioactive and stimulated steroidogenic response at a high affinity and in a dose–response manner. The extent of the stimulated androstenedione and progesterone secretion over basal level (2.5–3-fold increase) is in accordance with previous reports of the effect of native LH on primary bovine theca cells (for example, McNatty et al., 1984; Berndtson et al., 1995; Roth et al., 2001). This window of stimulated/unstimulated steroid ratio (especially for androstenedione) is probably smaller than the measurement of cAMP levels in these cells or the use of cells derived from inbread mice or rats for the estimation of the activity. In addition, to deal with slaughterhouse material, ovaries that come from cows of different ages, day of the cycle and health conditions, are likely to contribute to inter-assay variations and require a large number of repetitions to enable a reliable interpretation of the results. We measured steroid secretion in response to LH rather than cAMP, mainly because the concern of introducing additional variability originated from atretic follicles that may be present in the pools of the collected ovaries, since it has been shown that early atretic bovine follicle theca cells were capable of secretion of cAMP but not the androgen in response to LH (McNatty et al., 1985). Although the steroidogenic system used by us is not appealing, the advantages of measurement of steroids secreted from cells that are targeted by bovine LH in the ovary of the cattle surpass the above-mentioned disadvantages for the purpose of further evaluation of the bioactivity of the recombinant analogs. The general conclusion from these experiments is that, similar to the heterodimer, the bovine LH analogs induced the production of two main steroids (androstenedione and progesterone) in the same-species bioassay. A variation in the potency of LHboCTP␣, LHhuCTP␣, LH111␣ and LH␣ was detected, and the LHboCTP␣ analog was ranked as the most effective. The structural changes associated with the conversion of the native bovine LH into the tethered variants include: (a) fusion of two subunit treminii, (b) deletion of the carboxy-decapeptide of the bovine LH domain in the construction of the LHboCTP␣, LHhuCTP␣ and LH111␣ analogs and (c) in two of the engineered proteins, the addition of either the huCTP or boCTP linker. These modifications are likely to result in variants with differences in
the quaternary structure that, although tolerated, may affect the fruitful interaction with the bovine LH receptor. The reason why the steroidogenic response in the bovine theca cells is not identical for LHboCTP␣, LHhuCTP␣, LH111␣ and LH␣ is currently unknown. In the future, systematic receptor-binding/signal-transduction in the primary theca cultures should reveal the basis for the differences in steroid production observed among these single-chain analogs. The LHboCTP␣ variant showed the highest potency in the species-specific bioassay. This suggests that this design, that the tethered analog is spaced and has the sequence entirely derived from the bovine genome, activates the bovine LH receptor more strongly than the alternative analogs. Hence, the conformation of the LHboCTP␣ analog may be favorable for the activation of the bovine LH receptor or the boCTP contains determinants contributing to the steroidogenic response. Interestingly, the progesterone production profiles that we previously observed in immortalized rat granulosa cells expressing the rat LH receptor were nearly identical for the same four variants (Nakav et al., in press). That the steroidogenic response to the analogs is literally the same in the heterologous bioassay, but not in the homologous bioassay, implies an action which is different in the rat and bovine LH receptors. Possibly, the rat receptor tolerates the mutagenesis of the bovine LH as induced in the single-chain analogs to a greater extent than the bovine LH receptor. This hypothesis is in accordance to previous studies that showed that, unlike the human LH receptor, the rat receptor efficiently binds LHs originated from various species (Jia et al., 1991). The production of recombinant bovine LH in the single-chain format is advantageous, because it simplifies expression and circumvents the subunit assembly which is a delicate process and a limiting factor in the biosynthesis of heterodimeric LH (Corless et al., 1987; Kaetzel and Nilson, 1988). Activation of the immune system is a concern, when protein analogs are designed for potential in vivo application. To minimize possible antigenicity, the primary sequence of the bovine genes is maintained in the single-chain analogs, except for the introduction of a CTP domain; the huCTP as well as the boCTP sequences are not expressed in the bovine genes and thus could be immunogenic. However, the antigenicity of hCG-CTP domain is low (Birken et al., 1980; Berger et al., 1996) and since the boCTP is derived from the bovine DNA and, similar to the huCTP, the amino acid composition of the boCTP predicts a lack of secondary structure, it is also expected to be a weak antigen. On this basis, the cryptic boCTP linker is not anticipated to stimulate a strong immune response in cattle. The bovine LH single-chain analogs were bioactive, although the dose–response curves and effects over unstimulated – control – theca cells, especially at a low analog dose, were not identical for all four. These results provide a basis for the development of LH analogs that exhibit a spectrum of steroidogenic profiles for potential veterinary application in assisted reproduction procedures. Bioactive analogs capable of eliciting various levels of steroidogenesis in vivo could be beneficial for the purpose of fine tuning of the hormonal milieu at different stages of follicular growth in order to reduce inconsistency and undesired effects
R. Braw-Tal et al. / Molecular and Cellular Endocrinology 252 (2006) 136–141
associated with the contemporary superovulation procedures in cattle. Acknowledgements The authors thank Dr. Albert Parlow and the National Hormone & Peptide Program (NHPP) for the generous provision of the bovine LH and Dr. Fortuna Kohen for antibodies essential to the study. This research was supported by a grant from the Israel Science Foundation – The Charles H. Revson Foundation (448/03). We thank Angela Cohen for her excellent editorial assistance. References Ben-Menahem, D., 2004. Single chain variants of the glycoprotein hormones and their receptors as tools to study receptor activation and for analogue design. J. Neuroendocrinol. 16, 171–177. Berger, P., Bidart, J.M., Delves, P.S., Dirnhofer, S., Hoermann, R., Isaacs, N., Jackson, A., Klonisch, T., Lapthorn, A., Lund, T., Mann, K., Roitt, I., Schwarz, S., Wick, G., 1996. Immunochemical mapping of gonadotropins. Mol. Cell. Endocrinol. 12, 33–43. Berndtson, A.K., Vincent, S.E., Fortune, J.E., 1995. Low and high concentrations of gonadotropins differentially regulate hormone production by theca interna and granulosa cells from bovine preovulatory follicles. Biol. Reprod. 52, 1334–1342. Birken, S., Canfield, R., Lauer, R., Agosto, G., Gabel, M., 1980. Immunochemical determinants determinants unique to human chorionic gonadotropin: importance of sialic acid for antisera generated to the human chorionic gonadotropin -subunit COOH terminal peptide. Endocrinology 106, 1659–1664. Corless, C.L., Matzuk, M.M., Ramabhadran, T.V., Krichevsky, A., Boime, I., 1987. Gonadotropin beta subunits determine the rate of assembly and the oligosaccharide processing of hormone dimer in transfected cells. J. Cell. Biol. 104, 1173–1181. Dirnberger, D., Steinkellner, H., Abdennebi, L., Remy, J.J., van de Wiel, D., 2001. Secretion of biologically active glycoforms of bovine follicle stimulating hormone in plants. Eur. J. Biochem. 268, 4570–4579. Fidler, A.E., Lin, J.S., Lun, S., Ng Chie, W., Western, A., Stent, V., McNatty, K.P., 2003. Production of biologically active tethered ovine FSH␣ by the methylotrophic yeast Pichia pastoris. J. Mol. Endocrinol. 30, 213–225. Galet, C., Le Bourhis, C.M., Chopineau, M., Le Griec, G., Perrin, A., Magallon, T., Attal, J., Viglietta, C., Houdebine, L.M., Guillou, F., 2001. Expression of a single ␣ chain protein of equine LH/CG in milk of transgenic rabbits and its biological activity. Mol. Cell. Endocrinol. 174, 31–40. Garcia-Campayo, V., Sato, A., Hirsch, B., Sugahara, T., Muyan, M., Hsueh, A.J., Boime, I., 1997. Design of stable biologically active recombinant lutropin analogs. Nat. Biotechnol. 15, 663–667. Garcia-Campayo, V., Boime, I., Ma, X., Daphna-Iken, D., Kumar, T.R., 2005. A single-chain tetradomain glycoprotein hormone analog elicits multiple hormone activities in vivo. Biol. Reprod. 72, 301–308. Grossman, M., Wong, R., Szkudlinski, M.W., Weintraub, B.D., 1997. Human thyroid-stimulating hormone (hTSH) subunit gene fusion produces hTSH with increased stability and serum half-life and compensates for mutagenesis-induced defects in subunit association. J. Biol. Chem. 272, 21312–21316. Heikoop, J.C., van Beuningen-de Vaan, N.M.M., van den Boogart, P., Grootenhuis, P.D., 1997. Evaluation of subunit truncation and the nature of the spacer for single chain human gonadotropins. Eur. J. Biochem. 245, 656–662.
141
Jia, K.C., Oikawa, M., Bo, M., Tanaka, T., Ny, T., Boime, I., Hsueh, A.J., 1991. Expression of human luteinizing hormone (LH) receptor: interaction with LH and chorionic gonadotropin from human but not equine, rat, and ovine species. Mol. Endocrinol. 5, 759–768. Kaetzel, D.M., Nilson, J.H., 1988. Methotrexate-induced amplification of the bovine lutropin genes in Chinese hamster ovary cells. Relative concentration of the alpha and beta subunits determines the extent of heterodimer assembly. J. Biol. Chem. 263, 6344–6351. Kanda, M., Jablonka-Shariff, A., Sato, A., Pixley, M.R., Hiro’oka, T., BenMenahem, D., Boime, I., 1999. Genetic fusion of an alpha-subunit gene to the follicle-stimulating hormone and chorionic gonadotropin-beta subunit genes: production of a bifunctional protein. Mol. Endocrinol. 13, 1873–1881. Kasuto, H., Levavi-Sivan, B., 2005. Production of biologically active tethered tilapia LH␣ by the methylotrophic yeast Pichia pastoris. Gen. Comp. Endocrinol. 140, 222–232. McNatty, K.P., Heath, D.A., Lun, S., Fannin, J.M., McDiarmid, J.M., Henderson, K.M., 1984. Steroidogenesis by bovine theca interna in an in vitro perifusion system. Biol. Reprod. 30, 159–170. McNatty, K.P., Lun, S., Heath, D.A., Keiboom, L.E., Henderson, K.M., 1985. Influence of follicular atresia on LH-induced cAMP and steroid synthesis by bovine thecae interna. Mol. Cell. Endocrinol. 39, 209–215. Min, K.S., Kang, M.H., Yoon, J.T., Jin, H.J., Seong, H.H., Chang, Y.M., Chung, H.J., Oh, S.J., Yun, S.G., Chang, W.K., 2003. Production of biological active single chain bovine LH and FSH. Asian-Aust. J. Anim. Sci. 16, 498–503. Nakav, S., Jablonka-Shariff, A., Kaner, S., Chadna-Mohanty, P., Grotjan, H.E., Ben-Menahem, D., 2005. The LH gene of several mammals embeds a carboxyl-terminal peptide-like sequence revealing a critical role for mucin oligosaccharides in the evolution of lutropin to chorionic gonadotropin in the animal phyla. J. Biol. Chem. 280, 6676–16684. Nakav, S., Dantes, A., Pen, S., Braw-Tal, R., Chadna-Mohanty, P., Amsterdam, A., Grotjan, H.E., Ben-Menahem, D., Design of bioactive singlechain bovine LH analogs: homologous and heterologous carboxyl terminal peptides (CTP) linker-sequences enhance the secretion. Exp. Clin. Endocrinol. Diab., in press. Narayan, P., Wu, C., Puett, D., 1995. Functional expression of yoked human chorionic gonadotropin in baculovirus-infected insect cells. Mol. Endocrinol. 9, 1720–1726. Pierce, J.G., Parsons, T.F., 1981. Glycoprotein hormones: structure and function. Ann. Rev. Biochem. 50, 495–565. Roth, Z., Meidan, R., Shaham-Albalancy, A., Braw-Tal, R., Wolfenson, D., 2001. Delayed effect of heat stress on steroid production in medium-sized and preovulatory bovine follicles. Reproduction 121, 745–751. Sherman, G.B., Wolfe, M.W., Farmeric, T.A., Clay, C.M., Threadgill, D.S., Sharp, D.C., Nilson, J.H., 1992. A single gene encodes the -subunits of equine luteinizing hormone and chorionic gonadotropin. Mol. Endocrinol. 6, 951–959. Sugahara, T., Pixley, M.R., Minami, S., Perlas, E., Ben-Menahem, D., Hsueh, A.J.W., Boime, I., 1995. Biosynthesis of a biologically active single peptide chain containing the human common ␣ and chorionic gonadotropin  subunits in tandem. Proc. Natl. Acad. Sci. U.S.A. 92, 2041– 2045. Sugahara, T., Sato, A., Kudo, M., Ben-Menahem, D., Pixley, M., Hsueh, A.J.W., Boime, I., 1996. Expression of biologically active fusion genes encoding the common ␣ subunit and the follicle-stimulating hormone  subunit. J. Biol. Chem. 271, 10445–10448. Talmadge, K., Vamvakopoulos, N.C., Fiddes, J.C., 1984. Evolution of the genes for the beta subunits of human chorionic gonadotropin and luteinizing hormone. Nature 307, 37–40. Wolfenson, D., Inbar, G., Roth, Z., Kaim, M., Bloch, A., Braw-Tal, R., 2004. Follicular dynamics and concentrations of steroids and gonadotropins in lactating cows and nulliparous heifers. Theriogenology 62, 1042– 1055.