GENERAL AND COMPARATIVE ENDOCRINOLOGY 89, 28-38 (1993) immunocytochemical and immunogold Localization of Two Prolactin Isoforms in the Same Pituitary Cells and in the Same Granules in the Tilapia (Oreochromis mossambicus) Jennrrer L. Srecker, Mrtsuyo Kistipa, Livue Huana, Davin S. Kina,* Yosurraka NaGanama,t Hirosit UepA,f AND THOMAS R. ANDERSONS Department of Zoology. University of Rhode Island, Kingston, Rhode Island 02881-0816; *Howard Hughes Medical Instinute. Department of Molecular and Cell Biology, University of California, Berkeley, California 94720; {National Institute for Basic Biology, Okazaki 444, Jopan; #Toya Lake Station for Environmental Biology. Faculty of Fisheries, Hokkaido Unt Abuia, Hokkaido 049-57, Japan: and Berkeley Antibody Company, 4131 Lakeside Drive, Suite B, Richmond, Culifornia 94806-1965 Accepted August 17, 1992 ‘The tilapia pituitary secretes two forms of prolactin (PRL) and a single growth hormone AGA). The {PRLs share only 69% sequence identity and are designated {PRLj7 and (PRL gs to indicate the number of amino acid residues in each isoform. Our aim was to develop specific antisera for detection of these three related polypeptides. Ten peptides correspond- ing to unique epitopes on the tPRLs and two peptides of (GH were synthesized using solid-phase methods, conjugated to cartier proteins, and used as immunogens for antibody proxitction in rabbits, Select antisera for the PRLs were highly specific, exhibiting only 1% ity to the alternate (PRL under noncompetitive ELISA conditions at dilutions, munocytochemical analysis. The antiAGH specifically bound to cells in the prox- imal pars distalis. Production of both PRLS by a single cell type was indicated by the binding of both anti-tPREy77 and aatiAPRL yay fo the same cells cells stained sequei {PRI antibodies labeled with immunogold of two Ultastsuctutal analysis of PRL-prodw the rostral pars distafis. ly using the two different classes indicated that both (PRLs ‘appear in the same granules. These findings suggest that the biological significance of (wo forms of PRL in the adult tilapia is not a function of differential regutation of two different classes of PRL cells or differential release of res, Ie ‘The pituitary of the tilapia, Oreochromis mossambicus, served as the source for the first isolation of prolactin (PRL) and growth hormone (GH) from a teleost fish (Farmer et al., 1976, 1977) and for the first immuno- cytochemical confirmation of the identity of PRL- and GH-producing cells in the te- leost pituitary (Nagahama et al., 1981). Thus, the titapia has served as an important model system for understanding the regula tion and functions of PRL and GH in tele- osts (Bern, 1983; Nishioka ef al., 1988; Grau and Helms, 1989). Further analysis of the secreted products of the organ-cultured rostral lobe of the ti- lapia pituitary revealed that in fact two dis- 0016-6480193 $4.00 ‘pytaht ©1999 by Academic Press. I ahs of repreductin a any frm reserved. 1e secretory granules, ©1983 Academie tinct PRLs are produced (Specker et al., 198Sa,b). Whereas most hormonal isoforms are highly similar, these two tilapia PRLs represent an exaggerated example of hor- monal divergence on the molecular level. The two forms of this polypeptide hormone share only 69% sequence identity (Yamagu- chi et al., 1988). The larger tilapia PRL con- tains 188 residues and is designated {PRL gg, Whereas the smaller tilapia PRL. has two short deletions of five and six res- idues and is designated tPRLy77. Pairs of PRLs have been identified in the chum salmon (Oncorhynchus keta) and the com- mon carp (Cyprinus carpio), but in these other fishes the PRLs differ by only fourTILAPIA PRLS FROM SAME PITUITARY CELL residues and one residue, respectively (Ya- suda et al., 1986, 1987). The striking dissim- ilarity between the two tilapia PRL isohor- mones suggests that duplication on the ge- netic level took place long enough ago for divergence in their regulation and function to have also occurred. Progress in understanding the biology of these hormones has been hampered by lack of isoform-specific antibodies. To over- come the problems of generating highly specific polyclonal antibodies against struc- turally similar molecules, we generated an- tibodies by using synthetic peptides corre- sponding to sequences unique to each iso- form as antigens. Antibodies against synthetic fragments of tilapia GH (tGH) were also raised. Immunochemical analysis of pituitary cytology and cellular ultrastruc- ture confirmed the location of GH- producing cells in the proximal lobe and re- vealed that one cell type in the rostral lobe produces both tPRL ig and tPRLy7; and ‘stores the hormones in the same secretory granule. These results have been reported previously in abstract (Specker ef ai 1991). MATERIALS AND METHODS Animal care and maintenance. The tilapia used in the immunocytochemical studies and for hormone pro- duction were generously supplied by Dr. Rex Dun- ham, Auburn University. The tilapia were maintained in the Biological Sciences Center fish facility (URI) in freshwater under long days (14 hr lighG10 hr dark) in 30- or 50-gallon tanks with coarse gravel on the floor. The flow-through well water is heated to 27", aerated, ‘and filtered. The tilapia are fed Doromin and Cichlid flakes (Tetramin) twice daily. Adult tilapia collected in the vicinity of the Hawaii Institute of Marine Biology, generously supplied by Dr. Gordon Grau, were trans- ported 2 weeks later to the National Institute for Basic Biology (Okazaki, Japan), held in aerated, filtered freshwater at 28° for 2 weeks, and sampled for ultra- structural analysis. Peptide synthesis. Five pairs of peptides represent- ing sequences from the two tPRLs and two peptides from tGH were chosen for synthesis based on their having minimal sequence identity, maximal probabil- ity of being nonhelical connecting loops of high ther- ‘mal mobility (Abdel-Meguid e¢ al., 1987), and maximal 29 probability of corresponding to strongly immunogenic regions of human GH (Cunningham ¢7 al., 1989). Pep- tides (about 50 mg of each) were synthesized using solid-phase methodology (2-(benzotriazol-I-y!)- 1,1,3,3-tetramethyluronium hexafluorophosphate acti- vated 9-fluorenylmethoxycarbonyl amino acids (Fields ‘and Noble, 1990)) on an ABI 430A peptide synthesizer ‘employing user-devised cycles. Cleavage and depro- tection were accomplished with Reagent K (King et ai., 1990), and the crude peptides were purified to +>95% by reversed-phase HPLC witha gradient of ace- tonitrile in water (both containing 0.1% triuoroacetic acid). Structure and purity were confirmed in all cases by electrospray ionization-mass spectrometry. Antibody production. Immunogens were synthe- sized by individually linking each of the peptides to keyhole limpet hemocyanin (KLH). Screening anti- ‘gens were prepared by linking each peptide to bovine serum albumin (BSA). Conjugates were prepared by using the sulfonated derivative of m-maleimidoben- zoyl-N-hydroxysuccinimide to form thioether and ‘amide bonds between the peptide terminal cysteines of the peptide and primary amines on KLH or BSA, re- spectively. Polyclonal antibodies were induced in tab- bits (two rabbits/peptide) using @ perilymphnodal in- {jection strategy. Initial injections were made proximal to the axillary and inguinal lymph nodes, using an emulsion of 0.5 mg antigen in Freund's complete ad- Juvant. Subsequent boosts were made intramuscularly ‘with 0.25 mg immunogen emulsified with Freund’s in- complete adjuvant, Serum samples were drawn peri- odically from the immunized rabbits and antibody titer was determined by an indirect enzyme-linked immu nosorbent assay (ELISA). Microtiter plates were Coated with the BSA-conjugated peptides dissolved in phosphate buffered saline (PBS), pH 7.2-7-4, 50 wl, dried down with air overnight. The wells were blocked af room temperature for 2 hr (200 yl of 5% non-fat dry k, 0.1% sodium azide PBS) and washed with 0.5% BSA PBS, Serially diluted antisera was incubated in the wells, which were then washed, and bound anti- bodies were identified with 2 horseradish peroxidase (HRP)-conjugated second antibody as described be- low. Hormone preparation. Both tPRLs and 1GH were isolated as described previously (Specker ef al., 198Sa). Briefly, pituitaries were removed from fresh- water-adapted male and female tilapia weighing from 10 to 60 g. The pituitaries were cultured intact in hy- poosmotic Krebs bicarbonate supplemented with Ea- ale's minimum essential medium, Pituitaries inthis or- gan culture system can remain viable for 3 weeks; each Pituitary can produce up to 10 ug of each PRL and GH each day, The harvested culture media were filtered ‘through Millipore membrane (0.22 jm) to remove debris and concentrated by ultrafiltration using an Am- icon YMI0 membrane. The concentrated media were then subjected to reversed-phase HPLC using a gradi-