Gonadotropin-Releasing Hormone and Gonadotropin-Releasing Hormone Receptor Messenger RibonucleicAcids Expression in Nontumorous andNeoplastic Pituitaries*
NAOKO SANNO, LONG JIN, XIANG QIAN, R. YOSHIYUKI OSAMURA,BERND W. SCHEITHAUER, KALMAN KOVACS, AND RICARDO V. LLOYD
Department of Laboratory Medicine and Pathology (N.S., L.J., X.Q., B.S., R.V.L.), Mayo Clinic andMayo Foundation, Rochester, Minnesota 55905; Department of Pathology (N.S., R.Y.O.), TokaiUniversity School of Medicine, Isehara-City, Kanagawa, 259–11, Japan; and Department of Pathology(K.K.), St. Michael’s Hospital, Toronto, Ontario M5B 1W8, Canada
ABSTRACTIn the nontumorous pituitary, GnRH stimulates the release and
synthesis of LH and FSH by gonadotroph cells via the GnRH receptor(GnRH-R). Little is known, however, about expression of GnRH andGnRH-R messenger RNAs (mRNAs) in nontumorous pituitary tissueand in adenomas. To learn more about the distribution and regulatoryroles of GnRH and its receptor, we investigated the expression of bothGnRH and GnRH-R mRNAs in nontumorous human pituitary and invarious types of pituitary adenomas using the RT-PCR, in situ hy-bridization, and in situ hybridization in combination with RT-PCR (insitu RT-PCR). Using RT-PCR, GnRH mRNA was found to be ex-pressed in normal human pituitaries and in all types of adenomas.Similarly, GnRH-R mRNA was expressed in nontumorous humanpituitaries and in most, but not all, adenomas. These included 5
gonadotroph adenomas, 6 null cell adenomas, 1 of 2 GH-producingtumors, and 1 of 2 ACTH-producing adenomas, but not in the 2PRL-producing adenomas examined. In situ hybridization studiesshowed GnRH and GnRH-R mRNAs in all 3 nontumorous pituitariesand in 12 of 33 (36.4%) and 10 of 33 adenomas (30.3%), respectively.Using an indirect in situ RT-PCR technique to increase the sensitivityof the in situ localization, GnRH and GnRH-R mRNAs were detectedin 29 (87.9%) and 25 (75.8%) of 33 adenomas, respectively.
This is the first report of the localization of GnRH and GnRH-RmRNAs in individual pituitary adenoma cells using in situ RT-PCR.The frequent expression of GnRH and GnRH-R mRNAs in pituitarycells suggests that GnRH has autocrine/paracrine functions in non-tumorous and neoplastic pituitary tissues. (J Clin Endocrinol Metab82: 1974–1982, 1997)
IN THE NORMAL pituitary, GnRH is a hypothalamic hor-mone that stimulates the release and synthesis of LH and
FSH by gonadotroph cells via the GnRH receptor (GnRH-R)(1, 2). Although human pituitary adenomas have been shownto be monoclonal in nature (3, 4), it still is not known whetherhypothalamic hormones play a role in the genesis, growth,or differentiation of human pituitary adenomas. In vitro stud-ies have shown that GnRH induces a significant increase ingonadotropins and/or gonadotropic hormone subunit re-lease by human nonfunctioning pituitary adenomas (5, 6).However, a significant number of pituitary adenomas are notunder hypothalamic hormone regulation. In these adeno-mas, free gonadotropin subunit secretion has been observedafter stimulation by GnRH (7, 8). Abnormal responses toGnRH also have been observed in other types of adenomas.About 15–20% of acromegalic and Cushing’s disease patientshave abnormal increases in circulating GH and ACTH levelsafter GnRH administration (9, 10).
The presence of GnRH-producing cells (11) and of GnRH
messenger RNA (mRNA) (12) has been reported in rat an-terior pituitary, which suggests that endogenously synthe-sized GnRH may be involved in local regulatory mecha-nisms. Alexander et al. (13) detected GnRH-R mRNA inhuman pituitary adenomas in vitro. In addition, using PCRtechniques, Miller et al. (14) reported both GnRH andGnRH-R mRNAs in gonadotroph tumors, as well as in thenormal human pituitary. Little is known, however, aboutGnRH and GnRH-R gene expression in other types of pitu-itary adenomas.
In situ hybridization is a technique useful in demonstratinggene expression in individual cells, but it is limited in itsability to detect low copy numbers of mRNAs. Recently,combined in situ hybridization and reverse transcriptionPCR (in situ RT-PCR) methods were applied to pituitary cells(15). In the present report, we used this new technique toexamine the expression of GnRH mRNA and GnRH-RmRNA in nontumorous human pituitaries and in varioustypes of pituitary adenomas.
Materials and MethodsTissues and cells
Three normal autopsy pituitaries, each obtained within 8 h of deathfrom adult patients without endocrine abnormalities, and 33 pituitaryadenomas were used in this study. Of patients with pituitary adenomas,5 patients had GH-secreting tumors and acromegaly and elevated serumlevels of GH and/or insulin-like growth factor, 4 had PRL-secreting
Received January 6, 1997. Revised February 13, 1997. Accepted Feb-ruary 19, 1997.
Address all correspondence and requests for reprints to: Ricardo V.Lloyd, M.D., Department of Laboratory Medicine and Pathology, MayoClinic and Mayo Foundation, 200 First Street SW, Rochester, Minnesota55905.
* This work was supported in part by Grant NIH-CA-42951.
adenomas with elevated serum PRL levels, three had ACTH-secretingtumors and Cushing’s disease, and 21 had clinically nonfunctioningadenomas with no evidence of hormone hypersecretion and serum PRLlevels less than 100 ug/L. Tumors, in which 25% or more of adenomacells immunostained for FSH or LH b-subunits, were classified as go-nadotropic adenomas. The remaining 12 tumors, which did not showhormone immunoreactivity or in which less than 25% of cells stained forgonadotropin b-subunits, were classified as null cell adenomas.
Tumor tissue fragments were frozen immediately at 270 C for thepurpose of RNA extraction, immunohistochemistry, in situ hybridiza-tion, and in situ RT-PCR studies. As controls for RT-PCR, in situ hy-bridization, and in situ RT-PCR studies, two experimental cell lines wereused: 1) the GH3 rat PRL and GH-producing cell line (American TypeCulture Collection, Rockwille, MD); and 2) the aT3-1 mouse pituitarygonadotroph cell line that produces a-subunit (Dr. P. Mellon, Universityof California, San Diego, CA). Both cell lines were maintained in DMEM(Life Technologies, Grand Island, NY) with 15% horse serum, 2.5% FBS,5 mg/mL insulin and 1% antibiotics. Cells were harvested for RNAextraction, and cell aliquots were affixed to slides by cytocentrifugation,fixed for 20 min in 4% paraformaldehyde at pH7.2, as previously re-ported (15), and used for in situ studies.
Frozen sections of pituitary adenomas and nonneoplastic autopsypituitaries were cut at 8 microns, fixed in 4% paraformaldehyde, washedin 23 standard saline citrate (SSC), dehydrated in alcohol, and stored at270 C. Mounted on silane-coated glass slides, these tissues were usedfor immunohistochemistry, in situ hybridization, and in situ RT-PCRexperiments.
Oligonucleotide primers and probes
Oligonucleotide primers and hybridization probes were produced ona DNA oligonucleotide synthesizer (Applied Biosystems, Foster City,CA) (Table 1). Both primers and probes for human GnRH (GenEMBL:X15215) and human GnRH-R (GenEMBL: L07949) were synthesized onthe basis of published sequences (14, 16, 17). Mouse GnRH (GenEMBL:M14872) and GnRH-R (GenEMBL: M93108) primers also were synthe-sized (18, 19).
Solution PCR. Total RNA extraction was performed by the single-stepmethod (TRIzol reagent kit, Life Technologies) from 3 nontumorouspituitaries and 17 cases of adenomas (20, 21).
First-strand complementary DNA (cDNA) was prepared from totalRNA by using a first-strand synthesis kit (Stratagene, La Jolla, CA). TheRT reaction was performed at 37 C for 60 min in a final vol of 50 mL with
5 mg total RNA, 300 ng antisense primer, 13 RT buffer, 1.0 mmol/L ofeach deoxyribonucleotide (dATP, dCTP, dTTP and dGTP), 40 U ribo-nuclease (RNase) inhibitor, and 50 U Moloney murine leukemia virusRT. The reaction product was then heated at 95 C for 5 min and im-mediately placed on ice.
The PCR was performed in 100-mL final reaction volumes containing5 mL RT reaction product as template DNA, corresponding to cDNAsynthesized from 500 ng total RNA, 13 PCR buffer (Promega, Madison,WI), 1.5 mmol/L MgCl2, 0.2 mmol/L of each deoxynucleotide (Boehr-inger Mannheim, Indianapolis, IN), 300 ng of each sense and antisenseprimer for GnRH and GnRH-R, and 2.5 U Taq DNA polymerase (Pro-mega). Programmable temperature cycling (Perkin-Elmer/Cetus 480,Norwalk, CT) was performed with the following cycle profile: 95 C for5 min, followed by 94 C for 1 min, 60 C for 1 min, and 72 C for 2 min(30 cycles) for GnRH; and 94 C for 1 min, 62 C for 1 min, and 72 C for2 min (35 cycles) for GnRH-R, respectively. After the last cycle, theelongation step was extended at 72 C for 10 min.
A 20-mL aliquot of PCR product was analyzed by gel electrophore-sis, using a 2% agarose gel, and was stained with ethidium bromide.PHx174 DNA/HaeIII digest (Boehringer Mannheim) was used as thestandard. The separated PCR products were transferred to nylonmembrane filters. Southern hybridization, with a single internalprobe that hybridized to regions within the amplified sequences, wasperformed. Hybridization was performed with 1 3 106 cpm/mL [33P]deoxyadenosine diphosphate-labeled probe at 42 C for 18 h. Afterwashing with 63 SSC-0.1% SDS at 23 C for 20 min and at 42 C for 20min, autoradiography was performed at 270 C with Kodak Omat-ARfilm (Eastman Kodak, Rochester, NY) with intensifying screens. InRT-PCR experiments, total RNAs from the aT3-1 and GH3 cell lineswere included as respective positive and negative controls for GnRHand GnRH-R.
Immunohistochemistry. Immunostaining for anterior pituitary hormonesused the avidin-biotin peroxidase complex method (Vector Laborato-ries, Burlingame, CA). Primary antibodies were directed against the fullspectrum of anterior pituitary hormones, including GH (1:1000 dilution),PRL (1:1000), LHb (1:500), FSHb (1:500), TSHb (1:1000) (all rabbit poly-clonal and obtained from the National Pituitary Agency, Bethesda, MD),and rabbit polyclonal ACTH (1:1000) (Dako Corp., Santa Barbara, CA).The monoclonal antibody to a-subunit of glycoprotein hormones (1:250)was purchased from Biogenex (San Ramon, CA). Chromogranin A an-tibody (LK2H 10, 1:1000) was produced in our laboratory, as previouslydescribed (22). The reaction products were visualized by 3,39-diamino-benzidine tetrahydrochloride.
TABLE 1. Sequences of primers and hybridization probes for GnRH and GnRH-R mRNAs
Sequence Location in gene
HumanGnRH sense primer AGCTGGCCTTATTCTACTG 248–30GnRH antisense primer CTGCCCAGTTTCCTCTTCA 180–198GnRH hybridization probesa
MouseGnRH sense primer GGCCGGCATTCTACTGCTG 1607–1625GnRH antisense primer CTGCCTGGCTTCCTCTTCA 5021–5039GnRH hybridization probe TGCCATTTGATCCACCTCCTTGCCCATCTC 3561–3590GnRH-R sense primer TACAAAGCAACAGCAAGCTTG 488–508GnRH-R antisense primer ACGACAAAGGAGGTGGCAAAT 852–872GnRH-R hybridization probe ATTCATCTGTAGTTTGCG TGGGTCTTGATG 766–795
a Hybridization probes 1–4 used for both in situ hybridization and in situ RT-PCR.b Probe used for Southern hybridization.
GnRH AND GnRH-R mRNAs IN PITUITARIES 1975
In situ hybridization
A cocktail of four oligonucleotide probes for GnRH and for GnRH-R,internal to the amplified products from the primers, were labeled withdigoxigenin-deoxyuridine 5-triphosphate (Boehringer Mannheim) byterminal deoxyribonucleotidyl transferase reaction, as previously re-ported (23). The ISH procedure was performed as previously described(23, 24). In brief, after deparaffinization, the sections were treated with1 mg/mL proteinase K (Boehringer Mannheim) at 23 C for 15 min,followed by heat treatment, hydrochloride treatment, acetylation, andthen prehybridization. Thereafter, the sections were hybridized with 1ng/mL of the cocktail probe at 50 C for 18 h. After hybridization, im-munodetection was performed using antidigoxigenin at a 1:500 dilution(Boehringer Mannheim). The reaction product was visualized by ni-troblue tetrazolium salt and 5-bromo-4 chloro-3 indolyl phosphate(NBT-BCIP, Life Technologies). Control experiments were carried outusing sense probes that had complementary sequences to one of theantisense probes.
In situ RT-PCR
The in situ RT-PCR technique was performed according to a three-step protocol previously described (15). The same GnRH and GnRH-Rprimers and probes used for RT-PCR were employed for in situ RT-PCR.Briefly, tissue sections or aT3-1 cytospin cells were digested with 1mg/mL proteinase K at 23 C for 5 min and the enzyme was subsequentlyinactivated by heating to 80 C in PBS for 10 min. Then, after rinsing withH2O, the RT reaction was performed on the slides. The RT reactionmixture included: 10 mmol/L Tris-HCl, 50 mmol/L KCl, 1.5 mmol/LMgCl2, 10 mmol/L dithiothreitol, 0.2 mmol/L each of deoxyribonucle-otide (Boehringer Mannheim), antisense oligonucleotide primer (1 mg/100 mL), RNasin (75U/100 mL), and Superscript II RT (1000 U/100 mL;Life Technologies). The reaction solution (50 mL) was applied to eachslide and was covered by glass coverslips. After a 2-h incubation at 42
C, the slides were washed in 23 SSC, 13SSC, 0.53SSC, and water. ForPCR, a total vol of 100 mL PCR solution with 10 mmol/L Tris-HCl, 50mmol/L KCl, 2.5 mmol/L MgCl2, 1 mmol/L dithiothreitol, 0.2 mmol/Lof each deoxynucleotide, sense and antisense primers (1 mg each/100mL), and Taq DNA polymerase (10 U/100 mL; Promega) was applied toeach slide. The solution was sealed, using mineral oil (Sigma, St. Louis,MO), and then placed on the block of the thermocycler (OmniSlide;Hybaid, Middlesex, UK). After initial denaturation at 95 C for 5 min, PCRamplification was performed using a programmable cycle profile of 20cycles for GnRH mRNA and 30 cycles for GnRH-R mRNA of amplifi-cation, with denaturing at 94 C for 2 min, annealing at 60 C for 1.5 min,extension at 72 C for 1.5 min, and final extension at 72 C for 10 min. Afterthe PCR reaction, tissues were fixed in 4% paraformaldehyde for 5 min,followed by incubation in ethanol and a rinse in 23 SSC. Slides wereincubated with prehybridization solution at 23 C for 20 min and thenhybridized overnight with labeled probes at 42 C. Immunodetection ofthe hybridization signals was performed as above. Control experimentsperformed included: 1) omission of RT or Taq polymerase; 2) omissionof PCR primers; 3) pretreatment with RNase (Sigma), 100 mg/mL in PBSat 37 C for 2 h before reverse transcription; and 4) analyzing the in situRT-PCR amplified products from the supernatant by gel electrophoresisand Southern hybridization blotting (15).
Grading of the in situ hybridization and in situ RT-PCR reactions wasbased on signal intensity as follows: 2, negative; 11, weak; 21, mod-erate; 31, strong.
ResultsImmunohistochemical findings
Immunohistochemical analyses showed 4 adenomas pos-itive only for PRL; all 5 GH adenomas were positive for GH,with 4 of these cases also positive for PRL and for a-subunit
FIG. 1. RT-PCR and Southern hybrid-ization detection of GnRH mRNA andGnRH-R mRNA in normal human pi-tuitaries and human pituitary adeno-mas. Lanes 1–3, nontumorous humanpituitaries; lanes 4 and 5, PRL-secret-ing adenomas; lanes 6 and 7, GH-se-creting adenomas; lanes 8 and 9,ACTH-secreting adenomas; lanes 10–14, gonadotropin-secreting adenomas;lane 15–20, null-cell adenomas; lane 21,negative control without RT for nontu-morous pituitary (NP). The indicatedbands represent the expected 244-bpand 387-bp products for GnRH andGnRH-R, respectively. M, Molecularsize markers. The figures under theethidium bromide-stained gels showthe results of Southern hybridizationwith probes internal to the amplifiedproduct from the primers labeled with33P-detecting GnRH and GnRH-R-am-plified products.
1976 SANNO ET AL. JCE & M • 1997Vol 82 • No 6
of glycoprotein hormones (a-SU). There were 3 ACTH ad-enomas that were positive only for ACTH; 9 gonadotroph(GTH) adenomas, all of which were positive for FSHb anda-SU, 4 of which also expressed LHb. The 12 null cell ade-nomas were all diffusely positive for chromogranin A buthad less than 25% of cells in the adenomas positive for FSHbor LHb.
RT-PCR analysis
The results of RT-PCR are shown in Fig. 1. Analysis ofGnRH mRNA demonstrated that the expected 246-bp PCRproduct was detected by ethidium bromide staining in all 3nontumorous pituitaries and in all 17 tumors studied. Theexpected 387-bp PCR product for GnRH-R mRNA was de-tected in all 3 nontumorous pituitaries and in 5 gonadotropinsecreting adenomas, 6 null cell tumors, 1 of 2 GH-secretingadenomas, and 1 of 2 ACTH-secreting tumors analyzed. Itwas not seen, however, in 2 PRL-secreting adenomas. South-ern hybridization with GnRH and GnRH-R internal probesshowed strong bands for GnRH mRNA in all cases and forGnRH-R mRNA in 13 of the 17 adenomas. To further eval-uate the absence of GnRH-R mRNA in PRL-secreting ade-nomas, the film was exposed for 10 days; the PRL adenomasremained negative like the negative control lanes (data notshown). The control RT-PCR experiment, using RNA ex-tracted from aT3-1 gonadotroph cells, was positive for GnRHand GnRH-R mRNA, whereas RNA from the GH3 cell linewas negative for both (data not shown).
In situ hybridization and in situ RT-PCR
The in situ hybridization signal for GnRH mRNA wasdetected within the cytoplasm of both nontumorous pitu-itary cells and the aT3-1 cells. Compared with negative con-trols without probe or with sense probe, the signals werevariably positive. A positive, but weak, signal for GnRH-RmRNA also was detected in normal pituitaries and in aT3–1cells.
A positive hybridization signal for GnRH mRNA wasdetected in 12 of 33 adenomas (36.3%) by in situ hybridization(Table 2). In most cases, the signal was relatively weak (11).A weak cytoplasmic signal for GnRH-R mRNA also wasdetected in the cytoplasm of 10 of 33 adenomas (30.3%). Insitu hybridization controls with the sense probes were al-ways negative.
When in situ RT-PCR was performed with aT3-1 cells, thehybridization signal for GnRH mRNA was amplified signif-icantly. For the detection of GnRH-R mRNA, 30 cycles of PCRproduced a stronger signal than did 20 cycles (Fig. 2). Variouscontrols for in situ RT-PCR were used, including analysis ofthe supernatant from the reaction. When the products in thesupernatants were electrophoresed, after 20 or 30 cycles ofamplification, bands of expected size were observed forGnRH mRNA and GnRH-R mRNA. Using mouse primersets, in situ RT-PCR products of aT3-1 cells resulted in astronger band than that seen when using human primers.Southern blot hybridization, with internal oligonucleotideprobes labeled with 33P, confirmed the specificity of thebands for GnRH and GnRH-R from the supernatant solu-tions (Fig. 3). Although some nonneoplastic pituitary cells
were positive in the slide preparation, no band was detectedafter 30 cycles of in situ RT-PCR. In situ RT-PCR productsfrom a gonadotropin-producing adenoma showed strongbands for both GnRH and GnRH-R mRNAs. Other controls,including omission of PCR primers, RT, Taq polymerase, orRNase digestion before performing in situ RT-PCR, resultedin only background staining of tissue sections.
The results of in situ hybridization and in situ RT-PCRstudies are shown in Tables 2 and 3. GnRH and GnRH-RmRNAs were detected in 28 (84.8%) and 25 (75.8%) of the 33adenomas, respectively. Although GnRH mRNA was foundin all types of adenomas, GnRH-R mRNA was lacking in all4 PRL-secreting adenomas examined. GnRH and GnRH-RmRNAs were detected in several adenomas that were neg-ative by standard in situ hybridization (Table 3). The signalintensity also was increased for GnRH-R mRNA; moderate(21) to strong (31) signals were observed in tissues thatexpressed only weak signal by in situ hybridization. A pos-itive signal for GnRH mRNA was detected in most cells ofall types of adenomas (Fig. 4). A positive hybridization signalfor GnRH-R was detected in the majority of cells in gona-dotropin-secreting and null cell adenomas, whereas it was
TABLE 2. Results of in situ hybridization and in situ RT-PCRstudy for GnRH and GnRH-R mRNAs
ISH, In situ hybridization; IS-RT-PCR, in situ RT-PCR grading; 2,negative; 11, weak; 21, moderate; 31, strong.
GnRH AND GnRH-R mRNAs IN PITUITARIES 1977
present in 10 to 50% of adenoma cells in GH and ACTHadenomas and not in PRL adenomas.
Discussion
Using solution RT-PCR and in situ RT-PCR analyses, wedemonstrated GnRH and GnRH-R mRNAs in nontumorousand neoplastic pituitary cells. In this study, both GnRH andGnRH-R mRNAs were detected in nontumorous pituitaries.Detection of GnRH-containing cells (12) and of GnRH mRNAin the rat anterior pituitary has been reported (13) and sug-gests that endogenously synthesized GnRH may be involvedin local regulatory mechanisms.
GnRH-R mRNA has been characterized in the pituitariesof several species (25, 26). Alexander et al. (13) and Miller etal. (14) found GnRH and GnRH-R mRNAs in nontumorousand neoplastic human pituitaries by in vitro experiments andthe PCR technique. These reports, however, did not inves-tigate gene expression at the single-cell level.
In the present work, we applied in situ RT-PCR methodsfor the detection of GnRH and GnRH-R mRNAs. Using thistechnique, low amounts of mRNAs were detected in indi-vidual cells, and the signal intensity was found to be in-creased when compared with conventional in situ hybrid-
ization. An increase in the number of PCR cycles (from 20 to30 cycles) increased the percentage of cases positive forGnRH-R mRNA, a finding consistent with the results ofsolution RT-PCR. In situ RT-PCR and in situ PCR both havebeen useful in the amplification and localization of RNA andDNA when present in low copy numbers within cells (27–29).Specifically, this technique recently has been used to detectlow copy numbers of mRNAs in endocrine cells (15, 21, 30).One advantage of performing in situ RT-PCR with noniso-topic probes is the excellent resolution obtained with digoxi-genin or biotin labeling after amplification of low-abundancemRNAs by the PCR technique. The approach was successfulin the present study, because more cases were found to bepositive for GnRH and GnRH-R by in situ RT-PCR than bysimple in situ hybridization. Various controls for the in situRT-PCR procedure were used, including analysis of the su-pernatant of the reaction. After 20 or 30 cycles of amplifica-tion and electrophoresis of the supernatant, the expected sizebands for GnRH mRNA and GnRH-R mRNA were observed.In situ RT-PCR products from cultured aT3-1 cells, usingmouse primers, resulted in stronger bands for GnRH andGnRH-R mRNA than when the human primers were used.Performance of various controls supported the specificity of
FIG. 2. Expression of GnRH and GnRH-R mRNA in aT3–1 mouse gonadotroph cell line by in situ RT-PCR. The signals were detected withdigoxigenin-labeled probes internal to the amplified product from the primer set and visualized with NBT-BCIP (magnification, 3300). A, Thesignal for GnRH mRNA is detected in the cytoplasm of aT3–1 cells (left). Negative control experiment, pretreated with RNase digestion, showsno signal for GnRH mRNA (right). B, Strong signal for GnRH-R mRNA is observed in the cytoplasm of aT3–1 cells after 30 cycles of PCR reaction(left). Negative control, using in situ RT-PCR with ommision of the PCR primers, shows no signal for GnRH-R mRNA (right).
1978 SANNO ET AL. JCE & M • 1997Vol 82 • No 6
our in situ RT-PCR studies. These consisted of omission ofPCR primers, RT, Taq polymerase, and RNase digestion be-fore performing in situ RT-PCR and resulted in no stainingof the tissue sections. The GnRH-R mRNA analysis of thesupernatant products from the normal pituitary was nega-tive after 30 cycles. We have shown previously that the de-tection of bands by gel electrophoresis, using the superna-tants from the in situ RT-PCR reaction, is dependent on thenumber of amplification cycles (15). In the cases of normalpituitary, the supernatant was probably diluted by PRL cellsand some other cells that did not express GnRH-R mRNA.
With the RT-PCR method, GnRH mRNA was detected inall 17 adenomas of various types, and GnRH-R mRNA in 15of the 17. The high incidence of GnRH and GnRH-R geneexpression in tumorous pituitaries suggests that locally pro-duced hypothalamic hormones may play a role in the reg-ulation of tumor cell growth. The presence of GnRH-RmRNA expression in some GH and ACTH-producing ade-nomas could explain the abnormal response to GnRH in
some patients with acromegaly and Cushing’s disease (9, 10).GnRH-R mRNA expression was found in all types of ade-nomas, except PRL-secreting adenomas, by both RT-PCRand in situ RT-PCR. Although our series include only a smallnumber of tumors of varying type, it is of note that PRL-secreting adenomas did not express GnRH-R mRNA. In con-trast, expression of GnRH-R mRNA also has been reportedin aT3-1 mouse gonadotroph tumor cells (31).
In the present study, all 5 gonadotropin-producing ade-nomas and 6 null-cell adenomas expressed GnRH-R mRNA.Miller et al., using the same upstream and downstreamprimer set, reported that 9 of 10 gonadotroph adenomas werepositive for GnRH-R mRNA (14) and that only 6 of 10 ad-enomas were positive using another 59GnRH primer set. Onepossible reason for the difference between these and ourresults may be the selection of adenomas studied. There alsowere differences in the PCR conditions. Miller et al. used alower annealing temperature (at 53.2 C for GnRH and at 52.2C for GnRH-R) for PCR. Despite the fact that a higher an-nealing temperature increases the specificity of the PCR re-action, our experiments resulted in more positive cases, evenat a higher annealing temperature. Because, in the presentstudy, the negative control using RNA from GH3 cells wasconsistently negative for both GnRH and GnRH-R, contam-ination of RNA is highly unlikely.
Genetic studies carried out by X-chromosomal inacti-vation analysis revealed that the majority of pituitary ad-enomas are monoclonal in origin (3, 4). Nonetheless, thehypothalamic hypothesis that pituitary tumors are sec-ondary to hypothalamic dysregulation has not been com-pletely resolved. Indeed, transgenic mouse studies, inwhich GHRH overproduction led to the development ofpituitary adenomas in a time-dependent manner (32), dosuggest that hypothalamic hormones play a role in ade-noma development.
Evidence indicating that endogenous expression of hypo-thalamic hormone in the pituitary has been accumulating inrecent years. A number of hypothalamic neuropeptidesknown to influence anterior pituitary secretion, such asGHRH, TRH, CRH, vasoactive intestinal polypeptide, neu-ropeptide Y, substance P, and galanin have been shown to besynthesized by pituitary cells (33–39). With regard to pitu-itary adenomas, it is of note that using RT-PCR methods,GHRH gene expression has been reported in all somatotrophadenomas but is not evident in either nonfunctioning ornormal anterior pituitary (40). Furthermore, using RT-PCR,CRH transcript has been found in about 20% of pituitarytumors of different types (41).
FIG. 3. Analysis of in situ RT-PCR products. Top, In situ RT-PCRproducts from the supernatant were analyzed, using electrophoresis,followed by ethidium bromide staining; bottom, Southern blot hy-bridization with the 33P-labeled internal probes; left, in situ RT-PCRfor GnRH mRNA; lane 1, mouse aT3–1 cell line using mouse primersfor PCR; lane 2, aT3–1 cell line with human primers; lane 3, frozensection of a nontumorous human pituitary; lane 4, frozen section of anull cell adenoma; lane 5, gonadotroph cell adenoma; and lane 6,normal human pituitary control without RT. After 20 cycles of PCR,the expected 244-bp band for GnRH mRNA is detected in the super-natant. Right, In situ RT-PCR for GnRH-R mRNA; lane 1, aT3–1 cellline using mouse primers; lane 2, aT3–1 cell line with human primers;lane 3, nontumorous human pituitary; lane 4, a gonadotroph celladenoma; lane 5, a GH cell adenoma, and lane 6, normal humanpituitary without RT. After RT-PCR for 30 cycles, the expected 387-bpband for GnRH-R mRNA is detected in the supernatant. M, molecularsize markers.
TABLE 3. Expression of GnRH and GnRH-R mRNAs in pituitary adenomas
ISH, In situ hybridization; IS-RT-PCR, in situ RT-PCR
GnRH AND GnRH-R mRNAs IN PITUITARIES 1979
The finding of various releasing hormones in pituitaryadenomas sheds light on recent observations about pituitaryadenoma with neuronal choristoma lesions (42–44). Suchlesions, which consist of varying proportions of pituitary
adenoma, usually GH cell adenoma, and neurons were longconsidered manifestations of ectopic hypothalamic neuronswith induced adenoma (42, 43). The recent morphologicdemonstration of cells transitional between adenoma cells
FIG. 4. Expression of GnRH and GnRH-R mRNAs in human pituitaries by in situ RT-PCR. The signals were detected with a cocktail of internalprobes labeled with digoxigenin and visualized with NBT-BCIP. A, A positive signal for GnRH mRNA is detected in most cells of a nontumoroushuman pituitary (left). Omission of PCR primers results in no staining (right). (Magnification, 3250.) B, A positive signal for GnRH-R mRNAis present in the cytoplasm of nontumorous pituitary cells. A few cells are negative for GnRH-R mRNA (arrow). (Magnification, 3300.) C, Astrong (21) signal for GnRH mRNA is present in the cytoplasm of a null cell adenoma. (Magnification, 3300.) D, A control slide, with RNAsedigestion at 37 C for 2 h before RT, results in loss of the positive signal. (Magnification, 3300.) E, A positive signal for GnRH-R mRNA is detectedin an ACTH-secreting adenoma. (Magnification, 3300.) F, A gonadotroph adenoma is positive for GnRH-R mRNA in most of the tumor cells.(Magnification, 3300.)
1980 SANNO ET AL. JCE & M • 1997Vol 82 • No 6
and neurons (44) suggests a metaplastic origin of the neu-rons. This interpretation is supported by the present workand that of Wakabayashi et al. (40), which conclusively dem-onstrated releasing hormones in morphologically typical ad-enomas devoid of neuronal elements.
The expression of many hypothalamic hormone receptorgenes by pituitary cells also has been demonstrated. Forinstance, the pharmacologic and biochemical characteristicsof dopamine D2 receptors in PRL-secreting adenomas havebeen investigated extensively (45). Similarly, somatostatin(SS) receptors (SSTRs), which includes five separate sub-types, also have been detected in the pituitary; and SSTR2 andSSTR5 mRNA have been demonstrated (46). Given the ther-apeutic importance of bromocriptine and SS analogs bindingto these receptors (47), the presence of these receptors is ofclinical interest. Expression of GHRH-R mRNA and TRH-RmRNA in human pituitary and pituitary adenomas also hasbeen reported (48).
The processes of GnRH-producing, hypothalamic neuronsterminate not only in the median eminence but also on otherGnRH-producing nerve cells. It is no surprise, therefore, thatGnRH-producing neurons express abundant GnRH-R, anarrangement suggesting that autocrine regulation of GnRHrelease does occur within the brain (49). Thus, it seems thatGnRH acts as a neurohormone, neurotransmitter, or neuro-modulator, as well as a local hormone in the brain. It has beenproposed that pituitary tumors can produce a number ofsubstances that may have secretory, differentiating, and pro-liferative functions (50). Hypothalamic hormones may havejust such specific roles, acting not only as classical releasinghormones, but also as neuromodulators.
Although GnRH-R has been detected in the a-subunitproducing aT3-1 cell line (31, 51), expression of GnRH bythese oncogenically transformed cells has not been detectedpreviously by in situ RT-PCR or other in situ methods. On theother hand, our study showed that aT3-1 cells express bothGnRH and GnRH-R mRNAs by RT-PCR. Similar resultswere obtained using cultured cells by in situ RT-PCR. Theseobservations suggest that autocrine/paracrine regulatorymechanisms may be ongoing in these cells, as well as innormal and neoplastic human pituitary tissues.
The biological significance of GnRH expression by all pi-tuitary adenomas and of GnRH-R expression by most tumorsis uncertain. Recent studies, using endometrial carcinomacell lines, found low levels of GnRH and GnRH-R mRNAs,but secretion of GnRH and functional receptor activity wereuncommon findings in these cell lines (51). Thus, the func-tional significance of GnRH and GnRH-R mRNA expressionby pituitary adenomas will have to be investigated further byfunctional studies in vitro.
In conclusion, the presence of GnRH and GnRH-R mRNAin various types of adenomas, as detected by in situ tech-niques, indicates that endogenous production of the hypo-thalamic hormone GnRH and its receptor in pituitaries mayhave a role in autocrine/paracrine regulation. Further in-vestigations will be needed to clarify the mechanisms of suchregulation and the possible role of hypothalamic hormonesand their receptors in the genesis, growth, and differentiationof pituitary adenomas.
Acknowledgment
The authors wish to thank the National Hormone and Pituitary Pro-gram, Baltimore, MD, for the antibodies for pituitary hormones. Theauthors also thank Dr. Akira Teramoto, Department of Neurosurgery,Nippon Medical School, for advice and assistance.
References
1. Crowley WF, Filicori M, Spratt DI, Santoro NF. 1985 The physiology ofgonadotropin-releasing hormone (GnRH) secretion in men and women. Re-cent Prog Horm Res. 41:473–531.
2. Marshall JC, Kelch RP. 1986 Gonadotropin-releasing hormone: role ofpulsatile secretion in the regulation of reproduction. N Engl J Med.315:1459 –1468.
3. Alexander JM, Biller BM, Bikkal H, Zarvas NT, Arnold A, Klibanski A. 1990Clinically nonfunctioning pituitary tumors are monoclonal in origin. J ClinInvest. 86:336–340.
5. Surmont DWA, Winslof CLJ, Loizou M, et al. 1983 Gonadotropin and alphasubunit secretion by human functionless pituitary adenomas in cell culture:long term effects of luteinizing hormone releasing hormone and thyrotropinreleasing hormone. Clin Endocrinol (Oxf). 19:325–336.
6. Kwekkeboom DJ, De Jong FH, Lamberts SWJ. 1989 Gonadotropin release byclinically nonfunctioning and gonadotroph pituitary adenomas in vivo and invitro: relation to sex and effects of thyrotropin-releasing hormone, gonado-tropin-releasing hormone and bromocriptine. J Clin Endocrinol Metab.68:1128–1135.
7. Lamberts SWJ, Verleum T, Oosteroom R, et al. 1987 The effect of bromocrip-tine, thyrotropin-releasing hormone and gonadotropin releasing hormone onhormone secretion by gonadotropin-secreting adenomas in vivo and in vitro.J Clin Endocrinol Metab. 64:524–530.
8. Katzenelson LJM, Alexander JM, Klibanski A. 1993 Clinical review 45: clin-ically nonfunctioning pituitary adenomas. J Clin Endocrinol Metab.76:1089–1094.
10. Spada A, Lania A. 1996 Hormone receptors in pituitary adenomas. In: LandoltA, Vance ML, Reilly PL, eds. Pituitary adenomas. New York: ChurchillLivingstone; 59–71.
11. Bauer TW, Moriarty CM, Childs GV. 1981 Studies of immunoreactive gona-dotropin releasing hormone (GnRH) in rat anterior pituitary. J HistochemCytochem. 29:1171–1178.
12. Pagesy P, Li JY, Bertet M, Peillon F. 1992 Evidence of gonadotropin-releasinghormone mRNA in the rat anterior pituitary. Mol Endocrinol. 6:523–528.
13. Alexander JM, Klibanski A. 1994 Gonadotropin-releasing hormone receptormRNA expression by human pituitary tumors in vitro. J Clin Invest.93:2332–2339.
14. Miller GM, Alexander JM, Klibanski A. 1996 Gonadotropin-releasing hor-mone messenger RNA expression in gonadotropin tumors and normal humanpituitary. J Clin Endocrinol Metab. 81:80–83.
15. Jin L, Qian X, Lloyd RV. 1995 Comparison of mRNA expression detected byin situ PCR and in situ hybridization in endocrine cells. Cell Vision. 2:314–321.
16. Seeburg PH, Adelman JP. 1984 Characterization of cDNA for precursor ofhuman luteinizing hormone releasing hormone. Nature. 311:666–668.
17. Chi L, Zhou W, Prikhozhan A, et al. 1993 Cloning and characterization of thehuman GnRH receptor. Mol Cell Endocrinol. 91:R1–R6.
18. Mason AJ, Hayflick JS, Zoeller RT, et al. 1986 A deletion truncating thegonadotropin-releasing hormone gene is responsible for hypogonadism in thehpg mouse. Science. 234:1366–1371.
19. Tsutumi M, Zhou W, Millar RP, et al. 1992 Cloning of a functional mousegonadotropin-releasing hormone receptor. Mol Endocrinol. 6:1163–1169.
20. Chomczynski P, Sacchi N. 1987 Single step method of RNA isolation by acidguanidinium thiocyanate-phenol chloroform extraction. Anal Biochem.162;156–159.
21. Qian X, Jin L, Grande JP, Lloyd RV. 1996 Transforming growth factor-b andp27 expression in pituitary cells. Endocrinology. 137;3051–3060.
22. Jin L, Chandler WF, Smart JB, England BG, Lloyd RV. 1993 Differentiationof human pituitary adenomas determines the pattern of chromogranin/se-cretogranin messenger ribonucleic acid expression. J Clin Endorinol Metab.76:728–738.
23. Lloyd RV, Cano M, Chandler WF, et al. 1989 Human growth hormone andprolactin secreting pituitary adenomas analyzed by in situ hybridization. Am JPathol. 134:605–613.
24. Lloyd RV, Jin L. 1995 In situ hybridization analysis of chromogranin A andB mRNAs in neuroendocrine tumors with digoxigenin-labeled oligonucleotideprobe cocktails. Diagn Mol Pathol. 4:143–155.
25. Naor Z, Clayton RN, Catt KJ. 1980 Characterization of gonadotropin-releasinghormone receptors in cultured rat pituitary cells. Endocrinology.107:1144–1152.
GnRH AND GnRH-R mRNAs IN PITUITARIES 1981
26. Pal D, Miller BT, Parkening TA. 1992 Topographical mapping of GnRHreceptors on dispersed mouse pituitary cells by backscattered electron imag-ing. Anat Rec. 233:89–96.
27. Haase AT, Retzel EF, Staskus KA. 1990 Amplification and detection of len-tiviral DNA inside cells. Proc Natl Acad Sci USA. 37:4971–4975.
28. Nuovo GJ, Gallery F, MacConnell P, Becker J, Bloch W. 1991 An improvedtechnique for the in situ detection of DNA after polymerase chain reactionamplification. Am J Pathol. 139:1239–1244.
29. Bagasra O, Seshamma T, Hansen J, et al. 1994 Application of in situ PCRmethods in molecular biology: I. Details of methodology for general use. CellVision. 1:324–335.
30. Martinez A, Miller MJ, Unsworth EJ, Siegfried JM, Cuttitta F. 1995 Expres-sion of adrenomedullin in normal human lung and in pulmonary tumors.Endocrinology. 136:4099–4105.
31. Horn F, Bilezikjian LM, Perrin MH, et al. 1991 Intracellular responses togonadotropin-releasing hormone in a clonal cell line of the gonadotrope lin-eage. Mol Endocrinol. 5:347–355.
32. Asa SL, Kovacs K, Stefaneanu L, et al. 1992 Pituitary adenomas in micetransgenic for growth hormone-releasing hormone. Endocrinology131:2083–2089.
33. Li JY, Knapp RJ, Sternberger LA. 1984 Immunocytochemistry of a privateluteinizing hormone-releasing hormone system in the pituitary. Cell TissueRes. 235;263–266.
34. Vrontakis ME, Sano T, Kovacs K, Friesen HG. 1990 Presence of galanin-likeimmunoreactivity in non-tumorous corticotrophs and corticotroph adenomasof the human pituitary. J Clin Endocrinol Metab. 70:747–751.
35. Hsu DW, Hooi SC, Hedley-Whyte ET, Strauss RM, Kaplan LM. 1991 Co-expression of galanin and adrenocorticotropic hormone in human pituitaryand pituitary adenomas. Am J Pathol. 138:897–909.
36. Jones PM, Ghatei MA, Steel J, et al. 1989 Evidence or neuropeptide Y synthesisin the rat anterior pituitary and the influence of thyroid hormone status:comparison with vasoactive intestinal peptide, substance P and neurotensin.Endocrinology. 125:334–341.
37. Benlot C, Pagesy P, Peillon F, Joubert-Bression D. 1991 Growth hormone-releasing hormone and somatostatin in normal and tumoral pituitaries. Pep-tides. 12:945–950.
38. Castro MG, Brooke J, Bullman A, et al. 1991 Synthesis of corticotropin-releasing hormone (CRH) in mouse corticotrophic tumor cells expressing thehuman proCRH gene. Intracellular storage and regulated secretion. J MolEndocrinol. 7:97–104.
39. Pagesy P, Croissandeau G, Le Dafniet M, et al. 1991 Detection of thyrotropin-releasing hormone (TRH) mRNA by the reverse transcription-polymerase
chain reaction in the human normal and tumoral anterior pituitary. BiochemBiophys Res Commun. 182:182–187.
40. Wakabayashi I, Inokuchi K, Hasegawa O, Sugihara H, Minami S. 1992Expression of growth hormone (GH)-releasing factor gene in GH-producingpituitary adenoma. J Clin Endocrinol Metab. 74:357–361.
41. Levy A, Lightman SL. 1992 Growth hormone-releasing hormone transcriptsin human pituitary adenomas. J Clin Endocrinol Metab. 74:1474–1476.
42. Asa SL, Scheithauer BW, BiIbao JM, et al. 1984 A case for hypothalamicacromegaly: a clinicopathological study of six patients with hypothalamicgangliocytomas producing growth-hormone releasing factor. J Clin Endocri-nol Metab. 58:796–803.
43. Scheithauer BW, Kovacs K, Randall RV, Horvath E, Okazaki H, Laws Jr ER.1983 Hypothalamic neuronal hamartoma, and adenohypophyseal choristoma:their association with growth hormone adenoma of the pituitary gland. J Neu-ropathol Exp Neurol. 42:648–663.
44. Horvath E, Kovacs K, Scheithauer BW, Lloyd RV, Smyth HS. 1994 Pituitaryadenoma with neuronal choristoma (PANCH): composite lesion or lineageinfidelity? Ultrastruct Pathol. 18:565–574.
45. Wood DF, Johnston JM, Johnston DG. 1991 Dopamine, the dopamine D2receptor and pituitary tumors. Clin Endocrinol (Oxf). 35:455–466.
46. Miller GM, Allexander JM, Bikkal HA, et al. 1995 Somatostatin receptorsubtype gene expression in pituitary adenomas. J Clin Endocrinol Metab.80:1386–1392.
47. Levy A, Lightman SL. 1993 The pathogenesis of pituitary adenomas. ClinEndocrinol (Oxf). 38:559–570.
48. Kaji H, Xu Y, Takahashi Y, Abe H, Tamaki N, Chihara K. 1995 Human TRHreceptor messenger ribonucleic acid levels in normal and adenomatous pitu-itary: analysis by the competitive reverse transcription polymerase chain re-action method. Clin Endocrinol (Oxf). 42:243–248.
49. Krsmanovic LZ, Stojilkovic SS, Mertz LM, Tomic M, Catt KJ. 1993 Expressionof gonadotropin-releasing hormone receptors and autocrine regulation of neu-ropeptide release in immortalized hypothalamic neurons. Proc Natl Acad SciUSA. 90:3908–3912.
50. Faglia G, Spada A. 1995 The role of the hypothalamus in pituitary neoplasia.Baillieres Clin Endocrinol Metab. 9:225–241.
51. Windle JJ, Weiner RI, Mellon PL. 1990 Cell lines of the pituitary gonadotropelineage derived by targeted oncogenesis in transgenic mice. Mol Endocrinol.4:597–603.
52. Chatzaki E, Bax CMR, Eidne KA, Anderson L, Grudzinskas JG, GallagherCJ. 1996 The expression of gonadotropin-releasing hormone and its receptorin endometrial cancer and its relevance as an autocrine growth factor. CancerRes. 56:2059–2065.
