Localization and Regulation of the Growth Hormone Receptor and Growth Hormone-Binding Protein in the Rat Growth Plate

Localization and Regulation of the Growth HormoneReceptor and Growth Hormone–Binding Protein in the Rat

Growth PlateEVELIEN F. GEVERS,1,2 BRAM C.J. VAN DER EERDEN,1 MARCEL KARPERIEN,1,3 ANTON K. RAAP,4

IAIN C.A.F. ROBINSON,5 and JAN-MAARTEN WIT1

ABSTRACT

Growth hormone (GH) has direct effects on the growth plate to stimulate longitudinal growth, but it is not clearwhich chondrocyte populations GH acts on. The dual effector theory suggests that GH would act primarily on the“stem cells.” However, staining with a GH receptor (GHR) antibody is found in all layers of the growth plate inrabbits and humans. We now have investigated the localization and regulation of GHR and the related GH bindingprotein (GHBP) in the rat growth plate using a sensitive immunohistochemical method involving tyramide signalamplification (TSA) and antibodies specific for GHR or GHBP. Both GHR and GHBP were shown in the germinaland proliferative chondrocytes, but most clearly in early maturing chondrocytes at the interface between prolif-erative and hypertrophic cells. Staining for GHR and GHBP was located in both the cytoplasm and the nucleus.Expression of GHR mRNA and GHBP mRNA in the growth plate was confirmed by reverse-transcriptionpolymerase chain reaction (RT-PCR). Immunohistochemical staining for GHR and GHBP decreased with age; in12-week-old normal rats, only the early maturing chondrocytes were stained. In GH-deficient dwarf rats, stainingseemed less than in normal rats, and in hypophysectomized (Hx) rats, staining for GHBP was clearly reduced.Treatment of Hx rats with thyroid hormones (T3 � T4), via subcutaneously (sc) implanted osmotic minipumps,induced little growth and induced a small layer of GHR-positive and GHBP-positive early maturing chondrocytes.Treatment with GH and thyroid hormones (TH) resulted in greater growth and a broader layer of GHR-positiveand GHBP-positive cells, indistinguishable from normal rats. In contrast, dexamethasone treatment of normal ratsinhibited their growth and reduced GHR and GHBP staining in the growth plate. These results show that GHRand GHBP in the growth plate are under hormonal control. The localization of GHR/GHBP suggests that inaddition to actions on germinal and proliferative cells in young rats, GH also has effects on early maturingchondrocytes and may be involved in their differentiation to a fully hypertrophic chondrocyte. (J Bone Miner Res2002;17:1408–1419)

Key words: growth plate, chondrocyte, growth hormone receptor, growth hormone binding protein, tyra-mide signal amplification

INTRODUCTION

THE GROWTH plate is divided into distinct layers of chon-drocytes, resting (also referred to as “reserve,” “germi-

nal,” or “stem”) cells, proliferating cells, and differentiating

(maturing) hypertrophic chondrocytes.(1,2) The border be-tween the zone of proliferating cells and differentiating/maturing cells consists of a few cell layers, often called thetransition zone. The cells in this layer and the upper part ofthe hypertrophic cell layer are considered as “early maturingcells.” In addition to an indirect action via insulin-likegrowth factor (IGF)-I generated at the liver, there is strongThe authors have no conflict of interest.

1Department of Pediatrics, Leiden University Medical Center, Leiden, The Netherlands.2Present address: National Institute for Medical Research, Division of Molecular Neuroendocrinology, The Ridgeway, Mill Hill,

London, United Kingdom.3Department of Endocrinology and Metabolic Diseases, Leiden University Medical Center, Leiden, The Netherlands.4Laboratory of Cytochemistry and Cytology, Department of Molecular Cell Biology, Leiden University Medical Center, Leiden, The

Netherlands.5Division of Molecular Neuroendocrinology, National Institute for Medical Research, Mill Hill, London, United Kingdom.

JOURNAL OF BONE AND MINERAL RESEARCHVolume 17, Number 8, 2002© 2002 American Society for Bone and Mineral Research

1408

evidence that growth hormone (GH) acts locally at thegrowth plate to stimulate longitudinal growth.(3,4) This hy-pothesis has been strengthened recently by experimentsshowing that growth of transgenic mice with a targeteddisruption of the IGF-I gene in the liver is not different fromnormal mice.(5,6) There also is direct evidence that GH actson the stem cells(7) resulting in differentiation to a prolifer-ative cell type that produces IGF-I,(8) which then acts in anautocrine or paracrine fashion to induce further proliferationand clonal expansion of the chondrocytes.(9) Although thistheory implicates the presence of GH receptors (GHRs),primarily in the stem cells in the rat, immunostaining hasshown the presence of the GHR in all cell layers of thegrowth plate.(10) However, this study was performed in therabbit and used a monoclonal antibody (MAb 263) directedagainst the extracellular domain of the GHR, which there-fore does not discriminate between GHR and the GH bind-ing protein (GHBP), which corresponds to the GHR extra-cellular domain.(11,12)

Although GHBP is produced mainly in the liver,(5,6) italso is produced in other tissues expressing GHR.(13) Itsfunction is not clear because, on one hand, it prolongs thehalf-life of GH in vivo,(14) whereas, on the other hand, itcompetes with GHR for binding to GH in vitro and canreduce the effect of GH.(15,16) In rats, GHBP is produced byalternative splicing of the primary GHR RNA(17) and theGHR/GHBP mRNA ratio differs between tissues.(13,18,19)

Furthermore, different GHR/GHBP transcripts exist withmultiple 5� untranslated exons and these might be respon-sible for tissue-specific regulation of GHR and GHBP ex-pression.(20,21) Until now, regulation of GHR expression hasbeen studied mostly in the liver and brain. However, themore relevant target for the regulation of skeletal growth byGH is the expression of GHR and GHBP in the growth plateitself.

In this study, we present data on the expression of GHRand GHBP and mRNA in the growth plate at different ages.By using antibodies specific for GHR, GHBP, or both, wediscriminated between GHR and GHBP and compared theirlocalization in the rat growth plate to identify target cells forGH action.

A sensitive immunohistochemical method, using tyra-mide signal amplification (TSA) with biotin-labeled tyra-mides, readily detected GHR and GHBP. In the rat, GHBPis produced by alternative splicing of exon 8, and thisgenerates a GHBP product with a unique 17 residue hydro-philic tail.(17) Thus, an antibody specific for this peptide tailcould be used to localize GHBP specifically(22) while anantibody against the intracellular domain of the GHR couldbe used to identify GHR specifically. Reverse-transcriptionpolymerase chain reaction (RT-PCR) was used to confirmthe presence of GHR and GHBP mRNA transcripts in thegrowth plate. To investigate hormonal regulation of GHRand GHBP in the growth plate, immunohistochemistry wasperformed on sections of growth plates from normal rats,dwarf rats with a specific GH deficiency,(23) and hypophy-sectomized (Hx) rats with and without thyroid hormone(TH) and/or GH treatment. We also studied the effect oftreatment with dexamethasone in normal-growing rats toinvestigate whether growth inhibition by dexamethasone is

accompanied by changes of GHR expression or localizationin the growth plate.

MATERIALS AND METHODS

Animal experiments

Unless stated otherwise, rats were kept in a light- andtemperature-controlled room (12 h of light and 23–25°C)with food and water available ad libitum. Experiments wereapproved by the local ethical committee for animal exper-iments.

Experiment 1: Male and female normal rats (Lewis strain;Harlan, Zeist, The Netherlands) and GH-deficient dwarfrats(23) (Lewis strain; Harlan), 1, 4, 7, and 12 weeks of age(n � 5–6 per group), were killed by decapitation. Theproximal part of the tibia was isolated, cut in half longitu-dinally, fixed overnight in buffered picric acid/formaldehyde,and then decalcified for 3 weeks in 20% EDTA (pH 8.0) asdescribed by Barnard and colleagues.(10)

Experiment 2: Male rats were Hx at the age of 21–28 days(n � 6 per group; Iffa Credo, Brussels, Belgium). One weeklater osmotic minipumps (Alzet; Broekman Institute, Som-eren, The Netherlands), delivering either 100 �g/day ofrecombinant human GH (hGH; Pharmacia-Upjohn, Stock-holm, Sweden) or 1.0 �g/day of T4 plus 0.125 �g/day ofT3 (T4/T3) (Sigma, St. Louis, USA) or both GH and T4/T3,were implanted subcutaneously (sc) under halothane/O2/N2O anesthesia. Control Hx or normal rats were implantedwith teflon rods. Five percent glucose and 0.9% NaCl wereadded to the drinking water. Treatment was for 2 weeksafter which tibias were isolated as described previously.Body weight was recorded weekly.

Experiment 3: Four-week-old normal Lewis female rats(n � 5 per group) received dexamethasone in the drinkingwater in a dose of 0.3 mg/liter or 2 mg/liter of tap water with0.02% alcohol. Two control groups received tap water with0.02% alcohol. One of these groups was pair fed to matchthe food intake of the group that received the highest doseof dexamethasone. Treatment was for 1 week after whichtibias were isolated as described previously. Body weightand body length were recorded.

Results of growth are shown as mean � SEM. Data weresubjected to ANOVA, followed by the Student–Newman–Keuls test. Statistical significant difference was acceptedbelow p � 0.05.

Immunohistochemistry

Preparation of tibial growth plates and the immunohisto-chemical procedures were modifications of the method usedby Barnard and colleagues(10) to visualize GHR/BP in rabbitgrowth plate. Decalcified tibias were stored frozen. One daybefore immunohistochemical analysis, 10 �m of cryostatsections were mounted on gelatin-KCr(SO4)-glutaraldehyde–coated slides. Endogenous peroxidase activity was blockedin 1% H2O2 in PBS/40% methanol. Sections then weresubjected to pepsin degradation (0.05% pepsin [Sigma] inHCl [pH � 2]) for 7 minutes at 37°C and subsequently to

1409GHR AND GHBP IN THE GROWTH PLATE

1% hydroxyl ammonium chloride for 15 minutes at roomtemperature. Next, sections were incubated in 0.5% Boehr-inger milk powder (in 0.1M TrisHCl/0.15M NaCl/0.05%Tween 20 (Sigma; 60 minutes at 37°C) to block nonspecificbinding, followed by incubation with the following antibod-ies: MAb 263 (1:10; Sanver Tech, Boechout, Belgium),raised to the extracellular part of the rat GHR(11); MAb 4.3,raised to a peptide corresponding to the 17 amino acids tailof GHBP (1:10)(22); or a polyclonal antibody raised to theintracellular part of GHR (PAb iGHR; 1:20). The latter twoantibodies were kindly provided by Dr. W. Baumbach (Cy-anamid, Princeton, NJ, USA). A nonrelevant MAb (mouseantibromodeoxyuridine; Dako A/S, Glostrup, Denmark)was used as a control. Incubations with the primary anti-bodies was for 60 minutes at 37°C. Sections then wereincubated with peroxidase-labeled rabbit anti-mouse anti-body or swine anti-rabbit antibody (1:100 for 60 minutes at37°C; Dako). Next, sections were incubated with biotin-labeled tyramides (a generous gift of Perkin Elmer, Boston,MA, USA; 1:300 in 0.1 M of Tris/0.1 M of NaCl/10%dextran sulfate/10 mM of imidazol/1:1000 H2O2, for 30minutes at room temperature), followed by a peroxidase-labeled antibiotin antibody (1:500 for 45 minutes at 37°C;Dako). The last antibody was visualized with diaminoben-zidine (DAB; Sigma; 0.05% DAB in 50 mM of TrisHCl/1mM of imidazol/0.085% H2O2, for 14 minutes at roomtemperature). Finally, sections were washed in tap waterand dried and mounted in Fluoromount (BDH, Poole, UK).Only in some sections hematoxylin counterstain was usedbecause we found that counterstaining reduced the visibilityof the immunostaining. In control sections, the first antibodywas omitted. Also, in other sections, immunohistochemistrywas performed without TSA to show the level of enhance-ment by tyramides, using only a biotin-labeled second an-tibody followed by a peroxidase-labeled antibiotin anti-body. When comparing staining between different groups,immunohistochemistry of the sections of all animals inthose groups was performed at the same time. Representa-tive sections are shown in the figures.

RT-PCR

Growth plates of 1, 4, 7, and 12-week-old male andfemale rats and livers of 8-week-old female rats were dis-sected carefully. The edges of the growth plate were cut offto avoid contamination with other cell types. Total RNAwas extracted following the method of Chomczynski andSacchi.(24) Two hundred fifty nanograms of total RNA was

transcribed into cDNA with 50 U of Moloney murine leu-kemia virus (MMLV) reverse transcriptase, 7.5 pmol ofdeoxynucleoside triphosphate (dNTP), 50 ng of randomprimer, 0.1 �mol of dithiothreitol (DTT), and 0.5 �l ofRNasin in a total volume of 20 �l of buffer. The reaction wascarried out at 37°C for 1 h and then heated to 70°C for 10minutes. Fifty units of MMLV reverse transcriptase and 0.5 �lof RNasin were added again and the mixture was incubated for30 minutes at 37°C and heated to 70°C for 5 minutes. Finally,180 �l of 10 mM Tris (pH � 8.0) was added.

PCR was performed in a total volume of 50 �l, containingbuffer, 2.5 mM of MgCl2, 0.3 mM of dNTP, 10 pmol offorward and reverse primer, and 0.25 U of Taq polymerase(Eurogentec, Seraing, Belgium), using 30 cycles (HybaidOmnigene System [Biozyme, Landgraaf, The Nether-lands]). Each cycle consisted of 30 s of denaturation at96°C, 30 s annealing at 56°C, and 1 minute extension at72°C, preceded by an initial denaturation at 96°C for 3minutes and followed by a final step at 72°C for 10 minutes.In each PCR experiment, negative controls included tubeswith distilled water instead of cDNA or with 25 ng of RNAinstead of cDNA for each tissue. PCR products were ana-lyzed by gel electrophoresis.

Oligonucleotide primers were purchased from Isogen Bio-science BV (Maarssen, The Netherlands). For the amplifica-tion of GHR cDNA, a primer set was chosen from the intra-cellular sequence of the GHR: the sense and antisense primerswere 5�-GAGGAGGTGAACACCATCTTGGGC-3� and 5�-ACCACCTGCCTGGTGTAATGTC-3�, respectively. Senseand antisense primers for the amplification of GHBP cDNAwere 5�-TGGTGATTTGTTGGACGAAA-3�and 5�-GCTA-GGGATGGCAGATCCTC-3�, respectively.

RESULTS

Enhancement of signal by biotin-labeled tyramides

Figure 1 shows immunohistochemistry with PAb iGHRin the growth plate of a 12-week-old female rat, performedwith and without the use of TSA. Without the use of TSA,the staining was very weak (Fig. 1A). When TSA was used,specific staining was enhanced whereas background stain-ing in the growth plate was still very low (Fig. 1B). Stainingmainly was in early maturing chondrocytes (transition zone/upper hypertrophic zone) but with the use of TSA staining alsocould be visualized, using higher magnification, in germinal

FIG. 1. Immunohistochemistry withPAb iGHR in the growth plate of a12-week-old female rat performed(A) without and (B) with the use ofTSA. The line indicates 100 �m. Im-munohistochemistry was performedwith DAB as chromogen (hematoxy-lin background staining).

1410 GEVERS ET AL.

cells. There was no staining when the nonrelevant MAb wasused (antibromodeoxyuridine).

Comparison of MAb 263, PAb iGHR, and MAb 4.3

Figure 2 compares the staining obtained with the threedifferent antibodies recognizing either the common extra-cellular part of the GHR/GHBP (MAb 263), the internal partof the GHR (PAb iGHR), or the tail of GHBP (MAb 4.3) inthe growth plate of a 12-week-old female rat. Specificity ofthe antibodies and the tyramides was shown by the disap-pearance of staining when the first antibody was omitted(Fig. 2A). As can be seen, background staining in bonetissue (but not in the growth plate) was seen. This probablywas caused by cross-reaction with rat antigen of the rabbitanti-mouse second antibody used to detect the monoclonalprimary antibodies, because this background was not ob-served when a swine anti-rabbit second antibody was used(not shown). The specificity of these antibodies for theirrespective antigens has been shown in other experiments bycompetition with excess antigen.(22,25)

All three antibodies showed staining mainly in the tran-sition zone/upper hypertrophic chondrocyte layer (Figs. 2B,2C, and 2D showing staining with MAb 263, PAb iGHR,and MAb 4.3). Immunohistochemistry with MAb 263 andPAb iGHR also revealed staining in proliferative and ger-minal chondrocytes (Figs. 2E and 2F). Although not theprimary object of this study, we noted that immunoreactiv-ity also was intense in bone and bone marrow proximal anddistal to the growth plate (data not shown).

FIG. 3. RT-PCR showing presence of GHR mRNA and GHBPmRNA in growth plate tissue and liver extracts of male and female ratsat 1, 4, 7 and 12 weeks of age. Lane 1–8: growth plate tissue extractsin the following order: lane 1: 1-week-old female, lane 2: 4-week-oldfemale, lane 3: 7-week-old female, lane 4: 12-week-old female and lane5: 1-week-old male, lane 6: 4-week-old male, lane 7: 7-week-old male,lane 8: 12-week-old male. Lane 9: control using water instead of DNA.Lane 10: liver tissue extract of 8-week-old female. Lane 11–20: con-trols: PCR reactions in same growth plate tissue of lane 1–10 butwithout reverse transcription.

FIG. 2. Immunohistochemical stain-ing in the growth plate of a 12-week-old female rat with three differentantibodies: (A) omission of first anti-body; (B) MAb 263, recognizing spe-cifically the extracellular part ofGHR/GHBP; (C and E) PAb iGHR,recognizing the iGHR; and (D and F)MAb 4.3, recognizing the 17 aminoacids tail specific for GHBP. Stainingis mainly in the zone of early hyper-trophic chondrocytes (E and F). Athigher magnification, staining in cellsin the resting zone (arrows) and someproliferative cells is revealed. Theline indicates 100 �m. Immunohisto-chemistry was performed with DABas chromogen (hematoxylin back-ground staining).

1411GHR AND GHBP IN THE GROWTH PLATE

FIG. 4. (A–D) Immunohisto-chemistry for GHR, using PAbiGHR in normal male rat growthplates at (A) 1week, (B) 4 weeks,(C) 7 weeks, and (D) 12 weeks ofage. Staining in 1-, 4-, and7-week-old rats is in cells of theresting zone, proliferative cells,and early hypertrophic chondro-cytes. In 12-week-old rats, stain-ing mainly is in early hyper-trophic chondrocytes in thetransition zone between the pro-liferative and hypertrophic celllayer. The line indicates 100 �m.(E–H) Immunohistochemistry forGHBP, using a MAb recognizingthe 17 amino acids tail specificfor GHBP (MAb 4.3) in growthplates of the same rats as used inpanels A–D. Staining for GHBPis similar as for GHR. (I–L) Im-munohistochemistry for GHBP,using MAb 4.3, in growth platesof GH-deficient dwarf rats of (I) 1week, (J) 4 weeks, (K) 7 weeks,and (L) 12 weeks of age. GHBPstaining seems less in dwarf ratsthan in normal rats (cf. with pan-els E–H). The line indicates 100�m. Immunohistochemistry wasperformed with DAB as chromo-gen (no background staining).

FIG. 5. Presence of GHR andGHBP immunostaining in cells ofthe resting zone (arrows) in4-week-old male rats. (A) PAbiGHR, recognizing GHR; (B)MAb 4.3, recognizing GHBP.The line indicates 25 �m. Immu-nohistochemistry was performedwith DAB as chromogen (nobackground staining).

1412 GEVERS ET AL.

FIG. 6. GHBP immunostaining using MAb 4.3 in a 7-week-old (A)male and (B) female rat. The growth plate appears smaller and the stainingless in the female rat. The line indicates 100 �m. Immunohistochemistrywas performed with DAB as chromogen (no background staining).

FIG. 7. Nuclear localization of (A) GHR immunostaining (PAbiGHR) and (B) GHBP immunostaining (MAb 4.3) in hypertrophicchondrocytes in a 4-week-old normal rat; (C) GHBP immunostaining(MAb 4.3) in hypertrophic chondrocytes in a 4-week-old GH-deficientdwarf rat. Note the decrease in nuclear localization of GHBP in thedwarf rat. The line indicates (A) 25 �m or (B and C) 100 �m.Immunohistochemistry was performed with DAB as chromogen (nobackground staining).

FIG. 8. Induction of GHBP immunostaining in the growth plate afterT3/T4 and GH treatment for 2 weeks in 4- to 5-week-old Hx rats. (A)Saline-treated Hx rat, (B) T3/T4-treated, (C) GH-treated, (D) GH �T3/T4–treated, and (E) normal intact rat. See text for more details. Theline indicates 100 �m. Immunohistochemistry was performed withDAB as chromogen (no background staining).

1413GHR AND GHBP IN THE GROWTH PLATE

Detection of GHR- and GHBP mRNA in thegrowth plate

RT-PCR confirmed the presence of GHR mRNA andGHBP mRNA in the growth plate (Fig. 3). GHR and GHBPRNA were clearly present in growth plate and liver and noband was observed in the appropriate controls. GHR andGHBP RNA were shown in the growth plate of both maleand female rats at all ages studied.

Expression of GHR and GHBP in growth plates atdifferent ages (experiment 1)

In 1, 4, and 7-week-old rats, staining with all three anti-bodies was seen in stem cells, proliferative chondrocytes,and early hypertrophic chondrocytes with faint staining insome more mature chondrocytes. In 12-week-old rats, stain-ing in stem cells and proliferative chondrocytes was reducedgreatly and staining was seen mostly in the transition zone andthe early hypertrophic chondrocytes. This is illustrated in Fig.4, showing the antibody PAb iGHR, recognizing GHR only(Figs. 4A–4D), and MAb 4.3, recognizing GHBP only (Figs.4E–4H). At all ages, cells in the transition zone showed themost intense staining, although cells in the stem cell zone alsostained positive for both GHR and GHBP, shown at a highermagnification with PAb iGHR and MAb 4.3 in Fig. 5. In 1, 4,and 12-week-old rats there were no differences between gen-der but in 7-week-old rats, staining with MAb 4.3 in stem cellsand proliferative chondrocytes consistently appeared moreabundant in male rats than in female rats (Figs. 6A and 6B). Indwarf rats, localization of GHR/GHBP staining was similar tonormal rats, but intensity seemed less when using MAb 4.3(Figs. 4I–4L). This difference was less clear when using theother two antibodies.

Staining with all three antibodies was present in both thecytoplasm, and in many chondrocytes, the nucleus (Figs. 7Aand 7B). Nuclear staining was most prominent in the earlyhypertrophic chondrocytes. Interestingly, GHBP staining ingrowth plates of dwarf rats was almost exclusively cyto-plasmic (Fig. 7C).

Regulation of GHR and GHBP by GH and thyroidhormones (experiment 2)

To gain insight in the regulation of GHR and GHBP inthe growth plate, Hx rats were treated with GH, T3/T4, orboth, and normal rats were treated with dexamethasone,after which immunohistochemistry was performed on theirgrowth plates. Hx rats showed virtually no growth whentreated with saline for 2 weeks. T3/T4 induced a smallweight gain of 5.2 � 0.2 g/week, and GH was more effec-tive and the weight gain was not increased further byadditional T3/T4 treatment (Table 1).

Figure 8 shows staining for GHBP in the growth plate ofrats from these different groups. Compared with normal rats(Fig. 8E), staining with MAb 4.3 was greatly reduced in Hxrats (Fig. 8A). T3/T4 treatment alone induced a small row ofpositive cells in the transition zone between the proliferativeand hypertrophic zone (Fig. 8B). GH treatment resulted in awider growth plate and a broader band of GHBP-positivecells (Fig. 8C). Staining was indistinguishable in tibias fromGH-treated, GH � T3/T4–treated, and normal rats (Figs.8C–8E).

Similar results were obtained for MAb 263, but differ-ences between groups were less clear when PAb iGHR wasused, suggesting that GHBP and GHR are regulated sepa-rately in the growth plate.

Regulation of GHR and GHBP by dexamethasone(experiment 3)

Table 2 shows the effect of low-dose (0.3 mg/liter drink-ing water) and high-dose (2 mg/liter drinking water) dexa-methasone treatment on gain in weight and length, com-pared with pair-fed and normal-fed rats. Dexamethasone-treated rats gained less weight and length (p � 0.01) thanboth the ad libitum fed group and the pair-fed group (pairedto the high-dose group only). Figure 9 shows that GHBPstaining was clearly reduced in proliferative chondrocytesof the dexamethasone-treated rats and was hardly detectablein the rats treated with the highest dose, compared with bothad libitum and pair-fed rats. Staining in the early hypertro-phic cell layer also was clearly diminished in thedexamethasone-treated rats. Similar results were obtainedwhen PAb iGHR and MAb 263 were used (not shown).

DISCUSSION

There is an increasing amount of evidence that, in addi-tion to the hepatic effects on IGF-I generation, GH affectsthe growth plate directly to stimulate longitudinal growth inthe rat.(3–6,26) However, its exact cellular target in thegrowth plate is not clear. The dual effector theory(9) impli-cates GH action on the resting cells, but GHR is not exclu-sively limited to this zone in the growth plate, at least in therabbit, where GHR has been shown not only in resting cellsbut also in proliferative chondrocytes and in hypertrophicchondrocytes.(10) One problem with most earlier studies isthat the antibody used does not distinguish between GHRand GHBP, the soluble extracellular part of the GHR, which

TABLE 1. WEIGHT GAIN IN NORMAL AND HORMONE-TREATED 4- TO 5-WEEK-OLD RATS

Group Weight gain (g/week)

Hx � saline 1.6 � 0.5Hx � T3/T4 5.3 � 0.2Hx � GH 15.6 � 1.0*,†

Hx � GH � T3/T4 15.1 � 0.8*,†

Normal control 44.1 � 2.7*

* p � 0.01 versus Hx, † p � 0.01 versus Hx � T3/T4.Normal and hypophysectomized (Hx), 4- to 5-week-old male

rats were treated with saline, T3/T4 (0.125 �g; T3 � 1.0 �g of T4),GH (100 �g/day), or combined GH � T3/T4 treatment, adminis-tered via sc implanted osmotic minipumps for 2 weeks.

1414 GEVERS ET AL.

also might be present in chondrocytes. To readdress thesequestions in the rat, we have used a sensitive immunohis-tochemical method to localize GHR and GHBP specificallyin the rat growth plate, using antibodies directed to theinternal part of GHR and to the hydrophilic tail of GHBP,which enabled us to recognize GHR and GHBP separatelyin this species.

To enhance the sensitivity of immunostaining, we usedbiotin-labeled tyramides for amplification.(27,28) Tyramidesprecipitate on interaction with peroxidase and hydrogenperoxide. We used a peroxidase-labeled second antibodyand biotin-labeled tyramides so that a large deposit ofbiotins is formed, which we detected with a peroxidase-labeled antibiotin antibody. Specificity of this TSA wasshown by a complete disappearance of staining when eitherprimary or secondary antibody was omitted or when anonrelevant first antibody was used. The increased sensitiv-ity achieved may account for the fact that we could detectstaining with all three antibodies in rat growth plates at allages, up to 12 weeks, whereas Barnard and colleagues didnot detect GHR/BP in rabbits older than 50 days.(10)

Both GHR and GHBP protein were located in the ratgrowth plate. GHBP is produced mainly in the liver, butmany extrahepatic tissues also synthesize GHBP.(13) It islikely that GHBP is produced also locally in the chondro-cytes, but the immunostaining done could not exclude thepossibility that the GHBP in chondrocytes was obtainedfrom the circulation and then internalized by chondrocytes.However, RT-PCR of growth plate cDNA confirmed thepresence of GHR mRNA and GHBP mRNA transcripts inthe growth plate and thus suggests local production of GHRand GHBP in the growth plate.

The localization of both GHR and GHBP in the same cellis not surprising because both proteins are transcribed fromthe same gene(17) but the functional implications of the localGH-GHR-GHBP interactions in the growth plate are in-triguing. GHBP could be produced locally to capture freeGH to enable it to bind to the receptor or, alternatively,protect cells continuously exposed to GH by preventing it tobind to the receptor. It could also participate in the dimer-ization process and affect signaling.(29)

In young rats, immunoreactive GHR and GHBP werepresent in cells of the resting zone, as has been found beforein rat embryos,(30) young rats,(31) and young rabbits.(10)

Staining was especially clear in the groove of Ranvier,where cells migrate from the resting zone.(32) However, asalso discussed by Lupu et al.,(33) it is difficult to demarcatethe resting zone from epiphyseal chondrocytes involved inthe growth of the secondary ossification center and it isunclear whether cells in the resting zone are stem cells/precursors of proliferative chondrocytes. The localization of

FIG. 9. Reduction of GHBP immunostaining after 1 week of dexa-methasone treatment in two different doses in 4-week-old female ratscompared with both ad libitum–fed and pair-fed rats (pair fed with ratsreceiving the highest dexamethasone dose). (A) Controls, ad libitum–fed (B) low-dose dexamethasone treatment (0.3 mg/liter of drinkingwater), (C) high-dose dexamethasone treatment (2 mg/liter of drinkingwater), and (D) controls, pair-fed to rats receiving 2 mg/liter of dexa-methasone. The line indicates 100 �m. Immunohistochemistry wasperformed with DAB as chromogen (no background staining).

TABLE 2. WEIGHT GAIN AND BODY LENGTH GAIN DURING

DEXAMETHASONE TREATMENT FOR 1 WEEK IN

4-WEEK-OLD FEMALE RATS

GroupWeight gain

(g/week)

Gain inlength

(mm/week)

Low-dose dexamethasone 13.3 � 1.2* 12.6 � 0.8*High-dose dexamethasone 4.3 � 0.9† 5.7 � 1.5†

Control, pair-fed 17.1 � 0.8 13.4 � 0.9Control, ad libitum fed 36.5 � 0.7 20.8 � 0.8

Dexamethasone was mixed in the drinking water in 0.3-mg/L(low dose) or 2-mg/L (high dose) concentrations. Control rats weread libitum fed or pair fed with the rats receiving the highest doseof dexamethasone.

* p � 0.01 versus ad libitum–fed controls; † p � 0.01 versuspair-fed and ad libitum–fed controls.

1415GHR AND GHBP IN THE GROWTH PLATE

CCAAT/enhancer-binding protein (C/EBP), an antimito-genic factor involved in differentiation of other mesenchy-mal cells like preadipocytes,(34) in these cells(31) couldsuggest that these cells are chondrocyte precursors differ-entiating into chondrocytes. Thus, the localization of GHRin these cells supports a direct action of GH on resting cells.However, it does not exclude a role for IGF-I, either pro-duced in the liver or produced locally in these cells, in thedifferentiation of germinal chondrocytes to proliferativechondrocytes.(35,36)

More faint immunostaining was seen in proliferatingchondrocytes of rats up to the age of 7 weeks but not in the12-week-old rats. This is in agreement with the originalstudy of GHR/BP in rabbits, in which a similar localizationwas found in 20- and 50-day-old rabbits but not in olderrabbits,(10) and with a recent report of GHR/BP in uremicrats(37) and with GHR mRNA in situ hybridization data.(33)

In other studies in rat tibial growth plates, immunostainingwith MAb 263 was not found in proliferative chondro-cytes,(31) but this might reflect the lower sensitivity ofstandard immunochemistry. The localization of GHR inproliferative chondrocytes suggests that GH also might actdirectly on these cells in young animals to stimulate mito-genesis. In culture, GH stimulates growth of large chondro-cyte colonies thought to represent rapidly proliferatingcells.(38,39)

However, the most prominently stained layer was seen ina band of early maturing chondrocytes at the transitionbetween proliferative and hypertrophic zones. Staining inthe most distal layer of fully mature hypertrophic chondro-cytes was absent or faint, similar to findings in rat fetalgrowth plates(30) and localization of GHR mRNA.(33) Thisis the first time, maybe because of the ultrasensitive immu-nostaining technique, that the localization in this transitionzone is so apparent. In older rats, GHR and GHBP local-ization in this zone remained visible whereas GHR andGHBP staining disappeared from proliferative chondro-cytes. The difference in staining between the early maturingand the fully mature chondrocyte might indicate that GHreceptor signaling is involved in the maturation process ofchondrocytes.

Interestingly, we previously showed that GH treatment indwarf rats results in a decrease in alkaline phosphataseactivity in those chondrocytes(40); thus, this could be a directeffect of GH. In addition, in the fetal rat, these chondrocytesexpress GHR/BP mRNA but not IGF-I mRNA,(41) alsosuggesting that GH can act directly on these transitionalchondrocytes without IGF-I as mediator, at least in thefetus.

Although not the primary aim of this study, we noted thatall three antibodies stained GHR or GHBP in nuclei as wellas cytoplasm, as has been shown before for GHR(42) andGHBP(43) and also for GH.(44) The significance of GHR andGHBP in the nucleus is not clear. They might provide anadditional signal transduction pathway or act as transcrip-tion factors,(42–44) although this is still controversial. Trans-location of the GHR to the nucleus requires the intracellularpart of the GHR but also GH binding.(42) We observed areduction of nuclear staining in GH-deficient dwarf ratsmost clearly with MAb 4.3, which may suggest that GHBP

requires GH binding for translocation to the nucleus. Nu-clear staining was not abolished completely but this mightbe because of the residual GH in these dwarf rats,(23) be-cause in Hx rats staining completely disappeared.

Interestingly, nuclear staining was seen mainly in chon-drocytes in the transition zone between proliferation andhypertrophy and rarely in stem cells or proliferating chon-drocytes. This raises the possibility that proliferative chon-drocytes use a different intracellular signal transductionpathway than hypertrophic chondrocytes (e.g., plasmamembrane–associated protein phosphorylation vs. nucleartranslocation).

Having established the localization of GHR and GHBP inthe growth plate, we then wished to study its regulation. Inthe rat, hepatic GHR and plasma GHBP are developmen-tally regulated with a gender difference occurring at the ageof puberty.(45–47) However, plasma GHBP mainly reflectshepatic GHBP production and little is known about GHR/GHBP regulation in other target tissues. In the growth plate,GHR and GHBP showed clear developmental regulation.Throughout sexual maturation, staining was very strong andwas reduced as animals became older (and grew moreslowly), by which time staining was only detected in theearly maturing chondrocytes. A mild sex difference wasdetected only at the end of puberty (7 weeks of age) withstaining less intense in female animals than in male animals,but this suppression is difficult to quantitate. At 7 weeks ofage, growth velocity is less in female animals than in maleanimals; therefore, although we cannot quantify it fromthese data, it is tempting to speculate that there may be arelation between growth velocity and GHR/BP expressionin the growth plate in this period of sexually dimorphicgrowth velocity.

A logical candidate for the in vivo regulation of growthplate GHR expression is GH itself, because in vitro studieshave shown regulation of chondrocyte GHR by GH(8) andGH is a main regulator of hepatic GHR and plasma GHBPin vivo.(48) Like plasma GHBP,(47) GHBP staining appearedless intense in dwarf rats compared with normal rats and thisdifference in staining disappeared in the older animals whennormals and dwarves were growing at a more similar rate.However, the polyclonal antibody specific for GHR showedless difference between normal and dwarf growth plates.The relative expression of GHR and GHBP in the growthplate may vary, but this will require confirmation with morequantitative techniques.

In the multiple hormone-deficiency state of hypophysec-tomy, in which growth has completely ceased, GHR andGHBP were virtually undetectable in the growth plate. GHRand GHBP expression in chondrocytes was induced againby T3/T4 and GH treatment, especially in the transitionzone. Lewinson et al.(49) concluded from experiments inhypothyroid rats that condylar chondrocytes were compro-mised in their response to GH although they could not showdifferences in GHR staining using MAb 263.

The early maturing chondrocytes on the transition fromthe proliferative zone to the hypertrophic zone may be apivotal zone in chondrocyte differentiation. For example,Indian hedgehog (Ihh), parathyroid hormone–related pro-tein (PTHrP), PTH/PTHrP receptors, and basic FGF

1416 GEVERS ET AL.

(bFGF), which are involved in chondrocyte maturation, arelocalized to the same chondrocytes.(28,50–54) We now showhere that it is again these cells in which GHR/GHBP ex-pression is restored after GH/T3/T4 replacement therapy.

We found that dexamethasone acts as a negative regulatorof GHR/GHBP in the growth plate. Staining for GHR andGHBP had almost disappeared not only from resting zonecells and proliferative chondrocytes but also from earlymaturing chondrocytes. This agrees with earlier results ondexamethasone suppression of GHR expression(55,56) butseems in contrast to the results of Heinrichs et al., whoshowed that GHR mRNA in the rabbit growth plate wasup-regulated by dexamethasone.(57) Of course, posttran-scriptional regulation and species difference could be in-volved in this discrepancy. In the rabbit, GHR and GHBPare derived from the same mRNA, whereas in rats GHR andGHBP are derived from two separate mRNAs so they couldbe regulated independently. In line with our results, it hasbeen shown in vitro that dexamethasone decreases GHRmRNA in rat chondrocytes.(58) Thus, apart from acting viacorticosteroid receptors in the growth plate,(59) the growthretardation seen after corticosteroid treatment could be in-direct due to a down-regulation of GHR and GHBP, result-ing in a relative insensitivity of chondrocytes to directeffects of GH.

The localization of GHR in the growth plate suggests adirect effect of GH not only on germinal cells, but also onproliferation of chondrocytes and maturation into hypertro-phic chondrocytes. It is still unclear whether the effect ofGH can be replaced fully by IGF-I; mice transgenic for GHhave increased skeletal growth,(60) and mice transgenic forIGF-I do not.(61) On the other hand, GH-deficient micetransgenic for IGF-I have normal growth,(62) although micewith a deletion of hepatic IGF-I also grow normally.(5)

Double knockouts for GHR and IGF-I are dwarfed moreseverely than the single knockout mice, and growth plates ofIGF-I knockout mice, with high circulating GH levels, havea smaller resting zone, a higher proliferation rate, and agreater height of individual hypertrophic chondrocytes thanthe double GHR/IGF-I knockout mice.(33) This is in linewith the localization of GHR that we found.

Both GH and IGF-I are capable of shortening cell cycletime of resting cells and duration of the hypertrophic phaseof chondrocytes in Hx rats.(35) However, IGF-I treatment ofGHR knockout mice does not restore completely femorallength(63) and IGF-I treatment of GHR-deficient childrendoes not have sustaining effects on growth.(64)

The localization of GHR in the growth plate suggests thatthe zone of chondrocytes in transition from a proliferative toa hypertrophic phenotype has to be considered as a majorsite of direct action for GH. It remains to be studied whetherthe effect of GH on these cells can be replaced by IGF-I.

ACKNOWLEDGMENTS

We thank Dr. W. Baumbach (American Cyanamid Co.)for the antibodies against GHBP and the internal part ofGHR. We also thank Dr. R. Dirks and F. van der Rijke(Laboratory of Cytochemistry and Cytology, Leiden Uni-

versity Medical Center) for their help in setting up theimmunohistochemistry. We are grateful to Pharmacia-Upjohn for their continuing supply of recombinant hGH,and we thank Dr. C.W.G.M. Lowik (Department of Endo-crinology, Leiden University Medical Center) for helpfulcomments.

REFERENCES

1. Kember NF 1960 Cell divisions in endochondral ossification.A study of cell proliferation in rat bones by the method oftritiated thymidine autoradiography. J Bone Joint Surg Br42:824–839.

2. Ohlsson C, Isgaard J, Tornell J, Nilsson A, Isaksson OGP,Lindahl A 1993 Review paper: Endocrine regulation of longi-tudinal bone growth. Acta Paed Suppl 391:33–40.

3. Isaksson OGP 1982 Growth hormone stimulates longitudinalbone growth directly. Science 216:1237–1239.

4. Spencer G, Hodgkinson SC, Bass JJ 1991 Passive immuniza-tion against insulin-like growth factor-I does not inhibitgrowth hormone-stimulated growth of dwarf rats. Endocrinol-ogy 128:2103–2109.

5. Sjogren K, Liu JL, Blad K, Skrtic S, Vidal O, Wallenius V,LeRoith D, Tornell J, Isaksson OG, Jansson JO, Ohlsson C1999 Liver-derived insulin-like growth factor I (IGF-I) is theprincipal source of IGF-I in blood but is not required forpostnatal body growth in mice. Proc Natl Acad Sci USA96:7088–7092.

6. Yakar S, Liu JL, Stannard B, Butler A, Accili D, Sauer B,LeRoith D 1999 Normal growth and development in the ab-sence of hepatic insulin-like growth factor I. Proc Natl AcadSci USA 96:7324–7329.

7. Ohlsson C, Nilsson A, Isaksson O, Lindahl A 1992 Growthhormone induces multiplication of the slowly cycling germinalcells of the rat tibial growth plate. Proc Natl Acad Sci USA89:9826–9830.

8. Nilsson A, Isgaard J, Lindahl A, Dahlstrom A, Skottner A,Isaksson OGP 1986 Regulation by growth hormone of numberof chondrocytes containing IGF-I in rat growth plate. Science233:571–574.

9. Isaksson OGP, Lindahl A, Nilsson A, Isgaard J 1987 Reviewpaper: Mechanism of the stimulatory effect of growth hormoneon longitudinal bone growth. Endocr Rev 8:426–438.

10. Barnard R, Haynes KM, Werther GA, Waters MJ 1988 Theontogeny of growth hormone-receptors in the rabbit tibia.Endocrinology 122:2562–2569.

11. Barnard R, Bundesen PG, Rylatt DB, Waters MJ 1984 Mono-clonal antibodies to the rabbit liver growth hormone receptor;production and characterization. Endocrinology 115:1805–1813.

12. Leung DW, Spencer SA, Cachianes G, Hammonds RG, Col-lins C, Henzel WJ, Barnard R, Waters MJ, Wood WI 1987Growth hormone receptor and serum binding protein: Purifi-cation, cloning and expression. Nature 330:537–543.

13. Carlsson B, Billig H, Rymo L, Isaksson O 1990 Expression ofthe growth hormone-binding protein messenger-RNA in theliver and extrahepatic tissues in the rat—coexpression with thegrowth-hormone receptor. Mol Cell Endocrinol 73:R1–R6.

14. Fairhall KM, Carmignac DF, Robinson ICAF 1992 Growthhormone (GH) binding protein and GH interaction in vivo inthe guinea pig. Endocrinology 131:1963–1969.

15. Lim L, Spencer SA, McKay P, Waters MJ 1990 Regulation ofgrowth hormone (GH) bioactivity by a recombinant humanGH-binding protein. Endocrinology 127:1287–1291.

16. Dattani MT, Hindmarsh PC, Brook CG, Robinson IC, Mar-shall NJ 1994 Inhibition of growth hormone bioactivity byrecombinant human growth hormone-binding protein in theeluted stain assay system. J Endocrinol 140:445–453.

1417GHR AND GHBP IN THE GROWTH PLATE

17. Baumbach WR, Horner DL, Logan JS 1989 The growthhormone-binding protein in rat serum is an alternativelyspliced form of the rat growth hormone receptor. Genes Dev3:1199–1205.

18. Tiong TS, Herington AC 1991 Tissue distribution, character-ization, and regulation of messenger-ribonucleic-acid forgrowth hormone receptor and serum binding-protein in the rat.Endocrinology 129:1628–1634.

19. Walker JL, Moatsstaats BM, Stiles AD, Underwood LE 1992Tissue-specific developmental regulation of the messengerribonucleic-acids encoding the growth hormone receptor andthe growth-hormone binding-protein in rat fetal and postnataltissues. Pediaetr Res 31:335–339.

20. Baumbach WR, Bingham B 1995 One class of growth hor-mone (GH) receptor and binding protein messenger ribonu-cleic acid in rat liver, GHR1, is sexually dimorphic and regu-lated by GH. Endocrinology 136:749–760.

21. Domene HM, Cassorla F, Werner H, Roberts CJ, Leroith D1995 Rat growth hormone receptor/growth hormone-bindingprotein mRNAs with divergent 5�-untranslated regions areexpressed in a tissue-specific manner. DNA Cell Biol 14:195–204.

22. Sadeghi H, Wang BS, Lumanglas AL, Logan JS, BaumbachWR 1990 Identification of the origin of the growth hormone-binding protein in rat serum. Mol Endocrinol 4:1799–1805.

23. Charlton HM, Clark RG, Robinson ICAF, Porter-Goff AE,Cox BS, Bugnon C, Bloch BS 1988 Growth hormone-deficientdwarfism in the rat: A new mutation. J Endocrinol 119:51–58.

24. Chomczynski P, Sacchi N 1987 Single step method of RNAisolation by acid guanidium thiocyanate-phenol-chloroformextraction. Anal Biochem 162:156–159.

25. Barnard R, Thordarson G, Lopez MF, Yamaguchi M, Edens A,Cramer SD, Ogren L, Talamantes F 1994 Expression ofgrowth hormone-binding protein with a hydrophilic carboxylterminus by the mouse placenta: Studies in vivo and in vitro.J Endocrinol 140:125–135.

26. Madsen K, Friberg U, Roos P, Eden S, Isaksson OGP 1983Growth hormone stimulates the proliferation of cultured chon-drocytes from rabbit ear and rat rib growth cartilage. Nature304:545–547.

27. Raap AK, van de Corput MP, Vervenne RA, van Gijlswijk RP,Tanke HJ, Wiegant J 1995 Ultra-sensitive FISH usingperoxidase-mediated deposition of biotin- or fluorochrometyramides. Hum Mol Genet 4:529–534.

28. van der Eerden BC, Karperien M, Gevers EF, Lowik CW, WitJM 2000 Expression of Indian hedgehog, parathyroidhormone-related protein, and their receptors in the postnatalgrowth plate of the rat: Evidence for a locally acting growthrestraining feedback loop after birth. J Bone Miner Res 15:1045–1055.

29. Ross RJ 1999 Truncated growth hormone receptor isoforms.Acta Paediatr Suppl 88:164–166.

30. Garcia Aragon J, Lobie PE, Muscat G, Gobius KS, Norstedt G,Waters MJ 1992 Prenatal expression of the growth hormone(GH) receptor/binding protein in the rat: A role for GH inembryonic and fetal development? Development 114:869–876.

31. Vidal NO, Ekberg S, Enerback S, Lindahl A, Ohlsson C 1997The CCAAT/enhancer-binding protein-alpha is expressed inthe germinal layer of the growth plate: Colocalisation with thegrowth hormone receptor. J Endocrinol 155:433–441.

32. Langenskiold A, Videman T, Nevalainen T 1993 Vital stainingindicating cell migration towards the periphery in the growthplate. Studies of fibular heads in rabbits. Acta Orthop Scand64:683–687.

33. Lupu F, Terwilliger JD, Lee K, Segre GV, Efstratiadis A 2001Roles of growth hormone and insulin-like growth factor 1 inmouse postnatal growth. Dev Biol 229:141–162.

34. Lin FT, Lane MD 1994 CCAAT/enhancer binding proteinalpha is sufficient to initiate the 3T3–L1 adipocyte differenti-ation program. Proc Natl Acad Sci USA 91:8757–8761.

35. Hunziker EB, Wagner J, Zapf J 1994 Differential effects ofinsulin-like growth factor I and growth hormone on develop-mental stages of rat growth plate chondrocytes in vivo. J ClinInvest 93:1078–1086.

36. Reinecke M, Schmid AC, Heyberger-Meyer B, Hunziker EB,Zapf J 2000 Effect of growth hormone and insulin-like growthfactor I (IGF-I) on the expression of IGF-I messenger ribonu-cleic acid and peptide in rat tibial growth plate and articularchondrocytes in vivo. Endocrinology 141:2847–2853.

37. Edmondson SR, Baker NL, Oh J, Kovacs G, Werther GA,Mehls O 2000 Growth hormone receptor abundance in tibialgrowth plates of uremic rats: GH/IGF-I treatment. Kidney Int58:62–70.

38. Lindahl A, Nilsson A, Isaksson OGP 1987 Effects of growthhormone and insulin-like growth factor-I on colony formationof rabbit epiphyseal chondrocytes at different stages of matu-ration. J Endocrinol 115:263–271.

39. Lindahl A, Isgaard J, Carlsson L, Isaksson OGP 1987 Differ-ential effects of growth hormone and insulin-like growth factorI on colony formation of epiphyseal chondrocytes in suspen-sion culture in rats of different ages. Endocrinology 121:1061–1069.

40. Gevers EF, Milne J, Robinson ICAF, Loveridge N 1996 Singlecell enzyme activity and proliferation in the growth plate:Effects of growth hormone. J Bone Miner Res 11:1103–1111.

41. Edmondson SR, Werther GA, Russell A, LeRoith D, RobertsCT Jr, Beck F 1995 Localization of growth hormone receptor/binding protein messenger ribonucleic acid (mRNA) during ratfetal development: Relationship to insulin-like growth factor-ImRNA. Endocrinology 136:4602–4609.

42. Lobie PE, Wood TJ, Chen CM, Waters MJ, Norstedt G 1994Nuclear translocation and anchorage of the growth hormonereceptor. J Biol Chem 269:31735–31746.

43. Lobie PE, Garcia-Aragon J, Wang BS, Baumbach WR, WatersJM 1992 Cellular localization of the GH-binding protein in therat. Endocrinology 130:3057–3065.

44. Lobie PE, Mertani H, Morel G, Morales-Bustos O, Norstedt G,Waters MJ 1994 Receptor-mediated nuclear translocation ofgrowth hormone. J Biol Chem 269:21330–21339.

45. Maes M, de Hertogh R, Watrin-Granger P, Ketelslegers JM1983 Ontogeny of liver somatotropic and lactogenic bindingsites in male and female rats. Endocrinology 113:1325–1332.

46. Massa G, Mulumba N, Ketelslegers J-M, Maes M 1990 Initialcharacterization and sexual dimorphism of serum growthhormone-releasing binding protein in adult rats. Endocrinol-ogy 126:1976–1980.

47. Carmignac DF, Wells T, Carlsson LMS, Clark RG, RobinsonICAF 1992 Growth hormone (GH)-binding protein in normaland GH-deficient dwarf rats. J Endocrinol 135:447–457.

48. Robinson I, Carmignac DF, Fairhall KM 1993 Growth hor-mone (GH) receptors, GH binding protein and GH: An auto-regulatory system? Int J Paed Suppl 82:22–28.

49. Lewinson D, Bialik G, Hochberg Z 1994 Differential effects ofhypothyroidism on the cartilage and the osteogenic process inthe mandibular condyle: Recovery by growth hormone andthyroxine. Endocrinology 135:1504–1510.

50. Jingushi S, Scully SP, Joyce ME, Sugioka Y, Bolander ME1995 Transforming growth factor-beta1 and fibroblast growthfactors in rat growth plate. J Orthop Res 13:761–768.

51. Vortkamp A, Lee K, Lanske B, Segre V, Kronenberg HM,Tabin C 1996 Regulation of rate of cartilage differentiation byIndian Hedgehog and PTH-related protein. Science 273:613–621.

52. Lanske B, Karaplis AC, Lee K, Luz A, Vortkamp A, Pirro A,Karperien M, Defize LHK, Ho C, Mulligan RC, Abou-SamraA-B, Juppner H, Segre GV, Kronenberg HM 1996 PTH/

1418 GEVERS ET AL.

PTHrP receptor in early development and indian hedgehog-regulated bone growth. Science 273:663–666.

53. Kato Y, Iwamoto M 1990 Fibroblast growth factor is aninhibitor of chondrocyte terminal differentiation. J Biol Chem265:5903–5909.

54. Wroblewski J, Edwall AC 1995 Inhibitory effects of basicfibroblast growth factor on chondrocyte differentiation. J BoneMiner Res 10:735–742.

55. Gabrielsson BG, Carmignac DF, Flavell DM, Robinson ICAF1995 Steroid regulation of growth hormone (GH) receptor andGH binding protein messenger ribonucleic acids in the rat.Endocrinology 136:209–217.

56. King AP, Carter SC 1995 Dexamethasone-induced antagonismof growth hormone (GH) action by down-regulation of GHbinding in 3T3–F442A fibroblasts. Endocrinology 136:4796–803.

57. Heinrichs C, Yanovski JA, Roth AH, Yu YM, Domene HM,Yano K, Cutler GB Jr, Baron J 1994 Dexamethasone increasesgrowth hormone receptor messenger ribonucleic acid levels inliver and growth plate. Endocrinology 135:1113–1118.

58. Jux C, Leiber K, Hugel U, Blum W, Ohlsson C, Klaus G,Mehls O 1998 Dexamethasone impairs growth hormone (GH)-stimulated growth by suppression of local insulin-like growthfactor (IGF)-I production and expression of GH- and IGF-I-receptor in cultured rat chondrocytes. Endocrinology 139:3296–3305.

59. Abu EO, Horner A, Kusec V, Triffitt JT, Compston JE 2000The localization of the functional glucocorticoid receptor al-pha in human bone. J Clin Endocrinol Metab 85:883–889.

60. Kopchick JJ, Bellush LL, Coschigano KT 1999 Transgenicmodels of growth hormone action. Annu Rev Nutr 19:437–461.

61. Mathews LS, Hammer RE, Behringer RR, Dercole AJ, BellGI, Brinster RL, Palmiter RD 1988 Growth enhancement oftransgenic mice expressing human insulin-like growth factor I.Endocrinology 123:2827–2833.

62. Behringer RR, Lewin TM, Quaife CJ, Palmiter RD, BrinsterRL, D’Ercole AJ 1990 Expression of insulin-like growth factorI stimulates normal somatic growth in growth hormone-deficient transgenic mice. Endocrinology 127:1033–1040.

63. Sims NA, Clement-Lacroix P, Da Ponte F, Bouali Y, Binart N,Moriggl R, Goffin V, Coschigano K, Gaillard-Kelly M, Ko-pchick J, Baron R, Kelly PA 2000 Bone homeostasis in growthhormone receptor-null mice is restored by IGF-I but indepen-dent of Stat5. J Clin Invest 106:1095–1103.

64. Backeljauw PF, Underwood LE 2001 Therapy for 6.5–7.5years with recombinant insulin-like growth factor I in childrenwith growth hormone insensitivity syndrome: A clinical re-search center study. J Clin Endocrinol Metab 86:1504–1510.

Address reprint requests to:Evelien F. Gevers, M.D., Ph.D.

National Institute for Medical ResearchDivision of Molecular Neuroendocrinology

The Ridgeway, Mill HillLondon NW7 1AA, UK

Received in original form October 8, 2001; in revised form March6, 2002; accepted April 2, 2002.

1419GHR AND GHBP IN THE GROWTH PLATE