JOURNAL OF PATHOLOGY, VOL.
167: 291-296 (1992)
PARATHYROID HORMONE-RELATED PEPTIDE: EXPRESSION IN FETAL AND NEONATAL DEVELOPMENT P. B. J. BURTON*, C. MONIZt, P. QUIRKEZ, A. MALIK5, T. D. B u r t , H. JUPPNERII, G. V. SEGREll AND D. E KNIGHT*
*Division of Biomedical Sciences, Kings College, London W8 7AH, U.K.; tDepartment of Clinical Biochemistry, Kings College Hospital, Denmark Hill, London SES 9RS, U.K.; $Department of Pathology, University of Leeds, Leeds LS2 9JT, U.K.; $Department of Medicine, Kings College Hospital, Denmark Hill, London SE5 9RS, U.K.; 11 Endocrine Unit, Massachusetts General Hospital, Boston, M A 02114, U.S.A. Received 10 September 1991 Accepted 17 February 1992
SUMMARY Hypercalcaemia frequently complicates the clinical management of cancer. Many factors have been implicated in the pathogenesis of this humoral hypercalcaemia of malignancy, the most recent candidate being parathyroid hormonerelated peptide (PTHrP). Until now, this peptide has been detected only in some normal and transformed adult tissues. In recent years, it has become apparent that tumours are capable of expressing and secreting factors previously elaborated only during fetal life. Many of these factors act to stimulate the growth of both tumour and fetal cells in an autocrine manner. The data presented here demonstrate that PTHrP is expressed in the human and rat fetus throughout gestation. Immunocytochemistry reveals a gestationally related, changing pattern of expression which is paralleled by changes in mRNA transcription. These data support the hypothesis that PTHrP may function as a fetal growth factor. KEY woms-PTH-related
peptide, fetal development, immunocytochemistry, growth factors.
INTRODUCTION
differ at only two of their N-terminal 1 1 1 amino acid^.^,^ This high degree of conservation of It is now 50 years since Albright suggested a sequence between species (particularly within reparathyroid hormone-like factor to be the cause of gions known not to be required for binding to a tumour-related hypercalcaemia.' This is a common common PTH/PTHrP receptor6) suggests that oncological occurrence and affects an estimated portions of the PTHrP molecule are involved in 10-20 per cent of all cancer patients, the clinical metabolic processes that are independent of manifestations of this syndrome being termed the molecule's ability to induce systemic humoral hypercalcaemia of malignancy or HHM.2 hypercalcaemia. Some examples of the apparently diverse roles Recently, a peptide with limited N-terminal amino acid homology to PTH has been isolated from a performed by PTHrP may include mediation of variety of tumours associated with HHM.3 This calcium transfer into milk across mammary epifactor has been sequenced, characterized, and t h e l i ~ m ,a~ putative neurotransmitter,' and the termed PTH-related peptide or PTHrP. Although regulation of the fetal hypercalcaemic state.*Rodda there is a high degree of homology between N- et al. have suggested that PTHrP derived from the terminal portions of PTH and PTHrP, these two fetal parathyroid glands may be responsible for peptides are structurally and immunologically maintaining the positive fetal calcium balance.' We have reported the immunocytochemical distinct. The sequences of human and rat PTHrP detection of PTHrP and its messenger RNA Addressee for correspondence: Dr D. E. Knight, Division of throughout the urogenital and respiratory tracts of Biomedical Sciences, Kings College London, London W8 7AH, first- and early second-trimester human fetuses."." U.K. To determine the distribution of PTHrP throughout 0022-34 17/92/070291 4 6 $08.00 0 1992 by John Wiley & Sons, Ltd
292
P. B. J. BURTON E T A L .
the entire gestational period, we have undertaken studies of the fetal rat. This paper therefore describes experiments in which immunocytochemical analysis for PTHrP was performed on sagittal sections of fetal rats throughout the gestation of 22 days. Changes in immunocytochemical distribution of PTHrP in the fetal rat were compared with those in the human fetus, and changes in transcription of the PTHrP gene in human fetal kidney were examined using Northern blot analysis. MATERIALS AND METHODS
used in this study are also used in radioimmunoassay and Western blot analysis of PTHrP. In neither case have they been found to react with PTH . Northern blot analysis Total RNA was preparedl3-I5 from kidneys of human fetuses aged 10, 15, and 24 weeks of gestation (ascertained by crown-rump length) that were obtained after therapeutic or spontaneous abortions, and Northern blot analysis was performed as described previously.’’
Animals Female Wistar rats were time-mated. Fetuses obtained on the 13th, 16th, 18th, and 21st days of gestation and 2-day-old neonatal animals were killed by inhalation of carbon dioxide. Fetuses were placed in phosphate-buffered formalin (4 per cent) prior to embedding in paraffin wax. The age of human fetuses was assessed using the crown-rump length. Immunocytochemistry Immunocytochemistry was performed at the same time on sections obtained from a number of different fetuses at each gestational age using antisera from the same batch. It was performed as described previously.’0”’ Sections were not trypsinized but non-specific background staining was reduced by flooding the sections with undiluted fetal calf serum for 20 min. In this study, sections were incubated for 6 h at room temperature with antisera raised against PTHrP (1-34) and (56-86) at a dilution of 1:400 and 1 :200, respectively. Results were identical with both antisera. Following extensive washing, the sections were incubated for 1 h at room temperature with peroxidase-labelled goat anti-rabbit immunoglobulins (Dako, U.K.) diluted 1 :loo. Peroxidase activity was detected using diaminobenzidine/hydrogen peroxide.I2 Specificity of immunostaining was assesed as follows. (a) Pre-immune rabbit serum was substituted for PTHrP antiserum. (b) 1 ml of PTHrP (1-34) antiserum diluted 1:400 was incubated for 8 h at 4°C with 15pg of human PTHrP (1-34) (Peninsula, U.K.) or (c) with 2Opg of human PTH (1-34) (National Institute for Biological Standards and Controls, U.K.). (d) PTH antiserum diluted 1:lOO was substituted for PTHrP antiserum. The antisera
RESULTS Widespread staining of a variety of tissues was seen at all ages of gestation and extending into early neonatal life. Nervous system The brain, spinal cord, dorsal root ganglion, peripheral nerves, and developing eye were all strongly immunoreactive for PTHrP from 13 days to 2 days. Dorsal root ganglia cell bodies showed diffuse cytoplasmic staining, whereas peripheral and central nervous system tissue showed cell process staining.
+
Cardiorespiratory system The cardiac muscle stained intensely throughout gestation. However, whilst the myocardium was positive, cushion tissue of developing valves was negative (data not shown). Developing airways revealed marked immunoreactivity of the epithelium which became polarized towards the luminal epithelial membrane with increasing maturity. In the early developing fetal lung (13 days), the mesenchyme stained positively but this was lost by day 18, with the rest of the lung parenchyma becoming negative (Figs 1A and 1B). Gastrointestinal system Immunoreactivity was seen in the villus tips of the small intestine (Fig. 2) and patchy staining was present in the cdonic epithelium. In both the lung and small intestine, staining, confined exclusively to the nucleus, could be detected. Oesophageal and gastric mucosa was not seen, although the squamous epithelium of the tongue was positive. The liver showed only patchy staining, whereas the
PTHrP IN FETAL DEVELOPMENT
293
Fig. I-PTHrP immunoreactivity is present in fetal rat lung. (A) Section of lung from a 13-day fetal rat. PTHrP immunoreactivity is present in epithelial cells and surrounding mesenchyme ofimmature bronchi. (B) Section of lung from a 21-day fetal rat. Mesenchyme remains negative for PTHrP; however, a number of bronchial epithelial cells are intensely positive
pancreatic acini were intensely positive. Ductal epithelium of the salivary gland was positive, but the acini were negative (Fig. 3). Skin The skin showed marked immunoreactivity from 13 days of gestation to f 2 days. However, by $ 2 days the staining became confined to the superficial epithelium, with absence of staining in the basal layer. Early developing hair follicles initially stained
strongly but subsequently immunoreactivity was lost or markedly reduced. The mesenchyme around the developing hair follicles was also positive for PTHrP. Musculoskeletal system
The striated muscle was intensely positive at all gestational ages. The skeletal system, however, changed with age. At 13 days, the mesenchyme of
294
P. B. J. BURTON E T A L .
Fig. 2-PTHrP is seen in the villous tips of the small intestine from a 21-day-old rat fetus
Fig. 4-Immature chondrocytes at the periphery of the developing vertebral body show immunoreactivity from the first wave of expression which decreases towards the centre of the vertebral body. Inset: mature chondrocytes regain immunoreactivity for PTHrP and osteoblasts stain intensely
cartilage being strongly positive but the intensity reducing with maturity until ossification. Genitourinary system
Changes were observed in the pattern of immunocytochemical distribution in sections of human fetal kidney. At 10 weeks, PTHrP immunoreactivity was present in the glomerulus and in the epithelia of all renal tubules; however, by 25 weeks, staining was markedly reduced in the glomerulus, as demonstrated previously." Similarly, immunoreactivity was present in tubular epithelium of kidney from 2-day-old neonatal rat kidney but was absent from the glomerulus by that stage. Fig. 3-PTHrP immunoreactivity is seen in theductal epithelium of the salivary gland from a 2-day-old neonatal rat, but is absent from acini
the sclerotome and the immature cartilage of the early vertebral column were strongly positive for PTHrP. The staining of the cartilage of the spinal cord decreased in intensity until the onset of ossification, when chondrocytes and osteoblast-like cells became positive (Fig. 4 and inset). The developing limbs and tail followed the same sequential pattern, with early mesenchyme and immature
Northern blot analysis
The changing pattern of expression of PTHrP observed immunocytochemically in fetal rat kidney was paralleled by both the changing pattern of expression of the peptide and the PTHrP mRNA isoforms in lo-, 15-, and 24-week-old human fetal kidneys. Kidneys from 10- and 15-week-old fetuses displayed a complex pattern of transcripts ranging in size from 2.1 to 4.5 kb. However, by 24 weeks the 4.5 kb transcript had decreased in intensity while a species of approximately 1.8 kb appeared (data not shown).
295
P T H r P I N FETAL DEVELOPMENT
Sections treated with PTHrP antisera incubated with PTHrP failed to stain, as did those sections treated with PTH antisera. However, incubating PTHrP antisera with PTH altered neither the intensity nor the pattern of localization. DISCUSSION PTHrP has been shown to possess PTH-like bioactivity in a number of physiological systems including bone resorption assays and renal calcium handling e~periments.’~’’ Although PTHrP does possess potent calciotropic actions, the data presented here also support the hypothesis that PTHrP may be closely associated with the regulation of fetal cellular growth and differentiation. In order for specialized epithelia to form bronchial airways or renal tubules supported in a matrix of mesenchyme, blood vessels and connective tissue must proliferate, migrate, and differentiate. How this is directed and regulated is a central, yet basically unanswered, question in biology. While not the only agents involved, growth factors provide one of the big est clues as to the mechanisms of organogenesis.50.21 Although in vitro studies are essential to assess the biological activity of such molecules, analysis of their cellular localization can provide many clues as to their likely functions. The data presented here offer PTHrP as a possible candidate in the control of fetal growth and development. The temporal and spatial pattern of expression of PTHrP in the human and rat fetus is similar to that observed for a wide range of growth factors including TGF-b22and the oncogene product receptor c ~ b B - 2 . ~ These ’ factors share localization in common cell types and a similar pattern of expression throughout gestation. In both the lung and the developing skeletal system, a changing pattern of expression was seen with immature mesenchyme, which initially stained strongly and subsequently lost immunoreactivity. With regard to the nuclear staining observed in the lung and gut, nuclear targeting has been observed for platelet-derived growth factor,24 fibroblast growth factor,2sand is suggested to occur for TGF/14.26 Although unproved, it is possible that similar targeting occurs for PTHrP, which would explain the high levels of nuclear staining. Developing hair follicles showed a change in mesenchymal staining during their development with loss of mesenchymal staining with maturity. It has been realized for over 20 years that embryonal mesenchyme directs epithelial differentiation and m a t u r a t i ~ n .The ~ ~ exact ’~~ manner by which this occurs is uncertain.
PTHrP may29or may not3’ possess transforming growth factor beta-like activity in vitro; however, evidence suggests that PTHrP does have growth factor-like bioactivity. For example, addition of epidermal growth factor to both osteosarcoma cells (SaOS2 cells)3’ and human k e r a t i n o c y t e ~stimu~~ lates PTHrP secretion, suggesting that PTHrP might form part of a multifactorial growth factor loop. Furthermore, the peptide has been demonstrated to be mitogenic to bone cells in ~ i t r oIt. has ~~ also been shown that immunoneutralization of PTHrP with antisera almost totally inhibits the growth of a PTHrP-secreting renal carcinoma cell line,34suggesting that PTHrP stimulates the growth of these transformed cells in an autocrine manner. Recently, Senior et reported the presence of PTHrP mRNA throughout the fetal rat using in situ hybridization. While their study failed to detect PTHrP mRNA in the liver, pancreas, or intestine, they did detect it in the lung and the CNS and in the epithelia of other organs. It is beginning to emerge that PTHrP has an important place in mammalian physiology as well as in pathophysiology. While the data presented here suggest a putative role for PTHrP as a fetal growth factor, caution in their interpretation should be exercised. Much in vitro and in vivo experimentation is required to support this hypothesis. The role that PTHrP performs in the fetus remains unknown but it must be multifactorial and involved in development. ACKNOWLEDGEMENTS
This work was supported in part by a grant from the Nuffield Foundation.
REFERENCES 1. Albright F. Case records of the Massachusetts General Hospital (case 2746 I ). N Enpl J Med 197 1 ; 225: 189-19 1.
2. Mundy G R , Ibbotson KJ, D’Soura SM, Simpson EL, Jacobs JW, Martin TJ. The hypercalcemia of cancer. Clinical implications and pathogenic mechanisms. N Engl J Med 1984; 3 1 0 1718-1 727. 3. Orloff JJ, Wu TL, Stewart AF. Parathyroid hormone-like proteins: biochemical responses and receptor interactions. Endocrine Rev 1989; 1 0 476-495.
4. Thiede MA, Rodan GA. Expression of calcium mobilizing parathyroid hormone-like peptide in lactating mammary tissue. Science 1988; 234: 278-280. 5. Yasuda T , Banville D, Rahbani SA, Hendy GN, Goltzman D. Rat parathyroid hormone-like peptide: comparison with human homologue and expression in malignant and normal tissue. Mol Endocrinol 1989; 3 518-524. 6 . Juppner H, Abou-Samra AB, Uneno S, Gu WX, Potts J T Jr, Segre GV. The parathyroid hormone-like peptide associated with humoral hypercalcemia of malignancy and parathyroid hormone bind to the
296
7.
8. 9. 10. I I.
12.
13. 14. IS.
16. 17 18.
19.
20.
P. B. J. BURTON E T A L .
same receptor on the plasma membrane of ROS 17/22 cells. J Brol Chem 1988: 263 8557-8560. Weir EC, Brines ML, Ikeda K, Burtis WJ, Broadus AE, Robbin, RJ. Parathyroid hormone-related peptide gene is expressed in the mammalian central nervous system. Proc Natl Acad Sci USA 1990.87: 108-1 12. Care AD. Development of endocrine pathways in the regulation of calcium homeostasis. Bai/liere's Clin Endo Melob 1990; 3 673-688. Roddd CP, Kubota M, Heath JA. r / a/. Evidence for a novel parathyroid hormone-related protein implicated in human hypercaicernia ofmalignancy. J Endocrinol 1988; 117: 261-271. Burton PBJ, Moniz C, Quirke P, e f a/. Immunocytochemical localization of PTHrP in the human fetal urogenital tract. Mol Cell Endocrind 1990; 6 9 R 13-R 17. Moniz C, Burton PBJ, Malik A, e f a/. Parathyroid hormone related peptide in human fetal development. J M u / Endocrinol 1990; 5 259-266. Scopsi L, Larsson LI. Increased sensitivity in peroxidase immunocytochemistry. A comparative study of a number of peroxidase visualization methods employing a model system. Hisrochemistry 1986; 84: 221-230. Chomczyski P, Sacchi W. Single step method of RNA isolation by acid guanidium-thiocyanate-phenolchloroform extraction. Anal Biochem 1981; 162 156-159. Fineberg AP, Vogelstein B. A technique for radiolabelling DNA restriction endonuclease fragments to high specificity. Anal Brochcm 1983; 132: 61-63. Cleveland DW, Lopata MA, MacDonald RJ, Cowan NJ, Rutter WJ, Kirschner MW. Number of evolutionary conservation of a- and 8tubulin and cytoplasmic p- and y-actin genes using specific cloned cDNA probes. Cell 1980; 2 0 95-105. Horiuchi N, Caulfield MP, Fisher JE, et ul. Similarity of synthetic peptide from human tumor to PTH in vivo and in >*itro.Science 1988: 238 1566-1568. Stewart AF, Mangin M, Wu T , er ul. A synthetic human parathyroid hormone-like protein stimulates bone resorption and causes hypercalcemia in rats. J Clin Invest 1988; 81: 596-600. Yates AJP, Guttierez GE, Smolens P, e f ul. Effects of a synthetic peptide ofPTH-related protein on calcium homeostasis, renal tubular calcium reabsorption, and bone metabolism in vivo and in virro in rodents. J Clin Invest 1988; 81: 932-938. McKee R, Goldman M, Caulfield MP, et ul. The 7-34 fragment of human hypercalcemia factor is a partial agonist /antagonist for PTH stimulated CAMPproduction. Endocrinology 1988; 122: 3008-3010. Mercola M, Stiles CD. Growth factor superfamilies and mammalian embryogenesis. Development 1988; 102: 451-460.
21. CrossM, DexterTM. Growth factorsin development, transformation and tumorigenesis. Cell 1991; 64:271-280. 22. Heine UI. Munoz EF, Flanders KC, e / a/. Role of transforming growth factor beta in the development of the mouse embryo. J Ce// B i d 1987; 105 2861-2876. 23. Quirke P, Pickles NL, Tuzi NL, Mohamadee 0,Gullick WJ. Pattern of expression of c-erbB-2 oncoprotein in human fetuses. Br J Cancer 1989, M): 64- 69. 24. Lee BA, Maher DW, Hannink M, Donoghue DJ. Identification of a signal for nuclear targeting in platelet-derived growth-factor-related molecules. Mo/Cell Biol 1987; 7: 3527-3537. 25. Gospodarowicz D, Ferrara N, Schweigerer L, Leufeld G. Structural characterization and biological functions of fibroblast growth factor. Endocrine Re)' 1987; 8: 95-1 14. 26. Roberts A, Sporn MB. (eds). The transforming growth factorbetas. In: Peptide Growth factors and their Receptors. Handbook of Experimental Pharmacology. Heidelberg: Springer-Verlag, 1989; 419-472. 27. Grobstein C. Mechanisms of organogenetic tissue interaction. Nut/ Concer Inst Monogr 1967; 26: 279-299. 28. Slack JMW. From Egg to Embryo. Cambridge: Cambridge University Press, 1983: 241 29. lnsogna KL, Stewart AF. Morris CA, Hough LM, Milstone LM, Cantrella M . Native and a synthetic analogue of the malignancy associated parathyroid hormone like protein have in v i m transforming growth factor-like properties. J Clrn Invest 1989; 83: 1057-1062. 30. Kikuchi H, Shigeno C, Lee K, e f a/. On the transforming growth factor beta-like activity of synthetic polypeptides comprising the amino terminal sequence of human parathyroid hormone-related peptide. Endocrinology 1991: 128: 1229-1237. 31. Rodan SB, Wesolowski G, Ianacone J, Thiede MA, Rodan GA. Production of parathyroid hormone-like peptide in a human osteosarcoma cell line: stimulation by phorbol ester and epidermal growth factor. J Endocrine/ 1989; 1 2 2 219-227. 32. Kremer R , Karaplis AC, Henderson J, er a/. Regulation of parathyroid hormone-like peptide in cultured normal human keratinocytes. J Clrn Ini'e.v/ 1991; 87: 884-893. 33. Gallagher JA, Hyman J, Pennington SR. PTHrP stimulates the proliferation of human bone cells and mouse 3T3 fibroblasts. Bone 1990; 11: P20 (Abstract). 34. Burton PBJ, Moniz C, Knight DE. Parathyroid hormone-related peptidecan function as an autocrine growth factor in human renal cell carcinoma. Biochem BiophJs Res Commun 1990; 167: 1134-1 138. 35. Senior PV, Heath DA, Beck F. Expression of parathyroid hormonerelated protein m R N A in the rat before birth: demonstration by hybridization histochemistry. J Mol Endocrinol 1991; 6 281. 290.