REGULATION OF ADRENOCORTICOTROPHIN AND ITS

PRECURSORS IN SMALL CELL LUNG CANCER

Mary Felicity Stewart, BSc, MB BS

A Thesis submitted to the University of London

for the degree of

DOCTOR OF MEDICINE ProQuest Number: U553788

All rights reserved

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a com plete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest

ProQuest U553788

Published by ProQuest LLC(2017). Copyright of the Dissertation is held by the Author.

All rights reserved. This work is protected against unauthorized copying under Title 17, United States C ode Microform Edition © ProQuest LLC.

ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106- 1346 ABSTRACT

The ACTH precursor gene, pro-opiomelanocortin (POMC) is expressed in many non-pituitary tumours. In this thesis, small cell lung cancer (SCLC) cell lines have been studied as an in vitro model of the regulation of human, non-pituitary POMC. In contrast to the aggressive growth of this tumour in patients, SCLC cell lines are slow growing and difficult to establish and maintain in culture. ACTH and its precursors were quantitated by two- site immunoradiometric assays based on monoclonal antibodies. Ten cell lines (56%) secreted predominantly ACTH precursors (POMC and pro-ACTH) in the range 37-1405 pmol/1, with very little, if any, processing to ACTH.

POMC RNA was demonstrated in 5 cell lines and a major transcript of approximately approximately 1350 bases was identified by Northern blot analysis in the cell line, COR L103, slightly larger than normal human pituitary (approximately 1100 bases). Incubation of SCLC cells for 10 days with hydrocortisone (500 and 1000 nM) did not inhibit intracellular POMC mRNA or ACTH precursor secretion, while this protocol produced significant inhibition in pituitary adenoma cells (AtT20). However, the dopamine agonist, bromocriptine (lOpg/ml) markedly suppressed ACTH precursor secretion following 24h and 7 days incubation. Corticotrophin releasing factor (CRF, 1000 ng/ml) had little effect on SCLC cells but significantly stimulated AtT20 and primary cultures of human pituitary cells. ACTH precursor secretion by SCLC cells was increased by dibutyryl cyclic AMP indicating that this intracellular signalling pathway is functional. These results are consistent with the clinical features of the ectopic ACTH syndrome, where ACTH circulates in precursor forms and is relatively resistant to glucocorticoid inhibition or stimulation with CRF. Thus the

POMC gene is regulated in SCLC but it appears that the mechanisms differ from those operating in the anterior pituitary.

2 To: Stephen, Eleanor and Peter

3 The results presented in this thesis have not been submitted by the author for any other degree and to the best of her knowledge are her own work, except where due acknowledgement is made.

4 i

ACKNOWLEDGEMENTS

I am indebted to Dr. Anne White for her careful supervision and direction

during the preparation of this Thesis.

I wish to acknowledge the generous donations of small cell lung cancer cell

lines by the following:

Dr. Peter Twentyman (MRC Clinical Oncology and Radiotherapeutics Unit,

Cambridge, UK).

Dr. Frank Cuttitta (National Cancer Institute, Bethesda, Maryland, USA).

Dr. M. Brouwer (Netherlands Cancer Institute, Amsterdam, The Netherlands).

Dr. Lou de Leij (University of Groningen, The Netherlands)

Dr. Gill Duchesne (Ludwig Institute for Cancer Research, Sutton, UK).

I am grateful to the many clinicians at Hope Hospital, Salford and

Wythenshawe Hospital, Manchester, for providing tumour samples from their

patients.

During the course of this project a number of collaborative studies were

carried out and the contribution of others to the work described in this

Thesis is detailed below.

1. The immunoradiometric assay for ACTH precursors was developed by Dr.

Steven Crosby. My own contribution was to culture the pituitary and small

cell lung cancer cells from which we were able to purify a POMC standard

for the assay.

2. The chromatography of ACTH-related peptides was carried out jointly by

Dr. Steven Crosby and me.

3. Studies of the molecular biology of POMC in small cell lung cancer cell

lines were undertaken with Dr. Adrian Clark and Professor L.H. Rees

(Department of Chemical Endocrinology, St. Bartholomew’s Hospital, London).

5 !f Northern and slot blot analysis of POMC, glucocorticoid receptor and ii, tyrosine amino transferase RNA were carried out by Dr. Adrian Clark and

Dr. Paul Lavender (St. Bartholomew’s Hospital) and Dr. Bill Farrell (Hope

Hospital). I thank Dr. Adrian Clark and Paul Lavender for carrying out one

set of cell culture experiments on AtT20 cells. These experiments are

identified in the text.

4. The dexamethasone binding assay was carried out by Dr. Bill Farrell and

Dr. Adrian Clark with help and advice from Professor Gavin Vinson (St.

Bartholomew’s Hospital).

5. Immunoradiometric assays for ACTH precursors and ACTH and cortisol

radioimmunoassays were performed by Mrs Sarah Gibson, Ms Naomi Simpson and

Ms Eileen Glover for which I am very grateful.

6. I carried out statistical analysis using the ’Minitab’ package with

helpful advice from Sally Hollis (University of Manchester, Department of

Statistical and Computational Studies). I thank Peter Crowe and Robert

Weinkove for teaching me to use computer graphics.

I particularly wish to thank Sarah Gibson, Steven Crosby and Bill Farrell

for many useful discussions, and other staff in the Department of Medicine

(Clinical Biochemistry) at Hope Hospital. Mrs. Christine Atkinson’s expert

secretarial skill was invaluable in producing this manuscript and I also

thank the Department of Medical Illustration for their help with

photographs and figures.

Finally, I wish to acknowledge Professor John G. Ratcliffe for supporting

me at the outset of my project and Professor Pankaj Vadgama for continuing

that support.

6 CONTENTS

INTRODUCTION

PAGE NOS

I. ACTH IN NORMAL TISSUES 18

1. THE POMC GENE 18

2. BIOSYNTHESIS AND RELEASE OF ACTH AND RELATED PEPTIDES 20

i) Structure of the ACTH precursor, Pro-opiomelanocortin 20

(POMC)

ii) Processing of POMC in the Anterior Pituitary 23

iii) The Neurointermediate Lobe 27

iv) POMC peptides in non-pituitary tissues 28

3. BIOLOGICAL ACTIONS OF POMC DERIVED PEPTIDES 29

i) ACTH 29

ii) N-POC 31

iii) The Lipotrophins and Beta-endorphin 32

4. REGULATION OF POMC EXPRESSION IN THE PITUITARY 32

i) Glucocorticoids 34

ii) Corticotrophin Releasing Factor 36

iii) Other Regulators of POMC expression 38 iv) Regulation of POMC in the Neurointermediate Lobe 40

II. MEASUREMENT OF POMC PEPTIDES 42

1. ACTH 42

2. Precursors of ACTH: POMC and Pro-ACTH 45

3. Other POMC derived peptides 46

III. HORMONES IN TUMOURS 47

1. Origin of Hormones in Lung Cancer 47

2. Nature of Hormones in Lung Cancer 50

3. Tumorigenesis: Links between Oncogenes, Autocrine 52

Growth Factors and Tumour Hormones

7 PAGE NOS

IV. ACTH IN SMALL CELL LUNG CANCER (SCLC) 55

1. Clinical Aspects 55

2. Nature of ACTH related peptides produced by SCLC 56

3. The POMC gene in SCLC 58

4. Regulation of ACTH in SCLC 59

V. SCLC CELL LINES 61

1. Establishment and Characterisation of SCLC Cell Lines 61

2. Cytogenetic abnormalities and oncogene expression 65

3. Hormone Production 66

4. Autocrine growth factors 68

5. Advantages and limitations 69

VI. SUMMARY AND AIMS OF THIS THESIS 71

MATERIALS AND METHODS

I Primary culture of human lung tumour cells 73

II Culture of established SCLC cell lines 77

III Culture of pituitary cells 80

IV Freezing down and retrieving cells from liquid nitrogen 82

V Methods for quantitation of tumour cells 83

VI Extraction of cell pellets for ACTH and precursors 85

VII Immunoradiometric assays for ACTH and ACTH precursor 86

peptides

VIII Cortisol radioimmunoassay 91

IX Analysis of RNA 92

X Glucocorticoid binding assay 94

XI Statistical analysis 95

8 RESULTS

Primary culture of lung and pituitary tumours 99

Secretion of ACTH and its precursors by SCLC cell lines 109

POMC gene expression in SCLC cell lines 121

Regulation of POMC gene expression and peptide secretion 127

in SCLC cell lines

Glucocorticoids 127

Bromocriptine 167

Corticotrophin releasing factor (CRF) 177

Effect of glucocorticoid and CRF on human corticotroph 184

cells

Second messenger mechanisms in SCLC cells 188

DISCUSSION

Primary culture of human lung and pituitary tumours 204

ACTH precursors in SCLC cell lines 205

ACTH and ACTH precursor secretion by pituitary cells 206

Expression of POMC RNA in SCLC cells 208

Regulation of POMC gene expression and peptide secretion 209

in SCLC cells

Summary and conclusions 218

9 PAGE NOS

REFERENCES 220

APPENDIX I Materials 257

APPENDIX II Tables of data 272

Section 1. Peptides 273

Section 2. RNA 292

APPENDIX III Normal Probability Plots 295

APPENDIX IV Summary of results of statistical 317

analysis

APPENDIX V Publications arising from this Thesis 325

10 LIST OF TABLES

Page

Table 1 Regulators of POMC in the anterior pituitary. 33

Table 2 Peptide hormones associated with SCLC. 51

Table 3 Composition of HITES medium for SCLC. 62

Table 4 Classification of SCLC cell lines based on 64

in vitro biological properties.

Table 5 Hormone secretion by SCLC cell lines. 67

Table 6 SCLC cell lines. 78

Table 7 Primary culture of lung tumour samples. 100

Table 8 Levels of ACTH precursors and ACTH secreted 107

by human pituitary cells.

Table 9 Levels of ACTH and precursor peptides in 112

supernatant medium from SCLC cell lines.

Table 10 Levels of ACTH and precursor peptides in cell 113

pellet extracts from SCLC cell lines.

Table 11 Levels of precursors in medium following 24h 117

incubation of SCLC cells.

Table 12 ACTH precursors secreted by SCLC cells at different 119

stages of growth.

Table 13 POMC RNA, ACTH precursors and ACTH levels in 122

5 SCLC cell lines.

Table 14 Glucocorticoid experiments on SCLC and AtT20 cell 131

lines.

Table 15 Effect of bromocriptine on POMC mRNA content in 176

COR L24 cells.

11 Page

Table 16 Effect of dibutyryl cAMP on POMC mRNA content of 193

COR L24 cells during 48h incubation.

Table 17 Effect of the phosphodiesterase inhibitor, IBMX and 197

dibutyryl cAMP on POMC RNA content of COR L24 cells.

Table 18 Effect of forskolin on POMC RNA content of COR L24 201

and COR L103 cells.

12 LIST OF FIGURES

Page

Figure 1 Structure of the POMC gene. 19

Figure 2 Putative promoter regions in the POMC gene. 21

Figure 3 Processing of POMC in the human, adult 24

pituitary.

Figure 4 Amino acid sequence of human ACTH 1-39. 30

Figure 5 Sites of action of factors which regulate the 35

POMC gene.

Figure 6 Intracellular second messenger systems 39

mediating hormone action on the corticotroph cell.

Figure 7 Principle of a two-site immunoradiometric assay. 44

Figure 8 Suggested models for the differentiation of 49

endocrine cells.

Figure 9 Binding sites of the MAbs used in the IRMAs for 87

ACTH-related peptides.

Figure 10 Outline protocol for the ACTH and precursor 88

IRMAs.

Figure 11 B-lymphoblastoid transformation. 101

Figure 12 Adenocarcinoma cell line. 102

Figure 13 Pulmonary carcinoid tumour. 104

Figure 14 SCLC cells in culture. 105

Figure 15 Human pituitary tumour cells. 108

Figure 16 Secretion of ACTH precursors and ACTH 110

throughout cell growth in 5 SCLC cell lines.

Figure 17 Sephadex G-75 chromatography of cell culture 114

medium from SCLC and carcinoid cells.

13 Page

Figure 18 Sephadex G-75 chromatography of cell culture 116

medium from pituitary cells.

Figure 19 Effect of cell density on ACTH precursor 118

secretion by COR L103 cells.

Figure 20 Stability of ACTH precursors secreted by COR 120

L103 cells.

Figure 21 Northern blot showing POMC mRNA in SCLC and 123

human pituitary cells.

Figure 22 ACTH precursors and ACTH throughout growth of 124

COR L103 cells.

Figure 23 Level of POMC gene expression during growth of 125

COR L103 cells.

Figure 24 Stability of hydrocortisone in culture medium. 129

GLUCOCORTICOID EXPERIMENTS

SCLC CELL LINES : 10D INCUBATION

Figure 25 COR L103 Experiment 1 132

Figure 26 COR L103 Experiment 2 133

Figure 27 COR L103 Experiment 3 134

Figure 28 COR L103 Experiment 4 135

Figure 29 COR L24 Experiment 1 136

Figure 30 COR L24 Experiment 2 137

Figure 31 COR L31 Experiment 1 138

Figure 32 COR L31 Experiment 2 139

Figure 33 COR L31 Experiment 3 140

Figure 34 COR L42 Experiment 1 141

Figure 35 COR L42 Experiment 2 142

Figure 36 COR L27 Experiment 1 143

14 Page

SCLC CELL LINES : 24H INCUBATION

Figure 37 COR L103 Experiment 1 144

Figure 38 COR L103 Experiment 2 145

Figure 39 COR L24 146

Figure 40 COR L31 147

Figure 41 COR L42 148

Figure 42 COR L27 149

AtTIO CELLS : 10D INCUBATION

Figure 43 Experiment 1 : ACTH precursors 151

Figure 44 Experiment 1 : ACTH 152

Figure 45 Experiment 2 : ACTH precursors 153

Figure 46 Experiment 2 : ACTH 154

Figure 47 Experiment 3 : ACTH precursors 155

Figure 48 Experiment 4 : ACTH precursors 156

Figure 49 Experiment 5 : ACTH precursors 157

AtT20 CELLS : 24H INCUBATION

Figure 50 Experiment 1 : ACTH precursors 158

Figure 51 Experiment 1 : ACTH 159

Figure 52 Experiment 2 : ACTH precursors 160

Figure 53 Experiment 2 : ACTH 161

Figure 54 POMC mRNA content of cells incubated with 163

hydrocort i sone.

Figure 55 TAT mRNA and GR mRNA content of cells 164

incubated with hydrocortisone. 3 Figure 56 Displacement of H-dexamethasone with 165

unlabelled dexamethasone in nuclear

preparation from COR L103 cells.

15 Page

Effect of bromocriptine on COR L24 cells. 168

(24h incubation).

Effect of bromocriptine on COR L31 cells 170

(24h incubation).

Effect of bromocriptine on COR L42 cells 171

(24h incubation).

Effects of bromocriptine on COR L24 cells 172

(7 day incubation, Experiment 1).

Effect of bromocriptine on COR L24 cells 173

(7 day incubation, Experiment 2).

Effect of bromocriptine on COR L31 cells 174

(7 day incubation).

Effect of bromocriptine on COR L42 cells 175

(7 day incubation).

Effect of oCRF on ACTH precursor secretion 178 by COR L42 cells.

Effect of oCRF on ACTH precursor secretion 179 by COR L103 cells.

Effect of oCRF on ACTH precursor secretion 180 by AtT20 cells.

Effect of oCRF on ACTH secretion by AtT20 181 cells.

Effect of different media on response to oCRF 183 by SCLC cells.

16 Page

Figure 69 Effect of dexamethasone and oCRF on ACTH 185

precursor secretion by human pituitary cells.

Figure 70 Effect of dexamethasone and oCRF on ACTH 186

secretion by human pituitary cells.

Figure 71 Effect of dibutyryl cyclic AMP on COR L24 cells. 189

Figure 72 Effect of dibutyryl cyclic AMP on COR L103 cells. 190

Figure 73 Time course of the effect of dibutyryl cyclic 192

AMP on COR L24 cells.

Figure 74 Effect of the phosphodiesterase inhibitor, 195

IBMX on COR L24 cells.

Figure 75 Effect of forskolin on COR L24 cells. 198

Figure 76 Effect of forskolin on COR L103 cells. 199

Figure 77 Effect of the phorbol ester, phorbol 12, 203

13-dibutyrate on COR L24 cells.

17 INTRODUCTION

I. ACTH IN NORMAL TISSUES

1. THE POMC GENE

The ACTH precursor, pro-opiomelanocortin (POMC), is a 31 KD glycopeptide whose complete amino acid sequence was first elucidated in cattle by

Nakanishi in 1979. The single human POMC gene is located on the short arm of chromosome 2 (Owerbach et al 1981) and its structure, which is well conserved in many mammalian species, has been characterised in detail

(Takahashi et al 1981, Cochet et al 1982, Whitfeld et al 1982). The POMC gene is expressed at high levels in the corticotroph cells of the anterior pituitary, but also at low levels in a number of non-pituitary tissues

including testis (Pintar et al 1984), ovary, placenta, duodenum, liver, kidney, adrenal medulla, hypothalalmus, lung, thymus and lymphocytes (Chen et al 1986, Debold et al 1988, Jingami et al 1984, Buzetti et al 1988).

In common with most mammalian genes, the POMC gene consists of exons

(regions which appear in the mature RNA transcript) separated by introns

(regions that are spliced out of the primary RNA transcript during the processing of this molecule to a mature messenger RNA) (Figure 1). The first exon of 87 base pairs (bp) contains no coding sequence, its RNA transcript functioning as a leader sequence which binds the ribosome at the start of translation. The second exon (152bp in length) contains the tail end of this leader sequence, followed by the initiator ATG sequence encoding a methionine residue which becomes the most N-terminal amino acid residue of the translation product. The next 78bp code for the signal peptide whose function is to translocate the nascent peptide through the endoplasmic reticulum , thence to the Golgi apparatus and ultimately to the secretory vesicle (Blobel and Dobberstein 1975).Immediately downstream of this signal sequence is the POMC coding region which starts in exon 2 and

18 Non-coding exon Intron A Intron B Coding exon t GENE I ,__ i— DNA 3'

Transcription Intron Excision

// MATURE 5' RNA Coding region Translation

1° PROTEIN NH2> ]-C 0 0 H PRODUCT Signal N- fragment ACTH 0LPH Peptide

Figure 1 Structure of the POMC gene. Transcription of the POMC gene gives a primary RNA transcript which is processed to the mature mRNA by removal of introns A and B. Open and stippled areas represent the exons, coding exons (open areas) appear in the mature transcript. The mature mRNA is translated to the 1* protein product.

19 continues in the third exon (835bp). The third exon contains the translation termination codon, and a further 137bp downstream, the poly-A addition, denoting the end of the RNA transcript, and the addition of the poly-A tail.

Transcription is driven by a promoter, known as the pituitary promoter, which occupies the 5’ flanking region upstream of the gene (Figure 2). This region contains an AT rich sequence or typical "TATA box" located 27bp upstream of the transcription initiation site (also known as the "cap" site). This sequence binds a ubiquitous protein, the TATA box binding protein, which in turn positions RNA polymerase II on the gene, ready to start transcription (Breathnach and Chambon,1981). Several other important regulatory elements almost certainly lie within this region including sequences which confer tissue specificity on the function of the promoter, cyclic AMP response elements (CREs) and alternative promoter sequences (see section I, 4).

2. BIOSYNTHESIS AND RELEASE OF ACTH AND RELATED PEPTIDES i) Structure of the ACTH Precursor

ACTH was the first pituitary polypeptide hormone to be isolated and sequenced (Li et al, 1954) but its biosynthesis was not understood until some years later. The recognition that insulin was synthesised in the form of a precursor,pro-insulin (Steiner et al 1967) suggested that this might also be the case for ACTH. In 1971, Yalow and Berson described high molecular weight forms of ACTH in the normal human pituitary, pituitary tumours, an ectopic ACTH-producing tumour and human plasma. It was suggested that these might represent a precursor form of ACTH with little bioactivity.

ACTH had been thought for many years to be secreted together with a peptide known as beta-melanocyte stimulating hormone (beta-MSH). This molecule had

20 Upstream Pituitary Downstream

Promoters

GENOMIC EXON 1EXON 2 EXON 3 DNA

Figure 2 Putative promoter regions in the POMC gene.

21 first been isolated from sheep and other mammalian pituitaries as an 18 amino acid peptide, whereas human beta-MSH apparently contained 22 amino acids (Harris 1959). The lipotrophins (beta- and gamma-LPH) so called because of their weak lipolytic activity on adipocytes in vitro, were also isolated from sheep pituitaries (Li et al , 1964; Li et al,1965,Chretien and Li 1967). When their amino acid structure was elucidated it was recognised that there was sequence homology between beta-MSH and the larger lipotrophin molecules (Li et al 1965) as well as a common sequence with

ACTH 4-10. Subsequently it was shown that human beta-MSH was produced as an artefact of extraction and the link with ACTH was really an association with beta-LPH (Bloomfield et al 1974). This was consistent with the earlier observation that ACTH and beta-LPH were released in roughly equimolar amounts by normal pituitary tissue in vivo and by human tumour tissue in vivo and in vitro (Abe et al 1969). In parallel with this work the discovery of methionine enkephalin (Hughes et al 1975) and the recognition that the amino acid sequence of this potent opioid peptide was contained within the beta-LPH molecule led to the identification of beta-endorphin

(beta-EP) as the C-terminal portion of beta-LPH which also possessed opioid activity (Li and Chung 1976).

In an immunohistochemical study on human pituitary, antisera to ACTH 17-39, alpha-MSH and beta-LPH all showed positive staining in the same cell type

(Phifer et al 1974). Beta-LPH and ACTH were found to share common secretory granules in pituitary cells (Pelletier et al 1977), raising the possibility that both peptides were derived from a common precursor. This was confirmed in an extensive series of experiments by Eipper and Mains in 1980, using the mouse corticotroph adenoma cell line, AtT20. The mouse precursor molecule was shown to be a 31KD glycopeptide, subsequently named pro­ opiomelanocortin. Using ACTH and beta-endorphin antisera for

22 immunoprecipitation and radioimmunoassay techniques, they were able to identify four molecular forms of ACTH, three molecular forms of beta- endorphin and a third unknown glycopeptide derived from the same precursor.

The ACTH-containing material had apparent molecular weights of 31KD, 22KD,

13KD and 4.5KD (Mains et al 1977). The beta-endorphin containing material had apparent molecular weights of 31KD, 11.7KD and 3.5KD, while the unknown glycopeptide had an apparent molecular weight of 16KD.

Pulse chase experiments using radioactively labelled amino acids were used to demonstrate that these molecular forms of ACTH were part of a biosynthetic pathway. The 31KD species represented the full precursor molecule and the 22KD fragment a biosynthetic intermediate. The 13KD and

4.5KD were alternative products, the former being glycosylated ACTH and the latter non-glycosylated ACTH 1-39. Similarly, the beta-endorphin containing fragments corresponded to the precursor (31KD), beta-LPH (11.7KD) and beta-endorphin (3.5KD). The 16KD fragment was isolated from AtT20 cells using an antiserum to crude preparations of tumour cell secretory products after prior removal of ACTH and beta-endorphin related material by immunoprec ipi tation.

The spatial arrangement of ACTH, beta-LPH and the 16KD fragment within the precursor was deduced by carboxypeptidase A and tryptic digestion of radiolabelled 22KD peptide. (Eipper and Mains 1980). This was confirmed by

Roberts and Herbert (1977) using cell free synthesis of precursor directed by AtT20 mRNA. ii) Processing of Pro-opiomelanocortin in the Anterior Pituitary

A schematic representation of POMC and it’s processing pathway is shown in

Figure 3. The human ACTH precursor is synthesised with a 26 amino acid signal peptide whose function has been described above. The signal peptide is rapidly removed by a signalase enzyme, probably before complete

23 Proopiomelanocortin (POMC)

Pro-ACTH 0 L P H

NPOC JP ACTH I A 7 M S H

7 LPH )3 EP

/3MSH M-Enk

Figure 3 Processing of POMC in the human adult pituitary. M-enk: Met-enkephalin. Other abbreviations are defined in the text.

24 translation of POMC as the peptide enters the intracisternal space of the endoplasmic reticulum (Harwood 1980). This yields the 241 amino acid POMC molecule. All the processed peptides within POMC are flanked by dibasic amino acids and proteolysis to release the active sequences takes place probably within the secretory vesicle. The conversion of POMC to ACTH occurs by two tryptic-like endopeptidase cleavage reactions and although both steps occur rapidly, the order appears to be strict. POMC is initially cleaved to yield pro-ACTH (also known as the 22KD peptide) and beta-LPH.

ACTH is then released in the second step giving a fragment consisting of an N-terminal peptide known as N-pro-opiocortin 1-76 (N-POC 1-76) and the

31 amino acid joining peptide (JP). The N-terminal peptide corresponds to the mouse 16KD fragment. The fate of JP is at present uncertain but it has recently been suggested that JP 1-18 is the cortical androgen stimulating hormone (CASH) (Parker et al 1989) which would imply that it is further cleaved before being released into the circulation. N-POC appears to undergo little further processing in the anterior pituitary although it does contain dibasic amino acids flanking an MSH sequence (gamma-MSH).A proportion of human beta-LPH (89 amino acids) undergoes two tryptic cleavages. The initial cleavage yields gamma-LPH and beta-endorphin following which beta-endorphin is further cleaved to yield gamma (1-17) and alpha-endorphin (1-16) in small amounts. These peptides are not flanked by pairs of dibasic amino acids and require specific endopeptidases. There may also be an alternative pathway in which beta-endorphin is not formed and beta-LPH 1-89 is converted to beta-LPH 1-75 or 1-74 plus beta-EP 18-31

(Liotta et al 1982). Exopeptidase cleavages remove basic amino acids from the COOH terminus of the peptides generated by endopeptidase action.

Although a number of enzyme activities have been described which are

25 capable of carrying out these reactions, POMC processing enzymes await full characterisation.

Thus normal anterior pituitary corticotrophs synthesise the POMC precursor and by a series of proteolytic cleavages produce N-POC, ACTH and beta-LPH as major products with a small percentage of beta-EP formed, all being co­ secreted simultaneously. (Mains and Eipper 1981, Estivariz et al 1981, Chan et al 1983).

Post-translation modification of POMC-derived peptides

In the pituitary 20-60% of the precursor present binds to the lectin, concanavalin A (Orth and Nicholson 1977). There is a classic Asn-x-Ser sequence in the N-POC region, following the gamma-MSH sequence. The Asn residue is N-glycosylated in the endoplasmic reticulum. The oligosaccharide core is ’trimmed’ and further modified in the Golgi apparatus by attachment of N-acetylglucosamine, galactose and N-acetylneuraminic acid. Further, less complex, O-glycosylation of Ser and Thr residues may take place in the

Golgi. ACTH itself does not appear to be glycosylated in man and this explains the lack of 13KD ACTH in normal, tumourous and post-mortem pituitary tissue (Allen et al 1980). However, a significant proportion of human ACTH is phosphorylated at Ser-31 (Eipper and Mains 1982). In the rat,

POMC also contains sulphate.

Glycosylation does not affect the biological activity of ACTH (Gasson 1979) or the processing of POMC (Loh and Gainer 1982, Jenks et al 1985). The suggestion that glycosylation prolongs the biological half life of POMC peptides (Shibasaki et al 1980) requires corroboration. Thus at present, the physiological significance of this modification remains unclear.

Other modifications include amidation and acetylation, which along with the proteolyic cleavage reactions, take place in the secretory vesicle. In general, these reactions are more a feature of the intermediate lobe. Many

26 bioactive peptides require amidation of a C-terminal glycine residue for

full activity and protection against degradation by extracellular

carboxypeptidase. Alpha-MSH was thought to be the only amidated peptide

derived from POMC but recently, the JP has also been shown to be C-

terminally amidated (Eipper et al 1986), supporting the view that JP may

be a functionally important product.

Both beta-endorphin and alpha-MSH are substantially alpha-N-acetylated in

the intermediate lobe but beta-endorphin is <1% N-acetylated in the human

anterior lobe. This modification affects their biological activity in

opposite ways, reducing opiate receptor binding and analgesic potency of

beta-endorphin (Smyth et al 1979) but greatly enhancing the activity of

alpha-MSH on melanocytes (Guttmann and Biossonnas 1961).

iii) The Neurointermediate Lobe

In man the neurointermediate lobe (NIL) of the pituitary is well defined

only in foetal life, whereafter it involutes although scattered

intermediate lobe cells may remain in variable numbers in the mature gland.

However many species, for example the rat, have a prominent intermediate

lobe in which the POMC gene is expressed in significant quantities and the

peptide product undergoes different processing. Furthermore, the regulatory

mechanisms are quite different. This is important because similarities

have been observed between the NIL pathway and that found in extra-

pituitary tissues. The initial processing is the same but whereas beta-LPH

is a major product in the anterior pituitary, in the NIL almost all the

beta-LPH is cleaved to beta-EP and gamma-LPH.

ACTH itself is an intermediate product in the NIL, being further processed

to alpha-MSH (alpha-N-acetyl-ACTH-l-13-amide) and ACTH 18-39. The latter molecule is known as Corticotrophin- Like Intermediate Lobe peptide (CLIP).

Amidation and acetylation of the products is a feature more characteristic

27 of the intermediate lobe. CLIP may be phosphorylated or glycosylated. The

N-POC region is more resistant to complete processing with approximately

70% being converted to N-terminal 1-48 or 1-49 and gamma-MSH (N-POC 50-74) in the rat (Jackson et al 1983). Thus the processing of the POMC precursor in the NIL can yield more than 10 small peptides but their significance and physiological role remains uncertain, iv) POMC peptides in non-pituitary tissues

The detection of POMC mRNA (as referred to above) is direct evidence for the expression of the gene although it does not necessarily imply that translation and secretion actually occurs. Indeed on northern blot analysis the principal POMC mRNA in these tissues is 800-900 bases in length, i.e. significantly shorter than that in the pituitary (Debold et al 1988). The peptide translation product of this short mRNA lacks a signal sequence and therefore should not be secreted . In the testis, and possibly in other tissues, very small quantities of full length message coexist with these short transcripts, but it is not yet clear whether an individual cell can produce both species or whether they arise in different cells.

POMC peptides have been detected by immunocytochemistry or radioimmunoassay

(RIA) in extracts from many different human tissues (De Bold et al 1988) including placenta (Liotta et al 1977), pancreas (Bruni et al 1979), stomach (Tanaka et al 1982), adrenal medulla (Evans et al 1983), brain

(Emson et al 1984) and testis (Chen et al 1986). The placenta appears to process the precursor extensively, producing beta-EP and alpha-MSH ,thus resembling the NIL pathway (Liotta and Krieger 1980). In human gastric antral mucosal cells, gamma-MSH, ACTH and beta-EP have been found with a predominance of the latter peptide (Tanaka et al 1982).

28 3. BIOLOGICAL ACTIONS OF POMC RELATED PEPTIDES i) ACTH

The actions of ACTH on the synthesis of adrenal steroids are well established. ACTH binds to specific receptors on the adrenocortical cell and thereby acts at several levels to stimulate steroidogenesis. The primary structure of human ACTH 1-39 is shown in Figure 4. A major active site has been localised to amino acids 5-10 (Schwyzer et al 1971), and in particular, the sequence His-Phe-Arg-Trp (positions 6-9) has strict structural requirements. The N-terminal sequence 1-3 may act as a secondary binding site and potentiate the steroidogenic effect, the amino group of

Ser-1 being extremely important in this context (Schwyzer 1980). A second active site has been identified in the 11-24 region. Amino acids 15-18 are basic and hence vulnerable to tryptic degradation and it is thought that the acidic region 25-32 may act to protect this basic sequence (Goverde et al 1980).

The current view is that there are at least two classes of ACTH receptor which recognise different sequences in the ACTH molecule and activate different effector mechanisms (Schwyzer 1980). One class binds ACTH 1-10 and stimulates steroidogenesis by cAMP production. The other class of receptor recognises the 11-24 sequence and probably acts via calcium mobilisation (Li et al 1989). ACTH increases adrenal uptake of cholesterol

(the precursor of steroid synthesis) in the form of low density lipoproteins (LDL). It stimulates cholesterol ester hydrolase and may also lead to the synthesis of proteins which appear to be involved in the transport of cholesterol from the cytoplasm to mitochondria (Bondy 1985).

ACTH also influences aldosterone production, being particularly important in compensatory increases following sodium depletion (Brownie and Pedersen

1987), an effect which may be potentiated by other peptides derived from

29 POTENTIATES PRINCIPAL PUTATIVE ACTH 5-10 ACTIVE SITE SECOND ACTIVE SITE

Ser Tyr Ser Met Glu His Phe Arg Trp Gly Lys Pro Val Gly Lys Lys Arg Arg 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Pro 20 Val DIBASIC A A ’S 21 Lys 22 Val 23 Tyr 24 Pro 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 Asn Phe Glu Leu Pro Phe Ala Glu Ala Ser Glu Asp Glu Ala Gly

ACIDIC SEQUENCE WHICH MAY PROTECT ACTH 15-18 FROM TRYPTIC DEGRADATION

Figure 4 Amino acid sequence of h u m a n A C T H 1-39

30 POMC. However the earliest function ascribed to ACTH was its ability to

prevent adrenal atrophy following hypophysectomy in animals although it has

recently emerged that other POMC derived peptides may stimulate adrenal

growth.

There are a number of extra-adrenal actions of ACTH including the

melanocyte stimulating activity attributable to the alpha-MSH sequence

within the molecule. Other actions proposed include effects on behaviour,

learning and memory (Born et al 1986), regulation of nerve growth and

repair (Strand and Smith 1980) and modulation of the immune system (Blalock

1985).

ii) N-POC

N-POC has been investigated for its ability to potentiate ACTH-induced

steroidogenesis in the adrenal. It can activate cholesterol ester hydrolase

and make more free cholesterol available for the steroid synthetic pathway

(Pederson et al 1980).

Human aldosteronoma cells cultured in the presence of N-POC showed

increased aldosterone output (Schiffrin et al 1983) and this study also

showed that the steroidogenic activity resided in the gamma-MSH region. Al-

Dujaili and colleagues (1981) reported that N-POC potentiated ACTH

stimulation of corticosterone and aldosterone in isolated, perfused rat and

‘ human adrenal cells but N-POC alone did not show this action. N-POC was 1 i | found to stimulate RNA synthesis in adrenal cell incubations (Al-Dujaili

| et al 1982). N-POC may also be involved in the compensatory adrenal growth

| seen after unilateral adrenalectomy. The N-terminal N-POC 1-48/49

(excluding the gamma-MSH sequence) has mitogenic activity on the rat

adrenal gland (Estivariz et al 1982) while gamma-MSH appears to stimulate

adrenal hypertrophy but to have no mitogenic activity (Lowry et al 1983).

31 iii) The Lipotrophins and Beta-endorphin

The lipotrophins, beta- and gamma-LPH, have been shown to be weakly lipolytic in vitro but there is no evidence that this effect is physiologically important in vivo. Beta-LPH contains the sequences of beta-

EP (beta-LPH 61-91), alpha-EP (61-76) and gamma-EP (61-77). Although there is a large literature on the opioid activity of these peptides, exactly how they act to modulate pain and stress responses in vivo remains to be established. In any event, the source of opioid peptides for this role may not be the anterior pituitary in view of the fact that beta-EP may not be a major product of the corticotroph processing pathway. In the brain and central nervous system beta-EP may function as a neurotransmitter or neuromodulator.

4. REGULATION OF POMC EXPRESSION

POMC is subject to regulation at multiple levels both in terms of gene transcription, RNA processing and turnover, as well as translation, processing and secretion of the peptide products. The principal factors capable of exerting regulatory effects on the production of POMC related peptides are listed in Table 1. It is important to note that regulation in the NIL is different to the anterior pituitary and will be described separately. In the anterior pituitary, the major factors acting to suppress and stimulate POMC production are glucocorticoids and corticotrophin releasing factor (CRF) respectively and these are described in some detail.

Although for the sake of simplicity, their effects are described individually, in reality, coordinated control of ACTH secretion is achieved not through these factors acting in isolation but by their interaction together with many other modulators. Secretion of ACTH occurs in a pulsatile fashion with superimposed diurnal variation and in addition, the TA B L E 1. Regulators of P O M C in the anterior pituitary

STIMULATION REFERENCE

Corticotrophin Releasing Factor Vale et al, 1981.

Vasopressin Gillies et al, 1982.

Vasoactive Intestinal Polypeptide Oliva et al, 1982

Angiotensin II Plotsky et al, 1988.

Cholecystokinin C-terminal Reisine and Jensen octapeptide 1986

Bombesin/Gastrin Releasing Peptide Hale et al, 1984.

Catecholamines Plotsky et al, 1985.

Interleukin 1 Bernton et al, 1987.

INHIBITION

I Glucocorticoids Keller-Wood and Dallman 1984.

Somatostatin Reisine, 1985.

Oxytocin Page et al, 1990

f i

33 corticotroph cell is able to respond on a minute by minute basis to complex physiological stresses such as hypovolaemia, hypotension and hypoxia,

i) Glucocorticoids

Glucocorticoids suppress POMC gene expression through at least two mechanisms. They can act directly on the corticotroph cell to reduce the levels of POMC mRNA (Roberts et al 1979). They also act on the hypothalamus to reduce the synthesis and release of CRF. Adrenalectomy raises the levels of CRF immunoreactivity in the paraventricular nucleus of the hypothalamus

(Swanson et al 1983) and raises POMC mRNA levels and transcription rate in the anterior pituitary (Birnberg et al 1983, Eberwine and Roberts 1984), this latter observation suggesting that glucocorticoids tonically inhibit

POMC gene activity in the anterior pituitary. The effects on transcription rate appeared to be rapid (occuring within 15 minutes) whereas maximal inhibition of POMC mRNA was not observed until 36-48h.

The effects of glucocorticoids are mediated via the classical steroid- receptor interaction. The glucocorticoid receptor (GR) is expressed at high levels in the corticotroph cell and is one of a family of receptors having homologous sequences. At the molecular level, much remains to be determined about the action of the glucocorticoid-receptor complex on POMC DNA. Four binding sites for GRs known as glucocorticoid inhibitory elements (GIEs) have been identified in the promoter region of the rat gene. Functional analysis suggests that only one of these is needed in vivo (Drouin et al

1987), this being the one that is conserved at position -62bp in the human gene. This binding site lies within the so-called pituitary promoter region

(represented in Figure 5). This overlaps a sequence known as COUP which has been identified as an upstream regulatory element (URE) for the chick ovalbumin gene (Wang et al 1987). The protein which binds to COUP has been purified and cloned and has some features similar to the steroid/thyroid

34 GLUCOCORTICOIDS CRH

cAMP

CRE COUP TATA \BPJ v BP J

POMC

- 1 CAP SITE • d x XD

PITUITARY 'FACTORS'

Figure 5 Sites of action of factors which regulate the POMC gene in promoter regions upstream of the CAP site. Tiss spec (1) and (2) confer pituitary specificity on the POMC gene

35 hormone group of nuclear receptors (Wang et al 1989). It appears to act as a permissive regulatory factor. The ligand for this protein has not yet been identified but it needs to associate with a second protein known as

S300-II to be fully functional (Tsai et al 1987). The closeness of the GIE and the putative COUP site in the human POMC has provoked the hypothesis that GR might exert its inhibitory effects by displacement of or association with the COUP protein (Drouin et al 1990). Precedents for this exist in other genes (Akerblom et al 1988). Cyclic AMP (cAMP) is an activator of the POMC gene and cAMP response elements (CREs) have been identified in the rat gene. It has also been suggested that binding of GR could induce a conformational change interfering with the ability of cAMP regulatory proteins to stimulate the POMC gene (Reisine and Affolter

1987). It is interesting to note that POMC gene transcription is not completely blocked in the presence of glucocorticoids. This is explicable if the action of GR is to interfere with binding of a transcription factor that modulates basal expression (Lundblad and Roberts 1988).

In addition to the transcriptional effects, glucocorticoids can act rapidly to inhibit hormone-stimulated POMC peptide secretion. Inhibition of CRF- stimulated peptide release by glucocorticoids appears to be biphasic, with an early phase which does not require protein synthesis (Abou-Samra et al

1986b). Second messenger pathways are potential targets but the precise steps influenced by glucocorticoids are undefined, ii) Corticotrophin Releasing Factor (CRF)

In 1955, two groups presented evidence for the existence of a humoral hypothalamic factor that stimulated ACTH secretion (Saffran et al 1955,

Guillemin and Rosenberg 1955) and named it Corticotrophin Releasing Factor

(CRF). CRF was finally isolated, purified and characterised as a 41 amino acid peptide (Vale et al 1981) and is now recognised as the most potent and

36 effective physiological stimulator of ACTH secretion. In common with many

neuropeptides, CRF is widely distributed within the CNS and other tissues

(Petrusz et al 1985) but the highest concentrations are found in

hypothalamic nuclei and in particular, the median eminence. Parvocellular

| neurons arising from the paraventricular nuclei abut onto pericapillary

spaces and from here, CRF is released into the portal circulation. I i CRF both stimulates peptide release and increases the synthesis of POMC in

[ the anterior pituitary. POMC mRNA levels are raised in vivo by continuous \ | administration of CRF to rats ( Bruhn et al 1984) and in vitro , an

[ f increase in POMC mRNA levels was detected after 4h application of CRF to

AtT20 cells (Affolter and Reisine 1985). This involves an increase in POMC

gene transcription (Herbert et al 1986, Roberts 1986). CRF acts through a

cytoplasmic, membrane-bound receptor coupled via a stimulatory guanine

nucleotide binding protein (Ns or Gs) to adenylate cyclase (Aguilera et

al 1983). The cAMP mediated actions of CRF on both peptide release and POMC

gene transcription have been shown to be, at least in part, calcium

dependent. The interactions of second messenger pathways within the

corticotroph are further discussed in section 4(i i i). As with

glucocorticoid inhibition, the effect of CRF on gene transcription rate is

seen rapidly (within minutes) in comparison with the time taken to observe

a change in POMC mRNA levels. This is because there is a relatively large

pool of cytoplasmic mRNA in comparison with the amount of primary

transcript in the nucleus. Transcriptional regulation of the POMC gene by

CRF is presumably mediated by binding of a regulatory factor to the

putative cAMP response element in the promoter region (Figure 5).

The current consensus is that while CRF is the principal natural ACTH

secretagogue, other factors must participate. Antisera to CRF do not

completely abolish ACTH increases caused by adrenalectomy or other stresses

37 (Rivier et al 1982) and furthermore ACTH responses to stresses such as

haemorrhage or insulin hypoglycaemia are greater than those observed during

maximal stimulation with CRF (Gillies et al 1982; Plotsky et al 1985).

There are many putative candidates but with the exception of vasopressin,

their individual significance has not been conclusively established,

iii) Other Regulators of POMC expression.

Table 1 lists the other factors which can influence POMC expression and the

principal second messenger pathways through which they are believed to act

are summarised in Figure 6. Arginine vasopressin (AVP) was suspected for

many years to be an ACTH releasing agent and this has been recently

confirmed, although its intrinsic activity is much weaker than CRF-41. It

appears to act synergistically with CRF, via phosphatidyl inositol (PI)

turnover, amplifying the effects of CRF-41 2-3 fold (Gillies et al 1982).

CRF and AVP have been found to be co-localised in paraventricular neurons

of the hypothalamus that project to the median eminence (Sawchenko et al

1985). The physiological role of other ACTH secretagogues is less well

defined. Oxytocin has been shown to exert an inhibitory effect on the ACTH

response to CRF-41 and may modulate the neuroendocrine response to stress

in man (Page et al 1990). Co-localisation of Cholecystokinin C-terminal

octapeptide (CCK-8) with CRF has been shown in some neurons in the

paraventricular nucleus (Mezey et al 1985) and this peptide can stimulate

ACTH release from both AtT20 cells and primary cultures of rat anterior

pituitary (Reisine and Jensen 1986). Vasoactive Intestinal Peptide (VIP)

is also present in the hypothalamus and can stimulate ACTH secretion from

AtT20 cells (Reisine et al 1982) and human anterior pituitary cells (Oliva

et al 1982). Angiotensin II (All) mediated ACTH release has been demonstrated in vivo in humans (Rayvis and Horon 1971) but reports on its ability to act alone as an ACTH secretagogue in vitro are conflicting.

38 AVP CRF Bombesin VIP All B-adrenerglc CCK-8 ♦ i OUT

plasma membrane

pl-c y —

Ca‘ POE DAG ATP CAMP 6‘ AMP

Protein Kinase C Protein Kinase A Ca'* fc. ♦ © * ▼ Phophoprotelns Phoaphoprotelns

Figure 6 Schematic representation of the intracellular second messenger systems implicated in mediating hormone action on the corticotroph cell. H: hormone, R: receptor, G: G-protein, PL-C: phospholipase C, IPj: inositol 1, 4,5-triphosphate; PIP2: phosphatidylinositol-4,5 -bisphosphate;DAG:diacylglycerol; PDE: phosphodiesterase; other abbreviations are defined in the text.

39 Recent studies on rat pituitary cells suggest that it acts to enhance CRF- stimulated ACTH release (Suda et al 1989). In vivo it may act mainly at the hypothalamic level, stimulating release of CRF into the portal circulation

(Plotsky et al 1988). Recently, bombesin, a 14 amino acid amphibian peptide, whose sequence shows considerable homology with the C-terminal of mammalian gastrin releasing peptide (GRP), has been shown to have weak intrinsic CRF-releasing activity and also to be capable of potentiating the action of CRF-41 (Hale et al 1984).

AVP, oxytocin, bombesin, All, CCK-8 and alpha-adrenergic agonists act via

PI turnover, mobilisation of intracellular calcium stores and activation of protein kinase C (Miyazki et al 1984, Abou-Samra et al 1986a). Figure

6 illustrates the interdependence of the second messenger pathways in the corticotroph cell. AVP can enhance CRF stimulated cAMP levels, probably by inhibiting phosphodiesterase although it does not significantly affect cAMP when acting alone (Vale et al 1983, Abou-Samra et al 1987). CRF and beta- adrenergic agonists which act to raise cAMP also raise cytosolic calcium levels in AtT20 cells by activation of protein kinase A (Guild and Reisine

1987). Somatostatin, another hypothalamic peptide, can prevent the accumulation of cAMP in response to these agents but also reduces cytosolic calcium levels independently of its effects on cAMP (Reisine 1985, Luini et al 1986). The importance of calcium is illustrated by the observation that the calcium ionophore A23187 can directly stimulate POMC gene transcription in vitro (Eberwine et al 1987). Precisely how cAMP-dependent and calcium-dependent processes interact at the molecular level to stimulate POMC production remains to be determined, iv) Regulation of POMC in the Neurointermediate Lobe

CRF and beta-adrenergic agonists can stimulate POMC production in NIL cells but glucocorticoids have very little effect apparently because the GR is

40 not expressed in this tissue (Schacter et al 1982). The major pegative regulator in NIL cells is dopamine which acts through a D2 receptor (Cote et al 1985) to inhibit the activation of adenylate cyclase. In ^addition gamma-aminobutyric acid (GABA-ergic) innervation inhibits the release of

POMC peptides in the NIL (Tomiko et al 1983).

41 II. MEASUREMENT OF POMC PEPTIDES

The assays available for the measurement of POMC peptides are based on biological or immunological techniques. Bioassays have the advantage that they measure the specific biological activity of a peptide and do not suffer from interference due to inactive fragments or precursor forms.

However, they are often complex and suitable for only limited numbers of samples. There may be difficulties in interpretation due to species specificity. Immunoassays are, in general, more simple, sensitive and appropriate for routine use. Nevertheless it has not proved easy to develop reliable, specific radioimmunoassays (RIAs) for POMC derived peptides because of the heterogeneity of molecular forms present in clinical samples, several of which contain common epitopes. Most clinical studies have relied on immunoassays and the remainder of this section is confined to a discussion of these techniques.

1) ACTH

In the past it has been difficult to achieve a radioimmunoassay of sufficient sensitivity to measure the very low levels of ACTH circulating in normal plasma. ACTH is poorly immunogenic and therefore high avidity antisera are not generally available. This problem can be overcome by extraction of plasma prior to RIA (Rees et al 1971) but this adds to the complexity of the procedure. Cross-reaction with other peptides, particularly the precursor forms of ACTH, can result in poor assay specificity. It has not been possible to ascertain the extent of such cross-reactivity since no suitable precursor standards have been obtained.

Differential adsorption of peptides during extraction procedures and differences in the in vivo and in vitro half lives of ACTH containing peptides are factors which further complicate the interpretation of

42 results. Therefore it is not surprising to find a dissociation between bioassay and immunoassay data (Yalow and Berson, 1971; Fleischer et al,1974).

These problems have to a great extent been overcome by the development of a two-site immunoradiometric assay (IRMA) based on monoclonal antibodies

(MAbs) for ACTH (White et al 1987). The principle of this type of assay is illustrated in Figure 7. The two-site IRMA employs two separate antibodies directed against different epitopes on the antigen. The first antibody is labelled and present in excess so that all the antigen in the sample is effectively bound in a labelled complex. The second antibody is bound to a solid phase and binding of this antibody to a different site on the peptide separates the peptide-labelled antibody complex from free, excess labelled antibody. There are a number of theoretical advantages to this approach. Firstly, immunometric assays in general offer improved sensitivity and working range in comparison with RIAs. The use of two MAbs offers improved specificity and ensures a continuous supply of homogeneous reagents.

The production and characterisation of MAbs to the N- and C-terminal of

ACTH (White et al 1985) led to the development of the ACTH IRMA referred to above. This assay detects intact ACTH 1-39 with no interference from fragments although it does retain a degree of cross-reactivity with the precursors, pro-ACTH (22KD peptide) and POMC. It is simple to perform and suitable for large numbers of samples. The normal range for plasma ACTH has been defined with this assay. In normal subjects, mean plasma ACTH levels are 30ng/l at07.30h and 15ng/l at 16.30h at rest (White et al 1987)

43 ANTIGEN LABELLED ANTIBODY

- ★

ANTIBODY LINKED TO SOLID PHASE •»-*

Separate

^ \ I% > Wash and count radioactivity

Figure 7 Principle of a two site immunoradiometric assay.

44 2) PRECURSORS OF ACTH: POMC AND PRO-ACTH

Detection of the ACTH precursors was, until recently, only possible by chromatographic separation followed by RIA of the fractions using antisera to component peptides such as ACTH, gamma-MSH or beta-LPH. This approach is technically difficult, labour intensive and requires a large sample volume. Most importantly, quantitation is variable, depending upon antiserum specificity and the peptide used as standard. Studies using this method reported high molecular weight precursor forms of ACTH circulating in pathological conditions, in particular, the ectopic ACTH syndrome

(Ratter et al 1980). Precursors could not be detected in the plasma of normal sujects except following insulin hypoglycaemia (Hale et al 1986).

We have recently developed a sensitive two-site IRMA for the precursors of

ACTH (Crosby et al 1988). This assay employs a MAb to gamma-MSH in combination with an ACTH MAb. A positive signal in the assay requires that both peptide sequences be present in the molecule(s) detected. Thus both

POMC and pro-ACTH are recognised but ACTH, N-POC or beta-LPH do not cross- react. We obtained a POMC standard by partial purification of growth medium collected from cultured human pituitary corticotroph tumour cells

(provided as part of a collaborative study with Dr. A. Clark, Professor L.

Rees and Professor G. M. Besser, St. Bartholomew’s Hospital, London). The assay is direct, precise, fully quantitative and simple to perform. The improvement in sensitivity obtained with this assay enabled the detection of low levels of ACTH precursors in the plasma of normal subjects (5-

34pmol/l at 09.30h, n=ll).

45 3) OTHER POMC DERIVED PEPTIDES

Radioimmunoassays have been described for the N-terminal peptide, N-POC

(Chan et al 1983, Bertagna et al 1983,) but pre-extraction and concentration of plasma was required to detect gamma-MSH immunoactivity in normal subjects (Hale et al 1984). These studies reported that N-POC is present in plasma at molar concentrations which are 2-5 fold higher than

ACTH in patients with various disorders of the hypothalamic-pituitary- adrenal axis. Determination of the physiologically relevant forms of beta-

LPH is complicated by heterogeneity in plasma. Antisera to beta-MSH, beta-

EP and gamma-LPH may recognise each sequence in beta-LPH therefore chromatographic separation is required prior to RIA. Different physiological stimuli, such as insulin hypoglycaemia, appear to alter the relative proportions of beta-LPH, gamma-LPH and beta-EP although all appear to circulate in normal plasma (Tanaka et al 1978, Yamaguchi et al 1980,

Wilson et al 1981, Suda et al 1982).

Little is known about circulating levels of the JP although RIAs are available (Bertagna et al 1988). This is likely to become more important if the suggestion that JP gives rise to CASH (JP 1-18) is confirmed (vide supra).

Further characterisation of these peptides requires the development of more sensitive and specific methods for their measurement. New two-site IRMAs based on different combinations of MAbs to specific epitopes in POMC potentially offer a solution to this problem.

46 III. HORMONES IN TUMOURS

1) ORIGIN OF HORMONES IN LUNG CANCER

It has been recognised for many years that hormones may be produced by tumours in tissues other than the normal physiological site of synthesis, leading on occasion to the so-called "ectopic" hormone syndromes. Small cell lung cancer (SCLC) is particularly noted for this phenomenon. Whether these tumours are truly producing "ectopic" hormones is now in doubt in view of the emerging evidence that hormone genes may be expressed in many normal tissues other than the primary endocrine organ. It is also clear that the relatively unusual clinical syndromes represent the most extreme manifestation of an underlying phenomenon which is actually quite common.

This is a field of considerable interest in terms of gene regulation and tumour biology.

The propensity of SCLC to elaborate peptide hormones together with histological features such as the presence of dense core granules and argyrophilic staining properties was thought to indicate that the disease was quite distinct from other forms of lung cancer.The APUD theory (Pearse

1969) suggested that hormone-secreting tumours were derived from endocrine cells present in the normal tissue which shared a common embryological origin in the neural crest. These cells all possessed Amine Precursor

Uptake and Decarboxylation (APUD) properties. SCLC along with phaeochromocytoma, medullary thyroid cancer , islet cell tumours of the pancreas and carcinoid tumours of the gut and lung were considered to be

APUD tumours.

The normal tracheo-bronchial epithelium of the adult lung contains small numbers of scattered endocrine cells which produce peptides such as calcitonin (CT), calcitonin gene related peptide (CGRP), bombesin-like immunoreactivity/gastrin releasing peptide (BLI/GRP), growth hormone

47 releasing factor (GRF), somatostatin (SS), antidiuretic hormone or vasopressin (AVP) and ACTH. Foetal lung contains higher hormone concentrations and relatively more endocrine cells than adult lung, suggesting a possible paracrine regulatory role for these peptides in development and maturation (Becker 1984). Pulmonary endocrine cells contain cytoplasmic dense core granules, enzymes such as L-dopa decarboxylase and neurone specific enolase, and other APUD features (Hage 1980). However, in the 1970’s it was shown that the endocrine cells of the gut and lung derived not from the neural crest but from endodermal stem cells (Pictet et al 1975, Sidhu 1979). Thus the APUD hypothesis was modified so that these scattered endocrine cells were considered to be part of a diffuse neuroendocrine system which could give rise to hormone-secreting tumours, although not necessarily exhibiting APUD staining characteristics. It has been suggested that these cells were hormonally pluripotential at an early stage of development with much of this endocrine activity being switched off during differentiation. Transformation then induces dys-differentiation such that some of their pluripotential characteristics are regained (Baylin and Mendelsohn 1980).

Two models of differentiation in normal bronchial mucosa have been suggested by Baylin (1985) and are schematically represented in Figure 8.

In model (A), the basal cell progresses to a totipotent cell capable of differentiating along several pathways, ultimately giving rise to endocrine, mucous or ciliated cell types. In model (B), a transient endocrine cell stage exists from which arise a small number of terminally differentiated endocrine cells, while the majority become large indifferent cells and eventually mucous or ciliated types. This latter model might explain the relatively greater proportion of endocrine cells in foetal life compared with their sparse occurence in the adult.

48 A B Mucous cell Ciliated cell Mucous cell Ciliated cell ,00 Terminally differentiated 00 endocrine cell Indifferent cell Terminally differentiated Young endocrine endocrine cell cell Indifferent Transient cell endocrine cell

Basal (reserve) Basal (reserve) cell cell

Figure 8 Suggested models for the differentiation of endocrine cells. From Baylin (1985) with permission.

49 Both models can be extended to encompass neoplastic changes and are compatible with the theory that all lung tumours are linked in a differentiation spectrum (Yesner and Carter 1982). This hypothesis explains many observations which were at variance with the suggestion that SCLC had a separate embryological origin from non-SCLC (NSCLC). Histologically, up to 6% of SCLC show evidence of heterogeneity with respect to cell type in the tumour and at post mortem, 25-40% of tumours from treated patients show evidence of another histological type, usually large cell carcinoma but occasionally squamous or adenocarcinoma (Gazdar et al 1981). Although production of peptide hormones, biogenic amines and enzymes is most frequently associated with SCLC and pulmonary carcinoids, it is well documented in lung tumours of all histological types (Yamaguchi et al

1985). This is more explicable on the basis of these models. The use of such markers may prove important in delineating these differentiation relationships more clearly, by analogy with human leukaemias and lymphomas.

While these models of differentiation explain why certain tumours have the capability to produce hormones, the question of why a particular hormone gene is expressed, sometimes to excess, in an individual tumour remains unresolved. Hormone gene expression may be regarded as a phenotypic consequence of the transforming events themselves. The detailed nature of these events remains obscure in most human neoplasia but some unifying themes are emerging and are discussed in the following section.

2) NATURE OF HORMONES IN SCLC

An enormous number of polypeptide hormones have been described in association with SCLC (Table 2) and those now known to be present in the normal tracheo-bronchial epithelium are all represented. ACTH and vasopressin have received most attention because of the clinical syndromes they produce, and together with calcitonin, have been extensively evaluated

50 TABLE 2

Peptide hormones associated with SCLC

In vivo

Comment Reference

ACTH HHW forms and other POMC-derived Hansen et al 1980 peptides co-secreted. Ratcl iffe et al 1982 Causes ectopic ACTH syndrome.

ADH (vasopressin) Neurophysins may be co-secreted. Eagan et al 1974 Causes syndrome of inappropriate Hansen et al 1980 ADH secretion (SIADH)

Calcitonin Calcitonin gene related peptide Luster et al 1982 (CGRP) may be co-secreted

CRF or CRF-1ike May cause Cushing’s syndrome Upton & Amatruda 1971 activity in SCLC or bronchial carcinoid Kirkland & Ellison 1984

Bombesin/GRP May cause Cushing’s syndrome by Moody 1984 virtue of CRF-1ike activity Hewlett et al 1985

Parathormone Unusual in SCLC; more common Schmelzer et al 1985 PTH like peptides in NSCLC. Causes hypercalcaemia

Tissue extracts

SCLC tumours examined

GRP 23/40 Yamaguchi et al 1985 Calcitonin 11/40 Neurotensin 6/40 ) Somatostatin 7/40 CGRP 4/40 Growth hormone releasing factor 10/40 Met-enkephalin 4/40 Neuropeptide Y 3/40 VIP 3/40 ACTH 8/40 ADH 3/40

51 as potential tumour markers (Hansen et al 1980). However no hormonal marker has yet been found which is of sufficient sensitivity or specificity to be useful in the detection of SCLC and they are of limited usefulness in monitoring for early relapse after treatment. Enzyme markers such as creatine kinase BB (CK-BB) and neuron specific enolase (NSE) appear to be of more clinical value (Bork et al 1988). A number of studies attest to the high frequency of hormone production in SCLC. A large study of 157 primary lung tumour extracts found that 83% of SCLC tumours produced at least one peptide (Yamaguchi et al 1985). Multiple hormone production is frequently found (Havemann et al 1985) and inappropriate secretion of antidiuretic hormone (ADH, vasopressin) occurs together with the ectopic ACTH syndrome in 20% of cases (Eagan et al 1974). Other studies report raised levels of several hormones in the plasma of individual patients, with and without clinical syndromes.

Processing of peptide hormones in lung and other tumours is characteristically incomplete so that a large proportion is in the form of high molecular weight prohormones. This has been described for ACTH (Yalow and Berson 1973), calcitonin (Luster et al 1982), and GRP (Cuttitta et al

1988). Much interest has focussed on GRP/Bombesin-like peptides in SCLC because they are found in a very high proportion of tumours and more importantly, there is compelling evidence that they can function as autocrine growth factors. These studies are described in more detail in section V (iv).

3) TUMORIGENESIS : LINKS BETWEEN ONCOGENES. AUTOCRINE GROWTH FACTORS AND

TUMOUR HORMONES

Most human neoplasia seem to result from a multistep process whereby the transformed cells acquire their ability to replicate, invade and metastasise. Much has been learned recently about the genetic basis of

52 these changes following the discovery of oncogenes and anti-oncogenes

(tumour-suppressor genes). Oncogenes are modified forms of a subset of highly conserved cellular genes (proto-oncogenes) which are normally present in all cells, although not necessarily constitutively expressed.This class of genes is believed to encode regulatory proteins which are critical to normal development and differentiation. Chromosomal alterations, including translocation and deletion, mutations and gene amplification (increased copy number) can lead to inappropriate expression of the oncogene product, which may be abnormal, in the case of a mutated gene. Loss of anti-oncogenes, such as the tumour suppressor gene Rb, in retinoblastoma and Wilm’s tumours, can also lead to oncogene activation.

Oncogene products include growth factors, growth factor receptors, signal transduction proteins, protein tyrosine kinases, DNA and nucleotide binding proteins. All of these can act in regulatory cascades. In parallel with the work on oncogenes, many growth factors capable of stimulating tumour cell replication have been identified in recent years. Cancer cells are able to replicate relatively autonomously and they can synthesise, secrete and respond to their own endogenous growth factors. This process has been termed "autocrine growth control" (Sporn and Todaro 1980). Tumour hormones are obvious candidates to act as autocrine growth factors and their synthesis could be influenced by oncogene expression in several ways.

Oncogene expression at a particular stage in the cell cycle could redirect differentiation so that the cell expresses endocrine features (Figure 8).

Alternatively, oncogene activation might lead to concomitant derepression of other genes in the same chromosomal segment. This idea is supported by the recent observation that the genes for many known tumour markers are clustered with cellular proto-oncogenes (Hozier et al 1987). Similarly an altered anti-oncogene might lead to derepression not only of the oncogene(s) it controls,but also other linked genes.

Thus the two apparently disparate areas of oncogene and growth factor research are now intimately linked and in the process, a potentially fundamental role for tumour hormones in the progression of neoplastic disease has been uncovered.

54 IV ACTH IN SCLC

1) CLINICAL ASPECTS

SCLC is the commonest non-pituitary neoplasm associated with ACTH production. Although the "ectopic" ACTH syndrome is relatively unusual in that it affects 2.5-7% of patients with SCLC (Kohler and Trump 1986), this disease accounts for more than 50% of cases overall. Typically patients present acutely ill, with severe hypercortisolism and hypokalaemic metabolic alkalosis. While there may be profound muscle weakness, weight loss, oedema, hyperglycaemia and hypertension, the other classical features of Cushing’s syndrome namely truncal obesity, moon facies and cutaneous striae are often absent. This is presumed to relate to the rapidity of onset. These patients have a particularly poor prognosis since they may rapidly succumb to the effects of the gross metabolic derangement.

Plasma ACTH

Plasma ACTH levels are often, although not invariably, grossly elevated.

ACTH secretion from "ectopic” tumours may be episodic resulting in marked variation in plasma levels throughout the day (Crosby - unpublished observation). Studies investigating the potential role of ACTH as a tumour marker for SCLC report that 20-30% of untreated patients have raised plasma

ACTH levels. There does not appear to be a good correlation with stage of disease or clinical course (Hansen et al 1980, Ratcliffe et al 1982). ACTH does not appear to be particularly useful in monitoring response to therapy or for detecting relapse. Only in the few patients with elevated pre­ treatment ACTH is a reasonable correlation found between plasma ACTH and clinical response (Havemann et al 1985).

Immunocytochemistry

There is little published data on immunocytochemical detection of ACTH and other POMC derived peptides in SCLC, partly because the majority of

55 patients are treated with chemotherapy rather than surgery and therefore it is difficult to obtain suitable tissue. This type of study is critically dependent on antiserum specificity and it has also been suggested that tumours causing a clinical syndrome may have a higher rate of secretion of peptide hormones with intracellular levels too low to detect by immunocytochemistry (Coates et al 1986).

Tumour Extracts

Published data on tumour extracts give variable estimates of the prevalence of ACTH related peptides. A study of tumour extracts from patients with

SCLC and carcinoid tumours, in whom there was no clinical or biochemical evidence of the ectopic ACTH syndrome, suggested that more than 90% of these tumours contain significant levels of ACTH (Bloomfield et al 1977) but the number of patients studied was small. In contrast, the much larger study of 157 primary lung tumours showed that 20% of SCLC tumour extracts contained raised ACTH levels in comparison with normal lung, but NSCLC and carcinoid tumours were negative (Yamaguchi et al 1985).

The true prevalence of production of ACTH in SCLC is therefore hard to ascertain from existing data. Interpretation and comparison of results between different studies is hampered by the problems of RIA sensitivity and specificity.

2) The NATURE OF ACTH RELATED PEPTIDES PRODUCED BY SCLC (AND OTHER NON-

PITUITARY TUMOURS)

The complex way in which ACTH is synthesised in the pituitary requires specific proteases and ordered glycosylation reactions and it seems unlikely that non-pituitary tumours would possess the full complement of enzymes needed for the post-translational processing of POMC. Published data suggests that tumour processing may differ both in terms of sites of proteolytic cleavage and post-translational modification. Some tumours

56 appear to process POMC extensively in a similar fashion to the NIL producing CLIP (Ratcliffe et al 1973) and N-terminal ACTH (Orth et al

1973). N-POC is found in glycosylated and non-glycosylated forms and may be further processed to gamma-MSH (Bertagna et al 1983). Beta-LPH, garoma-

LPH and beta-EP are all found in ectopic ACTH producing tumours but in varying proportions (Tanaka et al 1978). Multiple molecular forms of beta-

EP have been demonstrated in lung and medullary thyroid cancer tumour extracts with acetylated and non-acetylated species isolated from both

(Suda et al 1982). Elevated plasma levels of N-POC have been found in 62% of SCLC patients but also in 10% of patients with squamous carcinoma and

8% of patients with benign lung disease (Gilbert et al 1987).

A striking feature in SCLC is that ACTH is present in high molecular weight

(HMW) forms, both in the tumour tissue (Hirata et al 1983, DeBold et al

1988) and in plasma (Ratter et al 1983). These HMW forms can be shown by chromatographic analysis to correspond to both POMC (31KD) and pro-ACTH

(22KD).It has been estimated that pro-ACTH may constitute the majority

(more than 50%) of the circulating ACTH immunoreactivity in the ectopic

ACTH syndrome (Hale et al 1986). In these studies quantitation of precursors was approximate but recently their conclusions have been corroborated using the direct IRMA for the ACTH precursors (Crosby et al

1988). Thus 5 patients with the ectopic ACTH syndrome due to SCLC had grossly elevated plasma precursor levels (260-2340 pmol/1) which were at least 10 fold higher than normal subjects and 2-3 fold higher than patients with pituitary-dependent Cushing’s disease.In contrast, plasma ACTH levels in these 5 patients were much lower than precursor levels on a molar basis.

It appears that in SCLC, processing of POMC is often incomplete, with a substantial proportion of ACTH remaining in HMW form. This could be due either to a lack of specific proteases or to a switch from a regulated to

57 a constitutive secretory pathway in which precursors are packaged in

secretory vesicles and exported from the cell more or less intact (Gumbiner

and Kelly 1982). Precisely how the gross hypercortisolism characteristic of the ectopic ACTH syndrome is caused is uncertain since the bioactivity of the precursor forms is thought to be low. However, a peptide of low bioactivity secreted at grossly elevated levels with no diurnal variation might stimulate adrenal steroidogenesis to excess or alternatively, the precursor peptides might be cleaved to bioactive moieties at the level of the adrenal gland.

3) THE POMC GENE IN SCLC (AND OTHER NON-PITUITARY TUMOURS)

There are clearly important differences at the level of POMC peptide synthesis between the pituitary and SCLC and recently differences have also been shown at the level of gene expression. Northern blot analysis of POMC mRNA from "ectopic" ACTH producing tumours (SCLC and other sites) shows that the size of the predominant species is 1200bp, i.e. similar to that found in normal pituitary (de Keyzer et al 1985, Clark et al 1989). At least 20% of these tumours also produced a larger transcript (approximately

1400-1500 bases) and using very sensitive detection techniques, this larger species could be detected at low levels in most tumours as well as in normal pituitary. There is evidence that an alternative promoter, upstream of the normal pituitary promoter region, is responsible for this larger message (Clark et al 1989). Translation would still be initiated at the same codon, therefore the peptide product would not differ. Some tumours produce a POMC message with an extended poly-A tail although the functional significance of this is unclear (Clark et al 1989).

POMC mRNA can also be isolated from tumours (including SCLC) where there is no evidence of the "ectopic" ACTH syndrome and in these cases the predominant species is 800-900 bases in length (DeBold et al 1988). This

58 is probably identical to the short transcript found in many normal tissues and which is not thought to produce a secretable peptide. Transcription of this mRNA is thought to be driven by another promoter at the 3’end of intron 2 (see Figure 2).

Thus the heterogeneity found in POMC peptides in non-pituitary tumours is reflected in variation at the level of RNA. How the postulated promoter switches might be caused is completely unknown. It is also not certain whether individual cells can produce more than one type of POMC message.

SCLC tumours which do secrete POMC peptides contain a full length (and/or longer) message yet normal lung tissue contains the short transcript at only very low levels. Either a promoter switch must occur or there must be small numbers of cells in the pulmonary epithelium which produce the full length message. Such cells might differ in their transforming potential.

In this context, it is interesting to note that the POMC gene maps close to N-myc (Hozier et al 1987), an oncogene known to be expressed and amplified in a significant proportion of SCLC tumours.

4) REGULATION OF ACTH IN SCLC

There has been a presumption that ACTH production by tumours was autonomous and unregulated. Clinical data clearly supports the view that glucorticoids and CRF do not exert the same regulatory control as they do in the pituitary. Thus a hallmark of the "ectopic” ACTH syndrome is its failure to respond to negative feedback with high dose dexamethasone or to positive stimulation by CRF. Although normal diurnal variation is lost, secretion of ACTH by SCLC and other "ectopic" tumours is not constant and unvarying.

Wide variations over weeks or months have been reported in some cases of the "ectopic" ACTH syndrome.

Concomitant production of peptides with CRF-like activity,such as vasopressin and GRP/bombesin is well documented (Eagan et al 1974,

59 Yamaguchi et al 1985) although it is not known whether these endogenous peptides influence POMC gene expression in the tumour cells. Tumour secreted CRF and bombesin-like peptides may be bioactive at the level of the pituitary and occasional cases of Cushing’s syndrome have been attributed to the ectopic secretion of both these peptides (Howlett et al

1985; Jessop et al 1986).

There is little published data on the regulation of ACTH secretion by tumour cells in vitro. Monoamines and other agents which raise intracellular cAMP levels can stimulate ACTH release from ectopic tumour tissue fragments (Hirata et al 1975) and fresh dispersed tumour cells

(Hirata et al 1979). In the latter study, high concentrations of calcium ions also stimulated ACTH release from MTC and thymoma cells in vitro.

Essentially nothing is known about the molecular basis for the apparent resistance to glucocorticoids observed in the "ectopic” ACTH syndrome.

However, if ACTH and related peptides have an important functional role in the normal tracheo-bronchial epithelium, particularly during development when pulmonary endocrine cells are relatively abundant, synthesis and secretion must be subject to regulation, although the mechanisms may be different from those operating in the normal adult, anterior pituitary.

Thus, any surviving regulatory controls in SCLC cells would reflect those existing in the normal counterparts of those cells. If this is indeed an endodermal stem cell diverted towards malignancy at a neuroendocrine stage of differentiation (Baylin 1985)), then ACTH secretion in SCLC might be regulated in a manner more akin to the developing foetal lung than the anterior pituitary.

60 V SCLC CELL LINES

1) ESTABLISHMENT AND CHARACTERISATION OF SCLC CELL LINES

SCLC accounts for approximately 25% of all new cases of primary lung cancer and although major advances have been made in developing regimes of

intensive combination chemotherapy which can achieve initial clinical responses in 75% of patients, essentially the median survival rates for treated patients has remained unchanged at at 7-12 months in extensive disease and 12-16 months in patients with limited disease (Leonard 1989).

There is therefore an urgent clinical need to improve our understanding of the biological properties of SCLC tumours and to this end, intensive efforts have been made to establish and characterise SCLC cell lines.

Despite the virulence of the disease in vivo, growth of SCLC cells in culture initially proved to be extremely difficult. Until the 1980’s, very few established cell lines were reported (Oboshi et al 1971, Ohara and

Okamoto 1977, Ellison et al 1976). Early attempts to establish cell lines from clinical samples of SCLC tumours using basal media (RPMI 1640 or

Weymouth’s medium) supplemented with 10-20% foetal calf serum (Pettengill et al 1980, Gazdar et al 1980) produced overall success rates in establishing cell lines of 10-20% (Gazdar et al 1980). At about this time, it was suggested that individual tumour types had specific growth factor requirements (Barnes and Sato 1980) and the use of serum-free medium was advocated. The development of a serum-free, hormone supplemented medium for

SCLC was a major breakthrough (Simms et al 1980). The composition of this medium is shown in Table 3 and has been given the eponymous name HITES. Its use led to dramatically improved success rates (75%) and the establishment of large panels of SCLC cell lines. Although initially developed as a serum-free medium, it was noted that the addition of very small amounts of foetal calf serum (2.5%) improved results still further (Carney et al 1985,

61 TABLE 3. Composition of HITES medium for SCLC

Basal medium: RPMI 1640

Supplements: Hydrocortisone lOnM

Insulin 5ug/ml

Transferrin lOug/ml

Estradiol lOnM

Selenium 30nM

after Simms et al 1980

62 Baillie-Johnson et al 1985), suggesting a requirement for other, as yet undefined growth factors.

Characterisation of these large panels of SCLC cell lines led to a classification based on in vitro growth properties and morphological, biochemical and genetic criteria (Carney et al 1985). Thus SCLC cell lines can be subdivided into two distinct types, named "classic" and "variant" and the basis of this classification is outlined in Table 4. The majority of SCLC cell lines fall into the "classic" category and typically express four biochemical markers, L-dopa decarboxylase, the APUD enzyme (DDC),

Bombesin-like immunoreactivity (BLI), neuron-specific enolase (NSE) and the

BB-isoenzyme of creatine kinase (CK-BB). The classic cell lines have readily identifiable dense core granules on electron microscopy. In contrast, in the "variant" subtype, CK-BB is expressed but DDC and BLI levels are undetectable and NSE levels are significantly lower than in

"classic" cell lines. These four markers also serve to distinguish SCLC from NSCLC cell lines since none of the four are elevated in the latter

(Carney et al 1985).

The "variant" subtype exhibits more aggressive in vitro growth properties with greater colony forming efficiency and shorter doubling times. They are more radio-resistant and more likely to be resistant to chemotherapeutic agents (Carney et al 1983). Interestingly, these in vitro characteristics appear to be reflected in more malignant behaviour in vivo .

SCLC cell lines have distinct cultural morphology in vitro and Carney and colleagues (1985) describe four growth patterns. The majority of cell lines established in RPMI 1640 medium plus HITES grow as floating aggregates of cells in suspension either as tightly packed spheroids (Type I), or loose aggregates of cells (Type II) or in strings or chains of floating cells, sometimes with a "stacked coin" appearance (Type III). A few cell lines

63 TABLE 4. Classification of SCLC cell lines based on in vitro biological properties

CHARACTERISTIC SCLC CELL LINES

Classic Variant

Morphology Floating Floating/Adherent

Cytology SCLC SCLC/Large cell

Cloning efficiency 1-5% 10-30%

Nude mouse Yes Yes tumorigenicity

Radiation Sensitive Resistant sensitivity c-myc amplification Occasional(<10%) Frequent(>80%) (DNA) c-myc expression Occasional (< 10%) Frequent(>80%) (RNA)

Dopa decarboxylase Elevated Absent

Bombesin Elevated Absent

Creatine kinase BB Elevated Elevated

Neuron specific enolase Elevated Low

Adapted from Carney et al 1985

64 grow as an adherent monolayer (Type IV). Culture morphology shows a tendency towards Type I and II growth for ’'classic” cell lines and Type III and IV growth for ’’variant" cell lines, but this is by no means invariable.

The pattern of growth in vitro does however seem to be linked with the method used to establish and culture the cells since the use of Waymouth’s medium plus 20% foetal calf serum resulted in a majority of cell lines growing as adherent monolayers (Pettengill et al 1980).

2) CYTOGENETIC ABNORMALITIES AND ONCOGENE EXPRESSION IN SCLC CELL LINES.

A number of chromosomal abnormalities have been demonstrated both in SCLC cell lines and fresh tumour specimens including rearrangements, deletions, homogeneously staining regions (HSRs) and the presence of double minutes, suggestive of gene amplification (increased copy number) . A specific abnormality has been identified and shown to be almost invariably present in SCLC as well as in a high proportion of NSCLC tumours. This is a deletion in the short arm of chromosome 3, 3p(14-23) (Whang-Peng et al

1982, Kok et al 1987). The precise gene(s) involved in this region are unknown but include the ErbA beta gene, which codes for a thyroid hormone receptor, and another homologous sequence which has yet to be characterised

(Weinberger et al 1986) Recently, an abnormality in the retinoblastoma gene

(Rb), which maps to chromosome 13q(14), has been identified in SCLC but not

NSCLC specimens (Harbour et al 1988). It is possible that these deleted genes act as tumour suppressor or anti-oncogenes.

A number of oncogenes have been found to be over-expressed or amplified in

SCLC cell lines, notably the myc family of nuclear-acting oncogenes, c-myc,

N-myc and L-myc. Amplification of c-myc is characteristic of ’’variant” cell lines and if the gene is transfected into a ’’classic” SCLC cell line, the transfected line takes on the characteristics of a ’’variant” (Johnson et al 1986). Other oncogenes reported in association with SCLC include the

65 Ras family, c-raf-1, c-myb, c-jun and P53 (Minna 1989). The precise roles of these genes and their protein products in the pathogenesis of SCLC is currently unknown.

3) HORMONE PRODUCTION BY SCLC CELL LINES

Secretion of a wide variety of hormones by SCLC cell lines has been reported and these studies are summarised in Table 5. ACTH-related peptides, CT and bombesin are especially prominent and several studies have reported the production of multiple hormones by individual cell lines

(Sorenson et al 1981, Luster et al 1985,Bergh et al 1985). The diversity and relative frequencies of the peptides produced is similar to that found in patient sera. As might be expected, bronchial carcinoids and NSCLC cell lines have also been shown to secrete hormones in vitro (Kirkland and

Ellison 1984, Luster et al 1985). The production of estradiol (E2)

(Sorenson et al 1981) is perhaps surprising, but this study did not investigate whether the SCLC cells were capable of de novo synthesis of steroids.

A number of the well-established SCLC cell lines have been reported to secrete ACTH, although in most cases, detailed analysis of the nature of the ACTH peptides produced has not been done. Culture medium from DMS 79 was however shown to contain HMW forms of ACTH, together with immunoreactivity corresponding to ACTH 1-39, beta-LPH, gamma-LPH and beta-

EP (Bertagna et al 1978). CT is secreted by SCLC cell lines in predominantly HMW forms, in keeping with the findings in patient sera and in contrast to medullary thyroid cancer, which tends to secrete monomeric

CT (Becker et al 1984). SCLC cell lines also secrete calcitonin gene related peptide (CGRP) although the relative proportions of CT and CGRP differ. Thus, DMS 53 produces high levels of CT with low levels of CGRP,

66 TABLE 5. Hormone secretion by SCLC cell lines

Reference Designation Hormones No. of cell lines

Ellison et al 1976 - ACTH 2/3 CT 2/3

Sorenson et al 1981 DMS ACTH 10/13 CT 5/13 Estradiol 13/13 OT-NP 9/13 AVP-NP 1/13 PTH 3/13 Glucagon 3/13 LH 3/13 HCG 1/13 SS 1/13 PRL 1/13 GH 1/13

Pettengill et al 1984 DMS Bombesin 13/13

” NCI ACTH 5/18 CT 11/18 Bombesin 16/18

Luster et al 1985 M R ACTH 1/4 CT 3/4 Bombesin 3/4 Estradiol 2/4 Neurotensin 1/4

Bergh et al 1985 U ACTH 7/7 CT 4/7 Bombesin 7/7 Neurotensin 1/7

Macaulay et al 1988 HC12 IGF-1 1/1

CT:calcitonin, OT-NP:oxytocin-neurophysin, SS:somatostatin. while in DMS 153 the ratio of the two peptides is reversed (Craig et al

1985).

A high incidence of bombesin or bombesin-like immunoreactivty (BLI) was found in both the DMS and NCI series of SCLC cell lines (Pettengill et al

1984). High levels of intracellular BLI are characteristic of the "classic" subtype, as described above. The function of bombesin in SCLC is discussed in the next section.

4) AUTOCRINE GROWTH FACTORS IN SCLC

The elegant studies of Cuttitta and colleagues (1985) demonstrated that bombesin-like peptides can function as autocrine growth factors in SCLC and were the first to ascribe a definitive role to one of the numerous peptides secreted by this neoplasm. The mammalian equivalent of amphibian bombesin is gastrin releasing peptide (GRP). The two peptides share a common heptapeptide carboxy-terminal sequence, the region which is essential for receptor recognition and biological activity . GRP is produced by the human foetal lung and is a potent mitogen both in-vivo and in vitro (Cuttitta et al 1985). Bombesin/GRP is synthesised and secreted by SCLC cells, which express high affinity receptors for bombesin-like peptides (Moody 1984).

These receptors, when activated, stimulate calcium mobilisation and phosphatidyl inositol turnover. Exogenous bombesin/GRP stimulates the clonal growth of SCLC cells in vitro and the growth of SCLC xenografts in vivo. Monoclonal antibodies to bombesin markedly inhibited the growth of

SCLC cells both in vivo and in vitro (Cuttitta et al 1985). These findings form the basis of a phase I/I I clinical trial using a MAb to the biologically active region of GRP (Mulshine et al 1988).

There is some also evidence that insulin-like growth factor I (IGF-I) be an autocrine growth factor for SCLC (Nakanishi et al 1988, Macauley et al

1988). Attempts to demonstrate a similar role for other peptides secreted

68 by SCLC have produced conflicting results. ACTH has been claimed to

stimulate growth of SCLC cells in a soft agar clonogenic assay (Luster et al 1985) whereas a similar system used to investigate a range of peptides

found ACTH to be inneffective (Bepler et al 1987). The possibility that other peptides derived from POMC might be mitogenic to SCLC, either alone or in combination, awaits exploration.

5) ADVANTAGES AND LIMITATIONS OF SCLC CELL LINES

The availability of large numbers of well characterised SCLC cell lines has greatly facilitated the systematic investigation of cytogenetic changes and oncogene expression, and has been crucial to the study of the production and response to growth factors. These cell lines have enabled the development of in vitro drug and radiation sensitivity testing assays for lung cancer cells which may have direct clinical application in the design of new therapeutic regimes for individual patients. Potentially they should provide a good in vitro model for the production and regulation of tumour hormones. It would appear that the phenotypic characteristics of

SCLC cell lines are fairly stable in long term culture and xenotransplanted cells result in tumours which faithfully reflect the histological and ultrastructural features of the parent tumour (Pettengill et al 1980a).

However, there are disadvantages to the use of any in vitro model of human disease. The very act of attempting to culture tumour cells in the laboratory applies a strong selective pressure and those that survive may not be representative. For example, the vast majority of SCLC cell lines have been established from metastatic sites rather than the primary tumour and differences may exist between these two populations of tumour cells.

Despite improved techniques and success rates in recent years, establishment and maintenance of SCLC cell lines remains difficult and

69 requires carefully optimised conditions, although in vivo the disease is

highly virulent.

The classification of cell lines into '’classic" and "variant" subtypes has

highlighted not only the features in common but also the phenotypic heterogeneity which is observed. Therefore observations on a single SCLC cell line cannot necessarily be extrapolated to others. The prevalence of

a particular characteristic in vitro may only reflect the ease with which

that subset of tumour cells can be cultured and may be quite different from the prevalence in vivo. Culture of tumour cells might induce or change the

level of expression of certain genes. Changes in culture morphology, chromosome number and levels of hormone production have been reported in the DMS cell lines (Pettengill et al 1985) despite careful standardisation of conditions.

Finally, study of the behaviour of isolated, pure populations of tumour cells offers a simple, relatively stable and reproducible experimental system. It must nevertheless be borne in mind that tumours in patients are subject to many complex influences, including stromal cells, innervation, vascularisation and the host immune system.

70 VI SUMMARY AND AIMS OF THIS THESIS

In the past decade it has been confirmed that adrenocorticotrophin (ACTH)

is synthesised in the normal pituitary as part of a large peptide

precursor, pro-opiomelanocortin (POMC). The amino acid sequence and

mechanism of post-translational processing of POMC has been elucidated and

the POMC gene has been isolated, sequenced and cloned. At the same time,

corticotrophin releasing factor (CRF) has been isolated from the

hypothalamus, sequenced and identified as the major physiological

stimulator of ACTH synthesis and secretion. Much has also been learned

about the actions of glucocorticoids in feedback inhibition of ACTH. Thus

our understanding of normal pituitary ACTH production and regulation has

rapidly advanced.

ACTH is also produced by a multiplicity of tumours and advances in molecular biology have demonstrated expression of the POMC gene in many normal tissues. However, little is known of the function and regulation of

ACTH in these extra-pituitary sites. Small cell lung cancer is perhaps the

most clinically important non-pituitary site of ACTH synthesis in man, but

there are obvious difficulties in attempting to study the phenomenon in

patients. SCLC cell lines are now widely used in the study of tumour biology and have already allowed the identification of autocrine growth

factors. Potentially they offer a suitable in vitro model for the study of

ACTH production by SCLC.

The aims of this thesis are therefore:

1. To establish a panel of SCLC cell lines which express POMC (using both primary cultures and previously characterised cell lines).

2. To establish culture conditions and experimental protocols which optimise basal peptide secretion.

71 3. To characterise and quantitate ACTH and its precursor peptides using novel, precise immunoradiometric assays (IRMAs) based on monoclonal antibodies (MAbs).

4. To measure expression of POMC RNA using Northern and slot blot analysis.

5. To study regulatory mechanisms in SCLC cells and compare, where possible, to pituitary cell function.

72 MATERIALS AND METHODS

Sources of equipment, reagents and culture media are listed in the

Appendix. All cell culture manipulations were carried out in a Class II laminar flow hood, in accordance with safety regulations governing the handling and disposal of human tissue.

I. PRIMARY CULTURE OF HUMAN LUNG TUMOUR CELLS

Culture Media:

Two growth media were used routinely : i) RTISS(2.5),consisting of RPMI 1640 supplemented with 2.5% foetal calf serum (FCS), lOug/ml human transferrin, 5ug/ml bovine insulin and 3 x 10 -8M sodium selenite. This medium was used for all SCLC samples. ii) RS(10), consisting of RPMI 1640 supplemented with 10% FCS.

If cell yields were sufficiently high, tumour samples were split between flasks containing RS(10) and RTISS(2.5), particularly if the histological diagnosis was unknown.

Both media were supplemented with 4mM glutamine, ImM sodium pyruvate and lOmM N-2-hydroxyethyl-piperazine-N’-2-ethanesulphonic acid (HEPES) buffer.

Antibiotics were only added to collection and wash media but were not used in routine growth media.

Further details of medium and supplement preparation, sources of reagents and tissue culture equipment are given in the Appendix.

Tissue Samples:

Four types of tumour sample were obtained from patients suspected of harbouring lung tumours, namely: i) Bronchoscopic biopsies ii) Surgically resected solid tumours iii) Bone marrow aspirates iv) Pleural effusions

73 Approval for the collection of patient samples for tissue culture was obtained from the hospital Ethical Committee. Collection and dispersion

techniques adopted for each sample type are described separately.

i) Bronchoscopic Biopsies

Collection: An additional biopsy was taken for the purpose of cell culture at the time of diagnostic bronchoscopy, thus samples were obtained from unselected patients with suspected lung tumours. The biopsy was collected

into a sterile universal containing 10ml of RS(10) + gentamicin 0.2mg/ml).

Dispersion: These samples consisted of several small fragments of tissue

(<5mm diameter) and were therefore treated as explants. The specimen was

2 + centrifuged at lOOOrpm for 5min at room temperature (RT), washed x2 in Ca

and Mg2f free phosphate buffered saline (PBS), then resuspended in 4ml of

RTISS(2.5) and transfered to a 25cm tissue culture flask.

ii) Surgically resected solid tumours

Collection: Lung tumour tissue obtained at the time of thoracotomy was

placed in a sterile pot containing RS(10) + gentamicin 0.2mg/ml).

Mechanical Dispersion: Tumour tissue was transfered to a sterile Petri dish

and dissected as finely as possible. The cells which dispersed into the medium by this means were strained through a sterile steel wire gauze and washed through with fresh culture medium. These cells were then centrifuged

at lOOOrpm for 5min at RT, washed x2 in PBS or fresh medium, resuspended

in RTISS(2.5) and transferred to tissue culture flasks. Any remaining

pieces of tumour tissue were either treated as explants, as described above

or subjected to enzymatic dispersion.

Enzymatic Dispersion: Dissected fragments of tumour tissue were resuspended

in fresh serum free medium containing 0.2% collagenase and tranferred to

a sterile dispersion apparatus. The apparatus consisted of a sterile 50ml

pot fitted with a rotating Teflon paddle. The tissue was mechanically

74 agitated in the enzyme solution for 15min at 37°C. The supernatant containing released tumour cells was then removed and the reaction stopped by addition of an equal volume of ice cold RS(10). The cells were then washed and resuspended in a suitable volume of RTISS(2.5) and transferred to tissue culture flasks. The procedure was repeated until no further viable tumour cells were obtained. iii) Bone marrow aspirates;

Collection: Bone marrow samples were obtained from untreated patients with pathologically confirmed SCLC, the investigation being carried out as part of the routine staging procedure. l-3ml bone marrow aspirate was collected from the sternum into a syringe containing 500IU preservative-free heparin

(Leo Laboratories) and mixed gently to prevent clotting. The syringe was then capped aseptically.

Recovery of Tumour Cells: Clotted samples were discarded. Bone marrow fluid was transferred to a sterile universal and diluted 1:10 in RTISS(2.5). 15ml of diluted sample was carefully layered onto 10ml of Lymphocyte Separation

Medium (LSM,Flow Laboratories)in a sterile centrifuge tube. This was then centrifuged at lOOOrpm for 20min at RT. Nucleated cells were harvested from the interface between the LSM and diluent, washed x2 in PBS and resuspended in 4ml RTISS(2.5). The resuspended cells were transferred to a 25cm tissue culture flask. iv) Pleural Effusions:

Collection: Pleural aspiration was performed either as a diagnostic procedure or for symptomatic relief, therefore histological diagnosis was often retrospectively obtained. Pleural fluid was collected into sterile

500ml bottles containing 5000IU preservative free heparin.

Recovery of Tumour Cells: Unless very heavily bloodstained, all the fluid was centrifuged at lOOOrpm for 5min at RT and the cell pellets combined,

75 washed x2 and resuspended in a suitable volume of RTISS(2.5). In the event of heavy contamination with red blood cells, these were removed by following the procedure described for bone marrow aspirates.

Primary cultures: Where the total cell yield was sufficiently high and with a sample of unknown histology, tumour cells were split between flasks containing RTISS(2.5) and RS(10). A high cell density (>5 xlO^ cells/ml) was maintained in early cultures as this appeared to be important for their continuing viability. Following initial harvesting of tumour cells, flasks were gassed with a 5% CC^, 95% air mixture, sealed and incubated at 37°C in a 5% COg atmosphere. Cultures were examined daily for the presence of tumour cells, in particular, the multicellular spheroids typical of SCLC.

Cultures were initially fed every 1-3 weeks, depending on the rapidity of cell proliferation, or more frequently if indicated by a reduction in the pH of the medium between feeds. Feeding was carried out after allowing cell clumps to settle to the bottom of the flask and then replacing approximately two-thirds of the medium. Cells were regularly pipetted to reduce the size of the aggregates and avoid central necrosis. In the event of excessive fibroblast proliferation, serum-free RTISS was used until the overgrowth was inhibited.

Subculture was not performed until flasks were crowded with cell aggregates and the pH of the medium was reduced between feeds. Cells were passaged by splitting the contents of the flask 1:2. No attempt was made to establish frozen stocks in liquid nitrogen until the "cell line” had been stabilised by several passages.

76 II. CULTURE OF ESTABLISHED SCLC CELL LINES

A panel of SCLC cell lines established in four different institutions was obtained through the generous collaboration of the following:

i) Ten "COR" cell lines (COR L24, COR L27, COR L31, COR L32, COR L42, COR

L47, COR L51, COR L88, COR L99, COR L103) were donated by Dr P. Twentyman,

MRC Clinical Oncology and Radiotherapeutics Unit, Cambridge, UK, (Baillie-

Johnson et al 1985).

ii) Four NCI cell lines (NCI H82, NCI H128, NCI H209, NCI N417) were made available for study courtesy of Dr. F. Cuttitta, National Cancer

Institute, Bethesda, Maryland, USA (Carney et al 1985).

iii) HC12 and HX149 were kindly donated by Dr. G. Duchesne, Ludwig

Institute for Cancer Research, Sutton, UK (Duchesne et al 1987).

iv) GLC-1 and GLC-1-M13 were a gift from Dr. M. Brouwer, Netherlands Cancer

Institute, Amsterdam, The Netherlands, and Dr. L. de Leij, University of

Groningen, The Netherlands (de Leij et al 1985)

All the cell lines were derived from patients with pathologically confirmed

SCLC with the exception of COR L32 which was obtained from a patient with a poorly differentiated squamous carcinoma of the lung. The cell line, however, exhibited the characteristics of SCLC (Baillie-Johnson et al

1985). The characteristics of these cell lines are summarised in Table 6.

77 TABLE 6. SCLC cell lines

Cell line Reference Source of Subtype Culture tumour Classic(C) morphology cells Variant(V) Type I-IV

COR L24 Baillie- LN V II COR L27Johnson et al BM (T) V II-III COR L31 1985 BM (A) ? I COR L32 BM (T) C II COR L42 BM (A) C I COR L47 LN C I COR L51 PE C I COR L88 PE C IV COR L99 ? ? I COR L103 ? V II

NCI H82 Carney et al PE III NCI H128 1985 PE I NCI H209 BM (A) II ,o NCI N417 tumour III

HC12 Duchesne et al PE II HX 149 1987 tumour II

GLC-1 de Leij et al PE II GLC-1-M13 1985 sub-clone I of GLC-1

BM (A): bone marrow aspirate, BM (T): bone marrow trephine, LN; lymph node, PE: pleural effusion.

78 Growth and Maintenance of SCLC Cell Lines

Materials, reagents and the formulation of media used are described in the

Appendix. Cell lines were routinely grown in RTISS(2.5) and passaged by diluting the cultures 2-fold. The majority of the cell lines grew as floating aggregates of cells in suspension. Where cells grew adherent to the flask they were released prior to subculture by trypsinisation. Cells were regularly pipetted to prevent the formation of excessively large aggregates. Growth rates of the cell lines were extremely variable therefore frequency of feeding and subculture differed for individual cell lines. Cultures were judged to be "confluent" when crowded with cell aggregates causing the medium to change to an acid pH. This corresponded

e with an approximate cell density of 1 x 10 cells/ml.

79 Ill CULTURE OF PITUITARY CELLS i) Primary culture of human pituitary tumour cells

Collection: Human pituitary tumour tissue was obtained at hypophysectomy from patients with Cushing’s disease or Nelson’s syndrome. Tissue was collected into sterile pots containing RS(10) medium plus gentamicin

0.2mg/ml.

Dispersion: Depending on the size and consistency of the tumour tissue obtained, the pituitary cells were dispersed mechanically as described for

SCLC samples or subjected to enzymatic dispersion. For enzymatic dispersion, tumour tissue was washed in Hank’s buffered saline solution

(HBSS), transferred to a sterile Petri dish and minced finely with crossed scalpels. The pieces were transferred to a sterile, siliconised glass conical flask containing a sterile magnetic "flea” and this was placed in a 37°C water bath on a heated stirrer apparatus. The dispersion mixture contained 0.2% collagenase, 2ug/ml DNAse and 0.1% hyaluronidase in DMEM.

The tumour tissue was gently stirred in this enzyme solution for 30min, triturating the cells in sterile, siliconised glass pasteur pipettes at lOmin intervals, until a fine cell suspension was obtained. Following enzymatic dispersion, cells were washed through sterile fine gauge nylon mesh and gently centrifuged, washed and resuspended in serum supplemented growth medium.

The choice of growth medium was determined according to whether the aim was to attempt to initiate a cell line from the tumour tissue or to perform short term hormone release experiments on the primary cultures. In the former case, RS(10) was additionally supplemented with Tri-iodothyronine

(Tp 5 x 10 ^M) and epidermal growth factor (EGF, 40ng/ml). The pituitary

n cells were cultured in 25cm flasks, incubated at 37°C in a 5% COg atmosphere and observed at regular intervals for signs of proliferation of tumour

80 cells. Feeding was carried out weekly initially or more frequently if necessary. Fibroblast overgrowth was treated by reducing the serum concentration to 2.5% or zero, if required.

Primary cultures destined for short term hormone release experiments were cultured in DMEM supplemented with 10% horse serum, 2.5% FCS, 4mM glutamine, ImM sodium pyruvate, lOOIU/ml penicillin, O.lmg/ml streptomicin and 2.5pg/ml fungizone. At the end of the dispersion procedure, cells were counted and resuspended in growth medium at a density of 5 x 10 cells /ml.

The resulting cell suspension was dispensed as 1ml aliquots in 24 well tissue culture plates. All incubations were at 37°C in a 5% CC^ atmosphere.

The wells were observed daily until the cells had formed an adherent monolayer at which point, they were washed and fed with serum-free DMEM supplemented with 4mM glutamine, ImM sodium pyruvate and antibiotics as above. Following a basal incubation period of 24h, cells were washed x3 in serum-free medium prior to the experimental incubation period, ii) Culture of the mouse corticotroph adenoma cell line. AtT20.

AtT20 cells (CCL 89) were obtained from the American Tissue Culture

Collection (ATCC) and were cultured as previously described (Orth et al

1973) in Ham’s F10 medium supplemented with 15% horse serum, 2.5% FCS, 4mM glutamine and ImM sodium pyruvate. The cells grew in small clusters floating in suspension.

81 IV FREEZING DOWN AND RETRIEVING CELLS FROM LIQUID NITROGEN

Actively proliferating cells at high density (i.e. approaching confluence) were chosen. Cells were spun down at lOOOrpm for 5min and resuspended in

Freezing Mixture to give an approximate cell density of 5 x 10 6 -1 x 10 7 cells/ml. 0.9ml of the cell suspension was added to pre-labelled screw top ampoules containing 0.1ml Dimethyl Sulphoxide (DMSO). The ampoules were then quickly transferred to the slow freezer attachment (see Appendix) for

2h. After this time ampoules were transferred to storage containers and immersed in liquid nitrogen.

To retrieve cells, ampoules were quickly thawed in a warm water bath (35°C) and the contents transferred to 10ml of RS(10) at 4°C. The cells were then centrifuged at lOOOrpm for 5min and then resuspended in the appropriate culture medium. Some of the SCLC cell lines were initially grown in RS(10) immediately after retrieval from liquid nitrogen as they did not appear to tolerate the low serum concentration in RTISS(2.5). Once re-established in culture, most cell lines could readily be gradually adapted to RTISS(2.5).

82 v. METHODS FOR QUANTITATION OF TUMOUR CELLS i) Cell Counts

Single cell suspensions were prepared by trituration and viable counts obtained using trypan blue exclusion. However.for many of the SCLC cell lines this provides only an approximate guide since it is difficult to disaggregate the cell clumps without a significant effect on their viability. ii) DNA Assay

Cellular DNA was assayed according to a fluorometric method (West et al

1985), the fluorochrome dye being Hoechst 33258. The standard was salmon testis DNA type III.The assay was carried out in glass TCI tubes. 25ul of cell suspension or standard was added to 1.25ml lOmM EDTA, pH 12.3 and incubated for 20min at 37°C. The tubes were then rapidly cooled on ice and the contents neutralised to pH 7.0 with 1M KH^POj, the exact volume required having been determined by titration. 1.6ml Hoechst reagent was added rapidly to samples and standards, ensuring adequate mixing and the fluorescence measured in a Perkin Elmer LS-5 spectrofluorimeter.

Wavelengths for excitation and emission were 350nM and 455nM respectively, with slit width set at lOnm. iii) Protein Assay

Cellular protein was assayed using the Bio-Rad method. 1ml of cell suspension was transferred to an LP3 tube, centrifuged at 2000rpm for 5min and washed x2 in PBS. The supernatant was discarded and if necessary cell pellets could be frozen at this stage for assay at a later date. Cells were solubilised by addition of 1ml 0.1M NaOH and incubation for 40min at 37°C.

Samples were then vortexed thoroughly. Bovine serum albumin, Fraction V in

0.1M NaOH was used as standard. lOOul of sample or standard was added to

2.5ml of Bio-Rad dye reagent (diluted 1:5 in deionised water), mixed

83 thoroughly and incubated for 5min at RT. Optical density was read at 595nm in a spectrophotometer (Pye-Unicam SP6-500 UV).

The DNA assay ,used in initial experiments, was reliable but reagent preparation was complex and the technique was time-consuming. The Bio-Rad protein assay proved to be quicker and simpler to perform and was therefore adopted in later experiments. Comparison of results obtained in the two assays gave good correlation.

84 V I. EXTRACTION OF CELL PELLETS FOR ACTH AND PRECURSORS

Cell pellets were immmediately frozen on dry ice and stored at -20°C until

extraction. Pellets were thawed and 1ml of 0.01M HC1 added. They were sonicated for 2min, centrifuged at 3000g for lOmin at 4°C and the supernatants frozen and stored at -70°C before ACTH and precursor assay.

These procedures were optimised for preservation of ACTH by extraction of exogenous ACTH added to non-secreting cell lines, and by extraction of endogenous ACTH from rat pituitary cells.

85 VII. IMMUNORADIOMETRIC ASSAYS FOR ACTH AND ACTH PRECURSOR PEPTIDES

ACTH and the ACTH precursors were measured using two-site IRMAs based on

monoclonal antibodies. The binding sites of the MAbs employed in the two

assays are illustrated schematically in Figure 9 and an outline protocol

is shown in Figure 10.

i) ACTH

The ACTH IRMA was developed and optimised as previously described (White et al 1987) and employs two MAbs: MAb 1A12 (specific for ACTH 10-18) was radioiodinated and MAb 2A3 (specific for ACTH 25-39) was coupled to

Sephacryl S300 as solid phase. Human ACTH standards (NIBSC Code 74/555) were prepared at concentrations between 1 and 1110pmol/l (4.5-5000ng/l).

The assay sensitivity (2.5 x standard deviation (s.d.) at zero ACTH) is

0.8pmol/l (3.5ng/l) and the within and between batch coefficients of variation (c.v.) are <10% at 5-1110pmol/l (22-5000ng/l) and 6-1110pmol/l

(27-5000ng/l) respectively. The assay measures ACTH 1-39 and there is no interference from fragments of ACTH such as a-MSH, ACTH 18-39 and ACTH 1-

24. Using this combination of ACTH MAbs, the assay retains some cross reactivity with the ACTH precursors (POMC <1% and pro-ACTH <10%). ii) ACTH PRECURSORS

The development of the precursor assay was described in detail by Crosby and colleagues (1988). MAb 1A12 (specific for ACTH 10-18) was radioiodinated and MAb 1C11 (specific for gamma-MSH) was coupled to

Sephacryl S300 as solid phase. A partially purified POMC standard was prepared from growth medium of a cultured human pituitary tumour by

Sephadex G75 chromatography under acid dissociating conditions. Further details of the method used are given below. Standards were prepared at concentrations of 2.6-2600pmol/l. The assay sensitivity (2.5 x s.d. at zero

POMC) is 2.6pmol/l and the within and between batch CVs are <10% between

86 a) PRECURSOR IR M A

■ < ------POMC >•

ACTH

-<------P ro -A C T H ------>-

ACTH

1C11 1A12

b) ACTH IRM A

ACTH

1A12 2 A3

Figure 9 Binding sites of the MAbs used in the IRMAs for ACTH- related peptides. In the recursor IRMA (a),the *“i- MAb-lA12 recognises ACTH 10-18 and the solid phase- linked MAb-lCll recognises the gamma-MSH sequence. In the ACTH IRMA(b), the 1Z5I-MAb-lA12 recognises ACTH 10-18 and the solid phase-linked MAb-2A3 recognises ACTH 25-39.

87 STANDARD OR SAMPLE (lOOyL)

+

125I - MAb (lOOyl) I Incubate overnight at 4°C + MAb - coupled to solid phase Sephacryl-S-300 (lOOyl) 1 Shake on sucroagitator for 2h at RT. I Separate by Sucrose layering (2 pass system) I COUNT (LKB multiwell gamma counter Model 1260

Figure 10 Outline protocol for the ACTH and precursor IRMA’s

88 20-2600pmol/l and 37-2600pmol/l respectively. Using the POMC standard, pro-

ACTH cross-reacts 100% in this assay, thus it fully quantitates both ACTH

precursor peptides but does not distinguish between them. Other POMC derived peptides e.g. ACTH, beta-LPH and N-POC do not cross react at levels up to 1000pmol/l.

PREPARATION OF ACTH PRECURSORS FOR STANDARDISATION

i) Corticotroph adenoma cell culture

A pituitary tumour (pituitary tumour No. 2, see Table 8) was obtained at hypophysectomy from a 42 year old man with Cushing’s disease and cultured according to the methods described above. The tumour tissue was obtained as part of a collaborative study with Dr.A. Clark and Professor L. Rees,

St Bartholomew’s Hospital, London, UK. Histologically the tumour was an invasive corticotroph adenoma. The cells were cultured in RS(10) plus 80

KlU/ml aprotinin and 0.2mg/ml gentamicin. Medium changes were carried out weekly and supernatant medium removed from the cells was immediately flash

frozen and stored at -70°C before hormone assay or chromatography. The ACTH concentration in the cell culture medium after 7 weeks was 100,000ng/l.

ii) Sephadex G75 Chromatography

The chromatographic separation system developed by Ratter et al (1980) employs acid-dissociating conditions which prevent the aggregation of small peptides or binding of peptides to larger carrier proteins. It also gives good recovery and is therefore optimal for the identification and purification of the HMW precursors of ACTH. A Sephadex G-75 superfine column (1.5 x 90cm) was calibrated with 125 I-glyceraldehyde-3-phosphate dehydrogenase (36KD), ^I-prolactin (25KD), ^I-alpha-lactalbumin (14.2KD) and ^1 -ACTH 1-39 (4.5KD). Growth medium (2ml) from the tumour cells was acidified with formic acid (1% final concentration) and applied to the column at 5ml/h. Fractions (4ml) were eluted with 1% formic acid containing

89 Polypep (lmg/ml; added to reduce non-specific binding of POMC-derived peptides and improve recovery) and were immediately neutralised with 5M

NaOH (approximately 85ul) and 0.5M sodium phosphate buffer (400ul) pH 7.5 and flash frozen before assay. ACTH precursors in the fractions were detected using the ACTH IRMA, iii) Quantitation of the POMC standard

The reference standard was prepared from the peak eluting from the column which corresponded to POMC (31KD). The molar concentration of POMC in the peak fraction was 26nmol/l as determined by the fluorometric assay for N- terminal tryptophan (Hakanson and Sundler 1971) using N-POC as standard.

90 VIII. CORTISOL RADIOIMMUNOASSAY

Cortisol was assayed in culture media using incubation at low pH to

eliminate interference by corticosteroid binding proteins present in serum.

The cortisol antiserum was preincubated overnight at 4©C with donkey anti­

sheep antiserum and normal sheep serum (all supplied by the Scottish

Antibody Production Unit) diluted in 0.13M sodium chloride/phosphate buffer pH 4.0. This reagent (700pl) was incubated with cortisol-3-carboxy- 125 methyloxime I histamine (Amersham International) as tracer (250pl) and standard (Sigma, H4001) or sample (50pl) for 90 mins at 37©C before centrifugation, separation of the supernatant and counting of the precipitate (LKB multiwell gamma counter, model 1260).

91 IX. ANALYSIS OF RNA

Analysis of RNA was carried out by Dr.A. Clark and Dr. P. Lavender (St.

Bartholomew’s Hospital) and Dr. W.E. Farrell (Dept of Clinical

Biochemistry, Hope Hospital) as part of a collaborative study. Pellets of

SCLC cells were stored at -70°C until used for the preparation of total cellular RNA by the method of Chirgwin et al (1979). Cells were homogenised from the frozen state in 4M guanidine isothiocyanate, layered onto a cushion of 5.7M caesium chloride and centrifuged for 18h at 50,000g. The precipitate of RNA was resuspended in Tris-HCl (lOmM; pH 7.5) containing

EDTA (ImM) and extracted with an equal volume of chloroform:isobutanol

(4:1, v/v) before precipitation with ethanol. RNA was quantified by measurement of absorbance at 260nm. Aliquots of RNA (2.5-10ug) in 5 x SSC

(1 x SSC is 0.15M sodium chloride and 0.015M sodium citrate; pH7.0) were applied to pre-wet nitrocellulose paper (Schleicher and Schuell,

Dassell,FRG) using a slot-blot apparatus (GIBCO). The filter was then baked at 80°C under vacuum for 2h before pre-hybridisation and hybridisation as described by Clark et al (1989). The DNA probes used are described below.

Filters were washed (2 x 30min in 2 x SSC containing 0.2% (w/v) sodium dodecyl sulphate (SDS), followed by 2 x 30min in 0.2% SDS at 50°C) and subjected to autoradiography with Kodak XAR-5 film and intensifying screens at -70°C. Films were developed after intervals ranging from 4h to 5 days, in order to obtain a suitably contrasted image with each of the probes used. Slot intensity was quantified using a Parry DT1405 densitometer and all results were expressed relative to the beta-actin RNA level, as a control for the quantity and quality of RNA applied to the blot.

For Northern blot analysis (Thomas, 1980), poly(A)+ RNA was selected by affinity chromatography of total RNA with oligo(dT) Sepharose (Uniscience

Ltd) eluting the bound fraction in Tris-HCL (lOmM; pH 7.5) containing EDTA

92 (ImM). The poly(A)+ RNA (2ug) produced was separated by electrophoresis on

a 1.4% agarose, 6% formaldehyde gel. This RNA was then transferred to

nitrocellulose paper by Northern blotting for 18h, and the filter was then

baked, pre-hybridised and hybridised as before. After hybridisation for 18h

the filter was washed as before, except that the latter washes were at

55°C. The filter was exposed to film for 14 days.

Probes

The POMC probe used for slot blot analysis was the llOObp insert from the plasmid pSNAC20 (Nakanishi et al 1979) which contains the full length bovine POMC cDNA (a gift from Professor S. Numa, Kyoto University, Kyoto,

Japan). For Northern blot analysis, an 800 bp probe derived from exon 3 of

the cloned human POMC gene was used (a gift from Professor S. Cohen,

Stanford University, Stanford, Ca. , U.S.A.). The human glucocorticoid

receptor was a 1.9Kb Ava 1 fragment of the cloned cDNA (Hollenberg et al

1985), and the Tyrosine Aminotransferase (TAT) probe was the 1.7 Kb insert of the plasmid hcTAT2-16 (Nat et al 1986). The beta actin probe was derived

from the cloned chick beta-actin cDNA. All probes were labelled with

[ 32P]dCTP using oligonucleotide primed synthesis (Feinberg and Vogelstein

1983) and were purified by passage over Sephadex G50. Specific activity was usually 10 9 cpm/ug and approximately 10 7 cpm were used per hybridisation.

93 X GLUCOCORTICOID BINDING ASSAY

Cells used in this study were cultured by the author at Hope Hospital.

Dexamethasone binding assays were carried out by Dr. W. E. Farrell and Dr.

Adrian Clark with help and advice from Professor Gavin Vinson (St.

Bartholomew’s Hospital).

Cells grown to mid-log phase, were washed once and resuspended in fresh unsupplemented medium, and were aliquoted amongst several 1.5ml tubes to give approximately 5 x 10 5 cells per tube. 100 fmoles of 3 H-dexamethasone

(specific activity 100 Ci/mmol) was added to these tubes together with various concentrations of unlabelled dexamethasone (ranging from 10’^ - 10

5M) in a total volume of 300ul, each concentration being performed in duplicate. The cells were incubated for 60min at 37°C in an atmosphere of

5% COg. Cells were then centrifuged at 6500rpm for 2min at 4°C and washed twice in cold PBS. Cell pellets were lysed by incubation for 20min at 2°C in 500ul of 1.5mM magnesium chloride, lOmM Tris hydrochloride (pH 7.4).

Complete lysis was checked microscopically. Nuclei were separated by centrifugation at 12,000g for 5min. Supernatant was carefully removed and nuclear and cytoplasmic fractions were counted in 500ul of scintillant

(Cocktail G) for lOmin each. Displacement curves and Scatchard plots were derived from these results. Specificity of the receptor was confirmed by competing bound 3 H-dexamethasone with 10 -5 M progesterone and aldosterone.

94 XI STATISTICAL ANALYSIS

The data were computed using the "Minitab" statistical package with

reference to techniques described in Brown and Swanson Beck 1990.

a) Probability plot coefficient test for normality

In order to validate the expression of results as means and standard deviations, and then to perform analysis of variance or alternative parametric statistical procedures, it was necessary to check that the data conformed to a normal distribution. The technique used (Filliben 1975) is based on the assumptions of analysis of variance where it is assumed that values are distributed around group means with the same overall standard deviation.

Since absolute values for the end-point (e.g. measurement of peptide concentration) differed between experiments, the following procedure was adopted to standardise values before constructing normal plots.

Each experiment generated groups of data (controls, test groups 1-x,) for which group means and an overall standard deviation were calculated. From each value, the appropriate group mean was subtracted to give residuals and each residual was then divided by the overall standard deviation to give standardised residuals. The residuals allow an assessment of the variance about the group means, whilst calculation of standardised residuals corrects for differences in absolute values obtained in different experiments. The procedure therefore allows the combination of multiple experiments, either repeat experiments on a single cell line or experiments conducted on several cell lines. Normal scores (i.e. expected values of order statistics from a normal distribution) were calculated on standardised residuals and normal probability plots (standardised residual values versus expected residual values) constructed. The correlation coefficient, r, was calculated and compared to a table of critical values

95 (Filliben, 1975). Where sufficiently high correlation was obtained to accept the hypothesis of normality, parametric tests were applied to the data. Normal probability plots are displayed graphically in Appendix III, together with a full worked example of the procedure and a summary table of calculated correlation coefficients, b) Parametric tests

Analysis of variance (one-way)

Analysis of variance was performed on all experiments comparing more than two groups (e.g. controls, cells + secretagogue dose 1, cells + secretagogue dose 2 etc). This technique is appropriate for such multiple comparison experiments. It assumes that within each population, observed values are normally distributed and that the variability within each group is the same, i.e. that values are distributed around group means with the same overall standard deviation. The method calculates the variance ratio

(F ratio):

F ratio = variance of the set of sample means

’pooled’ within group variance and a p value for the F ratio, i.e. the probability that the observed variance ratio could occur by chance. A high F ratio with a corresponding low p value (p <0.05) indicates the strength of the evidence that differences exist between the group means but does not show which groups differ. Follow up analysis to answer this question was performed by calculating 95% confidence intervals for the group means using the equation:

95% confidence interval (C.I.) = mean ± (^crror d f x S.E.M.) where standard error of the mean (S.E.M.) = pooled standard deviation

f n

96 Thus the standard error is derived from information on variability obtained

from all samples. If the 95% confidence intervals of any two groups do not

overlap, then it can be concluded that there is a significant difference

between the means.

ii) Linear Regression Analysis

Linear regression was used to analyse data from the experiment where COR

L24 cells were incubated for different time intervals, both with and

without the addition of db-cAMP (Figure 73, page 192). The regression

equation was calculated for the two groups of data as follows:

y = A + Bx

where x = time,

y = ACTH precursors/mg protein,

A = intercept on the y axis,

B = slope of the fitted line.

The linearity of the relationship between the x and y variables was checked

in addition to assessment of the distribution of the data.

iii) Analysis of covariance

Analysis of covariance was performed on the data from the experiment where

COR L24 cells were incubated with dibutyryl cyclic AMP (db-cAMP) either

alone or in combination with iso-butyl-methyl-xanthine (IBMX) (Figure 74, page 195). This technique allows an assessment of the difference between

the observed mean values of y (i.e. ACTH precursors/mg protein) for a given value of x (i.e. concentration of db-cAMP) in cells treated with and without IBMX.

iv) Two-tailed unpaired t tests

Two-tailed unpaired t tests were applied in experiments involving comparison of a control group with a single test group. The Minitab statistical package does not assume equal variance in the two samples.

97 95% confidence intervals for the mean difference were calculated in addition to p values and are given in Appendix IV. A p value <0.05 was considered significant. In the graphical display of these experiments, 95% confidence intervals shown for the group means were calculated using the pooled standard deviation for all values as described in Section XI b) i) above. v) Outliers

In constructing normal probability plots a single obvious outlier was identified in one set of data and two outliers in a second set of data.

Reanalysis following exclusion of these values indicated that the remaining data conformed to a normal distribution. However, in subsequent tests, the data were analysed with and without the outliers to check that they caused no material effect on the conclusions. The outliers are noted in Appendix

II.

98 RESULTS

I. PRIMARY CULTURE OF LUNG AND PITUITARY TUMOURS

i) Lung Tumours

To study ACTH synthesis by human non-pituitary tumours, an in vitro model

was required and SCLC cell lines provided a suitable choice because of the

frequency of the association with ACTH production. To develop a panel of

cell lines, samples of bone marrow, bronchoscopic biopsies, pleural

effusions and surgically resected lung tumours were collected over an

18month period from patients with suspected or confirmed lung cancer.

161 lung tumour samples were cultured of which 30(19%) were confirmed as

SCLC histologically (Table 7). From this small number of SCLC samples,

3 proliferated in culture and were successfully frozen down and recovered

from liquid nitrogen. Unfortunately none of these secreted ACTH and were therefore considered unsuitable for further characterisation or experiments

in this project.

In addition, one bone marrow sample, in which the corresponding smear was negative for tumour cell infiltration, appeared to show active proliferation of cells in culture. However, these cells were subsequently shown to be B lymphoblastoid in type, as evidenced by the presence of cytoplasmic membrane projections, known as uropods. These could be observed on the surface of the cells by microscopical inspection of the culture

(Figure 11). They were also positive for surface immunoglobulin (courtesy of Dr. T. Roberts, Christie Hospital, Manchester). Emergence of B lymphoblastoid cell lines in cultures of this type is well recognised and is due to in vitro transformation by Epstein Barr virus.

One cell line was established from a bronchoscopic biopsy of an adenocarcinoma and grew as an adherent monolayer with a cultural morphology quite distinct from the SCLC cells (Figure 12). Another cell line was

99 TABLE 7. Priaary culture of lung ti— our samples

Sample type Total no. No.(%) Cell lines cultured SCLC established

Wo. Type

Bronchoscopic 45 16(36%) Adeno­ biopsy carcinoma

Lung tumour 35 5(14%) SCLC

Bone marrow 38 5(13%)* B- lympho- blastoid

Pleural effusion 42 3(7%) 2 i) SCLC ii) Adeno­ carcinoma

Lymph node 1 1(100%)

TOTAL 161 30(19%)

* All bone marrow aspirates were from SCLC patients. This figure refers to the number of samples in which tumour cells were observed microscopically in the corresponding smears.

100 Figure 11 B lymphoblastoid transformation occurring in cultured bone marrow cells from a patient with SCLC. The arrow indicates uropods.

101 Figure 12 Adenocarcinoma cell line established from a bronchoscopic biopsy.

102 established from a pleural effusion developing in a patient thought to be suffering from malignancy. However the primary tumour was not identifiable

in this patient.

Continuous proliferation of cells was initially observed in cultures of three pulmonary carcinoid tumours. Early cultures demonstrated high levels of ACTH in the medium (range 282-369 ng/1; 62.7-82 pmol/1) but it did not prove possible to establish cell lines from these tumour cells. The appearance of these and other tumour cells in culture is illustrated in

Figures 13 and 14.

A number of factors account for the relatively modest success rate in this attempt to establish SCLC cell lines including: i) Only 19% of the total number of tumour samples cultured were

SCLC histologically. ii) The quality and quantity of tumour cells obtained was not

optimal in that :

a) most samples were obtained from patients in distant

hospitals and therefore there was frequently a

considerable delay in transit to the laboratory.

b) 16 of the 30 SCLC samples cultured were bronchoscopic

biopsies. These had the potential advantage of providing

cells from the primary tumour, but were extremely small,

contained very few viable cells and were almost

invariably contaminated with bacteria or fungi.

In those cultures where the initial yield of viable tumour cells was apparently adequate, the major problems were overgrowth of fibroblasts and failure of tumour cells to regrow after freezing down in liquid nitrogen.

103 Pulmonary carcinoid tumour cells in culture- Type I (xlO) Type 11 (x2 0)

t *. *

Type III (x4 0) Type IV (x4)

Figure 14 SCLC cells in culture illustrating morphology in vitro. Type I: Tight spheroids in primary cultures from SCLC tumour. Type II: Cell aggregates from cell line GLC-1-M13. Type III: Chains of cells in cell line GLC-1. Type IV: Adherent cells in primary cultures from SCLC tumour.

105 ii) Pituitary Tumours

Seven pituitary tumours were cultured with the dual aims of establishing a human ACTH-secreting pituitary cell line and using primary cultures for short term experiments. The pituitary tumour cells were maintained in culture for periods ranging from 4 weeks to 9 months. Thus a continuous cell line could not be established, probably reflecting the fact that these tumours are adenomas with a very low rate of cell division in vivo and low malignant potential. High levels of ACTH precursors as well as ACTH were observed in the medium of several of the tumours although the relative proportions were variable (Table 8). The medium from pituitary tumour no.

2 was collected and used to obtain standards for the precursor IRMA as described above (Crosby et al 1988).

The appearance of the cells in culture is illustrated in Figure 15.

Although the aggregates of tumour cells did increase in number and size initially, this phase of active growth was not maintained and with time, the cells gradually deteriorated and were eventually outgrown by f ibroblasts.

106 TABLE 8. Level8 of ACTH precursors and ACTH secreted by hiwan pituitary cells.

Pituitary Diagnosis Peptide levels in medium tumour Precursors ACTH Molar pmol/1 pmol/1 Ratio

Pituitary dependent ND 180 Cushing's disease

Pituitary dependent >20,000 > 20,000 Cushing’s disease (large, aggressive tumour)

Pituitary adenoma, 447 15 30:1 clinically silent.

Nelson’s syndrome > 20,000 > 20,000

Pituitary dependent 1430 384 4:1 Cushing’s disease*

Pituitary dependent 2262 114 20:1 Cushing’ disease

Pituitary dependent 791 953 0.8:1 Cushing’s disease

Growth medium was assayed for ACTH and precursors within 1-3 weeks of initiating cultures

* In this patient, no histological evidence of adenoma was found in the anterior pituitary but the most numerous cells were ACTH positive, suggestive of hyperplasia.

107 Human pituitary tumour cells in culture (pituitary tumour No. 2, Table 8).

108 II. SECRETION OF ACTH AND ITS PRECURSORS BY SCLC CELL LINES

The objective of this section was to assess the prevalence of ACTH

secretion in a panel of SCLC cell lines from various sources. SCLC tumours

had for some time been thought to produce high molecular weight forms of

ACTH in vivo. In this study, the development of the new two-site IRMA for

the ACTH precursors (Crosby et al 1988) enabled precise quantitation of

these peptides for the first time. An attempt was made to optimise the

growth conditions for the detection of ACTH related peptides,

i) Detection of ACTH and Precursors

Using the two-site IRMA for ACTH, low levels of ACTH immunoreactivity

(range 1.4-16.7pmol/l) were initially detected in culture medium from the

SCLC cell lines COR L24. COR L27, COR L31, COR L42, COR L103. Markedly higher levels of the precursors of ACTH were detected in these cell lines

(range 62-3640pmol/l) with the precursor assay which directly quantitates

POMC and pro-ACTH without detecting ACTH 1-39. Since POMC and pro-ACTH

cross-react in the ACTH IRMA (<1% and <10% respectively), secretion of

these high levels of ACTH precursors would account for the ACTH

immunoreactivity.

To determine the optimal stage of growth to detect ACTH precursor peptides, levels were measured throughout the growth of these five cell lines over a 14 day period as shown in Figure 16a. ACTH precursor levels increased progressively in the medium as the cells proliferated. This was accompanied by a parallel increase in the peptide levels measured in the ACTH IRMA

(Figure 16b). In four of the five cell lines ACTH precursor levels began to decline by 14 days. The increase in precursors reflected accumulation of peptides in the medium since endogenously secreted ACTH precursors were relatively stable in culture medium. These experiments are described in section II v).

109 < z QOi 80- B o a 60- (/) k . V)o 3w 0)O CL

< 2 G O) E ^ 1.0- E

X o <

14 Days in Culture

Figure 16 Secretion of (a) ACTH precursors and (b) ACTH throughout cell growth in five SCLC cell lines. Growth medium = RTISS (2.5). The cell lines are COR L24 (X-X), COR L27 ( •-• ), COR L31 (0-0), COR L42 ( ) and COR L103 ( A-A ).

110 ii) Prevalence of ACTH Precursor and ACTH secretion.

The prevalence of secretion of both ACTH precursors and ACTH was formally

assessed in all the 18 cell lines (Table 9) (Stewart et al 1989). Cells

were seeded at approximately 0.5 x 10 cells/ml and allowed to proliferate C to high density (1 x 10 cells/ml) without medium change as this was most

likely to yield maximal peptide levels (Figure 16). Ten of the cell lines

(56%) secreted ACTH precursor peptides and low levels of ACTH

immunoreactivity were detected in medium from seven (39%) of the cell lines

using the ACTH IRMA (Table 9). In cell pellet extracts, ACTH precursors and

ACTH were measurable at only low levels from six (33%) and three (17%) of

the cell lines respectively (Table 10).

The levels of ACTH precursors secreted by the SCLC cell lines assessed in

this study, although highly significant, were very much lower than those produced by either the human or the mouse (AtT20) corticotroph cells.

iii) Chromatography of ACTH-related peptides

Sephadex G-75 chromatography of medium from confluent cultures of COR L103

(Figure 17(a)) showed that the precursor IRMA detected two significant

peaks of immunoreactivity, corresponding to 31KD POMC and 22KD pro-ACTH.

Measurement of fractions with the ACTH IRMA confirmed that no significant

ACTH could be detected. The peak in the region of 31KD was not recognised

with the ACTH assay consistent with the known cross-reactivities of POMC

and pro-ACTH in the ACTH IRMA (Stewart et al 1989).

Growth medium from primary cultures of a bronchial carcinoid tumour gave a similar chromatographic profile with significant peaks eluting in the positions of both POMC and pro-ACTH, but no significant immunoreactivity corresponding to ACTH 1-39 (Figure 17 (b)). In other SCLC cell lines ( COR

L24, COR L31 and COR L42), authentic ACTH 1-39 was not detected.

Ill TABLE 9. Levels of ACTH and precursor peptides in supernatant medium fros SCLC lines.

Cell line No. of ACTH ACTH (pmol/1) observations precursors (pmol/1) pmol/1 median (range) median (range)

COR L24 5 361 (224-658) 2.0 (1.4-4.7)

COR L27 7 738 (270-876) 3.8 (1.6-5.1)

COR L31 3 637;1092;1872 11.3;12.8;13.1

COR L42 7 767 (296-3120) 4.2(2.6-8.7)

COR L88 3 88;109;208 < 0.8

COR L99 1 44 < 0.8

COR L103 27 1248 (442-3640) 9.3 (4.5-14.9)

NCI H128 3 26;42;42 1.3 ;1.5;1.6

GLC-1-M13 3 75;104;416 1.1;1.3 ;1.8

HC12 3 81;88;122 < 0.8

ACTH precursors <2.6pmol/l and ACTH <0.8pmol/l were observed in the following cell lines on at least 3 occasions: COR L32 , COR L47, COR L51, NCI H82, NIC H209, NCI N417, GLC-1, HX149. The cell lines were sampled at a peak viable density of 1 x 10 cells/ml. Where the number of observations is n^J, the individual peptide levels are given.

112 TABLE 10 Levels of ACTH and precursor peptides in cell pellet extracts from SCLC lines.

Cell Line Precursors (nmol/mg DNA) ACTH (pmol

COR L24 5.40 ND

COR L27 3.30 0.06

COR L31 0.40 ND

COR L42 4.40 0.018

COR L103 0.10 ND

GLC-1-M13 0.40 0.003

ND = not detected

Cell pellet extracts in the 12 cell lines not listed were negative for both ACTH and precursors.

113 r 300

y n 34 K 24K 4.5 K 250- -250 □ 200- -200 ACTH ACTH Precursors pmol/L pmol/L 150- -150

100- -100

50- -50

0J L-0

20 60 Fraction Number

300-. i-300

250- VQ 34K 24K 4.5K -250 □ 200- ACTH -200 ACTH Precursors pmol/L pmol/L 150- -150

100- -100

50- -5 0

0J

10 20 30 40 50 60 Fraction Number

Figure 17 G-75 chromatography of the cell culture medium from (a) small cell lung cancer cells (COR L103) and (b) bronchial carcinoid cells (designated BC035). ACTH precursors (open bars) and ACTH (solid bars) were detected in the chromatographic fractions by two-site IRMA. The arrows indicate the elution positions of the void volume (Vo), the total column volume (Vt) and peptides with Mr of 34000, 24000 and 4500.

114 Chromatography of peptides from these cell lines showed that POMC was the major component (data not shown).

These findings were in marked contrast to the chromatography of medium from the human pituitary tumours (Crosby et al 1990). The total amounts of ACTH- related peptides secreted were much higher from the pituitary cells (nmol rather than pmol). Both precursors were present in significant quantities in addition to authentic ACTH 1-39 (Figure 18). iv) ACTH Precursor levels in a 24h Incubation

To study regulation of POMC synthesis, it was important to establish an experimental protocol whereby basal ACTH precursor levels were consistently measurable in the SCLC cell lines. ACTH precursor levels in medium C following a 24h incubation of cells at high density (1-5 x 10 cells /ml) in serum-free unsupplemented medium (RBSAT) were variable (Table 11).

However when SCLC cells were grown to confluence under carefully standardised pre-incubation conditions, washed and re-incubated for 24h at different densities, the precursor levels obtained suggested a fairly constant rate of secretion per cell (Figure 19). The amount of precursors secreted in a 24h incubation was not affected by whether cells were obtained in mid-log growth or at confluence (Table 12). v) Stability of ACTH Precursors in Culture Medium

The stability of ACTH precursors was assessed by separating peptide- containing medium from SCLC cells by centrifugation and re-incubating at

37°C. Aliquots of medium were flash frozen at time 0, 5, 10, 20, 30, 60min and 2, 4, 6, 24, 48, 72h. ACTH precursor concentrations at 72h remained at

62% and 74% of the initial values in RS(10) and RTISS (2.5) respectively

(Figure 20). These results indicated that the precursors were relatively stable under these conditions and that the increase in precursors observed during growth reflects steady accumulation of peptides in the medium.

115 VQ 34 K 24 K 4.5 K 3 0 -. tit it r30

25- -25 □ ACTH 20H _T •20 H Precursors ACTH pmol/L x103 pmol/L xIO3 15H ■15

10- M l r10

5 - -5

0-I L-0 I I I 10 20 30 40 50 60 Fraction Number

lOO-i r-100 VQ 34K 24K 4.5K Vt

80- 111 11 -80 □ ACTH 60- -60 Precursors ACTH pmol/L pmol/L 40- ru -40 yi -20 20- L /

0J LO i i I ' I 1" 10 20 30 40 50 60 Fraction Number

Figure 18 Sephadex G-75 chromatography of the cell culture medium from (a) human pituitary tumour 2, derived from an aggressive corticotroph adenoma (b) pituitary tumour 5, histologically corticotroph hyperplasia (see Table 8). ACTH precursors (open bars) and ACTH (solid bars) were detected in the chromatographic fractions by two-site IRMA. The arrows indicate the elution positions of the void volume (Vo), the total column volume (Vt) and peptides with Mr of 34000, 24000 and 4500.

116 TABLE 11 Precursor levels in medium following 24h incubation of SCLC cells

Cell Line ACTH precursors umol/mg DNA median (range)

COR L24 58.2 (29.1-109.0)

COR L27 10.7 ( 9.4- 33.0)

COR L31 6.8 ( 3.9- 13.0)

COR L42 56.2 (40.6-107.6)

COR L103 11 ( 5.8- 13.5)

25ml SCLC cells were cultured to confluence (approximate density 1 x 10 cells/ml) washed and resuspended in 5ml serum-free medium (RBSAT). Peptide levels were measured following a 24h incubation.

117 COR L103 4000-

*5 3000-

0 2,0 10 25 50 100 Cell Density (x105/ml)

Figure 19 Effect of cell density on ACTH precursor secretion by COR L103 cells incubated for 24h. The four experiments shown were carried out on separate occasions using cells of different passage number. Each data point represents peptide levels in a single bottle of cells.

118 TABLE 12

ACTH Precursors secreted by SCLC cells at different stages of growth.

Cell Line ACTH precursors (pmol/mg DNA)

mid-log confluence

COR L24 22.0 18.0

COR L31 6.6 7.0

COR L42 36.0 41.0

SCLC cells were harvested in mid-log growth or at confluence, washed and resuspended in fresh medium. Precursor levels were measured following 24h incubation.

119 ACTH PRECURSOR STABILITY

ACTH precursors (% time zero cone) 120

100 -r

80 -

60 -

40 J

20 J

0 20 40 60 80 INCUBATION TIME (HRS.)

B RTISS(2.5) + RS10

Figure 20 Stability of ACTH precursors secreted by COR L103 cells. Culture medium (RS (10) and RTISS (2.5)) was separated from cells by centrifugation and reincubated at 37°C for different time periods before being flash frozen for subsequent precursor assay.

120 III. POMC GENE EXPRESSION IN SCLC CELL LINES

Having identified SCLC cell lines secreting significant amounts of ACTH precursor peptides, expression of the POMC gene was examined by analysis of POMC RNA with particular emphasis on variation throughout the growth of the SCLC cells. RNA analysis was performed by Dr. Adrian Clark and Dr.

Paul Lavender (St. Bartholomew’s Hospital). i) Detection of POMC RNA

POMC RNA transcripts were detected by slot-blot analysis in five SCLC cell lines (COR L24, COR L27, COR L31, COR L42 and COR L103) using RNA populations probed with the bovine POMC cDNA probe (Table 13). The abundance of POMC RNA was related to the expression of beta-actin, which was used as a control for RNA quantity and quality. The ratio of POMC RNA to beta-actin RNA ranged from 1.0-4.2. The cells were grown for 7days without medium change and the ACTH precursor levels ranged from 5.3-34.2 pmol/mg protein.

Northern blot analysis demonstrated that POMC RNA from COR L103 cells was approximately 1350 bases in length, larger than the 1250 species found in normal human pituitary (Figure 21). Furthermore, the pituitary signal obtained with 2ug total RNA was considerably more intense than the signal obtained with 2ug poly (A+) RNA from COR L103 cells. ii) POMC expression during growth of SCLC cells

In order to characterise any variation in POMC gene expression during growth, COR L103 cells were seeded at approximately 0.5 x 10^ cells/ml and cultured for 17 days. Figure 22 (a) shows the increase in DNA throughout the growth of these cells, which is paralleled by a rise in the ACTH precursor levels in the medium (Figure 22 (b)), reflecting the accumulation of these relatively stable secreted peptides. At each time point, cells were harvested for slot-blot analysis of POMC RNA, once again using beta- actin as an internal control probe. Figure 23 shows the ratio of POMC to

121 TABLE 13 POMC RNA ACTH precursors and ACTH levels in 5 S C L C cell lines

Cell line Ratio of POMC RNA ACTH precursors ACTH to beta-actin RNA (pmol/mg protein) (pmol/mg protein)

COR L24 2.0 8.5 0.15

COR L27 2.1 5.3 0.09

COR L31 1.2 34.2 0.51

COR L42 4.2 30.9 0.27

COR L103 1.0 22.3 0.30

122 Bases 1353

1150

665

Figure 21 Autoradiograph of a Northern blot showing (a) RNA (2pg poly (A)*RNA) from the SCLC cell line COR L103 and (b) RNA (2pg total RNA) from normal human pituitary cells, both probed with exon 3 of the human POMC gene. The sizes indicated are based on the hybridizable size markers described by Clark et al (1989).

123 Figure 22 Figure

ACTH Precursors ACTH (pmol/mg DNA) (pmol/mg DNA) DNA (mg/l) ACTH precursors and ACTH throughout the growth of the uentn ei. Results shown theare means ofsupernatant media. cellline, CellsCOR L103. were cultured for 17 days fourseparate growth experiments. n TS (.) eim n ppie maue in measured peptides and medium (2.5) RTISS in 0.5- 0.4- 0.3- 0 0.7- 0 0 30- 60- 40- 20 20 50- 70- 30- 10 10- . . . 0- 0J 6 2 0- 1 ------2 6 8 6 4 2 0 as n Culture in Days 124 iue2 Thelevel of POMC gene expression Figure23during the course

POMC : j3-Actin 0.5 — 0.5 i - 5 . 1 1 . 0 * — 0 - 12 3 5 7 6 5 4 3 2 1 0 Resultsshown are the means separate twoof growth thegrowthof L103, curveCOR of expressed the as experiments. ai o PM RA ees o -ci RA levels. RNA B-actin to levels RNA POMC ratio of asi culture in Days 1 ! ! * 125 11 I 14 ! * I 17 0 \ \ beta-actin RNA during growth of COR L103 cells and indicates that the expression of POMC remains at a relatively constant level throughout.

126 IV. REGULATION OF POMC GENE EXPRESSION AND PEPTIDE SECRETION IN SCLC

CELLS

In this section, the effects of the major physiological regulators of POMC, namely glucocorticoids and CRF, were assessed in this model of human non- pi tu it ary ACTH production. In the absence of a human pituitary corticotroph cell line, the responses in SCLC cells were compared with mouse AtT20 cells.

1) GLUCOCORTICOIDS

In these experiments , the aim was to establish whether the expression of

POMC could be inhibited, either at the level of RNA or peptide secretion, by incubation of precursor-secreting SCLC cells in the presence of glucocorticoids. Hydrocortisone was selected since it had the advantage that levels in medium could be monitored by cortisol assay, a) Experimental design

SCLC cell lines

The effect of hydrocortisone was investigated in five SCLC cell lines: COR

L103, COR L24, COR L31, COR L42 and COR L27. SCLC cells were grown to stationary phase and then passaged in fresh medium (RTISS 2.5) to generate bottles each containing 25ml cells at the same stage of growth and at an approximate cell density of 0.5 x 10 cells/ml. Hydrocortisone was added to final concentrations of 0, 500 and 1000nm/l. Three separate bottles of cells were established at each concentration. Cells were cultured for 10 days without medium change, although each dose of hydrocortisone was repeated after five days in order to maintain adequate glucocorticoid levels (see Section (b) below).

To perform a 24h incubation, three additional bottles were established and treated as controls in the lOd incubation (thus n=6 for controls in some lOd experiments). Following the lOd incubation, these cells were passaged

127 in fresh medium to generate nine further bottles of cells to which hydrocortisone was added at final concentrations of 0, 500 and lOOOnm/1, three separate bottles being established at each concentration. The cells were then incubated for 24h. At the end of the lOd or 24h incubation periods, supernatant medium was collected for ACTH precursor assay and an aliquot of cell suspension was reserved for protein assay from all bottles.

Cell pellets were reserved from COR L103 cells (Experiment 1) for RNA studies.

AtT20 Cells

A similar protocol was adopted for AtT20 cells using doses of hydrocortisone ranging from 0-2000nm/l (Experiment 1), 0, 500 and 1000nm/l

(Experiment 2) and 0 and 1000nm/l (Experiments 3, 4 and 5). Cell culture for Experiment 2 was performed by Dr. Adrian Clark and Dr. Paul Lavender

(St. Bartholomew’s Hospital).

Repeat experiments carried out on different occasions using cells of different passage numbers are listed in Table 14, (Page 131), together with the relevant figure numbers, b) Stability of hydrocortisone in medium

Hydrocortisone-supplemented medium was incubated in bottles with and without COR L103 cells for 3, 7, 10, 14 and 17 days. Measured cortisol concentrations remained virtually unchanged in medium alone but declined significantly when incubated in the presence of cells (Figure 24). In the light of this experiment the dose of hydrocortisone was repeated at 5 days in all lOd incubations. Cortisol levels were measured in medium collected at the end of the incubation period in a number of the experiments described above. Results are tabulated in Appendix II. Degradation of cortisol was negligible in COR L42 but occurred to a variable extent for all the other cell lines. Interestingly, hydrocortisone levels were

128 iue2 Saiiy f yrcrioe n utr medium. culture in hydrocortisone of Stability Figure24 Cortisol

(% time zero concentration)100 120 n 120 40- 0 2 80- 60- Hydrocortisone-supplemented medium wasincubated at aa on rpeet te eut n separate a inbottle. result the represents point data 7C ln () n wt CR 13 el () Each 37°C alone (O) and with(A). L103 COR cells - - 0 S t a b i l i t y o f h y d r o c o r t i s o n e 4 D a y s i n c u l t u r e 129 8 12 16 20 reduced to a similar extent in both 24h and lOd incubations, presumably because the cells for 24h incubation were seeded at higher density, c) Peptide secretion

SCLC cell lines

No suppression of ACTH precursor peptides was seen in any of the five SCLC cell lines tested, either following lOd or 24h incubation with hydrocortisone. Analysis of variance was performed on the results from individual cell lines. This indicated there was a statistically significant increase in peptide levels in three instances: i) COR L103 cells following lOd incubation, Experiment 1.

F ratio = 8.77, p = 0.008 (Figure 25, page 132). ii) COR L27 cells following lOd incubation.

F ratio = 10.89, p = 0.004 (Figure 36, page 143). iii) COR L27 cells following 24h incubation.

F ratio = 5.83, p = 0.039 (Figure 42, page 149 ).

Calculation of 95% confidence intervals for each group mean indicated that the differences were due to an apparent increase in peptide secretion with

500nm/l hydrocortisone as compared with the controls (as judged by lack of overlap between the respective confidence intervals). However, peptide levels secreted by the cells treated with 1000nm/l hydrocortisone did not differ significantly from the control cells.

130 TABLE 14

Glucocorticoid experiments on SCLC and AtT20 cell lines

CELL LINE Experiment No. Figure No. Page

SCLC cell lines 10 dav incubation

COR L103 1 25 132 2 26 133 3 27 134 4 28 135 COR L24 1 29 136 2 30 137 COR L31 1 31 138 2 32 139 3 33 140 COR L42 1 34 141 2 35 142 COR L27 1 36 143

24h incubation

COR L103 1 37 144 2 38 145 COR L24 1 39 146 COR L31 1 40 147 COR L42 1 41 148 COR L27 1 42 149

AtT20 cells 10 dav incubation

1 ACTH precursors 43 151 ACTH 44 152 2 ACTH precursors 45 153 ACTH 46 154 3 ACTH precursors 47 155 4 ACTH precursors 48 156 5 ACTH precursors 49 157

24h incubation

1 ACTH precursors 50 158 ACTH 51 159 2 ACTH precursors 52 160 ACTH 53 161

131 iue2 Concentrationsof ACTH precursors Figurein culture25 medium ACTH precursors (pmol/mg protein) 100 -i 0 2 1 - 0 4 - 0 6 - 0 8 20 - - f O L0 cls nuae fr 0 as with days 10 for incubated cells L103 COR of yrcrioe Eachhydrocortisone. datapoint represents peptide h aast Resultof analysis of Fvariance:ratio = 8.77, p= 0.008. thedata set. and 95% confidenceinterval are shown to the right of levels in a single Thegroup bottle meanof cells. 0 O L0: xeiet (0 day) (10 1 Experiment L103: COR H y d r o c o r t i s o n e ( n m / l ) 132 0 0 5 O O 1000 iue2 Concentrationsof ACTH precursors Figurein culture 26 medium ACTH precursors (pmol/mg protein) 1 0 0 -] 0 0 1 5 - 75 0 - 50 5 - 25 f O L0 cls nuae fr 0 as with days 10 for incubated Each cellshydrocortisone.data point represents L103 peptide COR of differencesobserved. and 95% confidence interval are shown to the right of rto 02, = .9, .. o significant no i.e. 0.793, = p 0.24, = ratio F h aast Resulttheof analysisdata set. of variance: levelsin a single Thegroup bottle mean of cells. 0 O L0: xeiet (0 day) (10 2 Experiment L103: COR H y d r o c o r t i s o n e ( n m / l ) 133 500 1000 iue2 Concentrationsof ACTHprecursors Figure in culture 27 medium ACTH precursors (pmol/mg protein) -i 0 0 2 100 - f O L0 cls nuae fr 0 as with days 10 for incubated Each cellsdatahydrocortisone. point represents L103 peptide COR of p= 0.92 i.e. nosignificant differences observed. and 95% confidence interval are shown to the right of the data Result set. of two-tailed, unpaired t test: levels ina single Thegroup bottle mean of cells. O L0: xeiet (0 day) (10 3 Experiment L103: COR H y d r o c o r t i s o n e ( n m / l ) 0 134 1000 iue2 Concentrationsof ACTH precursors Figure in culture28 medium ACTH precursors (pmol/mg protein] 50 n 50 5 - 45 40 f O L0 cls nuae fr 0 as with days 10 for incubated cells L103 COR of the data Result set. of two-tailed, andunpaired 95% confidencet test:interval are shown to the right Eachhydrocortisone. data point representsof peptide p= 0.31 i.e. nosignificant differences observed. levelsin a singleThegroup bottle meanof cells. O L0: xeiet (0 day) (10 4 Experiment L103: COR H y d r o c o r t i s o n e ( n m / l ) 0 135 1000 iue2 Concentrationsof ACTH precursors in cultureFigure 29medium ACTH precursors (pmol/mg protein) 200 0 - 300 5 - 250 100 5 n 350 5 - 150 0 - 50 - - f O L4 el icbtd o 1 dy with days 10 for incubated cells Each datahydrocortisone. point represents L24 peptide COR of and 95% confidenceinterval are shown to the right of rto 11, = .5. .. o significant no i.e. 0.355. differences = observed. p 1.16, = ratio F h aast Resultoftheanalysis data set. of variance: levels in asingle Thegroup bottle mean of cells. 0 O L4 Eprmn 1 1 day) (10 1 Experiment L24: COR H y d r o c o r t i s o n e ( n m / l ) 136 500 1000 iue3 Concentrationsof ACTH precursors inFigure culture 30 medium ACTH precursors (pmol/mg protein) 200 0 - 300 5 - 250 5 -i 350 100 5 - 150 - f O L4 el icbtd o 1 dy with days 10 for incubated cells Each hydrocortisone.data point represents L24 peptide COR of and 95% confidenceinterval are shown to the right of rto 18, = .4. .. o significant no i.e. 0.241. = p 1.82, = ratio F differencesobserved. Resulttheof analysisdata set. of variance: levels in asingle The group bottle mean of cells. 0 O L4 Eprmn 2 1 day) (10 Experiment 2 L24: COR H y d r o c o r t i s o n e ( n m / l ) 137 500 1000 iue3 Concentrationsof ACTH precursors inFigure culture 31 medium ACTH precursors (pmol/mg protein) 20 30 -i 30 10 - - yrcrioe Each datahydrocortisone. point represents peptide f O L1 el icbtd o 1 dy with days 10 for incubated cells L31 COR of rto 01, = .5. .. o significant no i.e. 0.859. differences = observed. p 0.15, = ratioF and 95% confidence interval are shown to the right of h aast Resulttheof analysisdata set. of variance: levels in asingle Thegroup bottle mean of cells. 0 O L1 Eprmn 1 1 day) (10 1 Experiment L31: COR H y d r o c o r t i s o n e ( n m / l ) 138 500 1000 iue3 Concentrationsof ACTH precursors in cultureFigure 32 medium ACTH precursors (pmol/mg protein) 20 -i 0 3 - yrcrioe Each datahydrocortisone. point represents peptide rto 00, = .9. .. o significant no i.e. 0.990. =differences pobserved. 0.01, = ratioF with and days95% confidenceinterval 10 are for shown incubated to the cells right of L31 COR of h aast Resulttheof analysis data set. of variance: levels ina singleThe group bottle mean of cells. 0 O L1 Eprmn 2 1 day) (10 2 Experiment L31: COR H y d r o c o r t i s o n e ( n m / l ) 139 500 1000 iue3 Concentrationsof ACTHprecursors Figurein culture 33 medium ACTH precursors (pmol/mg protein) 20 10 - 30 5 - 25 5 -| 35 5 - 15 0 - 5 - - - -

yrcrioe Eachhydrocortisone. data point represents peptide f O L1 el icbtd o 1 dy with days 10 for incubated cells L31 COR of rto 01, = .4. .. o significant no i.e. 0.840. = p 0.18, = ratio F and 95% confidence interval are shown to the right of differencesobserved. h aast Resulttheof analysisdata set. of variance: levels in a single Thegroup bottle mean of cells.

O O 50 1000 500 0 O L1 Eprmn 3 1 day) (10 3 Experiment L31: COR H y d r o c o r t i s o n e ( n m / l ) 140 O O O O O iue3 Concentrationsof ACTH precursors in cultureFigure 34medium ACTH precursors (pmol/mg protein) 0 - 40 20 -i 0 5 30 - 30 - differencesobserved. f O L2 el icbtd o 1 dy with days 10 for incubated cells L42 COR of rto 16, = .6. .. o significant no i.e. 0.263. = p 1.68, = ratioF and 95% confidence interval are shown to the right Each hydrocortisone. data of representspoint peptide h aast Resulttheof analysis data set.of variance: levels in The asingle group meanbottle of cells. 0 O L2 Eprmn 1 1 day) (10 1 Experiment L42: COR H y d r o c o r t i s o n e ( n m / l ) 141 500 1000 Figure ACTH precursors (pmol/mg protein) 5 Concentrationsof ACTH precursors in culture 35 medium 0 - 40 -i 0 5 0 - 30 20 - f O L2 el icbtd o 1 dy with days 10 for incubated cells Each datahydrocortisone. point represents L42 peptide COR of and 95% confidenceinterval are shown to the right of differencesobserved. rto 38, = .8. .. o significant i.e. no 0.084. = 3.85, p = ratio F h aast Resulttheof analysisdata set.of variance: levels in asingle Thegroup bottle mean of cells. 0 O L2 Eprmn 2 1 day) (10 2 Experiment L42: COR H y d r o c o r t i s o n e ( n m / l ) 142 500 1000 Figure36

ACTH precursors (pmol/mg protein) 0 n 10 f O L7 el icbtd o 1 dy with and 95% days confidence 10interval are for shown incubated to the cells right Each data hydrocortisone. point representsof L27 peptide COR of Concentrationsof ACTH precursors in culture medium Fratio = 10.89, p= 0.004. h aast Resultofthe analysis data set. of variance: levels in a Thesingle group bottle mean ofcells. 0 O L7 Eprmn 1 1 day) (10 1 Experiment L27: COR H y d r o c o r t i s o n e ( n m / l ) 143 500 1000 Figure37 ACTH precursors (pmol/mg protein) 4 0 2 6 8 f O L0 cls nuae fr 4 with 24h for incubated cells L103 COR of Concentrationsof ACTHprecursors in culture medium yrcrioe Eachhydrocortisone. data point represents peptide differencesobserved. Resulttheof analysis data set. of variance: and 95% confidence interval are shown to the right of rto 05, = .2. .. o significant i.e. no 0.624. = p 0.53, = ratio F levelsin a single Thegroup bottle mean of cells. O L0: xeiet (4 incubation) (24h 1 Experiment L103: COR 0 H y d r o c o r t i s o n e ( n m / l ) 144 500 1000 Figure38 ACTH precursors (pmol/mg protein) 4 0 2 8 6 f O L0 cls nuae fr 4 with 24h for incubated cells L103 COR of Concentrationsof ACTH precursors culture in medium differencesobserved. Resulttheof analysisdata set. of variance: Eachhydrocortisone. data representspoint peptide rto 23, = .7. .. o significant no i.e. 0.179. = p 2.33, = ratio F and 95% confidenceinterval are shown to the right of levels in Thea singlegroup mean bottle of cells. O L0: xeiet (4 incubation) (24h 2 Experiment L103: COR 0 H y d r o c o r t i s o n e ( n m / l ) 145 500 1000 iue3 Concentrationsof ACTH precursors inFigure culture 39 medium ACTH precursors (pmol/mg protein) 0 - 40 0 5 0 - 30 20 10 0 - - f O L4 el icbtd o 2h with 24h for incubated cells L24 COR of differencesobserved. Eachhydrocortisone. datapoint represents peptide and 95% confidence interval are shown to the right of rto 03, = .0. .. o significant i.e.no 0.709. = p 0.36, = ratio F Resulttheof analysisdata set. of variance: levels ina single Thegroup bottle mean of cells. O O O L4 2h incubation 24h L24: COR H y d r o c o r t i s o n e ( n m / l ) 146 500 O O O — i— — 1000 O O O COR L31: 24h incubation 3

£ .1 O *-• 2 3 uo o CL

0 0 500 1000 Hydrocortisone (nm/l)

Figure 40 Concentrations of ACTH precursors in culture medium of COR L31 cells incubated for 24h with hydrocortisone. Each data point represents peptide levels in a single bottle of cells. The group mean and 95% confidence interval are shown to the right of the data set. Result of analysis of variance: F ratio = 1.08, p = 0.396. i.e. no significant differences observed.

147 Figure41 ACTH precursors (pmol/mg protein) 1 0 5 f O L2 el icbtd o 2h with 24h for incubated cells L42 COR of differencesobserved. Concentrationsof ACTH precursors in culture medium yrcrioe Each datahydrocortisone. point represents peptide rto 02, = .9. .. o significant i.e.no 0.798. = p 0.23, = ratio F Resultofthe analysis data set.of variance: and 95% confidenceinterval are shown to the right of levels in a Thesingle group meanbottle of cells. 0 O L2 2h incubation 24h L42: COR H y d r o c o r t i s o n e ( n m / l ) 148 500 1000 iue4 Concentrationsof ACTH precursors in cultureFigure medium 42 ACTH precursors (pmol/mg protein) 3 0 2 1 f O L7 el icbtd o 2h with 24h for incubated cells L27 COR of yrcrioe Eachdata hydrocortisone. represents point peptide Fratio = 5.83, p= 0.039. Resultof theanalysis dataset. of variance: and 95% confidence interval are shown to the right of levels in Theagroup single mean bottle of cells. 0 yrcrioe (nm/l) Hydrocortisone O L7 2h incubation 24h L27: COR 149 500 1000 AtT20 cells

The response of AtT20 cells to hydrocortisone was in marked contrast to that of the SCLC cells, with significant suppression of both ACTH precursors and ACTH being produced in a lOd incubation with doses as low as 50nmol/l hydrocortisone in experiment 1 (Figures 43 and 44). Analysis of variance gave F ratio = 36.66, p <0.0001 for ACTH precursors and F ratio = 77.75, p <0.0001 for ACTH. Highly significant suppression of ACTH peptide secretion in response to hydrocortisone was also observed in four further 10 day incubation experiments (Figures 45-49). Results of analysis of variance are given in the figure legends and Appendix IV.

In the 24h incubation (experiment 1, Figures 50 and 51) a different response was observed. While there was no significant change in ACTH secretion, there was an apparent increase in the level of ACTH precursors secreted (F ratio = 3.23, p = 0.045) although inspection of the 95% confidence intervals demonstrates that the difference resided between the cells treated with 2000nmol/l hydrocortisone and the control cells. This could indicate reduced processing of precursors to ACTH. The most likely explanation for the lack of suppression by glucocorticoids is that the high basal levels and relatively long half-life of ACTH and ACTH precursors in medium (Figure 20) obscured any suppression of ACTH after this short incubation period.

In Experiment 2, (24h incubation) (Figures 52 and 53) one series of medium samples was unsuitable for analysis. Therefore, peptide results were obtained for only two bottles of cells at each concentration. Statistical analysis was not performed but the results are given as the cell pellets were used for RNA studies. No differences were apparent between control and test samples.

150 iue4 Concentrationsof ACTH precursors in culture Figuremedium 43 ACTH precursors (pmol/mg protein) 00 - 4000 00 n 5000 2000 00 - 3000 1000 f t2 cls nuae fr 0 as with days 10 for incubated cells Eachdata hydrocortisone. represents point peptide AtT20 of and 95% confidence interval are shown to the right of Fratio = 35.66, p<0.0001. h aast Resultof analysisthe data set.of variance: levels in Thea groupsingle mean bottle of cells. - - 0 t2: xeiet (0 day) (10 1 Experiment AtT20: 50 yrcrioe (nm/l) Hydrocortisone CH precursors ACTH 151 0 50 00 2000 1000 500 100 iue4 Cnetain o AT i clue eim f AtT20 medium culture of Concentrations in ACTH of Figure44 ACTH (pmol/mg protein) 150 150 100 2 - 125 5 - 75 0 I 8 -I 50 5 - 25 - i - oteo el. Thegroup bottlemean and of95% cells. confidence p<0.0001. Resultof analysis of variance: Fratio =77.75, Each hydrocortisone. withdata dayspoint 10 represents for incubated cells peptide levelsin a single nevl r son o h rgt f h dt set. data the right of the tointerval shown are

ACTH 8 5 10 0 10 2000 1000 500 100 50 0 I T0 Eprmn 1 1 day) (10 1 Experiment tT20: A H y d r o c o r t i s o n e ( n m / l ) i o 152 X * '•iH iue4 Concentrationsof ACTH precursors in cultureFigure medium45 ACTH precursors (pmol/mg protein) 00 - 40000 00 -| 50000 20000 00 - 30000 10000 f t2 cls nuae fr 0 as with days 10 for incubated cells AtT20 of yrcrioe Eachdata hydrocortisone. representspoint peptide and 95% confidence interval are shown to the right of Fratio =6.58, p= 0.017. Resulttheof analysis data set.of variance: levels in The asingle group mean bottle of cells. - - 0 T0 Eprmn 2 1 day) (10 2 Experiment tT20: A H y d r o c o r t i s o n e ( n m / l ) 153 CH precursors ACTH 500 1000 Figure ACTH (pmol/mg protein) j Cnetain o AT i clue eim f AtT20 culturemedium of Concentrations in ACTH of |j> 1000 H 0 - 500 5 - 750 5 - 250 oteo el. Thegroup bottlemean of 95%and cells. confidence Each data point represents peptide levelsin a single Resultof analysis of variance: el icbtd o 1 dy wt hydrocortisone. with days 10 for incubated cells Fratio 18.20,= p=0.001. nevl r son o h rgt f h dt set. data the right of the tointerval shown are 0 - o © © o t2: xeiet (0 day) (10 2 Experiment AtT20: I H y d r o c o r t i s o n e ( n m / l ) 154 ACTH 0 1000 500 00 ■I Figure47 ACTH precursors (pmol/mg protein) 0 5 5 f t2 cls nuae fr 0 as with days 10 for incubated cells AtT20 of Concentrationsof ACTH precursors in culture medium p = 0.0037, 95% Clfor the difference in means(5.31- Eachhydrocortisone.data point represents peptide thedata set. Result of two-tailed unpaired ttest:and 95% confidenceinterval are shown to the right of 9.08). levels in asingle Thegroup meanbottle of cells. t2: xeiet (0 day) (10 3 Experiment AtT20: H y d r o c o r t i s o n e ( n m / l ) 0 155 1000 Figure48 ACTH precursors (pmol/mg protein) 5 0 f t2 cls nuae fr 0 days 10 for incubated cells AtT20 of unpaired t test: p= 0.0004, 95% Cl for the Thedifference cells. bottle singleof a inpeptide levels Concentrationsof ACTH precursors inculture medium h rgt f h dt st Rsl o two-tailed of Result set. data the of rightthe represents point data Each hydrocortisone.with groupmean and95% confidence interval shown areto inmeans (5.95-7.02). t2: xeiet (0 day) (10 4 Experiment AtT20: yrcrioe (nm/l) Hydrocortisone 0 156 1000 Figure49 ACTH precursors (pmol/mg protein) 4 0 2 6 8 f t2 cls nuae fr 0 as with days 10 for incubated cells AtT20 of Concentrationsof ACTH precursors in culture medium yrcrioe Eachdata hydrocortisone. pointrepresents peptide p = 0.0054, 95% Cl for the Result the ofdata set.difference two-tailed in unpaired means (1.72-t test: and 95% confidenceinterval are shown to the right of 3.32). levels in Thea single group mean bottle of cells. t2: xeiet (0 day) (10 5 Experiment AtT20: H y d r o c o r t i s o n e ( n m / l ) 0 157 1000 iue5 Concentrationsof ACTH precursors Figurein culture50 medium ACTH precursors (pmol/mg protein) 0 -i 400 200 0 - 300 100 - - Resultof analysis of variance: Fratio = 3.23, Thebottlegroup ofmean cells. 95%and confidence of AtT20 cellsincubated for 24h with hydrocortisone. Each data point represents peptide levels in a single p= 0.045. nevl r son o h rgt f h dt set. data the right of theinterval to shown are 0 t2: xeiet (4 incubation) (24h 1 Experiment AtT20: 50 H y d r o c o r t i s o n e ( n m / l ) CH precursors ACTH 0 50 00 2000 1000 500 100 158 iue5 Cnetain o AT i clue eim f AtT20 medium culture of Concentrations in ACTH of Figure51 ACTH (pmol/mg protein) 0 - 40 0 n 50 20 0 - 30 10 - - Resultofanalysis of variance: Fratio 1.69,= Thegroupbottle ofmean and cells.95% confidence =020 i.e. 0.210. nop= significant differences observed. single a in levels represents peptide point data Eachcells incubated for 24h hydrocortisone.with nevl r son o h rgt f h dt set. data the right of the to shown interval are O © 5 10 0 10 2000 1000 500 100 50 0 t2: xeiet (4 incubation) (24h 1 Experiment AtT20: 4 k O . O yrcrioe (nm/l) Hydrocortisone < ► 8 1 8 - O ACTH 159 i e o i o > o o o O O iue5 Concentrationsof ACTH precursors incultureFiguremedium 52 ACTH precursors (pmol/mg protein) 00 - 4000 2000 2000 H 00 -i 8000 00 - 6000 Each data point represents peptidelevels in a single bottleof cells. of AtT20 cellsincubated for 24h with hydrocortisone. T0 Eprmn 2 2h incubation) (24h 2 Experiment tT20: A n O O 50 1000 500 0 H y d r o c o r t i s o n e { n m / l ) ------CH precursors ACTH 160 8 @ 1 ------1— iue5 Cnetain o AT i clue eimo AtT20 medium culture Concentrationsof in ACTH of Figure53 ACTH (pmol/mg protein) 0 - 400 0 | ACTH -| 00 5 0 O 200 H 0 - 300 100 bottleof cells. aa on rpeet ppie ees n single a in levels represents point peptide data Eachcellsincubated for 24hwith hydrocortisone. -

t2: xeiet (4 incubation) (24h 2 Experiment AtT20: T O O O 50 1000 500 0 H y d r o c o r t i s o n e ( n m / l ) ------161 O 1 ------O 1 -- d) POMC RNA levels

No statistically significant change in POMC RNA was observed in COR L103 cells after 24h or 10 days of hydrocortisone treatment (Figure 54A and C) whereas in AtT20 cells there was a significant reduction in POMC RNA at 10 days (analysis of variance : F ratio = 18,63, p = 0.002) but not at 24h

(Figure 54B and D). e) Expression of other glucocorticoid regulated genes

This demonstration that the POMC gene was relatively resistant to inhibition by hydrocortisone prompted an assessment of two other glucocorticoid regulated genes in the COR L103 cell line. Glucocorticoids normally suppress the gene for the glucocorticoid receptor (GR) and activate the gene for tyrosine amino transferase (TAT, a liver enzyme).

Both the glucocorticoid receptor (GR) and tyrosine amino transferase (TAT) genes were expressed in the COR L103 cells. GR mRNA levels remained unchanged in COR L103 cells and there was no significant stimulation of TAT mRNA in COR L103 cells after either 24h or 10 days hydrocortisone treatment

(Figure 55). f) Dexamethasone binding studies

A possible reason for resistance to glucocorticoids is failure to express

GR. Although GR RNA was detected in the COR L103 cells it was necessary to confirm that the receptor was expressed at the protein level. Displacable dexamethasone binding was measured in COR L103 and AtT20 cells on several occasions. The displacement curves for nuclear binding from these two cell lines are shown in Figure 56. Scatchard analysis (data not shown) gave

_q similar K^’s of 5 x 10 M in each cell type. This result is in good agreement with values reported elsewhere (Rousseau et al 1972). The estimated nuclear binding capacity was 640 fmol/mg protein for COR L103

162 POMC mRNA _l 0 I 1600 2B21000

10 Days Myorocortiscne irM) A. COR L103 B. AtT20

POMC.Q-Actin mRNA POMC:Q-Actin mRNA

24 Hours

C. COR L103 D. AtT20

POMC:Q-Actln mRNA POMC:S-Actin mRNA

Figure 54 POMC mRNA content of cells incubated with hydro­ cortisone. Results are expressed as the fractional change in the ratio of POMC mRNA densitometric units to the 13-actin densi tometric units. Each bar represents the mean from four experiments. Individual values are given in Appendix II, Section 2. Results of analysis of variance are given in Appendix IV. Statistically significant differences were observed only for AtT20 cells in a 10 day incubation (B): F ratio = 18.63, p = 0.002.

163 TAT mRNA □ o L ) too ES21000

Hydrocortisone (r>M) A. 10 Days B. 24 Hours

TAT.Q-Actin mRNA TAT;G-Actm mRNA 1.6r 1 2r

GR mRNA

C .10 Days D. 24 Hours

GR:Q-Actm mRNA GR:Q-Actm mRNA 1.2r 1 [

0.8 h

0.6r

0.4 j-

0.21- o1-

Figure 55 Cellular content of TAT mRNA (A, B) or GR mRNA (C, D) in COR L103 cells treated with hydrocortisone. Results are expressed as the fractional change in the ratio of each mRNA to the 13-actin mRNA and each bar is the mean of four experiments (see Appendix II, Section 2). No changes are statistically significant (see Appendix IV).

164 iue5 Dslcmn f H-dexamethasone Displacementof with unlabelled Figure56

Counts Bound / mg of Protein 10-1° 10-9 normalisedper rag protein. dexamethasone in nuclear preparations from COR L103 el () r t2 cls 0. ons on are bound Counts (0). cells AtT20 or (•) cells eaehsn ( olar) (M Dexamethasone 3 165 cells and 710 fmol/mg protein in AtT20 cells (Clark et al 1990). The specificity of the receptor for glucocorticoids was demonstrated by the absence of binding of progesterone and aldosterone.

166 2) EFFECT OF THE DOPAMINE AGONIST. BROMOCRIPTINE. ON ACTH

PRECURSOR SECRETION BY SCLC CELLS

In the neuro-intermediate lobe of the pituitary, POMC is negatively regulated by dopamine, not glucocorticoids. Since SCLC has features consistent with transformation at a neuroendocrine stage of differentiation, the possibility that dopaminergic regulatory mechanisms might operate in these cells was investigated. Since dopamine itself is not stable, the dopamine agonist bromocriptine was used in these experiments. a) Experimental Design

The effect of bromocriptine on ACTH precursor secretion was assessed in three SCLC cell lines (COR L24, COR L31 and COR L42).

SCLC cells were cultured to stationary phase and then passaged in fresh medium (RTISS 2.5). Cultures were inoculated at a density of approximately

1 x 10 fi cells/ml for a 24h incubation and at 1 x 10 5 cells/ml for a seven day incubation. For the 24h incubation, bromocriptine (dissolved in DMSO) was added to final concentrations ranging from 0.1-10 ug/ml. Equivalent volumes of DMSO were added to control cultures. The final concentration of

DMSO did not exceed 0.2% (v/v). For the seven day incubation, bromocriptine was added to a final concentration of lOug/ml and this dose was repeated at 48h intervals. Three bottles of cells were established at each concentration. b) Peptide secretion

Following 24h incubation, ACTH precursor levels were reduced in a dose dependent fashion in COR L24 cells (Figure 57) and analysis of variance indicated that the differences were statistically significant (F ratio =

17.16, p = <0.0001). Although significant F ratios and p values were also obtained for COR L31 (F ratio =6.99, p = 0.013) and COR L42 (F ratio =

19.30, p = 0.001), inspection of the results indicated that the difference

167 iue5 Effectof bromocriptine on ACTH precursor Figuresecretion57 ACTH precursors (pmol/mg protein) 100 40- 0 2 60- 80- variance: Fratio 17.16,= p<0.0001. (2.5) supplementedRTISS (final withbromocriptine Cells incubatedwere in for 24h by L24cells. COR groupmean and95% confidence interval shownareto concentration 0-10pg/ml). Each data point represents etd lvl i a ige ote f el. The cells. singlebottle of a inpeptide levels h rgt f h dt e. eut f analysis of of Result data theset.the rightof - - 0 . 0.50.1 B r o m o c r i p t i n e ( p g / m l ) COR L24 COR 168 1 5 10 resided between the control group and the cells treated with the highest dose of bromocriptine (lOpg/ml) in both cases (Figures 58 and 59 respectively). Results of 7d incubation experiments were analysed by two- tailed, unpaired t tests and are shown in Figures 60-63. After seven days, ACTH precursor levels were markedly lower in all the bromocriptine treated cells, however it was noted that the drug treated cells proliferated relatively poorly in comparison with the control cells, reflected in lower protein concentrations in the treated cultures, c) POMC RNA levels

The ratio of POMC RNA to beta-actin RNA was assessed by slot blot analysis in COR L24 cells (Table 15). Interestingly, the POMC RNA levels were markedly increased in the cells treated with lOpg/ml bromocriptine for 24h.

This could be interpreted as an inhibitory effect at the level of translation, resulting in accumulation of untranslated POMC message and a consequent reduction in the amount of secreted peptides. The three cell pellets obtained for each concentration were pooled before extraction of

RNA, therefore statistical analysis of the single results was not possible.

RNA analysis in this and succeeding experiments was performed by Dr. W.E.

Farrell, Hope Hospital.

169 i Figure58 ACTH precursors (pmol/mg protein) 2 TS (2.5) supplementedRTISS (final with bromocriptine Cells incubatedwere in for 24h L31by CORcells. Effectof bromocriptine on precursor ACTH secretion aine Fratio =6.99, pvariance:= 0.013. groupmean and95% confidence interval shownareto concentration 0-10pg/ml). Each data pointrepresents etd lvl i a ige ote f el. The cells. singlebottle of a inpeptide levels h rgt f h dt st Rsl o analysis of of Result theset. data rightthe of 0

B r o m o c r i p t i n e ( p g / m l ) 0.1 O L31 COR

170 1.0

10

iue5 Effectof bromocriptine on precursorACTH Figuresecretion 59 ACTH precursors (pmol/mg protein) 10 15 yCR 4 el. Cells incubatedwere for in 24h by L42 CORcells. TS (2.5)RTISS supplemented (final with bromocriptine variance: Fratio = 19.39, p= 0.001. groupmean and95% confidence interval shownare to concentration 0-10pg/ml). Each data point represents etd lvl i a ige ote f el. The cells. bottle singleof a inpeptide levels h rgt f h dt st Rsl o analysis of of Result theset. datathe right of 0 B r o m o c r i p t i n e ( f j g / m l ) 0.1 O L42 COR 171 1.0 10 iue6 Effectof bromocriptineon ACTH precursor secretionFigure 60 ACTH precursors (pmol/mg protein) 0 - 400 0 - 500 n 600 200 0 - 300 100 oteo el. Thegroup mean andbottle 95% ofconfidence cells. Each data point represents peptideRTISS (2.5) withlevelsand lOpg/mlwithout in bromocriptine.a single by CORL24 cells. Cellswere incubatedfor 7 days in eut f w tie, nard ts: = 0.036. = unpairedtest: tailed,p t twoResult of means: 49-567.7) nevl r son o h rgt f h dt set. data the right of the to showninterval are - - 9% ofdne nevl o te ifrne in difference the for interval confidence(95% O L4 Eprmn 1 7 incubation) (7d 1 Experiment L24: COR B r o m o c r i p t i n e ( p g / m l ) 0 172 10 iue6 Effectof bromocriptine on ACTH precursor secretion Figure61 ACTH precursors (pmol/mg protein) 200 200 n 100 5 - 150 0 - 50 byCOR L24 cells. Cellswere incubated for 7 fa days TS (2.5)RTISS withand without lOpg/ml bromocriptine. - oteo el. Thegroup bottle andmean of95%cells. confidence Each data point represents peptidelevels in a single ofdne nevl o te ifrne n means: difference in (95% <0.0001.confidence interval for the p test: t unpaired tailed, two of interval areshown tothe right Result ofthe data. 133.5-152.5) O L4 Eprmn 2 7 incubation) (7d 2 Experiment L24: COR B r o m o c r i p t i n e ( j j g / m l ) 0 173 10 Figure62 ACTH precursors (pmol/mg protein) 5 5 0 byCOR L31 cells. Cellswere incubated for 7 it days Each data point represents peptide levelsin a single Effectof bromocriptine on ACTH precursor secretion oteo el. Thegroup bottlemean 95%andofconfidence cells. RTISS (2.5) withand without lOpg/ml bromocriptine. ofdne nevl o te ifrne n means: difference in confidence interval the for (95% 0.0051. = test: p t unpaired tailed, twoof 5.96-11.24) interval areshown tothe right Result ofthe data. O L1 7 incubation 7d L31: COR B r o m o c r i p t i n e ( p g / m l ) 0 174 10 iue6 Effectof bromocriptineon precursorACTH secretionFigure 63 ACTH precursors (pmol/mg protein) 0 - 40 0 n 80 20 - 60 - TS (2.5)RTISS withand lOpg/ml without bromocriptine. by CORL42 cells. Cellswere incubated for 7days h means: 43.2-72.59) tailed,Result twoof 0.0035. test: unpaired t = p Thegroup bottlemean 95%andof confidence cells. Each data point represents peptide levelsin a single nevl r son o h rgt f h dt set. data the right of the to showninterval are 9% ofdne nevl o te ifrne in difference the for interval confidence(95% O L2 7 incubation 7d L42: COR B r o m o c r i p t i n e ( j j g / m l ) 0 175 10 TABLE 15

Effect of bromocriptine on POMC RNA content in COR L24 cells

Fractional change in Bromocriptine POMC : B-actin ug/ml mRNA

0 1.0

0.1 1.1

0.5 0.8

1.0 1.2

5.0 1.65

10.0 3.0

Three separate bottles of cells were incubated at each concentration of bromocriptine. The three cell pellets were pooled before extraction of RNA. Results are expressed as the fractional change in the ratio of POMC mRNA densitometric units to the B-actin mRNA densitometric units, the control value being designated 1.0.

176 3) CORTICOTROPHIN RELEASING FACTOR

The introduction of the CRF test into clinical practice has shown that in the ectopic ACTH syndrome, no rise in cortisol levels is seen following an intravenous bolus of CRF, suggesting that ACTH secretion by SCLC cells is not responsive to the normal hypothalamic releasing hormone. The response of SCLC cells in vitro to CRF was therefore assessed in succeeding experiments. a) Experimental Design

The effect of CRF on ACTH precursor secretion was assessed in two SCLC cell lines (COR L103 and COR L42) and AtT20 cells. Cells were cultured to confluence and passaged 1:3 in fresh medium (RTISS (2.5) for SCLC cells and

Ham’s F10 supplemented with 0.5% BSA for AtT20 cells). Ovine CRF was added to final concentrations of 10, 100 and 1000 ng/ml to test cultures.

Equivalent volumes of the respective diluent were added to control cultures. Cells were incubated at 37°C for 72h. This protocol was selected in the light of the studies of May and Eipper(1986) on rat pituitary cells.

Three separate bottles of cells were incubated at each concentration. b) Peptide secretion

Following a 72h incubation in the presence of oCRF, no increase was observed in ACTH precursor secretion by COR L42 or COR L103 cells (Figures

64 and 65). In contrast, increases in peptide secretion were observed in

AtT20 cells. Analysis of variance indicated that the increases were statistically significant for both ACTH precursors (F ratio = 9.98, p =

0.004) and ACTH (F ratio = 18.57, p = 0.001). The response was more pronounced for ACTH (Figure 67) than ACTH precursors (Figure 66). c) Effect of different media

In order to assess whether the presence of foetal calf serum affected the response to oCRF in SCLC cells, COR L24 cells were cultured for 72h in

177 iue6 EffectofoCRF onACTH precursor secretionFigure by64 COR ACTH precursors (pmol/mg protein) 40- 20 60- - TS (.) upeetd ih CF (final oCRF with supplemented (2.5) RTISS nosignificant differences observed. Cellswere incubated for 72hin L42cells. r hw otergto h aast Resultof areshown tothe right groupof the Themean confidence point 95% intervaldata and set. data cells. Each ng/ml).concentrations 0-1000 analysis of variance: F ratio = 2.27, p= 0.157. of i.e. bottle single a in levels peptiderepresents 0

10 COR L42COR o C R F ( n g / m l )

178 100

1000 1000 COR L103 10-1

<0 k _ o A ' < =3 L_ o Q_ 6 - o8 i /% -- 0) 8 ._ o> c l E X 4 -J K“ o O E < CL

2 -

0 0 10 100 1000 oCRF (ng/ml)

Figure 65 Effect of oCRF on ACTH precursor secretion by COR L103 cells. Cells were incubated for 72h in RTISS (2.5) supplemented with oCRF (final concentrations 0-1000 ng/ml). Each data point represents peptide levels in a single bottle of cells. The group mean and 95% confidence interval are shown to the right of the data set. Result of analysis of variance: F ratio = 0.21, p = 0.886. i.e. no significant differences observed.

179 Figure ACTH precursors

66 (pmol/mg protein) 4000 -i 4000 2000 3000- 1000 fet f CF n CH rcro scein by secretion precursor ACTH on oCRF of Effect .% oie eu abmn splmne wt oCRF supplemented albumin, serum with Bovine 0.5% AtT20, Cellswere incubated for 72h in Ham’sF10+ r hw otergto h aast Resultof areshown tothe right of the group mean 95% Theconfidencedataand intervalset. of bottle cells. single a in levels peptide represents analysisofvariance: Fratio = 9.98, p= 0.004. (final concentrations 0-1000 ng/ml). Each data point - - A t T 2 0 : A C T H p r e c u r s o r s 0

10 o C R F ( n g / m l ) 180

100

1000

iue6 Efc o oR o AT scein y AtT20. by secretion ACTH on oCRF of Effect Figure67 ACTH (pmol/mg protein) 40- 0 2 60- 80 n bovineserum withalbumin, supplementedoCRF (final el wr icbtd o 7h n a’ F0 0.5% + F10 in Ham’s 72h incubated for wereCells r hw otergto h aast Resultof areshown to the right ofthe data set. point data Each ng/ml). 0-1000 concentrations ersns etd lvl i a ige ote of bottle single a in levels peptide represents el. h group The mean 95% confidenceand interval cells. analysisof variance: Fratio = 18.57, =0.001. p - 0

10 AtT20: ACTH o C R F ( n g / m l )

181 100

1000 three different media: RTISS (2.5), serum free RTIS and RBSAT (RPMI 1640 supplemented with 0.5% BSA and trasylol). oCRF was added to a final concentration of lOOng/ml to test cultures. Three separate test and control bottles were established in each medium. Results were analysed by two- tailed unpaired t tests. No significant change in ACTH precursor secretion was observed in any of the oCRF treated cultures (Figure 68). In addition, the absolute levels of precursors secreted was much lower in both the serum free media. This was particularly marked in the case of serum free RTIS, and the final protein concentrations confirmed that COR L24 cells did not proliferate in this medium.

The failure of COR L24 cells to respond to an adequate stimulatory dose of oCRF cannot therefore be attributed to the presence of serum in the culture medium.

182 Figure ACTH precursors

68 (pmol/mg protein) Effect of different media on response to oCRF by SCLC n 0 2 unpairedttests are given in Appendix IV. oCRF added towas final culturestotest a RBSAT. different : (2.5), RTISSmedia free and RTLS serum significantly point stimulate dataACTH precursor secretion in are Each shown to groupthe mean Theconfidenceinterval right 95%and of thecells. oCRF datadidset. not lOOng/ml. of concentration n o te he mdaRsls f w tailed, two of media.Results three the of any COR cells.L24 cells wereincubated for24h in three ersns etd lvl i a ige ote of bottle single a in levels peptiderepresents TS 25 TS F RBSAT SF RTIS 2.5 RTISS 0

100 o C R F ( n g / m l ) 183 0

O L24 COR 100

* © 0

100 8

* 4) EFFECT OF GLUCOCORTICOID AND CRF ON HUMAN CORTICOTROPH CELLS a) Experimental Design

Primary cultures of human pituitary corticotroph cells were established as described in the methods section. The cells were derived from a pituitary tumour causing Cushing’s syndrome (see Table 8, tumour no. 7). Cells were innoculated into 24 well tissue culture plates at a density of 0.5 x 10 cells per well and cultured in growth medium (DMEM + 2.5% FCS + 10% HS

+G+P) for 12 days, by which time the cells had formed an adherent monolayer. The cells were then washed and the growth medium was replaced by serum free DMEM +G+P. On day 15, cells were washed several times in serum free medium and individual wells treated as follows : i. Controls: 1ml serum-free DMEM (n=7) ii. 1ml serum-free DMEM + lpm/1 Dexamethasone (n=8) iii. 1ml serum-free DMEM + lOOOng/ml oCRF (n=7)

Cells were then incubated for 72h at 37°C before collecting the medium for peptide analysis. The protocol for this experiment was selected in the light of the studies of May and Eipper (1986) on rat pituitary cells. b) Peptide secretion

The effects of dexamethasone and oCRF are illustrated in Figures 69 and 70.

ACTH and precursors were secreted in approximately equimolar amounts in the control wells suggesting that processing pathways operated adequately in these pituitary tumour cells. Analysis of variance indicated that highly significant differences existed between the groups for both ACTH precursors

(F ratio = 69.23, p = <0.0001) and ACTH (F ratio = 110.0, p = <0.0001).

Inspection of the results showed that the response to dexamethasone was limited, as might be expected from the clinial behaviour of such tumours.

ACTH levels were slightly, though not statistically significantly lower in

184 iue6 Effect dexamethasoneof ACTH precursor oCRF andon Figure69 ACTH precursors (pmol/mg protein) 1000 1500 -i 500- p< 69.23, = ratio F variance: analysisof of Result el (. 1 cells (0.5wells well)per 10 were incubated x for secretion by hunLan pituitary Control cells. and test ige el Te ru ma ad 5 confidence 95% and mean group The well.single Each data point72h. represents peptide levels in a nevl r son o h rgt f h dt set. data the right of the to showninterval are 0 - . o t o s E oCRF DEX Controls H u m a n p i t u i t a r y c e l l s : ACTH p r e c u r s o r s 0001 . 185 p 1000ng/ml 1pM iue7 Effect of dexamethasoneFigure and70 oCRF on ACTH secretion by ACTH (pmol/mg protein) 4000 - 2000 5000 - 3000 - 1000 ua iutr el. Controlhumanpituitaryand (0.5test wells cells. e l The groupwell. singlemean and a in 95% levels representspeptideconfidence point interval data are 0 cellsincubatedwell) 72h. per Each for were x10 hw t te ih o te aa e. eut of Result set. data the analysis ofof rightvariance: F ratio the= 110.0, p<0.0001. to shown - - H u m a n p i t u i t a r y c e l l s :ACTH 186 DEX f M 1000ng/ml 1fiM oCRFControls the dexamethasone treated cells but this was accompanied by a significant increase in precursors, as judged by the lack of overlap between the 95% confidence intervals for the respective group means. This could be interpreted as indicating reduced processing to ACTH 1-39. ACTH levels were dramatically stimulated in response to oCRF, mean levels being more than four-fold higher than in the controls. Interestingly, this was also accompanied by a significant reduction in the levels of precursors which suggests that processing to ACTH 1-39 may be enhanced by oCRF.

187 5) SECOND MESSENGER MECHANISMS IN SCLC CELLS

A. Cyclic Adenosine Monophosphate (cAMP)

The SCLC cell lines tested (COR L24, COR L42 and COR L103) exhibited no

significant stimulation of POMC peptide secretion in response to oCRF. The

next step was therefore to assess the integrity of the principal

intracellular signalling system through which this hormone acts on anterior

pituitary cells, namely the cAMP pathway,

i. Dibutyryl cAMP

Dibutyryl cAMP (db-cAMP) is a stable analogue of cyclic AMP which

penetrates the cell membrane to act intracellularly.

i.a) Experimental design

The effect of a single high dose (ImM) of db-cAMP was initially assessed

in two SCLC cell lines (COR L24 and COR L103). db-cAMP was solubilised in

DMSO and equivalent volumes of DMSO were added to control cultures. The

final concentration of DMSO did not exceed 0.5%. Cells were incubated in

RTISS(2.5) at 37°C for 48h. Three separate bottles of cells were

established for both the test and control incubations.

i.b) Peptide secretion

Following 48h incubation, ACTH precursor levels were significantly higher

in db-cAMP treated cultures of both COR L24 and COR L103 cells (Figures 71

and 72). Results were analysed by two-tailed unpaired t tests. For COR

L24 p = 0.030, 95% confidence interval for the mean difference (-71.3,-9.8)

and for COR L103 p = 0.026, 95% confidence interval for the mean difference

(-6.06,-1.01).

ii. Time course of the effect of db-cAMP on SCLC cells

To assess whether stimulation of ACTH precursors by db-cAMP could be detected acutely, COR L24 cells were incubated with and without db-cAMP for

intervals ranging from 30min to 48h.

188 iue7 ACTHprecursor secretion by CORcellsL24 incubated Figure71 ACTH precursors (pmol/mg protein) 100 150-1 125- 50- 75- 25- oteo el. Thegroup mean 95% andconfidencebottle ofcells. Each data point represents peptide levelsin a single Resultof two-tailed, unpairedttest: p= 0.030. for 48h in (2.5)RTISS with and without lmM db-cAMP. 71.3-9.8) intervalareshown to theright ofthe data set. (95% confidenceinterval for the difference in means: - 0 d b - c A M P ( m m o l / I )

O L24 COR 189 1

Figure72 ACTH precursors (pmol/mg protein) CH rcro scein y O L0 cells L103 COR by secretion precursor ACTH 0.026.(95% = p confidence intervalfor thethe difference dataset. Result of two-tailed, unpairedand t95%test: confidence interval are shown to the right of inmeans: -6.06, - levels The ingroup meanasingle bottle of cells. Eachdata pointlmMrepresentsdb-cAMP. peptide nuae o 4h nRIS 25 wt ad without and with incubated in (2.5)for 48h RTISS 1 0 d b - c A M P ( m m o l / l ) O L103 COR 1 . 01 190 ) ii.a) Experimental design

COR L24 cells were seeded at a density of 2 x 10^ cells/ml in RTISS(2.5). . 3 db-cAMP was added to test bottles to a final concentration of 10 M. Test

(n=18) and control bottles (n=18) were established and incubated at 37°C.

At 30min, lh, 2h, 5h, 24h and 48h, three test and three control bottles of cells were harvested for RNA and protein and the medium was flash frozen for ACTH precursor assay, ii.b) Peptide secretion

The data were analysed by linear regression with the following results:

Control cells: slope = 1.08 (SE 0.05), intercept = 2.52 (SE 0.92) db-cAMP treated cells: slope = 1.65 (SE 0.04), intercept = 1.10 (SE 0.91)

The difference in slopes was statistically significant

(p <0.001, 95% confidence interval for the difference in

slopes: 0.444, 0.7003).

There was no statistically significant difference in the

values of the y intercepts (p = 0.279).

The results are illustrated in Figure 73, using an arbitrary scale for the x axis (time). This demonstrates that the difference between the db-cAMP treated and control cells only becomes obvious after 24h. The data confirm the results obtained in the previous experiment and suggest that 24-48h is the critical time to demonstrate the stimulatory effect of db-cAMP. The increase in peptide levels at this stage may reflect the time taken for newly synthesised peptide to be secreted, c) POMC RNA levels

POMC RNA levels were higher in the db-cAMP treated cells than in the controls at 24 and 48h (Table 16). Each set of three test and control cell pellets was pooled before RNA extraction and therefore statistical analysis of the single result obtained was not possible.

191 Figure73 ACTH precursors (pmol/l) ifrne n h to lps p 001, 95% <0.001), (p slopes two the in difference is meanwithoutthe raM 1Eachpoint data db-cAMP. peptide level from three separate ofcells.bottles 0.444, 0.7003). were Cells cells. L24 CORprecursor secretion by Timecourse ofthe effect of db-cAMPon ACTH ofdne nevl o te ifrne n slopes: in difference the forconfidence interval ersin nlss niae a significant a indicated analysis regression Individual results are given in II. LinearAppendix nuae fr ., , , , 4 n 4h ih and with 48h and 24 5, 2, 1, 0.5, forincubated . 1 5 4 48 24 5 2 1 0.5 192 COR L24COR Time (h)

TABLE 16

Effect of dibutyryl cAMP on POMC RNA content of COR L24 cells during 48h incubation.

Incubation Fractional change in time (h) POMC : B-actin mRNA (Test/control cells)

0.5 0.6

1

2 1.0

5 1.3

24 2.8

48 4.0

Three control and three db-cAMP treated bottles of cells were incubated for each time interval. Each set of three cell pellets was pooled prior to extraction of RNA. Results are expressed as the fractional change (Test/control) in the ratio of POMC mRNA densitometric units to the B-actin mRNA densitometric units, the control values at each time point being designated 1.0.

193 iii). Effect of a Phosphodiesterase Inhibitor

Phosphodiesterase inhibitors such as isobutyl methyl xanthine, raise endogenous cAMP by preventing its enzymic breakdown. The effect of this agent on SCLC cells was assessed alone and in combination with exogenous db-cAMP. iii.a) Experimental Design

The effect of sub-maximal doses of db-cAMP was assessed in COR L24 cells, with and without the addition of the phosphodiesterase inhibitor, isobutyl methyl xanthine (IBMX). COR L24 cells were cultured to confluence, washed g and then passaged in fresh medium to a density of approximately 1 x 10 cells /ml. Bottles containing 5ml volumes of cells were established and db- cAMP was added to final concentrations of 10 M and 10 M, both with and without the addition of 0.5mM IBMX. Both reagents were solubilised in DMSO and equivalent volumes of DMSO were added to control cultures (final concentration = 0.5%). Cells were incubated at 37°C for 24h. Three separate bottles of cells were incubated at each concentration, iii.b) Peptide secretion

Results are illustrated in Figure 74. Analysis of covariance

showed:

Although there was an apparent upward trend in the means for

the db-cAMP treated cells, the differences were not

statistically significant (F ratio = 2.35, p = 0.14).

Mean levels of precursors were higher in the cells treated

with 0.5mM IBMX than those incubated without this agent and

the difference was statistically significant (F ratio = 38.33,

p = <0.0001).

There was no evidence of any interaction between the two

variables, i.e. db-cAMP and IBMX (F ratio = 0.24, p = 0.79).

194 iue7 Efc o te phosphodiesterase the of inhibitor, Effect IBMX on Figure74 ACTH precursors (pmol/mg protein) 0 0 2 100 150 50 0 CH rcro scein y O L4 el. Cells cells. L24 COR secretion precursor byACTH aast Analysis ofcovariance performed on was dataset. ee nuae fr 4 i RIS 25 wt 05 mM 0.5 (2.5) inwith RTISS incubated for 24h were h aa Resultsare discussed inthe thedata. text. the right of the to shown interval 95%confidence and are mean group The cells. of single bottlea Eachdata pointrepresents levelspeptidecAMP. in IBMX alone and in combination with 0, 1 and lOOpM db- 0 1 db-cAMP (mmol/l) 100 195 O L24 COR IM 05 ol/l m 0.5m IBMX + ■ ' L A iL a , A " A r A ^ A L A iii.c) POMC RNA levels

A marked increase in POMC RNA levels was observed in response to the combined effects of 10 ^M db-cAMP and 0.5mM IBMX (Table 17). The three cell pellets obtained with each concentration were pooled before RNA extraction, therefore statistical analysis of the single results was not possible. iv. Forskolin

Forskolin was used to determine whether POMC could be stimulated by direct activation of adenylate cyclase, iv.a) Experimental design

Forskolin, a direct activator of adenylate cyclase, was assessed in COR L24 and COR L103 cells. SCLC cells were incubated for 24h in RTISS 2.5 -7 containing forskolin at final concentrations of 10 -8 M, 10 M and 10 _6 M

(Seamon & Daly 1981). Forskolin was initially solubilised in ethanol and further diluted in medium. Three separate bottles of cells were incubated at each concentration. iv.b) Peptide secretion

ACTH precursor levels in medium were modestly increased in the medium of

COR L24 cells treated with forskolin (Figure 75). This increase was statistically significant (analysis of variance: F ratio = 5.04, p =

0.030). No increase was observed in the medium of COR L103 cells (Figure

76). Paradoxically, there was an apparent decrease in the precursor levels

(analysis of variance: F ratio = 5.23, p = 0.027). Inspection of the 95% confidence intervals indicated a difference between the control cells and cells treated with 10 M forskolin, yet no difference between controls and the higher dose of 10 M forskolin.

196 TABLE 17

Effect of the phosphodiesterase inhibitor. IBMX and db-cAMP on POMC RNA content of COR L24 cells. db-cAMP Fractional change in POMC : B-actin mRNA (UM/I)

- IBMX + IBMX (0.5mH/1)

0 1 1.6

1 1.6 1.8

100 0.85 4.0

Three separate bottles of cells were incubated at each concentration of db-cAMP +/~ IBMX (0.5mmol/l). The three cell pellets were pooled before extraction of RNA. Results are expressed as the fractional change in the ratio of POMC mRNA densitometric units to the B-actin mRNA densitometric units. The cells untreated with either agent were designated 1.0.

197 iue7 Effectofforskolin on ACTH precursor secretion Figureby75 ACTH precursors (pmol/mg protein) COR L24 Cells cells.wereincubated for24h in RTISS aast Resultof analysis of variance: data set. ige ote f el. h gop en n 95% meanand group The cells.singlea bottle of Fratio = 5.04, p=0.030. the right of theconfidence shown to interval are 0 ) Each data10’ point represents peptide M) levels in (2.5) supplemented withforskolin (0, 10" , and 10" O L24 COR F o r s k o l i n ( M ) -8 198 7 - Figure76 ACTH precursors (pmol/mg protein) etd lvl i a ige ote f el. The cells. of bottle single a in levels peptide (2.5)RTLSS supplemented forskolinwith (0, 10" , O L0 cls Cls ee nuae fr 4 in 24h for incubated were Cells cells. L103COR variance: Fratio 13.99,= p=0.002. Effect forskolinof on ACTH precursor secretion by groupmean and95% confidence interval shownareto h ih o h dt e. eut ofanalysis Result of theright ofdatathe set. 0 n 1 M Ec dt pit represents point data Each 10M) and 10 O L103 COR F o r s k o l i n ( M ) -8 199 7 - -6 iv.c) POMC RNA levels

An increase in POMC RNA levels was observed in response to all three doses of forskolin both in COR L24 and COR L103 cells (Table 18). The set of three cell pellets obtained for each concentration was pooled before extraction of RNA, therefore statistical analysis of the single results was not possible.

B. Protein Kinase C

AVP is believed to be an important physiological regulator of POMC in the anterior pituitary and acts predominantly via protein kinase C. The phorbol ester, phorbol 12,13-dibutyrate directly activates protein kinase C and was used to assess this intracellular signalling pathway in SCLC cells, i. Phorbol ester i.a) Experimental design

The effect of the protein kinase C activator, phorbol 12, 13-dibutyrate was assessed in COR L24 cells. Cells were incubated for 24h in RTISS (2.5) -8 -7 containing phorbol 12, 13-dibutyrate at final concentrations of 10 M, 10 M and 10 -fi M. This concentration range was based on previous published studies using other in vitro systems (Suda et al 1989). The phorbol ester was solubilised in DMSO and equivalent volumes of DMSO were added to control cultures (final concentration <0.5%). Three separate bottles of cells were incubated at each concentration, i.b) Peptide secretion

Analysis of variance indicated that statistically significant differences existed between the groups (F ratio = 13.99, p = 0.002). Calculation of

95% confidence intervals showed that this was due to an increase in ACTH -7 . precursor levels with 10 M phorbol 12, 13-dibutyrate. Levels in the

200 TABLE 18

Effect of forskolin on POMC RNA content of COR L24 and COR L103 cells

Forskolin Fractional change in POMC : B-actin RNA XMI

COR 124 COR L103

0 1.0 1.0

10‘8 1.7 2.2

10'7 1.85 2.25

106 1.65 1.7

Three separate bottles of cells were incubated at each concentration of bromocriptine. The three cell pellets were pooled before extraction of RNA. Results are expressed as the fractional change in the ratio of POMC mRNA densitometric units to the B-actin mRNA densitometric units, the control value for each cell line being designated 1.0.

201 groups treated with 10 ”8 and 10 *6 M phorbol 12,13-dibutyrate did not differ from the control cells (Figure 77). i.c) POMC RNA levels

The yield of RNA was insufficient for analysis in this experiment.

202 iue7 Efc o a hro etr n CH precursor ACTH on ester phorbol a of Effect Figure77 ACTH precursors (pmol/mg protein) 40- 50-. 20 30- iuyae(, 0 1 ad1 M. Each datapoint and10dibutyrate M). (0, , 10 10 analysisof variance: Fratio = 13.99, p= 0.002. group Themean confidence 95% intervaland cells. 24h inRTISs (2.5) supplemented with phorbol12, 13- r son o h rgt f h dt set. Result of rightare data shownthe the to of secretion by COR L24 Cells cells.wereincubated for ersns etd lvl i a ige ote of bottle single a in levelsrepresents peptide - ) 1 0 -8 -8 0 1 ) P h o r b o l 1 2 , 1 3 - d i b u t y r a t e ( M ) O L24 COR 203 7 - -6 DISCUSSION

There are serious limitations, both practical and ethical, on clinical

studies of the regulatory mechanisms affecting ACTH secretion by tumours.

No human corticotroph cell lines exist and this thesis presents the first

detailed study of POMC production and regulation in human non-pituitary

tumours, focusing on SCLC cell lines.

I. PRIMARY CULTURE OF HUMAN LUNG AND PITUITARY TUMOURS.

i) Small Cell Lung Cancer and Carcinoid Tumours.

The aim of this part of the project was to establish SCLC cell lines which

secreted ACTH-related peptides in order to use these as an in vitro model

of human non-pituitary POMC production. High success rates (50-90%) in

establishing cell lines from SCLC tumours have been achieved in recent

years following the development of the HITES-supplemented selective medium

(Carney et al 1985, Baillie-Johnson et al 1985) and essentially the

techniques employed were based on these two reports. In this study,

difficulties were experienced in establishing cell lines and a number of

factors were identified as being important for ensuring the continued

proliferation of tumour cells in vitro. These included minimal delay in

transit to the laboratory and establishing primary cultures and a high

initial yield of viable tumour cells from the sample. Since the majority

of samples were obtained from patients in distant hospitals, some delays

were inevitable and may have resulted in deterioration of the tumour cells.

More than 50% of the SCLC samples were bronchoscopic biopsies. It was

considered important to attempt to exploit this potential source of cells

from the primary tumour, since SCLC is rarely amenable to surgical

resection and therefore the majority of cell lines have been established

from metastatic sites. This study merely confirmed that these biopsies are

204 unrewarding for this purpose because they are too small and frequently

contaminated with bacteria or fungi.

The primary aim was unfulfilled insofar as the small number of cell lines

resulting from the work were either non-SCLC or non-POMC producers. This problem was circumvented by the success of other groups, leading to the availability of large panels of well characterised SCLC cell lines,

ii) Pituitary Tumours

The culture of pituitary corticotroph tumours, although not resulting in a permanent cell line, was crucial as the initial source of human POMC and pro-ACTH standards in sufficient quantities for the development of the precursor assay (Crosby et al 1988). and for a limited number of short term hormone release experiments. However the small size and short lifetime of these primary cultures mean that they are unsuitable for more complex or repeated experiments. Although the histological nature of these tumours would suggest that the chances of success are small, the potential value of a human corticotroph cell line justifies further effort in this area.

II. ACTH PRECURSORS IN SCLC CELL LINES

Using two immunoradiometric assays for the detection of ACTH precursors and

ACTH, this study has shown that in the panel of SCLC cell lines assessed,

ACTH precursors are secreted but ACTH 1-39 cannot be detected to any great extent. ACTH precursor peptides were synthesised and secreted at significant levels by 10 of the 18 cell lines (56%). The low levels of ACTH detected in the seven SCLC cell lines were measured in the presence of high levels of ACTH precursors and reflect the cross-reactivity of precursors in the ACTH IRMA rather than the presence of ACTH 1-39 itself. Thus the two-site IRMA for the ACTH precursors provides a more sensitive approach for the detection of POMC expression in SCLC cell lines and because it is a direct quantitative method has demonstrated that the overwhelming

205 majority of the ACTH secreted is in HMW precursor forms. Classical Sephadex

G-75 chromatography with measurement of fractions in both the ACTH IRMA and

the precursor IRMA confirmed that ACTH precursors predominated in culture medium and that COR L103 cells secreted almost equal amounts of POMC and pro-ACTH (Figure 17). No significant levels of ACTH were detected and while the presence of very small amounts of ACTH 1-39 cannot be ruled out

from this data alone, it is clear that processing of POMC beyond pro-ACTH occurs to a negligible extent, if at all, in these tumour cells.Measurement of ACTH and precursors in cell pellet extracts indicated that the peptides were not sequestered intracellularly and no non-secreting cell line contained measurable intracellular peptides.

It was noted that peptide levels secreted by SCLC cells were influenced by culture conditions and an attempt was made to optimise these for experimental purposes. The rise in ACTH precursor peptides in medium which occurred as the cells proliferated reflected accumulation of a stable peptide and the decline in precursor levels observed at 14 days coincided with increasing cell death as the nutrient medium became exhausted and possibly with increased release of proteolytic enzymes. The stability of the precursor peptides in culture medium maintained at 378C suggests that enzymatic degradation was minimal when the medium was recovered from actively growing cultures in which the proportion of dead or dying cells was low. Both ACTH and precursor peptides have been found to be relatively stable in blood or plasma at room temperature (White et al 1987, Crosby et al 1988).

The SCLC cell lines are difficult to culture because of their slow growth and because they are prone to low viability if they are grown at high density or if the cell aggregates are dispersed too fiercely. Therefore, extreme care has to be exercised in the management of the cells. Growth

206 of the cells from liquid nitrogen to give sufficient for 9 x 25ml cultures

takes several months and it is necessary to grow large numbers of different

cultures for different experiments which have to be planned many months in

advance. The variability in the levels of peptide produced by individual

cell lines in several 24h incubations of cells (Table 11) was in keeping

with similar observations by Pettengill and co-workers (1985) and was

thought to be due to subtle variations in medium composition rather than

"phenotypic drift". In order to minimise this problem, careful records of passage number were maintained and stocks were regularly frozen down in

liquid nitrogen so that cells could be retrieved from similar stages. To ensure consistency in experiments involving the use of secretagogues, a

large quantity of cells was generated over several passages and used as the source of all cells required to conduct triplicate test and control determinations. Cells for experimental incubations were harvested at a closely similar phase of growth and from cultures of high viability.

III. ACTH AND ACTH PRECURSOR SECRETION BY PITUITARY CELLS

Measurement of ACTH and precursor levels in culture medium from human pituitary cells highlighted a number of points of interest and contrast with the SCLC cells. Although the levels quoted in Table 8 are randomly sampled in terms of time and the precise number of corticotroph cells could not be determined in a heterogeneous primary culture, it is clear that these tumour cells secrete extremely high levels of both ACTH and ACTH precursors. In general, HMW forms constitute a smaller proportion of ACTH immunoreactivity in the plasma of patients with pituitary dependent

Cushing’s syndrome than in the ectopic ACTH syndrome (Hale et al 1986,

Crosby et al 1988), except where the pituitary tumour is large and aggressive.

207 The working range of both the ACTH and ACTH precursor assays is wide (4.9 -

>1110pmol/l and 29 - 2600pmol/l respectively, coefficient of variation

<10%) and no difficulties were encountered in measuring peptides in any of

the SCLC cell lines. In several of the pituitary tumours, peptide levels

in medium were extremely high and some samples did not dilute in parallel

in either the precursor or the ACTH IRMA. Although these assays use excess

reagents, it appeared that the labelled antibody (common to both assays) was effectively removed by precursors in the ACTH assay and vice versa.

These samples would require prior separation of peptides in order to quantitate them.

The precursor assay has also shown that the mouse corticotroph cell line,

AtT20, secretes large quantities of ACTH precursors. Although secretion of precursor peptides by AtT20 cells has been noted in earlier chromatographic studies, molar quantitation was not possible and the limited cross­ reactivity of the antisera employed has underestimated the amount secreted

(Mains and Eipper 1981). However the levels of both ACTH and precursor peptides secreted are approximately 1000-fold higher than produced by the human SCLC cells, presumably reflecting higher level expression of the POMC gene.

IV. EXPRESSION OF POMC RNA IN SCLC CELLS

Slot blot analysis of POMC RNA throughout the growth of the cell line , COR

L103, demonstrated that the expression of the POMC gene remains at a fairly constant level and is not influenced by the changes from log-growth to stationary phase. This is in keeping with the steady accumulation of peptides in medium (Figure 16) and the fairly constant secretion rate per cell observed in a timed incubation of SCLC cells at mid-log and confluence (Table 12). The mature POMC mRNA from COR L103 cells consists of approximately 1350 bases and is present at a very low level relative to

208 that in the pituitary. It is of slightly larger size than that found in

the normal human pituitary. A POMC transcript of similar size has been

described in a number of ectopic ACTH-producing tumours (DeBold et al 1983;

Steenbergh et al 1984; de Keyzer et al 1985; Hoppener et al 1986; Clark et

al 1989). It has been suggested that the origin of this larger transcript may be from a second upstream promoter (Clark et al 1989) although this

remains to be investigated in this case. Translation would be directed from the same initiation codon as the normal pituitary promoter and therefore no additional amino acid sequence would be predicted.

The SCLC cell lines thus far examined express POMC at a low level relative to the anterior pituitary. Endocrinological data is not available on the patients from whom these cell lines were derived but we are not aware that any exhibited the ectopic ACTH syndrome. The level of gene expression and peptide secretion might be expected to be higher in such cases.

V. REGULATION OF POMC GENE EXPRESSION AND PEPTIDE SECRETION IN SCLC CELLS. i) GLUCOCORTICOIDS

A hallmark of the "ectopic” ACTH syndrome is its failure to respond to negative feedback by glucocorticoids, and this feature is exploited clinically in the dexamethasone suppression test. This study of the effect of glucocorticoid on the expression of the POMC gene by SCLC cells in vitro , has demonstrated that the cell lines are relatively resistant to glucocorticoid suppression, exhibiting no reduction of POMC peptide secretion or of POMC mRNA levels in response to concentrations of hydrocortisone (500 and lOOOnM) that would normally suppress the expression and secretion of POMC (Keller-Wood and Dallman 1984). This relative resistance is emphasised by the more normal POMC suppression found in the

AtT20 cell line following a 10 day incubation with hydrocortisone, which has been described previously (Phillips and Tashjian 1982). No suppression

209 of ACTH was detected after 24h in the AtT20 cell line, probably because

this time interval was too short for the early effects of glucocorticoid on POMC transcription rate to be reflected in a change in the relatively

large pool of cytoplasmic mRNA and POMC peptides. However, the increase

in ACTH precursors observed in the 24h incubation of AtT20 cells

(Experiment 1, Figure 50) suggests that hydrocortisone might influence processing to ACTH.

The GR and TAT genes are recognised as being regulated by glucocorticoids

in other tissues and cell lines studied (Hollenberg et al 1985, Rosewicz et al 1988, Berkowitz et al 1988, Diesterhaft et al 1980). The finding that although both genes are expressed in COR L103 cells, neither responds significantly to the doses of hydrocortisone employed in these experiments, suggests that the resistance to glucocorticoids might be global and that any defect must be at a level in the regulatory pathway that is common to multiple genes.

The first common link is the glucocorticoid receptor itself and because expression of a gene at the RNA level does not necessarily imply that the protein product will be produced, identification of GR protein in the SCLC cells was essential. The glucocorticoid receptor was expressed at the protein level in the COR L103 cell line as shown by the nuclear dexamethasone binding study. The kinetics of binding were similar to those of GRs in other tissues and cells (Rousseau et al 1972). This demonstration of nuclear GR in a SCLC cell line is of interest since there is a previous report that SCLC solid tumours contained no cytosolic GR (Chaudhuri et al

1982). However, it is clear that absence of GR is not the explanation for the non-suppression of POMC in the SCLC cell lines. Northern blot analysis of GR mRNA from COR L103 cells did not reveal any obvious size abnormality to suggest a partial gene deletion (Dr A. Clark, data not shown), although

210 this does not exclude the possibility of a point mutation in a region not

required for steroid binding.

Lack of steroid hormone responsiveness in the presence of apparently normal

steroid binding has been described in cancer cells in relation to oestrogen

and progesterone receptors (Nawata et al 1981, Thorpe and Rose 1986).

Transfection of new steroid -responsive DNA into the genome of an

apparently steroid insensitive cell line resulted in normal steroid

responses (Darbre and King 1987) in the transfected sequence, implying that

the steroid signalling apparatus was fully functional. However, the cells became non-responsive following prolonged withdrawal of steroid. These observations led to the hypothesis that an irreversible change, such as DNA methylation, takes place in or around the GREs of the transfected gene. A

similar explanation could account for the lack of glucocorticoid responsiveness of the POMC gene in extrapituitary tumours. In the SCLC cell

line, the change, whatever its nature, affects multiple genes. It is unlikely that it is merely an in vitro phenomenon since the POMC gene is

clearly not suppressed by glucocorticoids in SCLC tumours in vivo.

Hydrocortisone was selected for these experiments since it is the most

significant natural glucocorticoid in man and had the advantage that the

levels in medium could be monitored by cortisol assay. For the future, the effects of more potent steroids such as dexamethasone should also be evaluated in order to explore the nature of the glucocorticoid resistance

in the SCLC cell lines, with the aim of establishing whether the failure of signalling is relative or absolute. The lack of availability of a suitable human control cell line is a limitation since comparison with mouse corticotrophinoma cells is perhaps not ideal.

Broadly, there are three major levels at which a defect in glucocorticoid signalling could occur: the endogenous receptor (which although expressed,

211 has not been definitively proven to be fully functional), the upstream elements of DNA involved in binding the GR complex and the tissue specific factors required for glucocorticoid responsiveness. The question of whether the endogenous GR is fully functional in these SCLC cells could be answered by transfecting with a glucocorticoid responsive sequence linked to a marker gene, for example, the mouse mammary tumour virus promoter linked to chloramphenicol amino-transferase (MMTV-CAT). If CAT could be switched on in the transfected cells by incubation with glucocorticoids, the functional integrity of the endogenous receptor would be confirmed. Failure to respond would suggest a receptor defect and co-transfection of MMTV-CAT with the hGR as an expression vector (Hollenberg et al 1985) should restore glucocorticoid responsiveness. In the event that these experiments demonstrated the endogenous GR to be normal, the next step would be to transfect in DNA sequences to which the GR complex binds, again linked to

CAT as a marker gene. Failure of these transfectants to respond to

glucocorticoids would suggest that tissue specific factors might be defective.

There has been a presumption that glucocorticoid resistance in these SCLC

cells represents a defective regulatory pathway. An alternative

interpretation which is perhaps consistent with the knowledge that the POMC

gene is expressed in normal lung, is simply that in this tissue, POMC is

subject to different regulatory factors than in the anterior pituitary.

ACTH levels have been reported to increase in response to glucocorticoids

in a human ectopic ACTH-secreting tumour (Roth et al, 1988). In this

study, two cell lines (COR L27 and COR L103) showed paradoxical increases

in ACTH precursor secretion in response to 500nM hydrocortisone, although

not lOOOnM. In a recent study of a human small cell lung cancer cell line,

the vasopressin gene could be both negatively and positively regulated by

212 glucocorticoid depending on the state of activation of the cyclic AMP second messenger pathway (Verbeeck et al, 1991). These pieces of evidence suggest considerable complexity in the action of glucocorticoids at a cellular level,

ii) BROMOCRIPTINE

Dopamine is the major negative regulator of POMC in the neurointermediate

lobe of the pituitary where glucocorticoids have very little effect, in

common with SCLC. In this study, bromocriptine, a dopaminergic agonist, produced a dose-dependent suppression in secreted ACTH precursor peptides

accompanied by an increase in POMC RNA in COR L24 cells after 24h

incubation. A 7d incubation reduced levels of secreted precursors to <33%

of control values in the three cell lines tested. The dose of lOug/ml bromocriptine is, however, very high and had clearly inhibited cell

proliferation in the cultures treated for 7d. However, a specific effect

is suggested by the accompanying increase in POMC RNA, indicating that

dopaminergic inhibition may be at the level of translation but not

transcription of the POMC gene. Experiments carried out after completion

of this Thesis have demonstrated that the inhibition of proliferation and

hormone synthesis was reversible when the cells were washed and reincubated

in bromocriptine-free medium (Farrell et al, submitted for publication;

data not shown). The specificity of the effect could be further explored

by simultaneous analysis of other hormones produced by the cell line, for

example oestradiol, whose synthesis is unlikely to be regulated by

dopamine.

Bromocriptine has proved to be of very limited use in pituitary dependent

Cushing’s syndrome, even where analysis of the peptides secreted into

plasma sugggested an intermediate lobe-Cell origin (Lamberts et al 1980,

Hale et al 1988). The response of the SCLC cells to bromocriptine in vitro

213 suggests a potential therapeutic application in the ectopic ACTH syndrome, a condition where the gross hypercortisolism can be difficult to control.

There is at least one case report where the drug was employed to good effect in a case of ectopic ACTH syndrome due to a bronchial carcinoid tumour (Reith et al 1987). iii) CORTICOTROPHIN RELEASING FACTOR

The CRF test is now well established in the differential diagnosis of

Cushing’s syndrome. Typically, patients with the ectopic ACTH syndrome show no cortisol response to an intravenous bolus of lOOug CRF, in contrast to pituitary-dependent disease in which 80% of patients show an excessive response in serum cortisol levels when compared with normal individuals

(Howlett and Besser, 1989). No increase in ACTH precursor secretion was observed in any of the three SCLC cell lines (COR L24, COR L42, COR L103) exposed to oCRF, in keeping with clinical observations.

Published data on the effect of releasing hormones on ’ectopic’ tumour cells in vitro is scanty. Hirata and coworkers (1979) reported that rat median eminence extract stimulated ACTH release from dispersed thymoma and

MTC cells, obtained from two patients with the. ’ectopic’ ACTH syndrome.

There are problems with the use of median eminence extract in that it is possible that K t or Ca 2+ ions present in the extract might be the true stimulus or indeed the presence of contaminating ACTH might have falsely raised the concentrations in medium. In this study other biogenic amines and releasing hormones were ineffective in MTC cells although ACTH secretion by thymic tumour cells was stimulated by TRH, noradrenaline and serotonin. Sorenson and colleagues reported that the secretion of bombesin

i and calcitonin by SCLC cell lines was modulated by cations (K , Ca ), biogenic amines, cholinergic and adenergic agonists. However, a variety of gastrointestinal and hypothalamic releasing hormones had no effect on

214 bombesin secretion by the SCLC cell line DMS 53 (Sorenson et al 1985).

In this study, primary cultures of human corticotroph cells showed a

greater than four fold increase in ACTH secretion in response to lOOOng/ml

oCRF, accompanied by a fall in precursors, suggesting enhanced processing.

CRF has been shown to alter the pattern of processing of POMC in rat

anterior pituitary, increasing the amount of beta-LPH secreted with a

concomitant reduction in beta-endorphin (Ham and Smyth 1986).

iv) SECOND MESSENGER MECHANISMS IN SCLC CELLS

a) Adenylate cyclase pathway

In the pituitary, the adenylate cyclase pathway mediates the stimulatory

effects of CRF on POMC. It would appear that POMC production by SCLC cells

can be stimulated by increasing intracellular levels of cyclic AMP, since

incubation of SCLC cells with the phosphodiesterase inhibitor, IBMX (0.5mM)

significantly increased ACTH precursor secretion. This agent in

combination with the stable analogue, db-cAMP (lOOnM) increased POMC RNA

in COR L24 cells although this was a single observation and interpretation

must be cautious. In COR L24 cells treated with ImM db-cAMP, the increase

in peptide secretion and POMC RNA was apparent after 24h incubation and was

more marked by 48h suggesting a sustained effect. These results suggest

that the intracellular signalling mechanisms distal to the activation of

adenylate cyclase are intact and that cyclic AMP is capable of activating

the POMC gene in SCLC cells.

Forskolin, a direct activator of adenylate cyclase, caused a small but

statistically significant stimulation of ACTH precursor secretion by COR

L24 at the highest concentration used (10 M) although no increase in

peptide secretion was evident with COR L103 cells. Indeed, in these cells,

there was a statistically significant reduction in ACTH precursor secretion

.7 with 10 M forskolin which is difficult to interpret. An increase in POMC

215 RNA was apparent in both cell lines at all three concentrations of forskolin used. This suggests that the 24h incubation period was insufficiently long for the change in POMC RNA levels to be reflected in an increase in secreted peptide.

The adenylate cyclase system has been shown to regulate other hormones in

SCLC cell lines. Sorenson and colleagues (1984) showed that bombesin secretion could be stimulated by 1 and 5mM db-cAMP in the SCLC cell line,

DMS 53. Lower concentrations were ineffective, b) Protein Kinase C pathway

The phorbol ester, phorbol 12,13-dibutyrate was used to investigate whether it could activate the protein kinase C pathway in SCLC cells as AVP is believed to stimulate POMC in anterior pituitary cells directly by this

.7 route. Phorbol 12, 13-dibutyrate at 10 M produced a modest but statistically significant stimulation of precursor levels in the medium of

COR L24 cells but this was not further increased by increasing the concentration of phorbol ester. Further experiments are required to establish whether the dose-response curve is bell-shaped. However, phorbol esters are known to be toxic to cells at concentrations in excess of 10 “ fiM.

Phorbol esters act acutely on ACTH secretion rather than biosynthesis in anterior pituitary cells (Wand et al 1988; Suda et al 1989) and the effect is not sustained chronically. Therefore it may be that the effect of phorbol esters on SCLC cells is similarly rapid. The reasons why such an effect might not be apparent in this experiment are discussed below.

In the majority of these experiments the incubation period was 24h. This was necessary to allow the secretion of adequate basal concentrations of

ACTH precursors as the POMC gene is expressed at a relatively low level in these SCLC cell lines (White et al 1989). We have shown that ACTH precursor peptides are relatively stable in culture medium (Stewart et al 1989) and

216 a rapid effect of secretagogue, e.g. within 30mins, could be masked by accumulation of peptide over 24h. Our data suggest that POMC peptides are not stored intracellularly to any great extent (Stewart et al 1989) thus the effect of an agent acting predominantly or solely on peptide release would be difficult to detect in this experimental protocol. This caveat also applies to the measurement of total POMC RNA. This end-point is not sufficiently sensitive to detect rapid effects on transcription rate since the number of copies of newly synthesised nuclear POMC transcript is small in comparison with the total pool of cytoplasmic POMC RNA which also has a relatively long half-life.

217 VI SUMMARY AND CONCLUSIONS

The principal findings of this investigation of POMC synthesis and

secretion in SCLC cells can be summarised as follows:

1. A high proportion of SCLC cell lines produce ACTH related peptides

(56% of a panel of 18 SCLC cell lines).

2. ACTH related peptides are secreted predominantly as the high

molecular weight precursor forms, POMC and pro-ACTH, with very

little, if any, processing to authentic ACTH 1-39.

3. There is relative resistance to glucocorticoids in that

hydrocortisone failed to suppress POMC expression in SCLC cells at

doses which significantly reduced POMC RNA and peptide secretion in

the pituitary corticotroph cell line, AtT20.

4. 3 SCLC cell lines failed to respond to CRF at doses which

significantly increased ACTH and precursor secretion in AtT20 cells.

5. POMC gene expression and peptide secretion are stimulated by

increasing intracellular cAMP.

6. POMC peptide secretion is significantly inhibited by the dopamine

agonist, bromocriptine.

These observations suggest that expression of POMC in non-pituitary tissues

may be subject to different controlling factors to those in the pituitary.

The relative resistance to glucocorticoids in SCLC, which has long been

recognised clinically in the ectopic ACTH syndrome, has also been demonstrated in cell lines. The molecular mechanisms underlying this

resistance can now be explored in this in vitro model, for example using

the techniques of transfection as outlined in section V(i). The novel

finding that bromocriptine can inhibit POMC in SCLC cells in vitro, if confirmed, offers a potentially useful therapeutic application in controlling hypercortisolism in the ectopic ACTH syndrome.

218 Many questions remain about the mechanisms regulating POMC gene expression

and peptide secretion in normal and neoplastic human non-pituitary tissues.

These studies confirm the value of SCLC cell lines as a model of human non- pituitary POMC production, while highlighting the need for a human pituitary corticotroph cell line. Future work will be directed at more detailed analysis of regulatory mechanisms governing POMC expression in

SCLC, with particular regard to the differences between secreting and non­ secreting cell lines. The functional role of POMC derived peptides in SCLC or its normal cellular counterpart in the lung remains to be determined.

In conclusion, this study of the expression and regulation of POMC in SCLC has demonstrated that the clinical features associated with ACTH production

in vivo are reflected in the in vitro characteristics of SCLC cell lines.

219 REFERENCES

Abe K., Nicholson W.E., Liddle G.W. , Orth D.N., Island D.P. (1969) Normal and abnormal regulation of beta-MSH in man. Journal of Clinical

Investigation 48, 1580-1585.

Abou-Samra A-B., Catt K.J., Aguilera G. (1986a) Involvement of protein kinase C in the regulation of adrenocorticotropin release from rat anterior pituitary cells. Endocrinology 118, 212-217.

Abou-Samra A-B., Catt K.J., Aguilera G. (1986b) Biphasic inhibition of adrenocorticotrophin release by corticosterone in cultured anterior pituitary cells. Endocrinology 119, 972-977.

Abou-Samra A-B., Harwood J.P., Manganiello V.C., Catt K.J., Aguilera G.

(1987) Phorbol 12-myristate 13-acetate and vasopressin potentiate the effect of corticotropin releasing factor on cyclic AMP production in rat anterior pituitary cells. Journal of Biological Chemistry 262, 1129-1136.

Affolter H-U., Reisine T. (1985) Corticotropin releasing factor increases proopiomelanocortin messenger RNA in mouse anterior pituitary tumor cells.

Journal of Biological Chemistry 260, 15477-15481.

Aguilera G., Harwood J.P., Wilson J.X., Morell J., Brown J.H., Catt K.J.

(1983) Mechanisms of action of corticotropin releasing factor and other regulators of corticotropin release in rat pituitary cells. Journal of

Biological Chemistry 258, 8039-8045.

220 Akerblom I.E., Slater E.P., Beato M., Baher J.D., Melcon P.L. (1988)

Negative regulation by glucocorticoid through interference with a cAMP

responsive enhancer. Science 241, 350-353.

Al-Dujaili E.A.S., Hope J., Estivariz F.E., Lowry P.J., Edwards C.R.W.,

(1981) Circulating human pituitary pro-gamma-melanotropin enhances the

adrenal response to ACTH. Nature 291, 156-159.

Al-Dujaili E.A.S. , Williams B.C. , Edwards C.R.W. , Salacinski P. , Lowry P.J.

(1982) Human gamma-melanotropin precursor potentiates corticotropin-induced

adrenal steroidogenesis by stimulating mRNA synthesis. Biochemical Journal

204, 301-305.

Allen R.G., Orvoll E., Kendall J.W., Herbert E., Paxton H. (1980) The

distribution of forms of ACTH and endorphin in normal, tumorous and autopsy

anterior pituitary tissue: virtual absence of 13K ACTH. Journal of Clinical

Endocrinology and Metabolism 51, 376-380.

Baillie-Johnson H., Twentyman P.R., Fox N.E., Walls G.A., Workman P.,

Johnson N., Reeve J.G., Bleehen N.M. (1985) Establishment and

characterisation of cell lines from patients with lung cancer

(predominantly small cell carcinoma). British Journal of Cancer. 52, 495-

504.

Baylin S.B., Mendelsohn G. (1980) Ectopic (inappropriate) hormone production by tumours: mechanisms involved and the biological and clinical

implications. Endocrine Reviews 1, 45-77.

221 Baylin S.B. (1985) The implications of differentiation relationships between endocrine and non-endocrine cells in human lung cancers. In: Hesch

R.D. (ed. ) Peptide Hormones as Mediators in Immunology and Oncology Vol 19, ppl-12, Raven Press. Serono Symposia Publications. New York.

Becker K.L., Snider R.H., Silva O.L., Moore C.F. (1978) Calcitonin heterogeneity in lung cancer and medullary thyroid cancer. Acta

Endocrinologica 89, 89-99.

Becker K.L. (1984) The Endocrine Lung. In: Becker K.L. & Gazdar A.F.(eds.)

The Endocrine Lung in Health and Disease pp3-46. W.B.Saunders Company.

Philadelphia, USA.

Bepler G. , Carney D.N., Gazdar A.F., Minna J.D. (1987) In vitro growth inhibition of human small cell lung cancer by physalaemin. Cancer Research

47, 2371-2375.

Bergh J. , Nilsson K. , Ekman R., Giovanella B. (1985) Establishment and characterisation of cell lines from human small cell and large cell carcinomas of the lung. Acta Pathologica Microbiolica et Immunologica

Scandinavica 93, 133-147.

Berkowitz G. D. , Carter K. M. , Migeon C. J., Brown T. R. (1988). Down- regulation of the glucocorticoid receptor by dexamethasone in cultured human skin fibroblasts : implications for the regulation of aromatase activity. Journal of Clinical Endocrinology & Metabolism 66, 1029-1036.

222 Bernton E.W. , Beach J.E., Holaday J. W. , Smallridge, R. C. Fein H.G.

(1987). Release of multiple hormones by a direct action of

interleukin-1 on pituitary cells. Science 238, 519-521.

Bertagna X.Y. , Nicholson W.E., Sorenson G.D., Pettengill O.S., Mount C.D.,

OrthD.N. (1978) Corticotrophin, lipotrophin and beta-endorphin production

by a human non pituitary tumor in culture: Evidence of a common precursor.

PNAS USA 75, 5160-5164.

Bertagna X.Y., Seurin D., Pique L. , Luton J-P., Bricaire H., Girard F.

(1983) Peptides related to the N^-terminal end of pro-opiocortin in man.

Journal of Clinical Endocrinology and Metabolism 56, 489-495.

Bertagna X.Y., Camus F., Lenne F. , Girard F. , Luton J-P. (1988) Human joining peptide: A proopiomelanocortin product secreted as a homodimer.

Molecular Encocrinology 2, 1108-1114.

Birnberg N.C., Lissitzky J.C., Hinnman M. , Herbert E. (1983)

Glucocorticoids regulate proopiomelanocortin gene expression in vivo at the

levels of transcription and secretion. PNAS USA 80, 6982-6986.

Blalock J.E. (1985) Pro-opiomelanocortin-derived peptides in the immune system. Clinical Endocrinology 22, 823-827.

223 Blobel G. , Dobberstein B. (1975) Transfer of proteins across membranes 1.

Presence of proteolytically processed and unprocessed nascent

immunoglobulin light chains on membrane-bound ribosomes of murine myeloma.

Journal of Cell Biology 67, 835-851.

Bloomfield G.A., Scott A.P., Rees L.H., Lowry P.J., Ratcliffe J.G, (1974)

A reappraisal of human beta-MSH. Nature 252, 492-493.

Bloomfield G.A., Holdaway I.M., Corrin B., Ratcliffe J.G., Rees G.M.,

Ellison M., Rees L.H. (1977) Lung tumours and ACTH production. Clinical

Endocrinology 6, 95-104.

BondyP.K., (1985) Disorders of the adrenal cortex. In: Wilson J.D., Foster

D.W. (eds.) Textbook of Endocrinology, pp816-890. W.B.Saunders.

Philadelphia, USA,

Bork E. , Hansen M. , Urdal P., Paus E. , Holst J. J. , Schifter S. , Fenger M. ,

Engbaek F. (1988) Early detection of response in small cell bronchogenic

carcinoma by changes in serum concentrations of creatine kinase, neuron

specific enolase, calcitonin, ACTH, serotonin and gastrin releasing

peptide. European Journal of Cancer and Clinical Oncology 24, 1033-1038.

Born J. , Fehm H.L., Voigt K.H. (1986) ACTH and attention in humans: A

review. Neuropsychobiology 15, 165-186.

Breathnach R. , Chambon P. (1981) Organisation and expression of eucaryotic split genes coding for protein. Annual Review of Biochemistry 50, 349-383.

224 Brown R.A., Swanson Beck J. (1990) Medical statistics on microcomputers.

BMJ Publications Ltd.. London, UK.

Brownie A.C., Pederson R.C. (1987) Effect of ACTH and other pro­ opiomelanocortin-derived peptides on aldosterone secretion. In: Li C.H.

(ed), Hormonal Proteins and Polypeptides, Vol XIII: ACTH. ppl27-146

Academic Press Inc (London) Ltd.

Bruhn T.O., Sutton R., Rivier C., Vale W. (1984) Corticotropin releasing factor regulates proopiomelanocortin messenger ribonucleic acid levels in vivo. Neuroendocrinology 39, 170-175.

Bruni J.F., Watkins W.B., Yen S.S.C., (1979) Beta-endorphin in the human pancreas. Journal of Clinical Endocrinology and Metabolism 49, 649-651.

Buzetti R., McLoughlin L., Lavender P.M., Clark A.J.L. Rees L.H. (1989)

Expression of pro-opiomelanocortin gene and quantification of adrenocorticotrophic hormone-like immunoreactivity in human normal peripheral mononuclear cells and lymphoid and myeloid malignancies.

Journal of Clinical Investigation 83, 733-737.

Carney D.N., Mitchell J.B., Kinsella T.J. (1983) In vitro radiation and chemosensitivity of established cell lines of human small cell lung cancer and its large cell variants. Cancer Research 43, 2806-2811.

225 Carney D.N., Gazdar A.F., Bepler G., Guccion J.G., Marangos P.J., Moody

T.W., Zweig M.H., Minna J.D. , (1985) Establishment and identification of small cell lung cancer cell lines having classic and variant features.

Cancer Research 45, 2913-2923.

Chan J.S.D., Seidah N.G., Chretien M. (1983) Measurement of N-terminal (1-

76) of human proopiomelanocortin in human plasma: Correlation with adrenocorticotrophin. Journal of Clinical Endocrinology and Metabolism 56,

791-796.

Chaudhuri P.K., Thomas P. A., Walker M. J., Briele H. A., Das Gupta T. K.,

Beattie C. W. (1982) Steroid receptors in human lung cancer cytosols.

Cancer Letters 16, 327-32.

Chen C-L. C., Chang C-C., Krieger D.T., Bardin C.W. (1986) Expression and regulation of pro-opiomelanocortin-like gene in the ovary and placenta:

Comparison with the testis. Endocrinology 118, 2382-2389.

Chirgwin J.M., PrzbylaA.E., MacDonald R.J., Rutter W.J. (1979) Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18, 5294-5299.

Chretien M. , Li C.H. (1967) Isolation, purification and characterisation of beta-lipotrophic hormone from sheep pituitary glands. Canadian Journal of Biochemistry 45, 1163-1174.

226 Clark A.J.L., Lavender P.M., Besser G.M., Rees L.H., (1989) Pro­ opiomelanocortin mRNA size heterogeneity in ACTH-dependent Cushing’s syndrome. Journal of Molecular Endocrinology 2, 3-9.

Clark A.J.L., Stewart M.F., Lavender P.M., Farrell W., Crosby S.R., Rees

L.H. , White A. (1990) Defective glucocorticoid regulation of proopiomelanocortin gene expression and peptide secretion in a small cell lung cancer cell line. Journal of Clinical Endocrinology and Metabolism 70,

485-490.

Coates P.J., Doniach I., Howlett T.A., Rees L.H., Besser G.M. (1986)

Immunocytochemical study of 18 tumours causing ectopic Cushing’s syndrome.

Journal of Clinical Pathology 39, 955-960.

Cochet M. , Chang A.C.Y., Cohen S.N. (1982) Characterisation of the structural gene and putative 5’-regulatory sequences for human pro­ opiomelanocortin. Nature 297, 335-339.

Cote T.E. , Felder R. , Kebabian J.W., Sekura R.D. , Reisine T. , Affolter H.U.

(1986) D-2 Dopamine receptor-mediated inhibition of proopiomelanocortin synthesis in rat intermediate lobe. Journal of Biological Chemistry 261,

4555-4561.

Craig R.K., Edbrooke M.R., Riley J.H., McVey J.H., Parker D. (1985)

Differential expression of the human calcitonin-CGRP gene in medullary thyroid carcinoma and lung carcinoma cell lines. In: Havemann K., Sorenson

G., Gropp C. (eds.) Recent Results in Cancer Research: Peptide Hormones in

Lung Cancer 99, pp71-78. Springer-Verlag. Berlin.

227 Crosby S.R., Stewart M.F., Ratcliffe J.G., White A. (1988) Direct

measurement of the precursors of adrenocorticotropin in human plasma by

two-site immunoradiometric assay. Journal of Clinical Endocrinology and

Metabolism 67, 1271-1277.

Crosby S.R., Stewart M.F., Farrell W.E., Gibson S.,White A. (1990)

Comparison of ACTH and ACTH precursor peptides secreted by human pituitary

and lung tumour cells in vitro. Endocrinology 125, 147-152.

Cuttitta F. , Carney D.N., Mulshine J. , Moody T.W., Fedorko J., Fischler A. ,

Minna J.D. (1985) Bombesin-like peptides can function as autocrine growth

factors in human small cell lung cancer. Nature 316, 823-826.

Cuttitta F., Fedorko J., Gu J. , Lebecq-Verheyden A-M., Linnoila R.I.,

Battey J.F. (1988) Gastrin-releasing peptide gene-associated peptides are

expressed in normal human fetal lung and small cell lung cancer : a novel peptide family found in man. Journal of Clinical Endocrinology and

Metabolism 67, 576-583.

Cuttitta F., Levitt M.L., Park J.G., Kasprzyk P., Nakanishi Y., Reeve J. et al (1987) Growth of human cancer cell lines in unsupplemented basal medium as a means of identifying autocrine growth factors. Proceedings of the American Association for Cancer Research 28, 27.

Darbre P. D., King R. J. B. (1987) Progression to steroid insensitivity can occur irrespective of the presence of functional steroid receptors.

Cell 51, 521-528.

228 DeBold C. R. , Schworer M. E., Connor T. B. , Bird R. E. , Orth D. N. (1983).

Ectopic pro-opiomelanocortin : sequence of cDNA coding for gamma-melanocyte stimulating hormone and beta-endorphin. Science 220, 721-723.

DeBold C.R., Menefee J.K., Nicholson W.E., Orth D.N. (1988) Pro­ opiomelanocortin gene is expressed in many normal tissues and in tumours not associated with ectopic ACTH syndrome. Molecular Endocrinology 2, 862-

870.

Diesterhaft M. , Noguchi T., Granner D. K. (1980). Regulation of rat liver tyrosine amino-transferase mRNA by hydrocortisone and by N 6 0 2- dibutyryladenosine 3,5- monophosphate. European Journal of Biochemistry

108, 357-365.

Drouin J., Charron J. , Gagner J.P. (1987) Proopiomelanocortin gene: a model for negative regulation of transcription by glucocorticoids. Journal of

Cellular Biochemistry 35, 293-304.

Drouin J. , Sun Y. L., Neraer M. (1990). Regulatory elements of the pro­ opiomelanocortin gene : pituitary specificity and glucocorticoid repression. Trends in Endocrinology and Metabolism March/April, 219-225.

Duchesne G.M., Eady J.J., Peacock J.H., Pera M.F. (1987) A panel of human lung carcinoma cell lines: Establishment, properties and common characteristics. British Journal of Cancer 56, 287-293.

229 Dunnett C.W. (1955) A multiple comparison procedure for comparing several treatments with a control. American Statistical Association 50, 1096-1121.

Eagan R.T., Maurer L.H., Forcier R.J., et al (1974) Small cell carcinoma of the lung: staging, paraneoplastic syndromes, treatment and survival.

Cancer 33, 527-532.

Eberwine J.H., Roberts J.L. (1984) Glucocorticoid regulation of pro­ opiomelanocortin gene transcription in the rat pituitary. Journal of

Biological Chemistry 259, 2166-2170.

Eberwine J.H. , Jonassen J.A. , Evinger M.J.Q.Y., Roberts J.L. (1987) Complex transcriptional regulation by glucocorticoid and corticotrophin-releasing hormones of pro-opiomelanocortin gene expression in rat pituitary cultures

DNA 6, 483-492.

Eipper B.A., Mains R.E. (1980) Structure and biosynthesis of pro- adrenocorticotropin/endorphin and related peptides. Endocrine Reviews 1,

1-27.

Eipper B.A., Mains R.E. (1982) Phosphorylation of pro- adrenocorticotropin/endorphin-derived peptides. Journal of Biological

Chemistry 257, 4907-4915.

Eipper B.A. , Park L. , Keutmann H.T. , Mains R.E. (1986) Amidation of joining peptide, a major pro-ACTH/endorphin-derived product peptide. Journal of

Biological Chemistry 261, 8686-8694.

230 Ellison M.L., Hillyard C.J., Bloomfield G.A., Rees L.H., Coombes R.C. ,

Neville A.M. (1976) Ectopic hormone production by bronchial carcinomas in culture. Clinical Endocrinology fSupplI 5, 397S-406S.

Emson P.C., Corder R., Ratter S.J., Tomlin S., Lowry P.J., Rees L.H. ,

Arregui A., Rosser M.N. (1984) Regional distribution of pro­ opiomelanocortin-derived peptides in the human brain. Neuroendocrinology

38, 45-50.

Estivariz F.E., Gillies G.E., Lowry P.J. (1981) Biosynthesis and control of secretion of corticotrophic, lipotrophic and melanotropic peptides.

Pharmacology and Therapeutics. 13, 61-67.

Estivariz F., Iturriza F. , McLean C., Hope J., Lowry P.J. (1982)

Stimulation of adrenal mitogenesis by N-terminal proopiocortin peptides.

Nature 297, 419-422.

Evans C.J., Erdelly E., Weber E., Barchas J.D. (1983) Identification of pro-opiomelanocortin-derived peptides in the human adrenal medulla. Science

221, 957-960.

Farrell W.E., Stewart M.F., ClarkA.J.L., Crosby S.R., White A. (1991)

Bromocriptine inhibits POMC mRNA and ACTH precursor secretion in small cell lung cancer cell lines. Journal of Clinical Investigation (submitted for publication).

Feinberg A.P., Vogelstein B.A. (1983) A technique for labelling DNA restriction endonuclease fragments to high specific activity. Analytical

Biochemistry 132, 6-13.

231 Filliben J.J. , (1975) The probability plot correlation coefficient test for normality. Technometrics 17, 111-117.

Fleischer M.R. , Glass D., Bitensky L., Chayen J., Daly J.R., (1974) Plasma corticotrophin levels during insulin induced hypoglycaemia: comparison of radioimmunoassay and cytochemical bioassay. Clinical Endocrinology. 3, 203-

298.

Gasson J. (1979) Steroidogenic activity of high molecular weight forms of corticotropin. Biochemistry 18, 4215-4224.

Gazdar A.F., Carney D.N., Baylin S.B. et al (1980) Small cell carcinoma of the lung: Altered morphological, biological and biochemical characteristics in long term cultures and heterotransplanted tumours. Proceedings of the

American Association for Cancer Research (Abstract) 21, 51.

Gazdar A.F., Carney D.N., Guccion J.G., Baylin S.B. (1981) Small cell carcinoma of the lung: cellular origin and relationship to other pulmonary tumors. In: Greco F.A., Bunn P.A., Oldham R.K., (eds) Small Cell Lung

Cancer, pp. 145-175. Grune and Stratton. New York.

Gilbert F., Walop W., ChanJ.S.D., Benjannet S., SeidahN.G., SpitzerW.G.,

Chretien M. (1987) The role of the Nl^-terminal portion of pro­ opiomelanocortin as a biomarker for human lung cancer. Clinical and

Investigative Medicine 10, 372-376.

Gillies G.E., Linton E.A., Lowry P.J. (1982) Corticotropin releasing activity of the new CRF is potentiated several times by vasopressin. Nature

299, 355-357.

232 Goverde Pesman G.J., Benraad T.J. (1980) Improved sensitivity to

adrenocorticotrophin after purification and pre-incubation of isolated rat

adrenal cells. Acta Endocrinologica 94, 221-228.

Guild S., Reisine T.D. (1987) Molecular mechanisms of corticotropin-

releasing factor stimulation of calcium mobilisation and

adrenocorticotropin release from anterior pituitary tumor cells. Journal

of Pharmacology and Experimental Therapeutics 241, 125-130

Guillemin R. , Rosenberg B. (1955) Humoral hypothalamic control of anterior pituitary: a study with combined tissue cultures. Endocrinology 57,599-607.

Gumbiner B. , Kelly R.B. (1982) Two distinct intracellular pathways transport secretory and membrane glycoproteins to the surface of pituitary tumor cells. Cell 28, 51-59.

Guttmann S., Biossonnas R.A. (1961) Influence of the structure of the N- terminal extremity of alpha-MSH on the melanophore stimulating activity of

this hormone. Experientia 17, 265-267.

Hage E. (1980) Light and electron microscopic characteristics of the various lung endocrine cell types Investigative and Cell Pathology 3, 345-

351.

Hakanson R. , Sundler F. (1971) Fluorimetric determination of N-terminal tryptophan-peptides after formaldehyde condensation. Biochemical

Pharmacology 20, 3223-3225.

233 Hale A. C. , Price J., Ackland J.C., Doniach I. , RatterS., Besser G.M., Rees

L.H. (1984) Corticotrophin releasing factor-mediated ACTH release from rat anterior pituitary cells is potentiated by C-terminal gastrin releasing peptide. Journal of Endocrinology 102, R1-R3.

Hale A.C., Ratter S.J., Tomlin S.J., Lytras N. , Besser G.M. Rees L.H.

(1984) Measurement of immunoreactive gamma-MSH in human plasma. Clinical

Endocrinology 21, 139-148.

Hale A.C., Besser G.M., Rees L.H. (1986) Characterisation of pro­ opiomelanocortin derived peptides in pituitary and ectopic ACTH secreting tumours. Journal of Endocrinology 108, 49-56.

Hale A. C., Coates P. J., Doniach I., Howlett T. A., Grossman A., Rees L.

H., Besser G. M. (1988). A bromocriptine-responsive corticotroph adenoma secreting «-MSH in a patient with Cushing’s disease. Clinical

Endocrinology 28, 215-223.

Hall P.F. (1987) ACTH and Corticosteroidogenesis. In: Li C.H. (ed.),

Hormonal Proteins and Polypeptides, Vol. XIII: ACTH dp89-125. Academic

Press Inc (London) Ltd.

Ham J., Smyth D. G. (1986). Chronic stimulation of anterior pituitary cell cultures with CRF leads to the secretion of lipotropin. Neuroendocrinology

44, 433-438.

Hansen M., Hansen H.H., Hirsch F.R., Arends J., Christensen J.D.,

Christensen J.M. , Hummer L. , Kuhl C. (1980) Hormonal polypeptides and amine metabolites in small cell carcinoma of the lung with special reference to stage and subtypes. Cancer 45, 1432-1437.

234 Harris J.I. (1959) Structure of a melanocyte-stimulating hormone from the

human pituitary gland. Nature 184, 167-169.

Harwood R. (1980) Protein transfer across membranes: the role of signal

sequences and signal peptidase activity. In: Freedman R.B., Hawkins H.C.

(eds) The Enzymology of Post-Translational Modification of Proteins, vol

1, pp3. Academic Press. London.

Havemann K., Holle R., Gropp C. (1985) Prospective multi-centre study of hormone markers in small cell lung cancer. In: Havemann K., Sorenson G.,

Gropp C. (eds.) Recent Results in Cancer Research 99, ppl94-208, Springer-

Verlag. Berlin.

Herbert E., Comb M., Thomas G., Liston D., Douglas J., Thorne B., Martin

M. (1986) Tissue specific regulation of expression of opioid peptides. DNA

5, 68.

Hirata Y. , Yamamoto H., Matsukura S., Imura H. (1975) In vitro release and biosynthesis of tumor ACTH in ectopic ACTH-producing tumors. Journal of

Clinical Endocrinology and Metabolism 41, 106-114.

Hirata Y. , Yoshimi H. , Matsukura S., Imura H. (1979) Effect of hypothalamic extract and other factors on release of adrenocorticotropin from and adenosine 3’,5 *-monophosphate levels in dispersed non-pituitary tumor cells. Journal of Clinical Endocrinology and Metabolism 49, 317-321.

Hirata Y. , Matsukara S., Fujita T. (1983) ACTH and related peptides in normal and abnormal human tissues. In: Fotherby K., Pal S. B. (eds.),

Hormones in Normal and Abnormal Human Tissues, Vol III ppl69-195, Walter de Gruyter & Co.. New York.

235 Hollenberg S.M. , Weinberger C. , Ong E.S., Cerelli G. , Oro A., Lebo R. ,

Thompson E.B., Rosenfeld M.G., Evans R.M. (1985) Primary structure and expression of a functional human glucocorticoid receptor cDNA. Nature 318,

635-641.

Hope J., Ratter S.J., Estivariz F.E., McLoughlin L., Lowry P.J. (1981)

Development of a radioimmunoassay for an amino-terminal peptide of pro- opiocortin containing the gamma-MSH region: measurement and characterisation in human plasma. Clinical Endocrinology 15, 221-227.

Hoppener J. W., Steenbergh P. H., Moonen P. J., Wagenaar S. S., Jansz H.

S., Lips C. J. (1986). Detection of mRNA encoding calcitonin, calcitonin gene related peptide and pro-opiomelanocortin in human tumours. Molecular and Cellular Endocrinology. 47, 125-30.

Howlett T. A., Price J., Hale A. C., Doniach I. , Rees L. H. , Wass J. A. H. ,

Besser G. M. (1985). Pituitary ACTH dependent Cushing’s syndrome due to ectopic production of a bombesin-like peptide by a medullary carcinoma of the thyroid. Clinical Endocrinology 22, 91-101.

Howlett T., Besser G. M. (1989). Cushing’s syndrome. Medicine

International 63, 2605-2611.

Hozier J.C., Mass M.J. Siegfried J.M. (1987) Genes for tumor markers are clustered with cellular proto-oncogenes on human chromosomes. Cancer

Letters 36, 235-245.

Hughes J. , Smith T.W., Kosterlitz H.W., Fothergill L.A., Morgan B.A.,

Morris H.R. (1975) Identification of two related pentapeptides from the brain with potent opiate agonist activity. Nature 258, 577-579.

236 Jackson S., Salacinski P., Hope J., Lowry P.J. (1983) An investigation of

N-terminal pro-opiocortin peptides in the rat pituitary. Peptides 4, 431-

438.

Jenks B.G., Ederveen A.G.H., Feyen J.H., Overbeeke A.P. (1985) The

functional significance of glycosylation of proopiomelanocortin in

melanotrophs of the mouse pituitary gland. Journal of Endocrinology 107,

365-374.

Jessop D. S., Cunnah D., Millar J. G. B. , Neville E. , Coates P. , Doniach

I., Besser G. M. , Rees L. H. (1986). A phaeochromocytoma presenting with

Cushing’s syndrome associated with increased concentrations of circulating

corticotrophin-releasing factor. Journal of Endocrinology 113, 133-138.

Jingami H. , Nakanishi S., Imura H., Numa S. (1984) Tissue distribution of messenger RNAs coding for opioid peptide precursors and related RNA.

European Journal of Biochemistry 142, 441-447.

Keller-Wood M. E. , Dallman M. F. (1984). Corticosteroid inhibition of ACTH

secretion. Endocrine Reviews 5, 1-24.

de Keyzer Y., Bertagna X. , Lenne F.,Girard F., Luton J.P., Kahn A. (1985)

Altered pro-melanocortin gene expression in ACTH-producing non-pituitary tumours. Comparative studies with corticotropic adenomas and normal pituitaries. Journal of Clinical Investigation 76, 1892-1898.

Kirkland S.C., Ellison M.L. (1984) Secretion of corticotrophin releasing factor-like activity by a human bronchial carcinoid cell line. Journal of

Endocrinology 103, 85-90.

237 Kohler P.C., Trump D.L. (1986) Ectopic hormone syndromes. Cancer

Investigation 4,543-554.

Kok K. , Osinga J., Carritt B. , Davis M.B., van der Hout A.H. , van der Veen

A.Y. , Landsvater R.M., de Leij L.F., Berendsen H.H., Postmus P.E., Poppema

5., Buys C.H.C.M. (1987) Deletion of a DNA sequence at the chromosome 3p

21 in all major types of lung cancer. Nature 330, 578-581.

Lamberts S. W. J., Klijn J. G. M., de Quijada M., Timmermans H. A. T.,

Uitterlinden P., Birkenhager J. C. (1980). Bromocriptine and the medical treatment of Cushing’s disease. In : Muller E. E. (Ed) : Neuroactive drugs in Endocrinology. Elsevier Science Publishing Co. Inc. New York, pp. 371-

382.

de Leij L., Postmus P.E., Buys C.H.C.M., Elema J.D., Ramaekers F., Poppema

5., Brouwer M.m Van der Veen A.Y., Mesander G., Hauw The T. (1985)

Characterisation of three new variant type cell lines derived from small cell carcinoma of the lung. Cancer Research 45, 6024-6033.

Leonard R.C.F. (1989) Small cell lung cancer. British Journal of Cancer 59,

487-490.

Li C.H., Geschwind 1.1., Levy A.L., Harris J.I., Dixon J.S., Pon N.G. ,

Porath J.O. (1954) Isolation and properties of corticotropin from sheep pituitary glands. Nature 173, 251-253.

Li C.H., Geschwind 1.1., Cole R.D., Raacke I.D., Harris J.I., Dixon J.S.

(1955) Amino acid sequence of alpha corticotropin. Nature 176, 687-689.

238 Li C.H., (1964) Lipotropin, a new active peptide from pituitary glands.

Nature 201, 924-925.

Li C.H., Barnafi L. , Chretien M., Chung D. (1965) Isolation and amino acid sequence of beta-LPH from sheep pituitary glands. Nature 208, 1093-1094.

Li C.H., Chung D. (1976) Isolation and structure of an untriakontapeptide with opiate activity from camel pituitary glands. PNAS USA 73, 1145-1148.

Li Z.G., Park D., LaBella F.S. (1989) Adrenocorticotropin (1-10) and (11-

24) promote adrenal steroidogenesis by different mechanisms. Endocrinology

125, 592-596.

Liotta A. , Osathanondh R., Krieger D.T., (1977) Presence of ACTH in human placenta: Demonstration of in vitro synthesis. Endocrinology 101, 1552-

1558.

Liotta A.S., Krieger D.T. (1980) In vitro biosynthesis and comparative post-translation processing of immunoreactive precursor corticotrophin/beta-endorphin by human placental and pituitary cells.

Endocrinology 106, 1504-1511.

Liotta A.S., Falaschi P., Krieger D.T. (1982) Characterisation of pro- poiomelanocortin (POMC)-related peptides in fresh human(H) anterior pituitary (AP). Proceedings of 64th Annual Meeting of the Endocrine Society

Abstract 221.

Loh Y.P. , Gainer H. (1982) Processing of normal and non-glycosylated forms of toad pro-opiocortin by rat intermediate (pituitary) lobe pro-opiocortin converting enzyme activity. Life Sciences 31, 3043-3050. Lowry P.J., Silas L., McLean C., Linton E.A., Estivariz F.E. (1983) Pro- gamma-melanocyte stimulating hormone cleavage in adrenal gland undergoing compensatory growth. Nature 306, 70-73.

Luini A., Lewis D., Guild S., Schofield G., Weight F. (1986) Somatostatin, an inhibitor of ACTH secretion, decreases cytosolic free calcium and voltage-dependent calcium current in a pituitary cell line. Journal of

Neuroscience 6, 3128-3132

Lundblad J. R., Roberts J. L. (1988). Regulation of pro-opiomelanocortin gene expression in pituitary. Endocrine Reviews 9, 135-158.

Luster W., Gropp C., Sostmann H. , Kalbfleisch H. , Havemann K. (1982)

Demonstration of immunoreactive calcitonin in sera and tissues of lung cancer patients. European Journal of Cancer and Clinical Oncology 18, 1275-

1283.

Luster W. , Gropp C., Kern H.F., Havemann K. (1985) Lung tumour cell lines established from tumours of four histological types: Characterisation of the cell lines and analysis of their peptide hormone production. British

Journal of Cancer 51, 865-875.

Macaulay V.M., Teale J.D., Everard M.J., Joshi G.P., Smith I.E., Millar

J.L. (1988) Somatomedin-C/insulin-like growth factor-1 is a mitogen for human small cell lung cancer. British Journal of Cancer 57, 91-93.

Mains R.E., Eipper B.A., Ling N. (1977) Common precursor to corticotropins and endorphins. PNAS USA 74, 3014-3018

240 Mains R.E. , Eipper B.A. (1981) Differences in the post-translational processing of beta-endorphin in rat anterior and intermediate pituitary. Journal of Biological Chemistry 256,5683-5688

May V. , Eipper B.A. (1986) Long term culture of primary rat pituitary adrenocorticotropin/endorphin-producing cells in serum-free medium.

Endocrinology 118, 1284-1295.

Mezey E., Reisine T., Skirboll L., Beinfeld M. , Kiss J. (1985) In:

Vanderhaegen J. , Crawley J. (eds), Neuronal Cholecystokinin ppl52-156, New

York Academy of Sciences. New York.

Moody T.W. (1984) Bombesin-like peptides in the normal and malignant lung.

In: Becker K.L., Gazdar A.F., (eds) The Endocrine Lung in Health and

Disease, pp328-335. W.B. Saunders Company Philadelphia, USA.

Mulshine J., Cuttitta F., Avis I., Treston A., Kasprzyk P., Carrasquillo

J., et al. (1988) Phase I evaluation of an anti-gastrin releasing peptide

(GRP) monoclonal antibody (MoAb) in patients with advanced lung cancer.

Proceedings of the American Society of Clinical Oncology 7, 213.

Nakanishi S., Inoue A., Kita T. , Nakamura M., Chang A.C.Y., Cohen S.N.,

Numa S. (1979) Nucleotide sequence of cloned cDNA for bovine corticotropin- beta-lipotropin precursor. Nature 278, 423-427.

Nakanishi Y., Mulshine J.L., Kaspryzyk P.G., Natale R.B., Maneckjee R. ,

Avis I. et al (1988) Insulin-like growth factor-1 can mediate autocrine proliferation of human small cell lung cancer cell lines in vitro. Journal of Clinical Investigation 82, 354-359.

241 Natt E. , Kao F.T. , Rettenmeier R. , Scherer G. (1986) Assignment of the human tyrosine aminotransferase gene to chromosome 16. Human Genetics 72,

225-228.

Nawata H. , Chong M. T. , Bronzert D. , Lippmann M. E. (1981). Estradiol- independent growth of a subline of MCF-7 human breast cancer cells in culture. Journal of Biological Chemistry 256, 6895-6902.

Oboshi S. , Tsugawa S., Seido T., Shimosato Y., Koide T., Ishikawa S. (1971)

A new floating cell line derived from human pulmonary carcinoma of oat-cell type. GANN 62, 505-516.

Ohara H., Okamoto T (1977) A new in vitro cell line established from human oat cell carcinoma of the lung. Cancer Research 37, 3088-3095.

Oliva D., Nicosia S., Spada A., Giannattasio G. (1982) VIP stimulates ACTH release and adenylate cyclase in human ACTH-secreting pituitary adenomas.

European Journal of Pharmacology 83, 101-105.

Orth D.N., Nicholson W.E., Mitchell W.M., Island D.P., Liddle G.W. (1973a)

Biologic and immunologic characterisation and physical separation of ACTH and ACTH fragments in the ectopic ACTH syndrome. Journal of Clinical

Investigation 52, 1756-1769.

Orth D.N., Nicholson W.E., Mitchell W.M., Island D.P., Shapiro M., Byyny

R.L. (1973b) ACTH and MSH production by a single cloned mouse pituitary tumor cell line. Endocrinology 92, 385-393.

242 Orth D.N., Nicholson W.E. (1977) High molecular weight forms of human ACTH

are glycoproteins. Journal of Clinical Endocrinology and Metabolism 44,

214-217

Owerbach D. , Rutter W.J., Roberts J.L. et al (1981) The pro-opiocortin

(adrenocorticotropin/beta-lipotropin) gene is located on chromosome 2 in

humans. Somatic Cell Genetics 7, 359-369.

Page S.R., Ang V.T.Y., Jackson R. , White A., Nussey S.S., Jenkins J.S.

(1990) The effect of oxytocin infusion on adenohypophyseal function in man.

Clinical Endocrinology 32, 307-313.

Parker L. , Lifrak E. , Shively J. , Lee T. , Kaplan B. , Walker P., Calcycay

J., Florsheim W. , Soon-Shiong S. (1989) Human adrenal gland cortical

androgen-stimulating hormone (CASH) is identical with a portion of the

joining peptide of pituitary pro-opiomelanocortin (POMC). In: Abstracts of

the 71st Annual Meeting of the Endocrine Society, Seattle, USA, 299.

Pearse A.G.E.,(1969) The cytochemistry and ultrastructure of polypeptide

hormone-producing cells of the APUD series and the embryologic, physiologic

and pathologic implications of the concept. Journal of Histochemistry

and Cytochemistry 17, 303-313.

Pedersen R.C., Brownie A.C., Ling N. (1980)

Pro-adrenocorticotrophin/endorphin-derivedpeptides: co-ordinate action on

adrenal steroidogenesis. Science 208, 1044-1046.

Pelletier G., Leclerc R., Labtrie F. , Cole J., Chretien M , Lis M. (1977)

Immunohistochemical localisation of beta-lipotropic hormone in the pituitary gland. Endocrinology 100, 770.

243 Petrusz P. , Merchenthaler I. , Maderdrut J.L. , Heitz P.U. (1985) Central and peripheral distribution of corticotropin-releasing factor. Federation

Proceedings 44, 229-235

Pettengill O.S., Curphey T.J., Cate C.C., Flint F., Maurer L.H., Sorenson

G.D. (1980a) Animal model for small cell carcinoma of the lung: effect of immunosuppression and sex of mouse on tumor growth in nude athymic mice.

Experimental Cell Biology 48, 279-297.

PettengillO.S. , Sorenson G.D., Wurster-Hill D.H., Curphey T.J., Noll W.W. ,

Cate C.C., Maurer L.H. (1980b) Isolation and growth characteristics of continuous cell lines from small cell carcinoma of the lung. Cancer 45,

906-918.

Pettengill O.S., Carney D.N., Sorenson G.D., Gazdar A.F. (1984)

Establishment and characteristation of cell lines from human small cell carcinoma of the lung. In: Becker K.L., Gazdar A.F. (eds.) The Endocrine

Lung in Health and Disease, pp460-468. W.B.Saunders Company Philadelphia,

U.S.A.

Pettengill O.S., Wurster-Hill D.H., Cate C.C., Sorenson G.D. (1985)

Morphological growth characteristics and hormone secretion of small cell carcinoma of the lung in vitro. In: Havemann K., Sorenson G. , Gropp C.

(eds.) Peptide Hormones in Lung Cancer. Recent Results in Cancer Research

99, ppl30-142, Springer-Verlag. Berlin.

Phifer J.E., Orth D.N., Spicer S.S. (1974) Specific demonstration of the human hypophyseal adrenocortico-melanotropic (ACTH/MSH) cell. Journal of

Clinical Endocrinology and Metabolism 39, 684-692.

244 Phillips M., Tashjian A. H. (1982). Characterisation of an early inhibitory effect of glucocorticoids on stimulated adrenocorticotrophin and endorphin release from a clonal strain of mouse pituitary cells.

Endocrinology 110, 892-900.

Pictet R.L., Rail L.B., Phelps P., Rutter W.J. (1975) The neural crest and the origin of the insulin-producing and other gastrointestinal hormone- producing cells. Science 191, 191-192.

Pintar J.E., Schachtter B. , Herman A.B., Durgerian S., Krieger D.T. (1984)

Characterisation and localisation of pro-opiomelanocortin mRNA in the adult rat testis. Science 225, 632-634.

Plotsky P.M., Bruhn T.O., Vale W. (1985) Evidence for multifactor regulation of the adrenocorticotropin secretory response to hemodynamic stimuli. Endocrinology 116, 633-639.

Plotsky P.M., Sutton S.W. , Bruhn T.O. , Ferguson A.V. (1988) Analysis of the role of angiotensin II in mediation of adrenocorticotropin secretion.

Endocrinology 122, 538-545.

Ratcliffe J.G. , Scott A.P., Bennett H.P.J., Lowry P.J., McMartin C. , Strong

J.A., Walbaum P.R. (1973) Production of a corticotrophin-like intermediate lobe peptide and of corticotrophin by a bronchial carcinoid tumour.

Clinical Endocrinology 2,51-55.

Ratcliffe J.G., Podmore J., Stack B.H.R., Spilg W.G.S., Gropp C. (1982)

Circulating ACTH and related peptides in lung cancer. British Journal of

Cancer 45, 230-236.

245 Ratter S.J., Lowry P.J., Besser G.M., Rees L.H. (1980) Chromatographic characterisation of adrenocorticotrophin in human plasma. Journal of

Endocrinology 85, 359-369.

Ratter S. J. , Gillies G. , Hope J., Hale A.C. , Grossman A. , Gaillard R. , Cook

D., Edwards C.R.W., Rees L.H. (1983) Pro-opiocortin Related peptides in human pituitary and ectopic ACTH secreting tumours. Clinical Endocrinology

18, 211-218.

Rayvis S.S., Horon R., (1971) Effect angiotensin II on adrenal and pituitary function in man. Journal of Clinical Endocrinology. 32, 539-546.

Rees L.H., Cook D.M., Kendall J.W., Allen C.F., Kramer R.M., Ratcliffe

J.G. , Knight R.A. (1971) A radioimmunoassay for rat plasma ACTH.

Endocrinology 89, 254-261.

Reisine T., Heisler S., Hook V., Axelrod J. (1982) Multireceptor-induced release of adrenocorticotropin from anterior pituitary tumor cells.

Biochemical and Biophysical Research Communications 108, 1251-1257.

Reisine T.D. (1985) Somatostatin inhibition of cyclic AMP accumulation and adrenocorticotropin release from mouse anterior pituitary tumor cells: mode of action and self-regulation. Advances in Cyclic Nucleotide Protein

Phosphorylation Research 19, 169-177

Reisine T. , Jensen R. (1986) Cholecystokinin-8 stimulates adrenocorticotropin release from anterior pituitary cells. Journal of

Pharmacology and Experimental Therapeutics 236, 621-626.

246 Reisine T. , Affolter H-U., (1987) Hormone receptor regulated proopiomelanocortin gene expression. Biochemical Pharmacology 36, 191-195.

Reith P., Monnor E. A., Bathija P. J. (1987). Prolonged suppression of a corticotropin-producing bronchial carcinoid by oral bromocriptine.

Archives of Internal Medicine 147, 989-991.

Rivier C. , Rivier J. , Vale W. (1982) Inhibition of adrenocorticotropic hormone secretion in the rat by immunoneutralisation of corticotropin releasing factor. Science 218, 377-379.

Roberts J.L., Herbert E. (1977) Characterisation of a common precursor to corticotrophin and beta-lipotropin: identification of beta-lipotropin peptides and their arrangement relative to corticotropin in the precursor synthesised in a cell free system. PNAS USA 74, 5300-5304.

Roberts J. , Budarf M. , Baxter J., Herbert E. (1979) Selective reduction of pro-adrenocorticotropin/endorphin proteins and messenger ribonucleic acid activity in mouse pituitary cells by glucocorticoids. Biochemistry 18,

4907-4915.

Roberts J. (1986) Transcription regulation of gene expresion in neuroendocrine cells. DNA 5, 68.

Rosewicz S., MacDonald A.R., Maddux B.A., Goldfine I.D., Miesfeld R.L.,

Logsdon C.D. (1988). Mechanism of glucocorticoid receptor down-regulation by glucocorticoids. Journal of Biological Chemistry 263, 2581-2584.

247 Roth K.A., Newell D.C., Dorin R.I., Eberwine J.H. , Hoffman A.R. , (1988)

Aberrant production and regulation of pro-opiomelanocortin derived peptides in ectopic Cushing’s syndrome. Hormone and Metabolic Research 20, 225-229.

Rousseau G. G. , Baxter J. D., Tomkins G. M. (1972). Glucocorticoid receptors : relations between steroid binding and steroid effects. Journal of Molecular Biology 67, 99-115.

Saffran M. , Schally A.V., (1955) The release of corticotrophin by anterior pituitary tissue in vitro. Canadian Journal of Biochemistry 33, 408-415.

Saffran M. , Schally A.V., Benfey B.G. (1955) Stimulation of the release of corticotropin from the adenohypophysis by a neurohypophysial factor.

Endocrinology 57, 439-444.

Sawchenko P.E., Swanson L.W. (1985) Localisation, colocalisation and plasticity of corticotropin releasing factor immunoreactivity in rat brain.

Federation Proceedings 44, 221-227.

Schacter B.S., Johnson L.K., Baxter J.D., Roberts J.L. (1982) Differential regulation by glucocorticoids of proopiomelanocortin mRNA levels in the anterior and intermediate lobes of the rat pituitary. Endocrinology 110,

1442-1444.

248 Schiffrin E.L., Chretien M. , Seidah N.G., Gutlowska Cantin M. ,

Gerest J. (1983) Response of human aldosteronoma cells in culture to the

N-terminal glycopeptide of pro-opiomelanocortin and gamma-j-MSH. Hormone and Metabolic Research 15, 181-184.

Schmelzer H-J., Hesch R-D., Mayer H. (1985) Parathyroid hormone and PTH mRNA in a human small cell lung cancer. In: Havemann K. , Sorenson G. ,

Gropp C. , (eds) Recent results in cancer research: Peptide hormones in lung cancer 99, pp 88-93. Springer-Verlag. Berlin.

Schwyzer R. , Schiller P., Seelig S., Sayer G. (1971) Isolated adrenal cells: log dose response curves for steroidogenesis induced by ACTH 1-24,

ACTH 1-10, ACTH 4-10 and ACTH 5-10. Federation of European Biochemical

Societies Letters 19, 229-231.

Schwyzer R. (1980) Structure and functions in neuropeptides. Proceedings of the Roval Society London (Biology) 210, 5-20.

Seamon K.B., Daly J.W.(1981) Forskolin : A unique diterpene activator of cyclic AMP generating systems. Journal of Cyclic Nucleotide Research 7, 201-

224.

Shibasaki T., Ling N., Guillemin R. (1980) Pituitaryimmunoreactivegamma- melanotropins are glycosylated oligopeptides. Nature 285,416-417.

Sidhu G.S. (1979) The endodermal origin of digestive and respiratory tract cells. American Journal of Pathology 96, 5-20.

249 Simms E., Gazdar A.F., Abrams P.G., Minna J.D. (1980) Growth of human small

cell (oat cell) carcinoma of the lung in serum-free growth factor-

supplemented medium. Cancer Research 40, 4356-4363.

Smyth D.G., Massey D.E., Zakarian S., Finnie M. D. (1979) Endorphins are

stored in biologically active and inactive forms: isolation of alpha-N- acetyl peptides. Nature 279, 252-254.

Sorenson G.D., Pettengill O.S., Brinck-Johnsen T. , Cate C.C., Maurer L.H.

(1981) Hormone production by cultures of small cell carcinoma of the lung.

Cancer 47, 1289-1296.

Sorensen G. D. , Pettengill 0. S., Cate C. C. , Ghatei M. A., Bloom S. R.

(1984). Modulation of bombesin and calcitonin secretion in cultures of small cell carcinoma of the lung. In : op. cit. Becker (1984) pp 596-602.

Sorenson G. D. , Cate C. C. , Pettengill 0. S. (1985). Regulation of hormone production in small cell carcinoma of the lung. op. cit. Craig et al

(1985). pp 143-156.

Sporn M.B., Todaro G.J. (1980) Autocrine secretion and malignant transformation of cells. New England Journal of Medicine 303, 878-880.

250 Steenbergh P. H. Hoppener J. W. M., Zandberg J. , Roos B. A., Jansz H. S.,

Lips C. J. M. (1984) Expression of the human pro-opiomelanocortin gene in

medullary thyroid carcinoma. Journal of Clinical Endocrinology and

Metabolism 58, 904-908.

Steiner D.F., Cunningham D., Spigelman L. et al (1967) Insulin

biosynthesis: evidence for a precursor. Science 157, 697-700.

Stewart M.F., Crosby S.R., Gibson S., Twentyman P.R., White A. (1989) Small

cell lung cancer lines secrete predominantly ACTH precursor peptides not

ACTH. British Journal of Cancer 60, 20-24.

Strand F.L., Smith C.M. (1980) LPH, ACTH, MSH and motor systems.

Pharmacology and Therapeutics 11, 509-533.

SudaT., Tozawa F. , Yamaguchi H., Shibasaki T. , Demur H., Shizume K. (1982)

Multiple forms of immunoreactive ©-endorphin are present in an ectopic

adrenocorticotrophin-producing tumour but not in normal pituitary or

pituitary adenomas. Journal of Clinical Endocrinology and Metabolism 54,

167-171.

Suda T. , Tozawa F. , Ushiyama T., Tomori N. , Sumimoto T. , Nakagami Y. ,

YamadaM., Demura H. , Shizume K. (1989) Effects of protein kinase-C-related

adrenocorticotropin secretagogues and interleukin-1 on proopiomelanocortin

gene expression in rat anterior pituitary cells. Endocrinology 124, 1444-

1449.

251 Swanson L., Sawchenko P., Rivier J.,Vale W. (1983) Organisation of ovine corticotropin-releasing factor immunoreactive cells and fibers in the rat brain : an immunohistochemical study. Neuroendocrinology 36, 165-186.

Takahashi H. , Teranishi Y. , Nakanishi S., Numa S. (1981) Isolation and structural organisation of the human corticotropin— lipotropin precursor gene. FEBS Letters 135, 97-102.

Tanaka K., Nicholson W.E., Orth D.N. (1978) The nature of immunoreactive lipotropins in human plasma and tissue extracts. Journal of Clinical

Investigation 62, 94-104.

Tanaka I., Nakai Y., Nakao S., Oki S., Masaki N. , Ohtsuki H., Imura H.

(1982) Presence of immunoreactive gamma-melanocyte stimulating hormone, adrenocorticotrophin and beta-endorphin in human gastric antral mucosa.

Journal of Clinical Endocrinology and Metabolism 54, 392-396.

Thomas P. S. (1980). Hybridisation of denatured RNA and small DNA fragments transferred to nitrocellulose. PNAS. USA. 77, 5201-5205.

Thorpe S. M. , Rose C. (1986). Oestrogen and progesterone receptor determinations in breast cancer : technology and biology. Cancer Surveys

5, 505-525.

252 Tsai S.Y., Sagami I., Wang H., Tsai M.J., O ’Malley B.W.(1987) Interaction between a DNA binding transcription factor (COUP) and a non-DNA binding factor (S300-11). Cell 50, 701-709.

Upton G.V., Amatruda T.T. (1971) Evidence for the presence of tumour peptides with corticotropin-releasing factor activity in the ectopic ACTH syndrome. New England Journal of Medicine. 285, 419-424.

Vale W., Speiss J., Rivier C., Rivier C. (1981) Characterisation of a 41- residue hypothalamic peptide that stimulates secretion of corticotropin and beta-endorphin. Science 213, 1394-1397.

Vale W., Vaughan J., Smith M., Yamomoto G., Rivier J., Rivier C. (1983)

Effects of synthetic oCRF, glucocorticoids, catecholamines, neurohypophysial peptides and other substances on cultured corticotrophic cells. Endocrinology 113, 1121-1131

Verbeeck M.A.E., Sutanto W. , Burbach J.P.H., (1991) Regulation of vasopressin messenger RNA levels in the small cell lung carcinoma cell line

GLC-8 : Interactions between glucocorticoids and second messengers.

Molecular Endocrinology 5, 795-801.

Wand G. S., May V., Eipper B. A. (1988). Comparison of acute and chronic secretagogue regulation of pro-adrenocorticotropin/endorphin synthesis, secretion, and messenger ribonucleic acid production in primary cultures of rat anterior pituitary. Endocrinology 123, 1153-1161.

253 Wang S.Y. , Sagami I., WangH., Tsai M.J., O ’Malley B.W. (1987) Purification and characterisation of chicken ovalbumin upstream promoter transcription factor from HeLa cells. Journal of Biological Chemistry 262, 16080-16086.

Wang L-H., Tsai S.Y., Cook R.G., Beattie W.G., Tsai M-J., O ’Malley B.W.

(1989) COUP transcription factor is a member of the steroid receptor superfamily. Nature 340, 163-166.

West D.C., Sattar A., Kumar S. (1985) A simplified in situ solubilisation procedure for the determination of DNA and cell number in tissue cultured mammalian cells. Analytical Biochemistry 147, 289-295.

Whang-Peng J. , Bunn P.A., Kao-Shan C.S., Lee E.C., Carney D.N., Gazdar

A.F., Minna J.D. (1982) A non-random chromosome abnormality, del 3p (14-

23), in human small cell lung cancer (SCLC) Cancer Genetics and

Cytogenetics 6, 119-134.

White A., Gray C., Ratcliffe J.G. (1985) Characterisation of monoclonal antibodies to adrenocorticotrophin. Journal of Immunological Methods 79,

185-194.

White A. , Smith H. , Hoadley M. , Dobson S.H., Ratcliffe J.G. (1987) Clinical evaluation of a two-site immunoradiometric assay for adrenocorticotrophin in unextracted human plasma using monoclonal antibodies. Clinical

Endocrinology 26, 41-52.

254 White A. , Stewart M.F., Farrell W.E., Crosby S.R. , Lavender P.M. , Twentyman

P.R., Rees L.H., Clark A.J.L. (1989) Pro-opiomelanocortin gene expression

and peptide secretion in human small-cell lung cancer cell lines. Journal

of Molecular Endocrinology 3, 65-70.

Whitfeld P.L., Seeburg P.H., Shine J. (1982) The human pro-opiomelanocortin

gene: organisation, sequence and interspersion with repetitive DNA. DNA 1,

133-143

Wilson R.E., Orth D.N., Nicholson W.E., Mount C.D., Bertagna X.Y. (1981)

Human beta-lipotrophin radioimmunoassay: Identification of immunoreactive beta-lipotrophin in human plasma and tissue. Journal of Clinical

Endocrinology and Metabolism 53, 1-9.

Yalow R.S., Berson R.A. (1971) Size heterogeneity of immunoreactive human

ACTH in plasma and in extracts of pituitary gland and ACTH-producing

thymoma. Biochemical and Biophysical Research Communications 44, 439-445.

Yamaguchi H., Liotta A.S. Krieger D.T. (1980) Simultaneous determination of human plasma immunoreactive beta-endorphin, gamma-1 ipotrophin and beta- endorphin using immune-affinity chromatography. Journal of Clinical

Endocrinology and Metabolism 51, 1002-1008.

255 Yamaguchi K. , Abe K. , Adachi I., Kimura S., Suzuki M., Shimada A. , Kodama

T., Kameya T. , Shimosato Y. (1985) Peptide hormone production in primary

lung tumours. In: Havemann K. , Sorenson G. , Gropp C. , (eds) Peptide

Hormones in Lung Cancer, Recent Results in

Cancer Research 99, ppl07-116. Springer-Verlag. Berlin.

Yesner R. , Carter D. (1982) Pathology of carcinoma of the lung:

changing patterns. Clinics in Chest Medicine 3, 257-289.

256 APPENDIX I

Materials MATERIALS

I. Major Equipment

Class II laminar flow Envair Ltd, Rossendale,

safety cabinet Lancashire, U.K.

C0£ Incubators Heraeus B5060 EK/O2

Heraeus Equipment

Ltd, 9 Waters Way,

Brentwood, Essex

CM15 9TB

Liquid nitrogen storage tank Union Carbide 35 HC

Inverted light microscope Olympus IMT-2, with photographic unit and Olympus Optical Co. Ltd., camera attachment. Tokyo, Japan.

Bench top centrifuge Fisons Instruments

Sussex Manor Park

Gatwick Road, Crawley,

West Sussex RH10 2QQ.

Refrigerated centrifuge Fisons Instruments

Sussex Manor Park

Gatwick Road, Crawley,

West Sussex RH10 2QQ.

Spectrophotometer SP6-500 UV, Pye Unicam,

Cambridge,U.K.

Spectrofluorimeter Perkin-Elmer LS-5,

Perkin-Elmer Ltd.

Post Office Lane

Beaconsfield

Bucks. HP9 1QA.

258 Gamma counter linked to LKB 1260 Multigamma. 11

BBC microcomputer

Sucroseparator Model SS12, Boots Celltech,

Slough, U.K.

Sucroagitator Boots Celltech, Slough, U.K.

Fraction collectors 1. LKB 2211 Superac,

LKB Produkter, Broma,

Sweden.

2. Pharmacia FRAC.300,

Pharmacia Fine Chemicals,

Uppsala, Sweden.

Chromatography columns Pharmacia Fine Chemicals. pH meter Jenway 3060, Jencons,

Leighton Buzzard,Beds,U.K.

Water bath Techne (Cambridge) Ltd.

Cambridge, U.K.

Hamilton automatic pipette Hamilton Microlab M,

Howe Ltd., London, U.K.

Slot blot apparatus GIBCO BRL,

Paisley, Scotland

Electrophoresis horizontal BioRad Laboratories Ltd, gel apparatus and powerpack Watford, Herts., U.K.

Densitometer Parry DT1405.

259 II. Minor Equipment

Sterile tissue culture Northumbria Biologicals,

plastic ware Cramlington,

Northumbria,U.K.

Plastic syringes Monosect, Crawley, Sussex,

U.K.

Syringe needles Yale Microlance,

Becton Dickinson,

Dun Laoghaire, Dublin,

Eire.

Dissection kit Raymond Lamb, London,

U.K.

Stainless steel gauze mesh J.Nichols and Son Ltd,

(120 mesh) Birmingham, U.K.

Plastic cuvettes Gallenkamp, Loughborough,

U.K.

Quartz cuvettes Gallenkamp, Loughborough.

Sterilising filters (0.2um) Flowpore, Flow Labs.

Irvine, Ayrshire, U.K.

Haemocytometer Weber Scientific

(Improved Neubauer) International Ltd, Lancing,

U.K.

Plastic LP3 and LP5 tubes LIP Ltd, Shipley for IRMA, RIA and protein Yorkshire, UK. assays.

Glass TCI tubes for DNA assay LIP Ltd, Shipley.

Liquid nitrogen freezing Union Carbide attachment

260 Liquid nitrogen freezing Sterilin Ltd, Feltham, U.K. ampoules

Nitrocellulose paper Schleicher and Schuell,

Dassell, F.R.G.

Kodak XAR-5 film and Eastman Kodak, Rochester, intensifying screens New Jersey, U.S.A.

261 III. Reagents for Cell Culture

Basal Media

RPMI 1640 GIBCO UK., Paisley,

Scotland

cat. no. 041-01875M

Dulbecco’s modification of Flow Laboratories, Irvine,

Eagle’s medium with HEPES Ayrshire, U.K. cat no. 12-334-54

H a m ’s F10 Flow Laboratories cat. no. 12-412-54

Foetal Calf Serum Flow Laboratories cat. no. 29-101-49

Horse Serum Gibco UK. cat. no. 034-06055M

Additives

Molar HC1 Add 3.67ml of concentrated

HC1 to 100ml double

distilled water. Autoclave

to sterilise

Molar NaOH Add 4g NaOH pellets to

100ml double distilled

water.

Filter sterilise (0.2um).

Glutamine (200mM) Flow Laboratories cat. no. 16-801-49

Sodium pyruvate (lOOmM) Flow Laboratories cat. no. 16-820-49

262 HEPES buffer (1M) : Flow Laboratories cat. No. 043-056304

Penicillin : Glaxo Laboratories Ltd., cat. no. J 83103 Greenford, U.K.

Streptomycin : Glaxo Laboratories Ltd. cat. no. J 82271

Gentamicin : Sigma Chemical Company, cat. no. G-7507 Poole, Dorset, U.K.

Fungizone : Gibco U K cat. no. 043-05290D

Salt solutions

Phosphate buffered saline : Flow Laboratories

Dulbecco’s formula (PBS) cat. no. 18-604-49

Hank’s balanced salt solution : GIBCO U.K.

(HBSS) cat. no. 041-4020

HITES constituents : Sigma Chemical Company

Hydrocortisone, cat. no. H4001

Insulin (bovine), cat. no. 15500

Transferrin (human), cat. no. T2252

Estradiol, cat. no. E8875

Sodium Selenite, cat. no. S1382

263 Hydrocortisone

Dissolve lOmg hydrocortisone in 10ml Ethanol \ 0.5ml

+9.5ml Ethanol (0.05mg/ml)

\ 0.7ml

50ml PBS (Ca2+, M g 2* free)

=2uM Hydrocortisone

Store 0.5ml aliquots at -20°C

Estradiol

Dissolve lOmg estradiol in 10ml Ethanol

| 0.5ml

+9.5ml ethanol (0.05mg/ml)

| 0.54ml

50ml PBS (Ca2+,Mg2+ free)

=2uM Estradiol

Store 0.5ml aliquots at -20°C

Transferrin. Insulin and Sodium Selenite (TIS)

DILUENT = 0.5% Bovine Serum Albumin in PBS (Ca2+,Mg2+ free)

Transferrin Insulin Sodium I I Selenite lOOmg 50mg lmg I / Dissolve in• Dissolve in Dissolve in

5ml diluent lml 0.1M NaHC03 20 ml diluent \ Add 9ml diluent \ 5ml 10mlI 1.04ml

Combine and make up to total volume of 50ml with diluent

Store 0.5ml aliquots at -20°C.

HITES supplements must be filtered (0.2um filter) after dilution in basal medium.

264 Media Formulations i) RTISS(2.5)

RPMI 1640 100ml

Foetal Calf Serum 2.5ml

HEPES buffer 1.0ml

Glutamine 2.0ml

Sodium pyruvate 1.0ml

Transferrin }

Insulin } 0.5ml aliquot

Sodium selenite } ii) RS(10)

RPMI 1640 100ml

Foetal Calf Serum 10ml

HEPES buffer 1.0ml

Glutamine 2.0ml

Sodium Pyruvate 1.0ml iii) Human pituitary cell culture medium

DMEM with HEPES 100ml

NaHC03 (7.5%) 0.9ml

Foetal Calf Serum 2.5ml

Horse Serum 10ml

Glutamine 2.0ml

Sodium Pyruvate 1.0ml iv) AtT20 growth medium

H a m ’s F10 100ml

Foetal Calf Serum 2.5ml

Horse Serum 15ml

Glutamine 2.0ml

Sodium Pyruvate 1.0ml

265 v) Freezing medium

D M E M with HEPES 100ml

Foetal Calf Serum 20ml

NaHC03 (7.5%) 0.9ml

Mercaptoethanol 0.05ml

Glutamine 2.0ml

Sodium Pyruvate 1.0ml

Gentamicin (50mg/ml) 0.04ml

Freezing Mixture was prepared by adding 6ml of the above medium to 3ml of foetal calf serum

266 IV. CHEMICALS AND REAGENTS

Chemicals were of Analar grade and were purchased from Sigma

Chemical Company, Poole Dorset, U.K. unless otherwise stated.

Reagents for DNA Assay

Glass distilled water is needed for the preparation of reagents and

should be filtered before use (0.2um filter). Scrupulously clean

glassware (not plastic) must be used.

1. DNA Standard: Salmon Testis DNA Type III

STOCK STANDARD = 2.5mg/ml in glass distilled water

Check exact concentration as follows:

i) Make a 1:50 dilution of the stock standard to give a 0.005%

solution.

ii) Read Absorbance at 260nm using U V spectrophotometer

(should=1.000).

iii) Calculate DNA concentration from:

E(l%)260n m = 200 (A x Aliquots of the stock standard may be stored at -20°C.

2. EDTA : lOmM Disodium salt, anhydrous (MW=336.2)

Adjust pH to 12.3 using ION NaOH.

Initial pH= 4.68 approx. Make fresh.

3. KH2P04 1M.

Store at 4°C.

4. Tris HC1 0.01M /NaCl 0.1M

Adjust pH to 7.0. Make fresh and store at 4°C.

267 5. Hoechst Reagent ( Hoechst 33258, Uniscience London U.K.)

Stock solution = 200ug/ml in glass distilled water. Stable at least 6 months stored at 4°C in a dark bottle, wrapped in foil.

Working reagent = 1:1000 of stock diluted in Tris/NaCl(reagent 4)

i.e. 200ng/ml.

Reagents 2-4 should be filtered prior to use (0.2um filter)

268 Reagents for Protein Assay

Bio-Rad Dye Reagent (Bio-Rad Laboratories Ltd, Watford,

Herts, U.K.) Dilute 1:5 with

deionised water and filtered

(Whatman No.l filter paper).

Stock Standard Bovine Serum Albumin, Fraction V

(Sigma), lmg/ml in 0.1M NaOH.

Aliquots may be stored at -20 °C.

Working Standards Dilute stock standard in 0.1M

NaOH from lmg/ml-0.0625mg/ml.

Use 0.1M NaOH as zero standard.

Reagents for Immunoradiometric Assays

Assay Diluent : 0.1M sodium phosphate pH 7.4

with 0.02% sodium azide.

Immediately prior to use add 0.5%

BSA and 0.1% Triton X100.

Iodinated antibody MAb diluted to give 100,000-

200,000 cpm with buffer+

0.5% normal sheep serum.

Solid-phase antibody MAb coupled to Sephacryl S-300

(Pharmacia) stored in assay

buffer, adjusted to a 5% slurry

and washed x3 with assay diluent

Sucrose solution 10% sucrose (granulated sugar),

0.1% sodium azide and 1%

Tween 20 in distilled water.

269 Reagents for cortisol radioimmunoassay

Assay diluent: 0.13M sodium citrate/phosphate buffer pH 4.0, containing gelatine (0.2%) and azide (0.05%)

Tracer cortisol: Amersham International, Amersham, Bucks, UK. Cortisol-3-CMO-'25I-histamine is dissolved in ethanol and aliquots stored at 4°C. For use, 1 aliquot is diluted in citrate buffer (20ml) to give *15,000 cpm/250pl.

Cortisol Anti serum: Scottish Antibody Production Unit (SAPU)

Second antibody: SAPU Donkey anti-sheep

Carrier: SAPU Normal sheep serum

Cortisol standards: Sigma Chemical Company Hydrocortisone (H4001) Prepared in charcoal stripped human serum

270 V. PEPTIDES. DRUGS AND MISCELLANEOUS

ACTH (Human purified) National Institute for

NIBSC code 74/555 Biological Standards and

(6.2 I.U. per ampoule) Control, Holly Hill,

London, U.K.

CRF (ovine) Bachem UK, Saffron Walden,

Essex, UK.

Hydrocortisone Sigma Chemical Company cat. no. H4001

Dexamethasone Organon Laboratories Ltd.

Cambridge Science Park

Cambridge, UK.

Bromocriptine Sandoz Pharmaceuticals,

Feltham, Middlesex, U.K.

Dibutyryl cyclic A M P Sigma Chemical Company cat. no. D0627

Isobutyl Methyl Sigma Chemical Company

Xanthine cat. no. 15879

Forskolin Novabiochem (UK) Ltd. cat. no. 3442-70 Nottingham, UK.

Phorbol 12, Sigma Chemical Company

13-dibutyrate cat. no. P1269

271 APPENDIX II

Tables of data from experiments in Section IV:

Regulation of POMC gene expression and peptide secretion in SCLC cells

272 SECTION 1 : PEPTIDES

SCLC cell lines: Ten day incubation with hydrocortisone

1 COR L103

Hydrocortisone ACTH precursors n mean C A 1 nM/1 pmol/mg protein

Experiment 1 (Figure 25, page 132)

0 66.0 49.2 35.0 41.0 61. 0 69.0 6 53.5 64.6)

500 95.0 74.0 95.0 3 88.0 103.6)

1000 76.3 70.0 71.5 3 72.6 88.2)

Experiment 2 (Figure 26, page 133)

0 19.0 21.0 50.0 3 30.0 56.7)

500 19.0 22.0 68.0 3 36.3 63.1)

1000 36.0 42.0 44.0 3 40.7 67.4)

Experiment 3 (Figure 27, page 134)

0 180 130 150 3 153 (: 184.1)

1000 160 150 145 3 152 (: 181.5)

Experiment 4 (Figure 28, page 135)

0 48.0 48.2 49.0 3 48.4 49.2)

1000 47.5 48.3 48.0 3 47.9 48.7)

273 SCLC cell lines: Ten day incubation with hydrocortisone

COR L24

Hydrocortisone ACTH precursors n mean (95% C.l) nM/1 omol/mg protein

ExDeriment 1 (Figure 29, page 136)

0 253. 1 270 .3 283.9 248.0 236 .1 265.7 6

500 227. 0 285 .1 297.1 269.7 (237.5, 302)

1000 268. 4 314 .0 275.9 286.1 (253.9, 318.3)

Experiment 2 (Figure 30, page 137)

0 265 282 292 3

500 273 280 275 3

1000 290 285 291 3

3 COR :L31

Hydrocortisone ACTH precursors n mean (95% C.l) nM/1 nmol/mg protein

Experiment 1 (Figure 31, page 138)

0 16.5 14.9 9.3 17.3 21.2 15.5 6

500 16.2 17.0 17.8 3

1000 17.4 13.3 18.8 3

Experiment 2 (Figure 32, page 139)

0 19.0 17.0 24.0 3

500 23.0 22.0 16.0 3

1000 16.0 18 23 3

Experiment 3 (Figure 33, page 140)

0 24.8 23.5 - 2

500 30.6 19.2 21.7 3

1000 20.5 22.1 24.1 3

274 SCLC cell lines; Ten day incubation with hydrocortisone

4 COR L42

Hydrocortisone ACTH precursors n mean (95% C.1) nM/1 pmol/mg protein

Experiment 1 (Figure 34, page 141)

0 34.7 33.2 40.8 3 36.2 (31.5, 41 )

500 30.7 30.6 32.4 3 31.2 (26.5, 36 )

1000 32.8 37.4 29.2 3 33.1 (28.4, 37.9)

Experiment 2 (Figure 35, page 142)

0 38 36 43 3 39.0

500 30 32 29 3 30.3

1000 33 24 36 3 31.0

COR L27 (Figure 36, page 143)

Hydrocortisone ACTH precursors n mean (95% C .1) nM/1 pmol/mg protein

0 3.3 2.9 3.6 4.4 4.7 4.5 6 3.9 (3.4, 4.4 )

500 5.6 6.2 5.7 3 5.8 (5.1, 6.6 )

1000 4.8 4.4 4.2 3 4.5 (3.7, 5.2 )

275 SCLC cell lines: 24h incubation with hydrocortisone

Hydrocortisone ACTH precursors n mean (95% C.1) nM/1 pmol/mg protein

1 COR L103 (Figure 37, page 144)

Experiment 1

0 3.2 3.3 4.2 3 3.6 (2.31, 4.82)

500 3.4 5.9 5.7 3 5.0 (3.74, 6.26)

1000 4.4 5.0 5.1 3 4.8 (3.58, 6.09)

Experiment 2 (Figure 38, page 145)

0 2.6 3.3 3.1 3 3.0 (1.67, 4.3 )

500 3.4 3.8 - 2 3.6 (1.97, 5.23)

1000 4.8 2.6 - 2 3.7 (2.07, 5.33)

2 COR L24 (Figure 39, page 146)

0 29.8 38.4 41.9 3 36.7 (28.1,45.3)

500 42.4 27.6 29.8 3 33.3 (24.7,41.9)

1000 34.1 39.9 37.4 3 37.1 (28.5,45.7)

3 COR L31 (Figure 40, page 147)

0 1.5 2.0 1.4 3 1.63 (1.27, 1.99)

500 1.6 2.0 2.1 3 1.9 (1.54, 2.26)

1000 1.5 1.6 1.8 3 1.63 (1.27, 1.99)

4 COR L42 (Figure 41, page 148)

0 8.4 8.0 13.1 3 9.8 (7.3, 12.4)

500 10.0 8.3 8.5 3 8.9 (6.4, 11.5)

1000 8.0 9.4 9.6 3 9.0 (6.5, 11.5)

5 COR L27 (Figure 42, page 149)

0 0.6 0.8 0.7 3 0.7 (0.36, 1.04)

500 1.8 1.2 1.1 3 1.4 (1.02, 1.71)

1000 0.8 1.1 0.9 3 0.9 (0.59, 1.28)

276 AtT20 cells : Ten day incubation with hydrocortisone

Experiment 1 (Figure 43, page 151)

Hydrocortisone ACTH precursors n mean (95% C.1) nM/1 pmol/mg protein

0 4188 3613 4793 3 4198 (3789, 4607)

50 3094 2744 2512 3 2783 (2374, 3192)

100 2481 2048 2050 3 2193 (1784, 2602)

500 1668 1563 1341 3 1524 (1115, 1933)

1000 1277 1452 1383 3 1371 ( 962, 1780)

2000 1288 1623 982 3 1298 ( 889, 1707)

(Figure 44, page 152)

Hydrocortisone ACTH mean (95% C.l) nM/1 pmol/mg protein

0 97.7 93.1 108.5 3 99.8 (93.3, 106

50 68.5 72.2 66.0 3 68.9 ( 62.4, 75.4

100 59.1 52.2 49.1 3 53.5 (47, 60

500 43.0 43.3 35.2 3 40.5 ( 34, 47

1000 31.4 33.9 33.6 3 33 (26.5, 39.5

2000 32.2 37.3 25.0 3 31.5 (25, 38

277 AtT20 cells : Ten day incubation with hydrocortisone

Experiment 2 (Cell culture performed by Dr. Adrian Clark and Paul Lavender)

(Figure 45, page 153)

Hydrocortisone ACTH precursors n mean (95% C.l) nM/1 nmol/mg protein

0 16.5 18.4 10.2 16.7 31.4 13.3 19.9 19.1 18.2 (13.9, 22.6)

500 6.8 4.8 5.8 (-3.0, 14.6)

1000 5.9 6.3 6.1 (-2.7, 14.9)

(Figure 46, page 154)

Hydrocortisone ACTH mean (95% C.l) nM/1 nmol/mg protein

0 0.46 0.51 0.37 0.33 0.52 0.33 0.32 0.33 8 0.40 (0.34,0.46)

500 0.14 0.14 2 0.14 (-0.02,0.26)

1000 0.07 0.11 2 0.09 (-0.03,0.21 )

Hydrocortisone ACTH precursors n mean (95% C.l) nM/1 pmol/mg protein

Experiment 3 (Figure 47, page 155)

0 10 9.98 11.3 10.4 (9.6,11.3)

1000 3.2 3.3 3.2 3.2 (2.4, 4.1)

Experiment 4 (Figure 48, page 156)

0 9.8 9.4 9.6 9.6 (9.4, 9.8)

1000 3.04 3.2 3.1 3.1 (2.9, 3.4)

Experiment 5 (Figure 49, page 157)

0 4.88 5.32 5.46 5.2 (4.9, 5.6)

1000 2.57 2.75 2.78 2.7 (2.3, 3.1)

278 AtT20 cells : 24h incubation with hydrocortisone

Experiment 1 (Figure 50 » page 158)

Hydrocortisone ACTH precursors n mean (95% C.l) nM/1 pmol/mg protein

0 226 186 201 3 204 (164, 244)

50 313 253 247 3 271 (231, 311)

100 219 278 221 3 239 (199, 279)

500 297 258 274 3 276 (236, 316)

1000 261 263 280 3 268 (228, 308)

2000 304 244 349 3 299 (259, 339)

(Figure 51 . page 159)

Hydrocortisone ACTH n mean (95% C.l) nM/1 praol/mff protein

0 13.1 14.5 16.6 3 16.0 (9.3, 22.8)

50 33.6 21.2 18.2 3 24.3 (17.6, 31.1)

100 17.3 22.5 18.2 3 19.3 (12.6, 26.1)

500 28.4 22.5 22.7 3 24.5 (17.8, 31.3)

1000 19.8 29.1 22.5 3 23.8 (17.0, 30.6)

2000 36.1 21.0 24.2 3 27.1 (20.3, 33.9)

279 AtT20 cells : 24h incubation with hydrocortisone

Experiment 2 (Cell culture performed by Dr. Adrian Clark and Paul Lavender)

(Figure 52, page 160)

Hydrocortisone ACTH precursors n mean (95% C.l) nM/1 pmol/mg protein

0 5.4 2.4 2 3.9 (1.14, 6.7)

500 3.0 2.9 2 2.95 (0.19, 5.7)

1000 3.1 2.9 2 3.0 (0.24, 5.8)

(Figure 53, page 161)

Hvdrocorti sone ACTH precursors n mean (95% C.l) nM/1 pmol/mg protein

0 0.22 0.17 2 0.195 (0.12, 0.27)

500 0.12 0.17 2 0.15 (0.07, 0.23)

1000 0.20 0.15 2 0.175 (0.10, 0.26)

Missing samples: a number of samples from this experiment proved unsuitable for analysis, due to loss of labelling occurring in transit.

280 SCLC cell lines: lOd incubation with hydrocortisone

Cortisol levels in medium at end of the incubation period

CELL LINE Experiment CORTISOL (nmol/1) No. CONTROLS M500nM" lOOOnM"

COR L103 1 <37 434 1208 <37 384 1205 <37 418 1271 <37 <37 <37

COR L103 2 <37 66 612 <37 61 317 <37 67 784

COR L24 <37 136 579 <37 170 566 <37 157 639 <37 <37 <37

COR L31 1 <37 797 1690 <37 592 1739 <37 678 1419 <37 <37 <37

COR L31 2 <37 370 821 <37 380 744 <37 320 767

COR L42 1 <37 1246 1891 <37 1175 1885 <37 1097 2406

COR L27 <37 629 1738 <37 673 1830 <37 715 1626 <37 <37 <37

281 SCLC cell lines: 24h incubation with hydrocortisone

Cortisol levels in medium at end of the incubation period

CELL LINE Experiment CORTISOL (nmol/1) No. CONTROLS ”500nM” "lOOOnM"

COR L103 1 <37 164 491 <37 167 424 <37 157 425

2 <37 42 107 <37 47 282 <37

COR L24 Assay failed

COR L31 <37 303 712 <37 307 691 <37 316 658

COR L42 <37 457 856 <37 471 904 <37 408 912

COR L27 <37 267 547 <37 264 563 <37 264 657

282 AtT20 cells:

Cortisol levels in medium at end of incubation period.

10 day Experiment

Hydrocortisone CORTISOL nmol/1

0 <37 <37 <37

50 <37 <37 <37

100 49 45 42

500 271 351 269

1000 649 656 782

2000 1496 1299 1381

24h Experiment

Hydrocortisone CORTISOL nmol/1

0 <37 <37 <37

50 <37 <37 <37

100 <37 <37 <37

500 234 237 213

1000 491 523 562

2000 1076 1101 955

283 Data for figure 57, page 168

Effect of bromocriptine on ACTH precursor secretion by COR L24 cells.

Bromocriptine ACTH precursors pmol/mg protein mean (95% C.l)

0 96.7 95.8 91.9 94.8 (86.1, 103

0.1 69.7 77.9 84.7 77.7 (68.7, 86,

0.5 83.4 88.7 78.6 83.6 (74.9, 92

1.0 88.7 70.9 65.8 75.1 (66.4, 83,

5.0 65.4 51.3 55.7 53.5 (48.8, 66

10.0 49.1 50.8 51.2 50.4 (41.7, 59,

Effect of bromocriptine on ACTH precursor secretion by COR L31 and COR L42 (24h incubation).

CELL LINE Bromocriptine ACTH precursors n mean (95% C.l) ug/1 pmol/mg protein

COR L31 Figure 58 0 1.45 1.80 2.45 3 1.90 (1.52, 2.28) page 170 0.1 1.58 1.47 1.57 3 1.54 (1.16, 1.92)

1 1.65 1.28 1.64 3 1.52 (1.15, 1.90)

10 0.89 0.73 0.97 3 0.86 (0.49, 1.24)

COR L42 Figure 59 page 171 0 11.1 9.2 10.8 3 10.4 ( 8.9, 11.8)

0.1 10.2 13.0 11.2 3 11.5 (10.0 , 12.9)

1 9.5 11.1 11.9 3 10.8 ( 9.4, 12.3)

10 5.2 5.5 5.8 3 5.5 ( 4.1, 6.9)

284 Effect of bromocriptine on SCLC cells in a 7 day incubation

CELL LINE Bromocriptine ACTH precursors n mean (95% C.l) HSZ.1 pmol/mg protein

COR L24 Experiment 1 Figure 60 page 172 0 328 528 479 3 445 (327, 563 )

10 139 130 141 3 137 (18, 255 )

Experiment 2

Figure 61 0 176 178 184 3 179 (173.5,185.2) page 173 10 33 39 37 3 36 ( 30.5, 42.2)

COR L31 Figure 62 page 174 0 10.6 12.3 10.4 3 11.1 ( 9.9, 12.3)

10 2.5 2.7 2.3 3 2.5 ( 1.3, 3.7)

COR L42 Figure 63 page 175 0 58.3 70.0 65.0 3 64.4 ( 55.6, 73.3)

10 7.0 6.1 - 2 6.6 ( -4.3, 17.4)

285 Effect of oCRF on ACTH precursor secretion by SCLC cells (72h incubation).

CELL LINE oCRF ACTH precursors n mean (95% C. I) (ng/ml) Dmol/mg protein

COR L42 Figure 64 page 178 0 42.0 43.4 44.3 3 43.2 (36.5, 50 )

10 49.0 64.0 45.9 3 53.0 (46.2, 59.7)

100 51.1 53.8 51.1 3 52.0 (45.2, 58.8)

1000 51.0 52.7 48.0 3 50.6 (43.8, 57.3)

COR L103 Figure 65 page 179 0 6.8 6.0 5.7 3 6.2 ( 5.1, 7.2

10 6.9 6.2 6.4 3 6.5 ( 5.5, 7.5)

100 5.7 5.4 7.7 3 6.3 ( 5.2, 7.3)

1000 6.6 6.7 - 2 6.7 ( 5.4, 7.9)

Effect of oCRF on ACTH precursor and ACTH secretion by AtT20 cells (72h incubation).

Figure 66 oCRF ACTH precursors n mean (95% C. I) page 180 (ng/ml) pmol/mg protein

0 2145 1800 2154 3 2033 (1703, 2363)

10 1704 1975 1977 3 1885 (1555, 2215)

100 2272 2266 1923 3 2154 (1824, 2484)

1000 2594 3318 2790 3 2901 (2571, 3231)

Figure 67 oCRF ACTH n mean (95% C. I) page 181 (ng/ml) pmol/mg protein

0 30.7 28.2 25.9 3 28.3 (21.4, 35.2)

10 27.7 28.8 30.9 3 29.1 (22.2, 36.0)

100 40.2 39.8 37.5 3 39.2 (32.3, 46.1)

1000 53.8 66.7 47.4 3 56 (49.1, 62.9)

286 Data for figure 68, page 183

Effect of different growth media on response to oCRF in COR L24 cells

MEDIUM oCRF ACTH precursors n mean (95% C.l) (ng/ml) pmol/mg protein

RTISS 2.5 0 13.3 14.5 15.7 3 14.5 (13, 16 )

100 16.0 16.7 15.4 3 16.0 (14.5, 17.6)

RTIS SF 0 1.73 0.58 0.65 3 0.99 ( 0, 1.98 )

100 0.88 2.05 1.52 3 1.48 (0.5, 2.5 )

RBSAT 0 5.69 5.52 5.70 3 5.64 (5.27, 6.0 )

100 5.8 5.9 5.32 3 5.67 (5.3, 6.0 )

287 Effect of dexamethasone and oCRF on ACTH and ACTH precursor secretion bv human pituitary cells

1. ACTH Precursors (pmol/1) (Figure 69, page 185)

CONTROLS DEXAMETHASONE oCRF luM 1000 ng/ml

809 910 424 749 1193 463 744 1180 421 798 1017 494 666 1144 432 881 894 380 889 1050 577 905

7 8 7 mean 791 1036 456

95% C.l. (715, 866) (966,1107) ( 380, 531)

2. ACTH (pmol/1) (Figure 70, page 186)

CONTROLS DEXAMETHASONE oCRF luM 1000 ng/ml

736 1010 4315 979 712 5342 694 822 4727 946 789 4894 1088 845 4197 1415 663 3864 813 780 ♦2695 598

7 8 7 mean 952 777 4291

95% C.l. (551,1352) (403,1152) (3890, 4691)

♦ outlier excluded from normal probability plot.

288 Effect of ImM db-cAMP on ACTH precursor secretion in a 48h incubation.

A. COR L24 (Figure 71, page 189)

ACTH precursors pmol/mg protein mean (95% Cl

CONTROLS 77.1 78.3 87 80.8( 66.8,94.8)

ImM db-cAMP 117 134 113 121.3(107.3,135.4)

B. COR L103 (Figure 72, page 190)

ACTH precursors pmol/mg protein mean (95% C.l.)

CONTROLS 3.8 4.5 4.3 4.2 (3.05, 5.35)

ImM db-cAMP 7.7 8.7 6.8 7.7 (6.58, 8.89)

Effect of db-cAMP on ACTH precursor secretion by COR L24 cells: 48h time course. (Figure 73, page 192)

ACTH precursors pmol/mg protein

TIME CONTROLS 10 J db-cAMP mean mean

30 min 1.98 1.28 2.98 2.29 2.06 1.71 1.91 1.79

60 min 3.1 3.0 2.66 2.64 3.51 3.56 2.16 4.17

2h 5.1 5.09 4.19 4.73 6.91 5.98 4.91 5.95

5h 10.6 10.61 11.21 10.47 12.67 9.93 10.03 6.5

24h 27.51 36.1 28.01 27.23 *62.3 42.73 26.16 29.8

48h *48.18 81.38 54.74 52.70 78.66 82.92 55.20 88.71

* outliers excluded from normal probability plot.

289 Data for figure 74, page 195.

Effect of db-cAMP on ACTH precursor secretion by COR L24 cells with and without IBMX (24h incubation).

db-cAMP (uM) ACTH precursors pmol/mg protein

- IBMX + IBMX (0.5 mM) mean mean

0 107 130 106 103.3 142 146 97 166

1 129 162 115 116.3 158 154 105 142

100 125 155 129 124.3 176 156.7 119 139

290 Effect of forskolin on ACTH precursor secretion by SCLC cells.

A. COR L24 (Figure 75, page 198)

Forskolin (M) ACTH precursors pmol/mg protein

mean 95% C.l.

0 35.9 38.3 37.2 37.1 (32.5, 41.7)

10'8 42.6 41.5 35.1 39.7 (35.1, 44.3)

io‘7 44.8 35.8 35.2 38.6 (34, 43.2)

10‘6 48.4 46.5 46.7 47.2 (42.6, 51.8)

A. COR L103 (Figure 76, page 199)

Forskolin (M) ACTH precursors pmol/mg protein

mean 95% C.l.

0 3.1 3.1 3.4 3.2 ( 2.95, 3.45)

10’8 2.8 3.2 2.6 2.9 ( 2.61, 3.12)

10'7 2.6 2.7 2.8 2.7 (2.45, 2.95)

10'6 3.1 3.3 3.2 3.2 (2.95, 3.45)

Effect of a Dhorbol ester on ACTH Drecursor secretion by COR: L24 cells. (Figure 77, page 203)

Phorbol 12, 13 dibutvrate ACTH precursors (M) (pmol/mg protein)

mean 95% C.l.

0 31.9 30.6 31.7 31.4 (29.3, 33.5)

10‘8 35.6 31.8 32.8 33.4 (31.3, 35.5)

10'7 36.5 39.3 41.0 38.9 (36.9, 41.0)

10*6 32.2 32.2 32.9 32.4 (30.4, 34.5)

291 SECTION 2 : RNA

RNA analysis in this section was performed by Dr. Adrian Clark and Paul Lavender (St. Bartholomew’s Hospital)

Data for figure 23, page 125

RELATIVE QUANTITATION OF POMC mRNA AND B-ACTIN mRNA THROUGHOUT GROWTH OF COR L103 CELLS.

RELATIVE INTENSITY OF POMC : B-ACTIN HYBRIDISATION

IN CULTURE EXPERIMENT 1 EXPERIMENT 2 MEAN

0 1.00 1.00 1.00

2 1.21 1.05 1.13

3 0.94 0.87 0.91

4 0.93 1.0 0.97

5 1.04 1.0 1.02

6 1.03 1.0 1.02

7 1.31 1.66 1.49

9 1.54 0.9 1.22

11 1.32 0.76 1.04

14 0.96 1.29 1.13

17 1.39 1.09 1.24

21 1.26 1.26

Slot blots prepared from COR L103 RNA were hybridised sequentially with the bovine POMC cDNA and B-ACTIN cDNA probes. Relative intensity of hybridisation was assessed by autoradiography followed by scanning densitometry.

292 Effect of glucocorticoids on RNA expression in COR L103 cells.

POMC/B-actin (Figure 54, page 163)

10 day incubation Experiment

Hydrocortisone nm/1 1 2 3 4 mean 0 1 1 1 1 1 500 0.97 2.40 2.00 0.94 1.43 1000 1.46 3.10 - 0.89 1.81

24h incubation Experiment

Hydrocortisone nm/1 1 2 3 4 mean 0 1 1 1 1 1 500 2.17 0.60 1.40 1.19 1.33 1000 0.52 - 2.28 1.40

GR/B-actin (Figure 55, page 164)

10 day incubation Experiment

Hydrocortisone nm/1 1 2 3 4 mean 0 1 111 1 500 0.69 0.40 0.45 1.16 0.68 1000 0.82 1.96 0.57 1.30 0.72

24h incubation Experiment

Hydrocortisone nm/1 1 2 3 4 mean 0 1 111 1 500 0.60 2.67 0.84 1.37 1000 0.53 0.79 1.56 0.91 0.95

TAT/B-actin (Figure 55, page 164)

10 day incubation Experiment

Hydrocortisone nm/1 1 2 3 4 mean 0 1 111 1 500 1.1 1.04 1.42 2.01 1.41 1000 0.94 1.16 1.46 2.20 1.44

24h incubation Experiment

Hydrocortisone nm/1 1 2 3 4 mean 0 1 111 1 500 1 0.42 0.83 0.86 0.78 1000 1.16 0.52 2.12 0.71 1.13

293 Effect of glucocorticoids on POMC RNA expression in AtT20 cells

P0MC/i3-actin (Figure 54, page 163)

10 day incubation Experiment

Hydrocortisone nm/1 1 2 3 mean 0 1 1 1 1 500 0.71 0.59 0.36 0.55 1000 0.73 0.78 0.79 0.76

24h incubation (Figure 54, page 163)

Experiment

Hydrocortisone nm/1 1 2 3 4 mean 0 1 1 1 1 1 500 0.95 0.89 0.67 0.85 0.77 1000 0.95 0.46 0.91 1.01 0.76

Source of cells for RNA experiments:

COR L103 cells were cultured by the author. AtT20 cells were cultured by Dr. Adrian Clark and Dr. Paul Lavender (St. Bartholomew’s Hospital).

Corresponding peptide results for Experiments 1, 2 and 3 are:

COR L103 Experiment 1 (Figures 25 and 37)

AtT20 Experiment 2 (Figures 45 and 46)

Experiment 4 was carried out on a separate occasion. Peptide results are not shown as they were not corrected for protein concentration and are therefore not directly comparable.

294 APPENDIX III

Normal probability plots

295 Probability plot correlation coefficient tests for normality (Filliben's test)

Summary of Results

APPENDIX DATA FROM EXPERIMENTS NO. OF CORRELATION CRITICAL VALUE FIG. NO. TEXT FIGS: OBSERVATIONS COEFFICIENT OF r AT 0.05 Irl SIGNIFICANCE LEVEL

Glucocorticoid experiments.

A2 25-36 SCLC lOd precursors 113 0.998 0.987 A3 37-42 SCLC 24h precursors 52 0.993 0.977 A4 43,45,47-49 AtT20 lOd precursors 48 0.983 0.975 A5 44,46 ACTH 30 0.986 0.964 A6 50,52 AtT20 24h precursors 24 0.980 0.957 A7 51,53 ACTH 24 0.966 0.957

Bromocriptine experiments

A8 57 COR L24 18 0.981 0.945 A9 58,59 COR L31 AND COR L42 24 0.985 0.957

oCRF experiments

A10 64-66 COR L24,COR L103 & AtT20 35 0.971 0.968

Human pituitary cells

All 69 Precursors 22 0.982 0.954 A12 70 ACTH excluding 1 outlier: 21 0.973 0.952 including 1 outlier: 22 0.896 0.954

db-cAMP experiments

A13 71,72 COR L24 AND COR L103 12 0.983 0.926 A14 73 COR L24 TIME COURSE excluding 2 outliers: 34 0.996 0.967 including 2 outliers: 36 0.914 0.968 A15 74 COR L24 db-cAMP +/- IBMX 18 0.990 0.945

Forskolin experiments

A16 75,76 COR L24 AND COR L103 24 0.985 0.965

Phorbol ester experiment

A17 77 COR L24 12 0.977 0.941

Pooled SCLC 24h controls A18 - 51 0.992 0.978

A19 55,56 Pooled RNA experiments 57 0.991 0.980 Probability plot correlation coefficient tests for normality (Filliben’s test)

Worked example

Data from figures 75 and 76.

Effect of forskolin on COR L24 and COR L103 cells.

The ’minitab* worksheet is shown in Figure Al.

The command sequence for performing the calculations is as follows:

>oneway C2 Cl C4 >oneway C3 Cl C5

The ’oneway’ command calculates group means and subtracts the appropriate group mean from each value to give the residuals.

>desc C4 C5

The ’describe’ command calculates an overall standard deviation (S.D.) for each set of residuals (i.e. columns 4 and 5).

>let C6 = C4/2.95 (overall S.D. for C4) >let C7 = C5/0.1614 (overall S.D. for C5)

The above commands calculate standardised residuals by dividing the residuals by the overall standard deviation.

>stack C6 C7 C8

Combines standardised residuals in a single column (C8)

>hist C8

Plots histogram of standardised residuals (shown in upper panel of figures A2-A19).

>nscores C8 C9

Calculates normal scores (i.e. expected values) for each standardised residual value (column 9).

>plot C8 C9

Plots observed versus expected values, i.e. the normal probability plot (shown in lower panel of Figures A2-A19).

>corr C8 C9

Calculates correlation coefficient (r) for observed versus expected values which is then compared to the table of critical values given in Filliben 1975.

297 C1 C2 C3 C4 C5 C6 C7 C8 C9 1 1 35.9 3.1 -1 .23333 -0.100000 -0.41808 -0.61958 -0.41808 -0.36854 2 1 38.3 3.1 1 .16666 -0. 100000 0.39548 -0.61958 0.39548 0.36354 3 1 37 .2 3.4 0.06667 0.200000 0.02260 1 .23916 0.02260 0.26024 4 2 42.6 2.8 2.86666 -0.066667 0.97175 -0.41305 0.97175 1 .03703 5 2 41 . 5 3.2 1.76667 0.333333 0.59887 2.06526 0.59887 0.60110 6 2 35. 1 2.6 -4.63334 -0.266667 -1 .57062 -1.65221 -1 .57062 -1.49944 7 3 44. 8 2.6 6.20000 -0.100000 2.10170 -0.61958 2.10170 1.95007 8 3 35.3 2.7 -2.80000 0 .000000 -0.94915 0.00000 -0.94915 -1.03703 9 3 35.2 2.8 -3.40000 0. 100000 -1.15254 0.61958 -1.15254 -1.23521 10 4 48.4 3.1 1.20000 -0.100000 0.40678 -0.61958 0.40678 0.48148 11 4 46.5 3.3 -0.70000 0. 100000 -0.23729 0.61958 -0.23729 -0.15498 12 4 46. 7 3.2 -0.50000 0.000000 -0.16949 0.00000 -0.16949 -0.05147 13 -0.61958 -0.66426 14 -0.61958 -0.66426 15 1.23916 1 .23521 16 -0.41305 -0.26024 17 2.06526 1.49944 18 -1.65221 -1.95007 19 -0.61958 -0.66426 20 0.00000 0.10309 21 0.61958 0.79964 22 -0.61958 -0.66426 23 0.61958 0.79964 24 0.00000 0.10309

Figure A1 Minitab worksheet

Column 1: Groups 1 = controls 2 = 10" M forskolin 3 = 10" M forskolin 4 = 10" forskolin

Column 2 Data for COR L24 Column 3 Data for COR L103 Column 4 Residuals for COR L24 Column 5 Residuals for COR L103 Column 6 Standardised residuals for COR L24 Column 7 Standardised residuals for COR L103 Column 8 Column 6 + Column 7 Column 9 Normal scores for Column 8.

298 SCLC cell lines lOd glucocorticoid experiments

Observed standardised residuals

3 0

2 6

20

10

-2 - 16 0.6 1 6

Normal probability plot

Exptcltd residual values

-1

-2

-3 - 3 -2 0 1 2 3

Figure A2: Data from text figures 25-36. Correlation coefficient (r) for normal probability plot = 0.998

299 SCLC cell lines 24h glucocorticoid experiments

Observed standardised residuals

14

12

10

■2 - 1 6 -1 -0.6 0 0.6 1 1.6 2

Normal Probability Plot

Expected residual value* 3

2

1

0

-1

-2

-3 ■3 •2 •1 0 1 2 3 Standardised residual values

Figure A3: Data from text figures 37-42. Correlation coefficient (r) for normal probability plot = 0.993

300 i j

AtT20: ACTH precursors 10d glucocorticoid experiments

Observed standardised residuals

20

16

10

6

o -2.0 -16 -10 -0.6 0 0.6 10 16 2.0 2.6

Normal probability plot

EiptcUd residual values

/ } . f

-3-2-10 1 2 3 Standardised residual values

Figure A4: Data from text figures 43, 45, 47-49. Correlation coefficient (r) for normal probability plot = 0.983

301 AtT20: ACTH 10d glucocorticoid experiments

Observed standardised residuals

7

6

6

4

3

2

1

0 -16 -10 -0.6 0 0.6 to 1 6 2.0

Normal probability plot

EzptcUd raaidual valuaa

- 1 0 1 2 SLandardia.d raaidual Ttluaa

Figure A5: Data from text figures 44 and 46. Correlation coefficient (r) for normal probability plot = 0.986

302 AtT20: ACTH precursors 24h glucorticoid experiments

Observed standardised residuals

10

0.6 1 0 1 6 2.0

Normal probability plot

Expected residual valuti

-2 -1 0 SU ndtrdiitd residua!

Figure A6: Data from text figures 50 and 52. Correlation coefficient (r) for normal probability plot = 0.980

303 AtT20: ACTH 24h glucorticoid experiments

Observed standardised residuals

10

-16 -10 -0.6 0 0.6 10 1 6 2.0

Normal probability plot

Expactad residual nluw

2

-1.5 - 1 -0.5 0 0.5 1 1.5 Standardised residual values

Figure A7: Data from text figures 51 and 53. Correlation coefficient (r) for normal probability plot = 0.966

304 COR L24 24h bromocriptine experiment

Observed standardised residuals

7

6

6

4

3

2

1 ^499999

0 -16 -10 -0.6 0 0.6 10 16 2.0 2.6

Normal probability plot

Expected residual n lu n

-2-10 1 2 3 Standardised residual values

Figure A8: Data from text figure 57. Correlation coefficient (r) for normal probability plot = 0.981

305 COR L31 and COR L42 24h bromocriptine experiments

Observed standardised residuals

7

8

6

4

3

2

1

0 -2.0 -16 -10 -0.6 0 0.6 10 16 2.0 2.6

Normal probability plot

Expected residual values

- 1 0 1 2 Standardised residual values

Figure A9: Data from text figures 58 and 59. Correlation coefficient (r) for normal probability plot = 0.985

306 COR L42, COR L103 a n d A tT 20 CRF e x p erim en ts

Observed standardised residuals

10

-1.6 - 1 0 -0.6 0 0.6 1.0 16 2.0 2.6

Normal probability plot

Expected residual values

2

Standardised residual values

Figure A10: Data from text figures 64-66. Correlation coefficient (r) for normal probability plot = 0.971

307 Human pituitary cells ACTH precursor secretion

Observed standardised residuals

-16 -10 -0.6 0 0.6 10 16

Normal probability plot

ExpecUd residual value*

-2 -1.5 -1 -0.5 0 0.5 1 1.5 2 Standardised residuals values

Figure All: Data from text figure 69. Correlation coefficient (r) for normal probability p l o t = 0.982

308 Human pituitary cells ACTH secretion

Observed standardised residuals 8

7

6

6

4

3

2

1

0 -2.0 -16 -10 -0.6 0 0.6 1.0 16 2.0 2.6

Normal probability plot

Ezp«ct«d residual values

1

- 2 ------3-2-10 1 2 3 Standardised residual values

Figure A12: Data from text figure 70. Correlation coefficient (r) for normal probability plot = 0.973

309 COR L24 and COR L103 db —cAMP experiment (48h incubation)

Observed standardised residuals 6

4

3

2

0 -10 - 0.6 0 0.6 10 16-16

Normal probability plot

Expected residual values

-i.5 -1 — 0.5 0 0.5 1 1.5 2 Standardised residual values

Figure A13: Data from text figures 71 and 72. Correlation coefficient (r) for normal probability plot = 0.983

310 COR L24 48h db —cAMP experiment

Observed standardised residuals

10

s

6

4

2

0 -2.0 -16 -10 -0.6 0 0.6 10 16 2.0 2.6

Normal probability plot

Expected residual vilusi

-3-2-10 1 2 3 Standardised residual values

Figure A14: Data from text figure 73. Correlation coefficient (r) for normal probability plot = 0.996

311 COR L24 db —cAMP + / — IBMX experiments

Observed standardised residuals e

6

4

3

2

1

0 -16 - 1.0 - 0.6 0 0.6 10 1 6 2.0

Normal probability plot

ExptcUd m idutl t i Iu m

- 2 1------2 — 1.5 -1 -0.5 0 0.5 1 1.5 2 Standardised residual values

Figure A15: Data from text figure 74. Correlation coefficient (r) for normal probability plot = 0.990

312 COR L24 and COR L103 Forskolin experiments

Observed standardised residuals

7

6

6

4

3

2

1

0 -16 -10 -0.6 0 0.6 10 1 6 2.0

Normal probability plot

Expected residual values

-2-10 1 2 Standardised residual values

Figure A16: Data from text figures 75 and 76. Correlation coefficient (r) for normal probability plot = 0.985

313 COR L24 Phorbol ester experiment

Observed standardised residuals

3.6

2.6

1.6

0.6

0.6 10 1.6

Normal probability plot

Expactad m idual n lu u

- 2 -1.5 -1 -0.5 0 0.5 1 1.5 2 Standardised rssidual value.

Figure A17: Data from text figure 77. Correlation coefficient (r) for normal probability plot = 0.977

314 SCLC cell lines Pooled 24h control values

Observed standardised residuals

12

10

8

6

4

2

0 ■2.0 - 1 6 - 1 0 - 0.6 0 0.6 1 0 1 6 2.0 2.6 3.0

Normal probability plot

Expected residual n lu tt

-2-10 1 2 Standardised residual values

Figure A18: Pooled control data from 24h incubations of SCLC cell lines. Correlation coefficient (r) for normal probability plot = 0.992

315 ooled RNA experiments

Observed standardised residuals

1 4

12

10

-2 - 1 6 -1 -0.6 0 0.6 11 6 2

Normal probability plot

Expactad rail dual vftluai

- 2

-3 -3 -2 -1 0 1 2 3

Pooled RNA experiments. Correlation coefficient (r) for normal probability plot = 0.991 APPENDIX IV

Summary of results of statistical analysis

317 Effect of glucocorticoids on ACTH precursor secretion bv SCLC cell lines

Results of Analysis of Variance

10 day incubations

CELL LINE Experiment No. Figure No. Page D.F. F Ratio £

COR L103 1 25 132 2,9 8.77 0.008 2 26 133 2,6 0.24 0.793

COR L24 1 29 136 2,9 1.16 0.355 2 30 137 2,6 1.82 0.241

COR L31 1 31 138 2,9 0.15 0.859 2 32 139 2,6 0.01 0.990 3 33 140 2,5 0.18 0.840

COR L42 1 34 141 2,6 1.68 0.263 2 35 142 2,6 3.85 0.084

COR L27 1 36 143 2,9 10.89 0.004

24h incubations

CELL LINE Experiment No. Figure No. Page D.F. F Ratio £

COR L103 1 37 144 2,6 0.53 0.624 2 38 145 2,4 2.33 0.179

COR L24 1 39 146 2,6 0.36 0.709

COR L31 1 40 147 2,6 1.08 0.396

COR L42 1 41 148 2,6 0.23 0.798

COR L27 1 42 149 2,6 5.83 0.039

Results of two-tailed unpaired t tests

10 day incubations

CELL LINE Experiment No. Figure No. Page D. F. t £ 95% C. I. for difference in means

COR L103 3 27 134 2 0.11 0.92 (-64, 67.0)

4 28 135 3 1.21 0.31 (-0.76,1.69)

D.F. = Degrees of freedom. C.I. = Confidence interval. Effect of glucocorticoids on ACTH peptide secretion Dv AtT20 cells.

Results of Analysis of Variance

10 day incubations

Experiment No. Figure No. Page D.F. F Ratio £

1. Precursors 43 151 5,12 35.66 <0.0001 ACTH 44 152 5,12 77.75 <0.0001

2. Precursors 45 153 2,9 6.58 0.017 ACTH 46 154 2,9 18.20 0.001

lcubations

Experiment No. Figure No. Page D.F. F Ratio £

1. Precursors 50 158 5,12 3.23 0.045 ACTH 51 159 5,12 1.69 0.210

Results of two-tailed unpaired t tests

Experiment No. Figure No. Page D.F. 95% C.I. for difference in means

3. Precursors 47 155 2 16.42 0.0037 (5.31,9.078) 4. Precursors 48 156 2 52.08 0.0004 (5.95,7,023) 5. Precursors 49 157 2 13.50 0.0054 (1.72,3.323)

319 Effect of glucocorticoids on RNA expression.

Results of analysis of variance.

COR L103 cells

Experiment Figure No. Page D.F. F ratio £

POMC/B-actin 54 163

10d incubation 2,8 0.96 0.422 24h incubation 2,7 0.39 0.689

GR/B-actin 55 164

10d incubation 2,9 1.11 0.371 24h incubation 2,9 0.44 0.658

TAT/B-actin 55 164

10d incubation 2,9 1.38 0.301 24h incubation 2,9 0.66 0.541

AtT20 cells

Experiment Figure No. Page D.F. F ratio £

POMC/B-actin 54 163

10d incubation 2,7 18.63 0.002 24h incubation 2,9 1.38 0.300

These data were also analysed by Dunnett’s multiple range test (Dunnett 1955) which gave the same results (Clark et al, 1990).

320 Bromocriptine experiments

Results of Analysis of Variance

CELL LINE Experiment No. Figure No. D.F. Page F Ratio p

COR L24 24h incubation 57 5,12 168 17.16 <0.0001 (0,0.1,0. 5,1,5,10|jg/ml)

COR L31 24h incubation 58 3,8 170 6.99 0.013

COR L42 24h incubation 59 3,8 171 19.30 0.001

Results of two-tailed, unpaired t tests

CELL LINE Experiment No. Fig.No. Page D.F. t p 95% Cl for difference in means

COR L24 7d incubation 1 60 172 2 5.11 0.036 (49,567.7) 2 61 173 3 47.96 <0.0001 (133.5,152.5)

COR L31 7d incubation 62 174 2 14.01 0.0051 (5.96,11.24)

COR L42 7d incubation 63 175 2 16.93 0 . 00 3 5 ( 43.2,72.59) Effect of oCRF on ACTH peptide secretion

Results of Analysis of Variance

CELL LINE Experiment No. Figure No. Page D.F. F Ratio

SCLC cell 72h incubation: 1 ines

COR L42 Precursors 64 178 3,8 2.27 0.157

COR L103 Precursors 65 179 3,7 0.21 0.886

AtT20 72h incubation:

Precursors 66 180 3,8 9.98 0.004 ACTH 67 181 3,8 18.57 0.001

Effect of different media on response to oCRF

Results of two-tailed, unpaired t tests

CELL LINE Figure No. Page Medium D.F. 95% Cl for difference in means

COR L24 68 183 RTISS 2.5 3 -1.95 0.15 (-4.04,0.97) RTIS S 3 -0.99 0.40 (-2.10,1.10) RBSAT 2 -0.19 0.86 (-0.847,0.77)

Human pituitary cells

Results of Analysis of Variance

Experiment Figure No. Page D.F. F Ratio

ACTH Precursor secretion 69 185 2,19 69.23 < 0.0001

ACTH 70 including 186 2,19 110.00 < 0.0001 outlier excluding 2,18 278.94 <0.0001 outlier

322 Effect of 1mm db-cAMP on SCLC cells (48h incubation)

Results of two-tailed unpaired t test

CELL LINE Figure No. Page t £ D.F. 95% C.I

COR L24 71 189 -5.67 0.030 2 (-71.3, -9.8)

COR L103 72 190 -6.02 0.026 2 (-6.06, -1.01)

Effect of 1mM db-cAMP on COR L24 cells (48h time course) (Figure 73. Page 192)

Results of linear regression analysis

Controls: slope = 1.08 (SE 0.05), intercept = 2.52 (SE 0.92) db-cAMP slope = 1.65 (SE 0.04), intercept = 1.10 (SE 0.91) treated cells:

Difference in intercepts : p = 0.279

Difference in slopes : p<0.001

95% Confidence interval for the difference in slopes = 0.44-0.7003.

The data were analysed excluding two outliers (identifiedin Appendix II). The effect of including the outliers would be to exaggerate the difference in slopes.

Effect of db-cAMP +/- IBMX on COR L24 cells. (Figure 74. Page 195)

Results of analysis of covariance

F ratio

Effect of db-cAMP 0-100pm/l 2.35 0.14

Effect of IBMX 38.33 < 0.0001

Interaction 0.24 0.79

323 Effect of forskolin on ACTH precursor secretion bv SCLC cells

Results of analysis of variance

CELL LINE Figure No. Page F Ratio jd D.F.

COR L24 75 198 5.04 0.030 3,8

COR L103 76 199 5.23 0.027 3,8

Effect of phorbol 12. 13-dibutyrate on ACTH precursor secretion b SCLC cel Is.

Result of analysis of variance

CELL LINE Figure No. Page F Ratio jd D.F.

COR L24 77 203 13.99 0.002 3,8

324 APPENDIX V

Publications and communications arising from this Thesis

325 Refereed publications

S.R. Crosby, M.F. Stewart, J.G. Ratcliffe and A. White. (1988)

Direct measurement of the precursors of adrenocorticotrophin in human plasma by two-site immunoradiometric assay. Journal of Clinical

Endocrinology and Metabolism 67(6) 1272-1277.

M.F. Stewart, S.R. Crosby, S. Gibson, P.R.Twentyman and A. White. (1989)

Small cell lung cancer cell lines secrete predominantly ACTH precursor peptides not ACTH. British Journal of Cancer 60, 20-24.

A. White, M.F. Stewart, W.E. Farrell, S.R. Crosby, P.M. Lavender, P.R.

Twentyman, L.H. Rees, and A.J.L. Clark. (1989)

Pro-opiomelanocortin gene expression and peptide secretion in human small cell lung cancer cell lines. Journal of Molecular Endocrinology 3, 65-70.

A.J.L. Clark, M.F. Stewart, P.M. Lavender, W.E. Farrell, S.R. Crosby, L.H.

Rees, A. White. (1990)

Defective glucocorticoid regulation of pro-opiomelanocortin gene expression and peptide secretion in a small cell lung cancer cell line. Journal of

Clinical Endocrinology and Metabolism 70, (2) 485-490.

326 International Communications

Variable patterns of ACTH secretion by small cell lung cancer (SCLC) cell lines.

M.F. Stewart, P.R. Twentyman and A. White.

First European congress of Endocrinology, Abs 227, Copenhagen, June 1987.

Small cell lung cancer cell lines : secretion of ACTH in culture.

A. White, M.F. Stewart and P.R. Twentyman.

European Tissue Culture Society Meeting, Glasgow, July, 1987.

Low level production of ACTH by small cell lung cancer (SCLC) cell lines.

A. White, P.R. Twentyman, A.J.L. Clark, P.M. Lavender, L.H. Rees and M.F.

Stewart.

International conference on hormones, growth factors and oncogenes in pulmonary carcinoma, Dartmouth, New Hampshire, USA, August 1987.

ACTH is secreted predominantly as its precursor by small cell lung cancer

(SCLC) cell lines in vitro.

A. White, S.R. Crosby, P.R. Twentyman and M.F. Stewart.

8th International Congress of Endocrinology, Kyoto, Japan, July 1988.

Glucocorticoid regulation of the synthesis and secretion of pro­ opiomelanocortin in human lung tumour cells.

A. White, M.F. Stewart, W.E. Farrell, S.R. Crosby, P.M. Lavender, L.H.

Rees and A.J.L. Clark.

American Endocrine Society, Seattle, Washington, U.S.A. June, 1989.

3277 Direct quantitation of ACTH precursors secreted by human pituitary and lung tumours.

S.R. Crosby, M.F. Stewart and A. White.

American Endocrine Society, Seattle, Washington, U.S.A. June, 198:9.

Regulation of pro-opiomelanocortin gene expression and peptide secretion in SCLC cell lines by the dopamine agonist bromocriptine.

Farrell W. E., Stewart M. F., Crosby S. R., Clark A. J. and White A.

73rd Annual Meeting of the Endocrine Society, Washington DC, USA.

June, 1991.

National communications

ACTH and its precursor peptides in small cell lung cancer.

M.F. Stewart, S.R. Crosby and A. White.

Endocrine section of the Royal Society of Medicine, London, March 1988.

Development of an immunoradiometric assay for the precursors of ACTH.

S.R. Crosby, M.F. Stewart, J.G. Ratcliffe and A. White.

7th Joint Meeting of the British Endocrine Societies, Exeter.

J Endocrinol (1988) 117, 74.

Production of adrenocorticotrophin and its precursor peptides by human small cell lung cancer (SCLC) cell lines in vitro.

M.F. Stewart, S.R. Crosby, P.R. Twentyman and A. White.

Focus *88 Annual Meeting of the Association of Clinical Biochemists,

Blackpool, May 1988.

328 Oestradiol in small cell lung cancer cell lines.

W.E. Farrell, M.F. Stewart and A. White.

Lung Cancer Workshop, Middlesex Hospital, London, October 1988.

ACTH precursor peptides in small cell lung cancer cell lines.

M.F. Stewart, W.E. Farrell, S.R.Crosby and A. White.

Lung Cancer Workshop, Middlesex Hospital, London, October 1988.

Comparison of ACTH and ACTH precursor peptides secreted by human pituitary and lung tumour cells in-vitro.

S.R. Crosby, F. Stewart, W. E. Farrell, S. Gibson, and A. White.

8th Joint Meeting of British Endocrine Society, April, 1989 Manchester.

J. Endocrinol. (1989) 121 : 209.

Pro-opiomelanocortin gene expression is not regulated by glucocorticoids in cultured small cell lung cancer cells.

M.F. Stewart, A.J.L. Clark, W.E. Farrell, S.R. Crosby, P.M. Lavender,

P.R. Twentyman, L.H. Rees and A. White.

8th Joint Meeting of British Endocrine Society, April, 1989 Manchester.

J. Endocrinol. (1989) 121 : 16.

Discordance between ACTH precursor peptide secretion and gene expression in small cell lung cancer cell lines.

W.E.Farrell, M.F. Stewart, S.R. Crosby, A.J.L. Clark & A. White.

9th Joint Meeting of British Endocrine Society, Glasgow, March 1990.

J Endocrinol (1990) 124: 54.

329 Relative resistance to glucocorticoid regulation of ACTH precursor peptide

secretion and gene expression in small cell lung cancer cell lines.

W.E. Farrell, M. F. Stewart, S. R. Crosby, A.J.L. Clark and A. White.

10th Joint Meeting of British Endocrine Societies, Brighton, April, 1991.

J Endoc Vol 129, Suppl, April 1991, Abstract 161.

330 Pro-opiomelanocortin gene expression and peptide secretion in human small-cell lung cancer cell lines

A. White, M. F. Stewart, W. E. Farrell, S. R. Crosby, P. M. Lavender*, P. R. Twentymanf, L. H. Rees* and A. J. L. Clark Department of Clinical Biochemistry, University of Manchester, Hope Hospital, Salford M 6 8 HD * Department of Chemical Endocrinology, St Bartholomew’s Hospital, London EC1A 7BE tM RC Clinical Oncology and Radiotherapeutics Unit, Hills Road, Cambridge CB2 2QH (A. J. L. Clark is now at Endocrine and Reproduction Research Branch, NICHD, NIH, Bethesda, Maryland 20892, U.S.A.)

r e c e iv e d 30 November 1988

ABSTRACT

Expression of the RNA coding for the ACTH- the cells, using a novel two-site immunoradiometric P-lipotrophin precursor, pro-opiomelanocortin assay based on monoclonal antibodies, which di­ (POMC), has been demonstrated in five human rectly quantifies both POMC and pro-ACTH but small-cell lung cancer (SCLC) cell lines. Using does not recognize ACTH. Levels of POMC in Northern and slot-blot hybridization analysis of medium accumulated throughout the growth of the RNA and a bovine POMC cDNA as probe, the pro­ cells, in contrast to POMC RNA which showed a cessed POMC RNA from SCLC cells was found to relatively constant level of expression. We conclude be approximately 1350 nucleotides in length, which that human SCLC cell lines are valuable models for is larger than that found in the normal human pitui­ studying the aberrant expression and regulation of tary. Expression of the POMC gene was confirmed the human POMC gene. by measurement of ACTH precursors secreted by Journal of Molecular Endocrinology (1989) 3, 65-70

INTRODUCTION Evidence for aberrant post-translational processing of POMC in SCLC rests on the demonstration of Adrenocorticotrophin (ACTH) is synthesized as high molecular weight precursor forms of ACTH part of a precursor, pro-opiomelanocortin (POMC), circulating in patients with the ectopic ACTH syn­ of approximate molecular weight 31 kDa, which is drome (Hale, Besser & Rees, 1986). Hitherto, there cleaved to yield an intermediate ACTH-containing has been no direct means of quantifying these pre­ peptide designated pro-ACTH. POMC is produced cursor peptides, and their measurement has relied primarily by the pituitary but can be detected in on chromatographic separation followed by radio­ lung, thyroid, brain, gastrointestinal tract, repro­ immunoassay using antisera to component peptides ductive tract and leucocytes (Krieger, 1983). POMC such as ACTH and y-melanocyte-stimulating is also found in tumours of non-pituitary origin, par­ hormone (y-MSH). This method is critically ticularly small-cell lung cancer (SCLC) which is dependent upon the cross-reactivity of the antisera well recognized for its ability to produce ACTH- employed in the detection assay and is unsuitable related peptides. In these tumours the expression of for large numbers of samples. We have overcome the POMC gene may be abnormal in comparison this problem by the development of a two-site with the pituitary in respect of both the size of the immunoradiometric assay (IRMA) for the precur­ POMC mRNA (Tsukada, Nakai, Jingami et al. sors, based on monoclonal antibodies to ACTH and 1981; DeBold, Schworer, Connor et al. 1983; Steen- y-MSH, which enables direct and precise quantifi­ bergh, Hoppener, Zandberg et al. 1984; de Keyzer, cation of both the precursor peptides POMC and Bertagna, Lenne et al. 1985; Hoppener, Steenbergh, pro-ACTH (Crosby, Stewart, Ratcliffe & White, M oonen et al. 1986; Clark, Lavender, Besser & 1988). With this assay we have found that ACTH Rees, 1989) and the post-translational processing of was secreted from SCLC cells in vitro predom i­ the POMC precursor molecule (Ratcliffe, 1985). nantly in precursor forms, with very little, if any, Journal of Molecular Endocrinology (1989) 3, 65-70 © 1989 Journal of Endocrinology Ltd Printed in Great Britain 0952-5041/89/003-065 802.00/0 ' 66 \. WHITE ;ind others • PO\l(' gene expression in cells

processing to ACTH(1 39) (Stewart, Crosby, Gib­ Growth experiments with SCLC cell lines son et al. 1989). At the start of each experiment SCLC cells were Study of the molecular biology of POMC in pipetted to reduce the size of the aggregates to five to SCLC has been restricted, in part, because of diffi­ ten cells, diluted in RTISS (2-5%) and replicate cul­ culty in obtaining adequate tissue samples, since tures established at an approximate cell density of this disease is rarely amenable to surgery. Major 0 5 x 105 cells/ml. The medium was not changed advances have, however, been made in recent years during the growth of the cells. At each time-point, in establishing SCLC cell lines (Baillie-Johnson, separate cultures were harvested and samples of Twentyman, Fox et al. 1985; Carney, Gazdar, supernatant media, cell suspension (for DNA or Bepler et al. 1985) and we aim to prove that these protein measurements) and cell pellets (for RNA provide suitable in-vitro models in which the phe­ measurements) were flash-frozen and stored at nomenon of hormone production can be studied. We — 20 'C before analysis. Protein was assayed using have examined the expression of POMC in five the Bio-Rad method (Bio-Rad Laboratories Ltd, SCLC cell lines, by comparing POMC gene ex­ Watford, Herts, U.K.). pression with secretion of ACTH and the precursor peptides, measured using specific IRMAs. Estimation of cellular DNA DNA was assayed according to a fluorometric method (West, Sattar & Kumar, 1985), the fluoro- MATERIALS AND METHODS chrome dye being Hoechst 33258 (Uniscience Ltd, London, U.K.). Salmon testes DNA (type III, SCLC cell lines Sigma) was used as standard. The SCLC cell lines (COR L24, COR L27, COR L31, COR L42 and COR L103) were established IRMA for ACTH from patients with pathologically confirmed SCLC The ACTH IRMA was developed and optimized as (Baillie-Johnson et al. 1985). The cell lines grow as previously described (White, Smith, Hoadley et al. floating aggregates of cells in suspension. Cells were 1987). The assay employed two monoclonal anti­ cultured in a growth medium, RTISS (2 5) consist­ bodies (MAbs): MAb 1A12 (specific for ing of RPMI 1640 (Gibco, Paisley, Strathclyde, ACTH(10—18)) was radioiodinated, and MAb 2A3 U.K.) supplemented with 2-5% (v/v) fetal calf (specific for ACTH(25-39)) was coupled to Sepha- serum (Gibco), 10 |ig human transferrin/ml (Sigma cryl S300 as solid phase. Human ACTH standards Chemical Co. Ltd, Poole, Dorset, U.K.), 5 fig (code 74/555; National Institute for Biological Stan­ bovine insulin/ml (Sigma), 30 n M sodium selenite dards and Control, London, U.K.) were prepared at (Sigma), 4 m\i glutamine (Flow Laboratories, concentrations between 1-1 and 1110 pM. The assay Irvine, Strathclyde, U.K.), 1 mM sodium pyruvate sensitivity (2-5 x s .d . at zero A C T H ) is 0-8 pM and (Flow) and 10 m M Hepes buffer (Flow). Antibiotics the within- and between-assay coefficients of varia­ were not used. All incubations were carried out at tion are <10% at 4-9—1110 pM and 6—lllOpM re­ 37 °C in a 5% C 0 2 atmosphere. spectively. The assay measures ACTH(l-39) and also recognizes POMC and pro-ACTH with cross- reactivities of <1% and <10% respectively. The Mouse pituitary cell line assay does not detect fragments of ACTH such as AtT20 cells (obtained from the American Tissue a-MSH, ACTH(18-39) and ACTH(l-24). Culture Collection, Rockville, MD, U.S.A.) were cultured as described previously (Buonassisi, Sato & IRMA for ACTH precursor peptides Cohen, 1962; Orth, Nicholson, Mitchell et al. 1973) The development of the precursor assay is described in Ham’s F10 medium (Flow) supplemented with in detail elsewhere (Crosby et al. 1988). MAb 1A12 15% (v/v) horse serum, 2*5% (v/v) fetal calf serum, (specific for ACTH(10-18)) was radioiodinated, and 4 m M glutamine, 100 IU penicillin/ml and 100 IU MAb 1 Cl 1 (specific for y 1-MSH) was coupled to streptomycin/ml. Sephacryl S300 as solid phase. A partially purified POMC standard was prepared from growth medium of a cultured human pituitary tumour by Sephadex Cell counts G-75 chromatography under acid dissociating con­ Single cell suspensions were prepared by trituration ditions. The POMC fraction was initially assigned and viable counts obtained using trypan blue an arbitrary potency of 10 000 precursor units/ 1, exclusion. corresponding to 26 n.M as determined by the fluoro- Journal of Molecular Endocrinology (1989) 3, 65-70 P O M C gene expression in S C L C calls \. win I U and others 67

metric assay for N-terminal tryptophan (Hakanson were expressed as the ratio of POMC signal to & Sundler, 1971). Standards were prepared at con­ P-actin signal. centrations of 0-2600 p.vi. The assay sensitivity For Northern blot analysis, poly(A)''' RNA was (2-5 xs.D. at zero POMC) is 2’6 p.vi and the within- selected by affinity chromatography of total RNA and between-assay coefficients of variation are with oligo(dT) Sepharose (Uniscience Ltd), eluting < 10% between 20 and 2600 p.vi and between 37 and the bound fraction in Tris-HCl (10m.\i; pH 7*5) 2600 p M respectively. The assay measures both containing EDTA (1 m.M). The poly(A)+ RNA POMC and pro-ACTH but does not distinguish be­ (2 gg) produced was separated by electrophoresis on tween them. Other POMC-derived peptides, e.g. a 1-4% agarose, 6 % formaldehyde gel. This RNA ACTH, P-lipotrophin and N-pro-opiocortin, are was then transferred to nitrocellulose paper by not recognized. Northern blotting for 18 h, and the filter was then baked, pre-hvbridized and hybridized as before, but using an 800 bp probe derived from exon 3 of the Analysis of RNA cloned human POMC gene (a gift from Professor S. Pellets of SCLC cells were stored at — 70 °C until Cohen, Stanford University, Stanford, CA, U.S.A.). used for preparation of total cellular RNA by the After hybridization for 18 h the filter was washed as method of Chirgwin, Przybyla, McDonald & Rutter before, except that the latter washes were at 55 JC. (1979). Briefly, cells were homogenized in 4 M guan­ The filter was exposed to film for 14 days. idine isothiocyanate, layered on to a cushion of 5 7 m caesium chloride and centrifuged for 18 h at 50 000 g. The precipitate of RNA was resuspended RESULTS in Tris—HC1 (10 m M ; pH 7-5) containing E D T A (1 m M ) and extracted with an equal volume of The presence of POMC RNA transcripts in five hu­ chloroform:isobutanol (4:1, v/v) before precipita­ man SCLC cell lines (COR L24, L27, L31, L42 and tion with ethanol. RNA was quantified by measure­ L103) was detected by slot-blot analysis using RNA ment of absorbance at 260 nm. Aliquots of RNA populations probed with the bovine POMC cDNA (2-5-10 pg) in 5 x SSC (lxSSC is 0-15 m sodium probe (Table 1). Figure 1 compares Northern blots chloride and 0.015 m sodium citrate; pH 7-0) were of COR LI03 and normal pituitary RNA. The applied to nitrocellulose paper (Schleicher & SCLC cell RNA was approximately 1350 bases in Schuell, Dassel, F.R.G.) using a slot-blot apparatus length, and clearly larger than the 1250 base pitui­ (Gibco). The filter was then baked at 80 °C under tary species. Furthermore, the pituitary signal ob­ vacuum for 2 h before prehybridization and hybridi­ tained with 2 pg total RNA was considerably more zation as described by Clark et al. (1989). The probe intense than the signal obtained with 2 pg COR used was the 1100 bp insert from the plasmid LI03 poly(A)+ RNA. Secretion of ACTH was de­ pSNAC20 (a gift from Professor S. Numa, Kyoto tected in the five SCLC cell lines grown for 7 days in University, Kyoto, Japan) which contains the full- culture without medium change (Table 1). The length bovine POMC cDNA. This probe was mean ACTH level was 0-26 pmol/mg protein, with a labelled with [32P]dCTP (Amersham International range of 0-09-0-51 pmol/mg protein. These levels pic, Amersham, Bucks, U.K.), using oligonucleo­ were relatively low and when the secretion of ACTH tide-primed synthesis (Feinberg & Vogelstein, precursors w-as compared with that of ACTH in the 1983), to a specific activity of approximately five cell lines, much higher levels of precursors were 108 c.p.m./pg DNA; 10 7 c.p.m. were used per observed (mean 20-2, range 5-3-34-2 pmol/mg pro­ hybridization. Filters were washed (2 x 30 min in tein). In contrast, the AtT20 mouse pituitary 2 x SSC containing 0-2% (w/v) sodium dodecyl sul­ tumour cell line secreted both ACTH and the pre­ phate (SDS), followed by 2 x 30 min in 0-2 x SSC cursor peptides at 1000-fold higher concentrations containing 0-2% SDS at 50 °C) and subjected to than found in the SCLC cell lines (mean precursor autoradiography with Kodak XAR-5 film and in­ levels 18 200 pmol/mg protein; mean ACTH levels tensifying screens at — 70 °C. Films were developed 400 pmol/mg protein). after 1—3 days and slot intensity was quantified using In order to use the SCLC cell line as an in-vitro a Parry D TI 405 densitometer. model to study human POMC gene expression, it The POMC probe was then washed off the filter was necessary to characterize any variation during using 0*2% SDS at 95 °C and, after air drying, the cell growth. Thus COR L103 cells were seeded at filter was rehybridized with the chick P-actin probe approximately 0-5 x 10 5 cells/ml and cultured for labelled in a similar way. After washing, the filter 17 days. Figure 2 a shows the increase in DNA was exposed to film as before but for 2—24 h. Slot throughout the growth of these cells, with a compar­ intensity was quantified in the same way, and results able increase in ACTH precursor levels (Fig. 2b) Journal of Molecular Endocrinology (1989) 3, 65—70

// / / A. WHlTl? and others ■ P ()M (' gene expression in S ( 'I X ' t ells

TABi.ii 1. Ratio of POMC RNA to P-actin RNA, determined by slot-blot analysis, and concentrations of ACTH precursors and ACTH in supernatant medium of small-cell lung cancer (SCLC) cell cultures established at 0-5 x IU5 cells/ml and grown for 7 days without a change of medium. Results are from a single experiment

Ratio of POMC RNA ACTH precursors ACTH to fPactin RNA (pmol/mg protein) (pmol/mg protein) C ell lin e COR L24 2 0 8-5 0-15 COR L27 2 1 5-3 009 CORL31 1-2 34-2 0-51 COR L42 4-2 30-9 0-27 COR L103 10 22-3 0-30

(a) (6)

40— 1 (a)

i 20—

10—

60- 50 ^ g y 50— Bases 40- au -£ iSB ■1353 - ^ 30- 5 20— ■1150 10-

0-7-| (c) ■665 0-6- < 0-5- X - 0-4- ac 0 3 - £ 02- 01- 0—1

Time in culture (days) f i g u r e 1. Autoradiograph of a Northern blot showing (a) RNA (2 pg poly(A)+ RNA) from the small-cell lung f i g u r e 2. ACTH precursors and ACTH throughout the cancer cell line COR L103 and ( b) RNA (2 pg total RNA) growth of a small-cell lung cancer cell line (COR L I03). from normal human pituitary cells, both probed with Cells were cultured for 17 days in RTISS (2 5) medium exon 3 of the human POMC gene. The sizes indicated and peptides measured in supernatant media. Results are based on the hybridizable size markers described by shown are the means of four separate growth Clark et al. (1989). experiments. Journal of Molecular Endocrinology (1989) 3, 65-70 1*0\I(' gene expression in S('I.(' cells • a. w hite ;ind others

1-5- line consists of approximately 1350 bases and is pre­ sent at a very low level relative to that in the pitui­ tary. It is of slightly larger size than that found in the normal human pituitary. A POMC transcript of similar size has been described in a number of ecto­ * pic ACTH-producing tumours (Tsukada et al. 1981; ~ 1-0— j S DeBold et al. 1983; Steenbergh et al. 1984; de Keyzer et al. 1985; Hoppener et al. 1986; Clark et al. * cs. 1989). The origin of such a transcript is probably < from a second upstream promoter, but this has not s Y s been shown in this case. Such a transcript directs s translation from the same initiation codon as the H 0-5-n * normal pituitary promoter, and thus no alternative or additional amino acid sequence would be predicted. In comparison with the AtT20 cell line, the cells from the human SCLC cell lines express the POMC gene at a much lower level and this is reflected in the 2 3 4 5 6 14i 1000-fold difference in the levels of ACTH precur­ Time in culture (days) sor peptides. The low levels of ACTH measured in the COR L I03 cell line can be accounted for by the fig u r e 3. The level of POMC gene expression during the course of the growth curve of a small-cell lung cancer cell cross-reactivity of the precursors in the ACTH line (COR L103) expressed as the ratio of POMC RNA IRMA. Thus authentic ACTH(l-39) is not pro­ levels to P-actin RNA levels. Results shown are the duced to any significant extent. In contrast, the means of two separate growth experiments. mouse pituitary tumour cells produce substantial amounts of both ACTH and precursors. This pre­ and ACTH (Fig. 2c), although the latter was at a dominance of precursors and absence of processing much lower level of secretion. Measurement of the has been observed in all the ten human SCLC cell stability of ACTH and ACTH precursors from lines found to produce ACTH-related peptides COR L I03 in medium maintained at 37 °C showed (Stewart et al. 1989). that 92 and 74% respectively remained after 72 h. From the growth experiments, detection of sig­ This would indicate that the increase in ACTH and nificant levels of precursor peptides in the SCLC precursors relates to a steady accumulation of the cell line appears critically dependent upon culturing peptides in the culture medium. At each time-point, the cells to high density (i.e. for 10 days) but the slot-blot analysis of RNA was performed and the main reason for this is that it is difficult to detect the abundance of POMC RNA related to the expression low levels of peptides unless they are allowed to of a P-actin internal control probe. The ratio of accumulate in medium. This is in keeping with the POMC to P-actin RNA throughout the growth of levels of POMC RNA (Fig. 3) which indicate that these cells (Fig. 3) indicated that expression of the gene expression is fairly constant and not influenced POMC gene remained relatively constant even when by the changes from log growth to stationary phase. the cells had ceased to divide. In conclusion, we have found that the human SCLC cell lines provide suitable in-vitro models for the study of POMC gene expression and peptide DISCUSSION secretion, and we have detected abnormalities in the processing of the POMC peptide and in the size of Dynamic studies on the regulation of POMC gene the mRNA consistent with a model for the ectopic expression and peptide secretion are limited to ani­ ACTH syndrome. mal models using either the AtT-20 mouse pituitary tumour cell line or primary cultures of rat anterior pituitary cells. No human model of either normal or ACKNOWLEDGEMENTS ectopic POMC expression is readily available. The detection of POMC RNA in cell lines derived from We thank Mrs S. Gibson for measurement of ACTH. patients with SCLC does, however, provide an This work was supported by a grant from the North opportunity to examine the inappropriate expression West Regional Health Authority. A.J.L.C. was of POMC in this neoplasm. supported by the MRC, and P.M.L. by the Joint The mature POMC mRNA from the SCLC cell Research Board, St Bartholomew’s Hospital. Journal of Molecular Endocrinology (1989) 3, 65-70 70 \. w hite and others POMC gene expression in S ('l.(' cells

REFERENCES I loppcncr, J. W. M., Steenhergh, P. 11., Moonen, P. J. J., Wugcnaar, S. S., Junsz, II. S. & Lips, C. J. M. (1986). Bail lie-Johnson, II., Twentyman, P. It., Fox, N. K., Walls, Detection of mRN A encoding calcitonin, calcitonin gene G. A., Workman, P., Watson, J. V., Johnson, N., Reeve, J. G. related peptide and proopiomelanocortin in human tumors. St Bleehen, N. M. (1985). Establishment and characterisation Molecular and Cellular Endocrinology 47, 125 - 130. of cell lines from patients with lung cancer (predominantly de Keyzer, Y., Bertagna, X., Lenne, F., Girard, F., Luton, J. P. small cell carcinoma). British Journal of Cancer 52, 495 504. Sc Kahn, A. (1985). Altered pro-opiomelanocortin gene Buonassisi, V., Sato, G. St Cohen, A. T. (1962). Hormone pro­ expression in adrenocorticotropin producing non-pituitary ducing cultures of adrenal and pituitary origin. Proceedings of tumors. Journal of Clinical Investigation 76, 1892-1898. the National Academy of Sciences of the U .S.A. 48, 1184-1190. Krieger, D. T. (1983). The multiple faces of pro-opiomelano­ Carney, D. N., Gazdar, A. F., Bepler, G., Guccion, J. G., cortin, a prototype precursor molecule. Clinical Research 31, Marangos, P. J., Moody, T. W., Zweig, M. II. Sc Minna, J. D. 342-353. (1985). Establishment and identification of small cell lung Orth, D. N\, Nicholson. W. E„ Mitchell, W. M„ Island, D. P., cancer cell lines having classic and variant features. Cancer Shapiro, M„ Bvyny, R. L. (1973). ACTH and MSH Research 45, 2913-2923. production by a single cloned mouse pituitary tumor cell line. Chirgwin, J. M., Przybyla, A. E., McDonald, R. J. 8c Rutter, Endocrinology 92, 385-393. W. J. (1979). Isolation of biologically active ribonucleic acid Ratcliffe, J. G. (1985). ACTH and related peptides in lung from sources enriched in ribonuclease. Biochemistry 18, cancer. Recent Results in Cancer Research 99, 46-55. 5294-5299. Steenbergh, P. H., Hoppener, J. W. \I., Zandberg, J., Roos, Clark, A. J. L., Lavender, P. M., Besser, G. M. St Rees, L. H. B. A., Jansz, H. S. St Lips, C. J. M. (1984). Expression of the (1989). Pro-opiomelanocortin mRNA size heterogeneity in human proopiomelanocortin gene in medullary thyroid ACTII-dependent Cushing's syndrome. Journal of Molecular carcinoma. Journal of Clinical Endocrinology and Metabolism Endocrinology 2, 3-9. 58,904-908. Crosby, S. R., Stewart, M. F., Ratcliffe, J. G. & White, A. Stewart, M. F., Crosby, S. R., Gibson, S., Twentyman, P. R. St (1988). Direct measurement of the precursors of adrenocorti- White, A. (1989). Small cell lung cancer cell lines secrete cotropin in human plasma by two-site immunoradiometric predominantly ACTH precursor peptides not ACTH. assav. Journal of Clinical Endocrinology and Metabolism 67, British Journal of Cancer. (In Press.) 1271-1277. Tsukada, T., N'akai, Y., Jingami, H., Imura, H., Taii, S., DeBold, C. R., Schworer, M. E., Connor, T. B., Bird, R. E. & Xakanishi, S. St Numa, S. (1981). Identification of the mRNA Orth, D. N. (1983). Ectopic proopiomelanocortin: sequence of coding for the ACTH-p-LPH precursor in a human ectopic cDNA coding for y-melanocyte stimulating hormone and ACTH producing tumor. Biochemical and Biophysical P-endorphin. Science 220, 721-723. Research Communications 98, 535-540. Feinberg, A. P. & Vogelstein, B. (1983). A technique for West, D. C., Sattar, A. & Kumar, S. (1985). A simplified in-situ labelling DNA restriction fragments to high specific activity. solubilisation procedure for the determination of DNA and Analytical Biochemistry 132, 6-13. cell number in tissue cultured mammalian cells. Analytical Hakanson, R. St Sundler, F. (1971). Fluorometric determination Biochemistry 147, 289-295. of N-terminal tryptophan peptides after formaldehyde White, A., Smith, H., Hoadley, M., Dobson, S. H. Sc Ratcliffe, condensation. Biochemical Pharmacology 20, 3223-3225. J. G. (1987). Clinical evaluation of a two-site immunoradio­ Hale, A. C., Besser, G. M. St Rees, L. H. (1986). Characteriza­ metric assay for adrenocorticotrophin in unextracted human tion of pro-opiomelanocortin-derived peptides in pituitary and plasma using monoclonal antibodies. Clinical Endocrinology ectopic ACTH-secreting tumours. Journal of Endocrinology 26, 41-52. 108, 49-56.

Journal of Molecular Endocrinology (1989) 3, 65—70

/ ./ 0()21JJ72X/9()/70f)2-0.|Hf>3()2.l)()/i) Journal of Clinical Endocrinology and Metabolism Vol. 70. No. 2 Copyright <:, iyy() by The Endocrine Society 1‘rinti'd in U.S.A.

Defective Glucocorticoid Regulation of Proopiomelanocortin Gene Expression and Peptide Secretion in a Small Cell Lung Cancer Cell Line*

A. J. L. CLARK, M. F. STEWART, P. M. LAVENDER, W. FARRELL, S. R. CROSBY, L. H. REES, and A. W HITE Departments of Endocrinology and Chemical Endocrinology, St. Bartholomew’s Hospital Medical College (A.J.L.C., P.M.L., L.H.R.), London EC1A 7BE; and the Department of Clinical Biochemistry, University of Manchester, Hope Hospital (M.F.S., W.F., S.R.C., A. W.), Salford M6 8HD, Great Britain

ABSTRACT. A human small cell lung cancer cell line (COR the same conditions. Two other genes that are regulated by L103) that actively expresses the proopiomelanocortin (POMC) glucocorticoids in other cell types, the tyrosine amino transferase gene has been used as a model of extrapituitary ACTH-secreting gene and the glucocorticoid receptor gene, were expressed in tumors to investigate the phenomenon of resistence of ACTH COR L103 cells. However, neither gene appeared to be regulated production to glucocorticoids. After both short term (24 h) and by hydrocortisone in this small cell lung cancer cell line. Further longterm (10 days) exposure to hydrocortisone at concentrations studies demonstrated that glucocorticoid receptor binding could of 500 and 1000 nM, the accumulation of intracellular POMC be detected in the nucleus and cytoplasm, with a Kd of 5 x 10~9 mRNA, ACTH, and ACTH precursor peptides in the culture M. It is concluded that nonsuppression of POMC by glucocorti­ medium was not suppressed. These finding contrast with those coids is probably part of a more global defect of glucocorticoid in the pituitary corticotroph cell line AtT20, in which POMC signaling in these cells, but that this defect lies distal to steroid mRNA, ACTH, and ACTH precursors were suppressed under binding in the nucleus. (J Clin Endocrinol Metab 70: 485,1990)

A CHARACTERISTIC of extrapituitary ACTH-se- ACTH are secreted into the culture medium, and the creting tumors is the persistent synthesis and se­ small amount of ACTH detected can be accounted for cretion of proopiomelanocortin (POMC)-derived pep­ by the cross-reactivity of the precursors in the ACTH tides, despite the patients’ own pathologically elevated assay. The presence of ACTH precursors is a feature cortisol levels ( 1 ). This feature is the basis of one of the often found in vivo in these tumors ( 6, 7). most reliable clinical tests used in the diagnosis of this Using this model we have investigated the regulation condition, the dexamethasone suppression test ( 2 ). by glucocorticoids of POMC and two other genes that In an attempt to understand the basis of this phenom­ are regulated by glucocorticoids in other tissues, the enon of glucocorticoid resistence we have exploited a tyrosine amino transferase (TAT) gene and the gluco­ model of human ectopic ACTH production that we have corticoid receptor (GR) gene. We demonstrate that in recently characterized in some detail (3, 4). This model this human SCLC cell line glucocorticoids do not.inhibit is provided by a human small cell lung cancer (SCLC) POMC, and that glucocorticoid regulation of the other cell line which expresses the POMC gene as a 1350-base genes is absent. However, binding studies demonstrate mRNA, a longer species than that produced in the nor­ that the glucocorticoid receptor is present and binds mal pituitary and similar in size to that occasionally dexamethasone with normal kinetic parameters. found in ACTH-producing extrapituitary tumors (5). We have found that expression of this gene is unaffected by Materials and Methods cell density and growth. In addition, processing of the Cell culture precursor peptide is minimal, in that POMC and pro- COR L103 cells were derived from a patient with pathologi­ Received May 23,1989. cally confirmed SCLC and were generously given to us by Dr. Address all correspondence and requests for reprints to: Dr. A. J. L. P. Twentyman (Cambridge, United Kingdom). The cells were Clark, Building 10, Room 8C407, Endocrine and Reproduction Re­ grown, as previously described ( 8), using RPMI-1640 (Gibco, search Branch, National Institute of Child Health and Human Devel­ opment, National Institutes of Health, Bethesda, Maryland 20892. Paisley, Scotland) supplemented with 2.5% fetal calf serum, *This work was supported in part by grants from the Medical human transferrin (10 #xg/mL), bovine insulin (5 jug/mL), so­ Research Council and the Northwest Regional Health Authority. dium selenite (3 x 10"a m ), HEPES buffer (10 mM ), glutamine

485 486 CLARK ET AL. JC'E & M * 1990 Vol 70 • No 2

(4 mM), and sodium pyruvate (1 mM). Antibiotics were not POMC .standard was prepared from growth medium of a cul­ used. Incubation was carried out at 37 C in an atmosphere of tured human pituitary tumor by Sephadex G-75 chromatogra­ 5% C 02. phy under acid-dissociating conditions. Standards were pre­ AtT20 cells (CCL 89) were obtained from the American Type pared at concentrations of 0-2600 pmol/L. The assay sensitiv­ Culture Collection (Rockville, MD) and cultured as described ity is 2.6 pmol/L, and the within- and between-assay previously (9) in Ham’s F-10 medium (Flow laboratories Rock­ coefficients of variation are less than 10% between 20-2600 ville, MD), supplemented with horse serum (15%), fetal calf and 37-2600 pmol/L, respectively. The assay measures both serum (2.5%), glutamine (4 mM), and penicillin (100 IU/mL)/ POMC and pro-ACTH, but does not distinguish between them. streptomycin (100 IU/mL) in an atmosphere of 5% C0 2 at Other POMC-derived peptides, e.g. ACTH, 0-lipotropin, and 37 C. N-proopiocortin, are not recognized.

Experimental design RNA studies

All experiments were performed in quadruplicate. For each Cell pellets were homogenized from the frozen state in 4 m experiment, cells were grown to the stationary phase and then guanidine isothiocyanate and centrifuged through a cushion of passaged (1:4 for COR L103 cells; 1:8 for AtT20 cells) in fresh 5.7 M cesium chloride (12). RNA pellets were resuspended in medium to generate four bottles (designated Bi-B4) containing TE buffer (10 m M Tris-HCl, and 1 m M EDTA, pH 7.5), 25 mL cells, all at the same density and state of growth. extracted once with chroroform-isobutanol (4:1), and precipi­ Hydrocortisone was added to a final concentration of 500 nM tated with ethanol. After washing and drying, RNA was resus­ (Bj) or 1000 nM (B2). A further bottle (B3) was retained as a pended in TE and quantitated by measurement of absorption control for this 10-day experiment. Each dose of hydrocortisone at 260 nm. was repeated after 5 days in culture in order to maintain Five-microgram aliquots of each RNA sample were applied glucocorticoid levels adequately, as monitored by cortisol RIA. to nitrocellulose filters using a slot blot apparatus (Gibco BRL, These three bottles (Bi-B3) were harvested after 10 days of Paisley, Scotland), and blots were hybridized at moderately culture. The remaining bottle (B4) was maintained undisturbed high stringency [50% formamide at 42 C, final wash at 50 C in in culture for 9 days before being passaged 1:3 with fresh 0.2 x SSC (1 x SSC is 0.15 M sodium chloride-0.015 M sodium medium (giving rise to bottles B41, B42, and B 4 3. Hydrocorti­ citrate)] to the DNA probes described below. Autoradiography sone was added to give final concentrations in B4.x and B 4.2 of was carried out at —70 C with image-intensifying screens and 500 and 1000 nM, respectively. The third bottle (B4.3) was Kodak XAR-5 film (Eastman Kodak, Rochester, NJ), for in­ retained as a control. All of these bottles were harvested after tervals ranging from 4 h to 5 days in order to obtain a suitably 24 h of steroid treatment. contrasted image with each of the probes used. Autoradiographs Harvesting of cells at the end of the experiment involved were quantitated with a Parry DT1405 densitometer,, and all removal of a 1-mL sample of the cell suspension for protein results are expressed relative to the /3-actin mRNA level as a analysis, followed by pelleting the cells, which were then frozen control for the quality and quantity of RNA applied to the blot. on dry ice and stored at —70C for RNA preparation. The culture medium was frozen and retained for measurement of Probes ACTH, ACTH precursors, and cortisol. The bovine POMC cDNA probe was derived from the full- length insert from the plasmid pSNAC 20 (13). The human GR Immunoradiometric assay (IRMA) for ACTH. probe was a 1.9-kilobase (Kb)Aval fragment of the cloned The ACTH IRMA (10) employed two monoclonal antibodies: cDNA (14), and the TAT probe was the 1.7-kb insert of the MAb 1A12 [specific for ACTH-(10-18)] was radioiodinated, plasmid hcTAT2-16 (15). The /3-actin probe was derived from and MAb 2A3 [specific for ACTH-(25-39)] was coupled to the cloned chick |3-actin cDNA. All probes were labeled using Sephacryl S300 as solid phase. Human ACTH standards the oligo-primed labeling technique (16) and were purified by (NIBSC code 74/555) were prepared at concentrations between passage over Sephadex G-50. Specific activity was usually 109 1.1-1110 pmol/L. The assay sensitivity (2.5 x sd at zero cpm/jig and approximately 107 cpm were used per hybridiza­ ACTH) is 0.8 pmol/L, and the within- and between-assay tion. coefficients of variation are less than 10% at 4.9-1110 and 6- 1110 pmol/L, respectively. The assay measures ACTH-(l-39) Glucocorticoid binding assay and also recognizes POMC and pro-ACTH, with cross-reactiv­ ities of less than 1% and less than 10%, respectively. The assay Cells, grown to midlog phase, were washed once, resuspended does not detect fragments of ACTH, such as aMSH, ACTH- in fresh unsupplemented medium, and aliquoted among several (18-39), and ACTH-(l-24). 1.5-mL tubes to give approximately 5 X 105 cells/tube. To these were added 100 fmol [ 3H] dexamethasone (Amersham; SA, 100 Ci/mmol) and various concentrations of unlabeled dexameth­ IRMA for ACTH precursor peptides asone ranging from 10_u - 10 “5 M in a total volume of 300 m L. The development of the precursor assay was described in All measurements were performed in duplicate. Incubation was detail previously (11). MAb 1A12 [specific for ACTH-(10-18)] carried out for 60 min at 37 C in an atmosphere of 5% C02. was radioiodinated, and MAb lC ll (specific for 7 tMSH) was Cells were then pelleted at 4 C by centrifugation at 6,500 rpm coupled to Sephacryl S300 as solid phase. A partially purified for 2 min and washed twice with cold phosphate-buffered saline. NONSUPPRESSION OF POMC IN SCLC CELLS 487

Pellets were lysed by incubation for 20 min at 2 C in 500 /zL ACTH Precursors 1.5 m M magnesium chloride-10 m M Tris hydrochloride (pH 24h Incubation 7.4). Complete lysis was checked microscopically. Nuclei were separated by centrifugation at 12,000 x g for 5 min. Superna­ C O R L103 AtT 20 tant was carefully removed, and nuclear and cytoplasmic frac­ B tions were counted in 500 nL scintillant (Cocktail G, Sigma St. c Louis, MO), for 10 min each. Displacement curves and Scat- *3*-> o 10 chard plots were derived from these results. Specificity of the Q_ receptor was confirmed by competing bound [ 3H]dexametha- t— 4000 sone with 10-5 M progesterone and aldosterone. b E o 2000 Statistical analysis E a Changes in levels of peptides and mRNA were analyzed using a one-tailed Dunnett’s multiple range test (17). In the i 0 m case of COR L103 ACTH precursor peptide levels, a two-tailed test was also used, and the quoted significance levels are based on this. 10d Incubation

Results C O R L103 AtT 20 Peptide secretion D COR L103 cells secrete significant levels of ACTH 2 100 20,000 precursor peptides, as has been described (3, 4), with Q_ relatively small amounts of ACTH-(1-39). Basal levels 0) of these peptides reflect these findings. No suppression E of precursors or ACTH was seen after either 24-h or 10- o 50 10,000 E * * day incubation with 500 or 1000 n M hydrocortisone (Figs. a 1, A and C, and 2, A and C).‘ Indeed, the increase in ACTH precursors at 10 days is statistically significant. In contrast, AtT20 cells incubated for 10 days with hydrocortisone show marked suppression of peptide lev­ F ig. 1. Concentrations of ACTH precursors in culture medium of cells els, although levels after short term suppression are incubated with hydrocortisone. A, Twenty-four-hour treatment of COR L103 cells; B, 24-h treatment of AtT20 cells; C, 10-day treatment of unchanged (Figs. 1, B and D, and 2, B and D). It must COR L103 cells; D, 10-day treatment of AtT20 cells. Each bar repre­ be remembered that any suppression in these cells after sents the mean ± SEM peptide concentration expressed per mg cell this shorter time would probably be masked by the high protein, from four separate experiments. □ No hydrocortisone; □, 500 levels and relatively long half-life of ACTH and ACTH n M hydrocortisone; ■, 1000 n M hydrocortisone. *, P < 0.05; **, P < precursors in the medium. Interestingly, AtT20 cells 0.01; nonstarred bars show no significant change (P > 0.05). secrete large quantities of POMC precursors, a feature TAT and GR mRNA that has not been noted preciously when ACTH RIA has been used. These peptides are processed to a much The COR L103 cell line expressed the TAT gene at greater extent than in the SCLC cell line, since large relatively low levels. After hydrocortisone exposure for quantities of ACTH are also present. either 24 h or 10 days there was no significant stimulation Cortisol levels confirmed that the hydrocortisone of TAT mRNA levels (Fig. 4A). Figure 4B shows that treatment regime was successful in maintaining fairly the GR mRNA was essentially unchanged in COR L103 constant steroid concentrations at the designated level cells in response to hydrocortisone. in both cell lines (data not shown). Dexamethasone binding studies POMC RNA levels Displacable dexamethasone binding was measured in Figure 3A shows that POMC mRNA levels in COR COR L103 cells and AtT20 cells on several occasions L103 cells were not suppressed after 24 h or 10 days of with similar findings. Figure 5 illustrates the displace­ hydrocortisone treatment. In AtT20 cells there is a sug­ ment curves for nuclear binding from these two cell lines. gestion of a small suppression of POMC mRNA after 24 Scatchard analysis (not shown) gave similar Kd values h, and this becomes statistically significant at 10 days of 5 x 10 -9 M in each cell type, which is in good agreement (Fig. 3B). with values reported preciously (18). The estimated nu- 488 CLARK ET AL. ■K.'K & M • 1990 Vol 7(J • No 1 ACTH C O R L103

24h Incubation 24 Hrs 10 Days AtT 20 COR L103 B c 500 0u . 0.3 Q. < 100 b 1 0.15 250 o E ill a flUL i AtT20 B 24 Hrs 10 Days

10d Incubation 200 200 u COR L103 AtT 20 < i C D I < z o 1.0 1000 100 100 CL O cc * 5 E o □) a. E o 0.5 500 I I E a F ig. 3. POMC mRNA content of COR L103 cells (A) or AtT20 cells (B) incubated with hydrocortisone for 24 h or 10 days. Results are expressed as the percent change in the ratio of POMC mRNA densi- i tometric units to the 0-actin mRNA densitometric units. Each result F ig. 2. Concentrations of ACTH in culture medium of cells incubated is the mean ± SEM of four experiments. □, No hydrocortisone; □, 500 with hydrocortisone. A, Twenty-four-hour treatment of COR L103 n M hydrocortisone; ■, 1000 n M hydrocortisone. * ,P < 0.05; nonstarred cells; B, 24-h treatment of AtT20 cells; C, 10-day treatment of COR bars show no significant change. L103 cells; D, 10-day treatment of AtT20 cells. Each bar represents the mean ± s e m ACTH concentration, expressed per mg cell protein, from since a fully responsive tumor would be unlikely to four separate experiments. □, No hydrocortisone; □, 500 n M hydrocor­ secrete excessive quantities of ACTH and, therefore, tisone; ■, 1000 n M hydrocortisone. **, P < 0.01; nonstarred bars show would not produce glucocorticoid excess (Cushing’s syn­ no significant change. drome), a condition associated with a high morbidity and clear binding capacity was 640 fmol/mg protein for COR a significant mortality. L103 cells and 710 fmol/mg protein in AtT20 cells. We have previously characterized a human SCLC cell Analysis of the cytoplasmic fraction also produced a line (COR L103) with a view to its use as an experimental displacement curve, but there was a sizeable nonspecific model of ectopic ACTH secretion (3, 4). Since adequate bound fraction in both cell lines. Approximately 50% endocrinological data on the patient from which this cell line was derived are not available, it is not clear whether displacement of [ 3H]dexamethasone was obtained with either progesterone or aldosterone at a concentration of the ectopic ACTH syndrome was apparent. However, it seems very probable that the parent tumor did express 10"5 M, confirming the specificity of this assay. the POMC gene. Nevertheless, we are aware that extrap­ Discussion olating from a cell line to clinically manifest ectopic ACTH syndrome may have limitations, and our results Ectopic ACTH-producing tumors exhibit a number of should be interpreted in this light. characteristics that serve to distinguish their endocrine In this study we have demonstrated that this cell line function both biologically and clinically from the normal behaves as predicted, exhibiting no suppression of corticotroph and corticotroph tumors. These features POMC peptide secretion or of POMC mRNA levels in include nonresponsiveness to both CRH and glucocorti­ response to concentrations of hydrocortisone that would coids and a tendency to process the POMC peptide normally suppress the expression and secretion of POMC product incompletely. Of these, nonresponsiveness to (19). This resistence to glucocorticoid effect is empha­ glucocorticoids is of the greatest clinical significance, sized by the more normal POMC suppression found in NONSUPPRESSION OF POMC IN SCLC CELLS 489

A 24 Hrs 10 D a y s that in the denervated rat neurointermediate lobe, glu­ cocorticoids may elevate POMC mRNA levels approxi­ 200 200 mately 2- to 3-fold (22). The mechanisms underlying .£ such effects remain unclear. o cS I < < There is no perfect control gene that is known to be <33. Z 100 100 ■—. CC expressed in SCLC and is responsive to glucocorticoids. H E We have found that both the GR and the TAT genes are < H expressed in COR L103 cells, and both genes are recog­ nized as being regulated by glucocorticoids in other tis­ sues and cell lines studied (14, 23-25). Therefore, the finding that in COR L103 cells neither gene responds B 24 Hrs 10 Da y s significantly to hydrocortisone is suggestive evidence of 200 r 200 r a global defect of glucocorticoid signaling. 0 Measurement of dexamethasone binding reveals that < < this cell line contains a moderate amount of receptor 1 100 100 03. DC that is present in the nucleus and binds glucocorticoid CC E with kinetics similar to those of GRs in other tissues and o ceils (18). This of itself is of interest since it had previ­ 1 1 i ously been reported that SCLC solid tumors contained Fig. 4. Cellular content of TAT mRNA (A) or GR mRNA (B) in COR no cytosolic GR (26). Clearly, this may still be the case L103 cells treated with hydrocortisone for 24 h or 10 days.. Results are in the solid tumor, but if so, it is not the sole explanation expressed as the percent change in the ratio of each mRNA to the 0- for the nonsuppression of POMC. actin mRNA, and each bar is the mean ± SEM of four experiments. □, No hydrocortisone; □, 500 n M hydrocortisone; ■, 1000 n M hydrocorti­ If glucocorticoid binds its receptor normally in the sone. No changes are statistically significant. nucleus, then the possibilities for a breakdown in this signaling system are limited. It seems highly unlikely 400 O— O AtT 20 that there is a mutation in part of the GR not required • COR L103 for binding steroid, although such mutants can be gen­ erated artificially (27). Moreover, Northern blot analysis o; 300 of the GR mRNA shows that there is no obvious size abnormality in COR L103 cells such as might be caused by a partial gene deletion (data not shown). Thus, it would seem probable that the proposed lesion should lie 200 "O either in some unidentified cofactor required for normal c 3 GR/DNA interaction or in some modification to the o CO glucocorticoid response element (GRE) itself. Failure of steroid hormone responsiveness in the pres­ c 3 ence of apparently normal steroid binding is a not infre­ o u quently described phenomenon in cancer cells and has been observed with estrogen and progesterone receptors (28, 29). Darbre and King (30) have shown that intro­ 10-10 10-9 10-8 10-7 10-6 duction of new steroid-responsive DNA into the genome Dexamethasone (Molar) of an apparently steroid insensitive cell line resulted in FlG. 5. Displacement of [3H]dexamethasone with unlabeled dexameth­ normal steroid responses in the transfected sequence, asone in nuclear preparations from COR L103 cells (•) or AtT20 cells implying that the glucocorticoid signaling apparatus is (O). Counts bound are normalized per mg protein. fully functional. However, withdrawal of the steroid for the AtT20 cell line, which has been described previously a prolonged period of time in these cells resulted in the (20). loss of that responsiveness. These observations have led There is a suggestion of some increase in secretion of to the hypothesis that an irreversible change, possibly ACTH precursors and in levels of POMC mRNA. This DNA methylation, occurs in or around the GREs of the is of potential interest since there has been at least one transfected gene. Such an explanation could account for published report of a human ectopic ACTH-secreting the nonresponsiveness of the POMC gene in extrapitui­ tumor that increased ACTH levels in response to gluco­ tary tumors associated with POMC peptide production corticoids ( 2 1 ). In addition, it has recently been shown and should be readily testable. However, if this is the 490 CLARK ET AL. .JCE & M • 199U Vol 70 • No 2 case, exactly why the GREs of multiple genes should ture. 1979;278:423-7. 14. Hollenberg SM, Weinberger C, Ong ES, et al. Primary structure become methylated remains an open question. and expression of a functional human glucocorticoid receptor cDNA. Nature. 1985;318:635-41. Acknowledgments 15. Natt E, Kao FT, Rettenmeier R, Scherer G. Assignment of the human tyrosine aminotransferase gene to chromosome 16. Hum We are most grateful to Dr. Peter Twentyman for permission to use Genet. 1986;72:225-8. the COR L103 cell line. We would also like to thank Prof. Gavin Vinson 16. Feinberg AP, Vogelstein B. A technique for labelling DNA restric­ for help and advice in the measurement of glucocorticoid binding, and tion endonuclease fragments to high specific activity. Anal we thank Dr. Gerd Scherer, Dr. Stan Hollenberg and Prof. S. Numa Biochem. 1983;132:6-13. 17. Dunnett CW. A multiple comparison procedure for comparing for provision of the plasmids from which our DNA probes were derived. several treatments with a control. Am Stat Assoc. 1955;50:1096- We also thank Dr. Guk Ooi for statistical advice. 121. 18. Rousseau GG, Baxter JD, Tomkins GM. Glucocorticoid receptors: References relations between steroid binding and biological effects. J Mol Biol. 1972;67:99-115. 1. Liddle GW, Nicholson WE, Island DP, Orth DN, Abe K, Lowder 19. Keller-Wood ME, Dallman MF. Corticosteroid inhibition of ACTH SC. Clinical and laboratory studies of the ectopic humoral syn­ secretion. Endocr Rev. 1984;5:1-24. dromes. Recent Prog Horm Res. 1969;25:283-305. 20. Phillips M, Tashjian AH. Characterization of an early inhibitory 2. Liddle GW. Tests of pituitary-adrenal suppressability in the diag­ effect of glucocorticoids on stimulated adrenocorticotrophin and nosis of Cushing’s syndrome. J Clin Endocrinol Metab. endorphin release from a clonal strain of mouse pituitary cells. 1960;20:1539-60. Endocrinology. 1982;110:892-900. 3. Stewart MF, Crosby SR, Gibson S, Twentyman PR, White A. 21. Roth KA, Newell DC, Dorin RI, Eberwine JH, Hoffman AR. Small cell lung cancer cell lines secrete high levels of ACTH Aberrant production and regulation of pro-opiomelanocortin de­ precursor peptides. Br J Cancer. 1989;60:20-24. rived peptides in ectopic Cushing’s syndrome. Horm Metab Res. 4. White A, Stewart MF, Farrell WE, et al. Pro-opiomelanocortin 1988;20:225-9. gene expression and peptide secretion in human small cell lung 22. Seger MA, van Eekelen AM, Kiss JZ, Burbach PH, de Kloet ER. cancer cell lines. J Mol Endocrinol. 1989;3:65-70. Stimulation of pro-opiomelanocortin gene expression by glucocor­ 5. Clark AJL, Lavender PM, Besser GM, Rees LH. Pro-opiomelano­ ticoids in the denervated rat intermediate pituitary gland. Neu­ cortin mRNA size heterogeneity in ACTH secreting tumours. J roendocrinology. 1988;47:350-7. Mol Endocrinol. 1989;2:3-9. 23. Rosewicz S, MacDonald AR, Maddux BA, Goldfine ID, Miesfeld 6. Ratter SJ, Lowry PJ, Besser GM, Rees LH. Chromatographic RL, Logsdon CD. Mechanism of glucocorticoid receptor down- characterization of ACTH in human plasma. J Endocrinol. regulation by glucocorticoids. J Biol Chem. 1988;263:2581-4. 1980;85:359-69. 24. Berkowitz GD, Carter KM, Migeon CJ, Brown TR. Down-regula- 7. Hale AC, Besser GM, Rees LH. Characterization of pro-opiome- tion of the glucocorticoid receptor by dexamethasone in cultured lanocortin-derived peptides in pituitary and ectopic adrenocorti- human skin fibroblasts: implications for the regulation of aroma- cotrophin-secreting tumours. J Endocrinol. 1986;108:49-56. tase activity. J Clin Endocrinol Metab. 1988;66:1029-36. 8. Baillie-Johnson H, Twentyman PR, Fox N, et al. Establishment 25. Diesterhaft M, Noguchi T, Granner DK. Regulation of rat liver and characterization of cell lines from patients with lung cancer. tyrosine aminotransferase mRNA by hydrocortisone and byN *02- Br J Cancer. 1985;52:495-504. dibutyrylandenosine 3,5-monophosphate. Eur J Biochem. 1980; 9. Orth DN, Nicholson WE, Mitchell WM, Island DP, Shapiro M, 108:357-65. Byyny RL. ACTH and MSH production by a single cloned mouse 26. Chaudhuri PK, Thomas PA, Walker MJ, Briele HA, Das Gupta pituitary tumor cell line. Endocrinology. 1973;92:385-93. TK, Beattie CW. Steroid receptors in human lung cancer cytosols. 10. White A, Smith H, Hoadley M, Dobson SH, Ratcliffe JG. Clinical Cancer Lett. 1982;16:327-32. evaluation of a two-site immunoradiometric assay for adrenocor­ 27. Miesfeld R, Godowski PJ, Maler BA, Yamamoto KR. Glucocorti­ ticotrophin in unextracted human plasma using monoclonal anti­ coid receptor mutants that define a small region sufficient for bodies. Clin Endocrinol (Oxf). 1987;26:41-51. enhancer activation. Science. 1987;236:423-7. 11. Crosby SR, Stewart MF, Ratcliffe JG, White A. Direct measure­ 28. Nawata H, Chong MT, Bronzert D, Lippman ME. Estrodiol- ment of the precursors of adrenocorticotrophin in human plasma independent growth of a subline of MCF-7 human breast cancer by two-site immunoradiometric assay. J Clin Endocrinol Metab. cells in culture. J Biol Chem. 1981;256:6895-902. 1988;67:1272-7. 29. Thorpe SM, Rose C. Oestrogen and progesterone receptor deter­ 12. Chirgwin JM, Przbyla AE, MacDonald RJ, Rutter WJ. Isolation minations in breast cancer technology and biology. Cancer Surv. of biologically active ribonucleic acid from sources enriched in 1986;5:505-25. \ ribonuclease. Biochemistry. 1979;18:5294-9. 30. Darbre PD, King RJB. Progression to steroid insensitivity can 13. Nakanishi S, Inoue A, Kita T, et al. Nucleotide sequence of cloned occur irrespective of the presence of functional steroid receptors. cDNA for bovine corticotrophin-beta-lipotrophin precursor. Na­ Cell. 1987;51:521-8. Small cell lung cancer cell lines secrete predominantly ACTH precursor peptides not ACTH

M.F. Stewart, S.R. Crosby, S. Gibson, P.R. Twentyman1 & A. White

University of Manchester, Department of Clinical Biochemistry, Hope Hospital, Salford M6 BHD, UK; and XMRC Clinical Oncology and Radiotherapeulics Unit, Hills Road, Cambridge CB2 2QH, UK.

Summary A panel of 18 well characterised human small cell lung cancer (SCLC) cell lines was assessed for the production of adrenocorticotrophin (ACTH) and its precursor peptides, pro-opiomelanocortin (POMC) and pro-ACTH. These precursor peptides were measured directly using a novel two-site immunoradiometric assay (IRMA) based on monoclonal antibodies, in conjunction with a similar IRMA for ACTH 1-39. Significant concentrations of ACTH precursors were secreted by 10 of the 18 cell lines (56%). The low levels of ACTH immunoreactivity detected in seven cell lines could be accounted for by the known cross-reactivity of precursors in the ACTH IRMA. This suggests there is little, if any, processing of ACTH precursors to ACTH. Cell pellet extracts contained undetectable or low levels of ACTH precursors and ACTH, indicating that these peptides are not stored intracellularly. During the growth of the SCLC ceils in vitro ACTH precursors accumulated progressively in the culture medium. Thus the combination of a direct assay for the ACTH precursors and the panel of SCLC cell lines provides a valuable in vitro model for the expression of POMC in human tumours.

Small cell lung cancer (SCLC) is a hormonally active a Precursor IRMA neoplasm associated with the secretion of a wide range of peptide hormones. Adrenocorticotrophin (ACTH) is a POMC notable example, giving rise to gross metabolic derangement with hypercortisolism and hypokalaemic alkalosis in the ACTH ectopic ACTH syndrome. While this is a relatively unusual clinical manifestation of SCLC, occurring in 2-3% of cases, Pro-ACTH- studies investigating the potential role of ACTH as a tumour marker have indicated that 20-30% of patients have elevated ACTH plasma levels of immunoreactive (ir)-ACTH, without overt clinical signs of glucocorticoid excess (Ratcliffe et al., 1982). Further, when lung tumours are extracted and assayed for ACTH, some 20% of SCLC tumours yield significant Y Y 1C11 1A12 amounts of peptide (Yamaguchi et al., 1985). Thus, the association of ACTH production with SCLC is sufficiently strong to address the question of a functional role in the b ACTH IRMA development or progression of this tumour. ACTH is synthesised as a high molecular weight (HMW) ACTH precursor, pro-opiomelanocortin (POMC, approximate molecular weight 31 kD), which is cleaved to an intermediate peptide, pro-ACTH (molecular weight approximately 22 kD) YY (Figure 1). It has been suggested that these precursor 1A12 2A3 peptides do not circulate in normal subjects but may Figure 1 Binding sites of the MAbs used in the IRMAs for circulate in pathological conditions, and particularly in the ACTH-related peptides. In the precursor IRMA (a), the l25I- ectopic ACTH syndrome (Hale et al., 1986). MAb-lA12 recognises ACTH 10—18 and the solid phase-linked Characterisation of the molecular species has hitherto only MAb-lCll recognises the gamma-MSH sequence. In the ACTH been possible by chromatographic separation followed by IRMA (b), the l25I-MAb-lA12 recognises ACTH 10-18 and the radioimmunoassay (RIA) of the fractions using antisera to solid phase-linked MAb-2A3 recognises ACTH 25-39. component peptides from the precursor such as ACTH or gamma-melanocyte stimulating hormone (gamma-MSH). We examined the influence of culture conditions and the have developed a new approach to quantitating ACTH and relationship between cell growth and peptide production. the precursor peptides directly in plasma samples. This involved the production of a range of monoclonal antibodies Materials and methods (MAbs) to ACTH and the development of a sensitive two- site immunoradiometric assay (IRMA) for ACTH (White et Small cell lung cancer cell lines al., 1987). Subsequently an IRMA for the two precursor peptides, POMC and pro-ACTH has been developed based Ten ‘COR’ cell lines (COR L24, COR L27, COR L31, on MAbs to ACTH and gamma-MSH (Crosby et al., 1988) CORL32, CORL42, COR L47, COR L51, COR L 88, (Figure 1). COR L99 and COR LI03) were established and SCLC cell lines are now widely used for the study of characterised as previously described (Baillie-Johnson et al., tumour biology and provide an appropriate in vitro model 1985). Four NCI SCLC cell lines (NCI H82, NCI HI28, since in vivo studies of hormone secretion are limited by the NCI H209 and NCI N417) were made available for study rapid clinical course of the disease. We have used these two courtesy of Dr F. Cuttitta (Carney et al., 1985). HCI2 and novel assays to establish the prevalence of secretion of HX149 were kindly donated by Dr G. Duchesne, Ludwig ACTH and its precursors in a large panel of cell lines which Institute for Cancer Research, Sutton, UK (Duchesne et al., were established in several different centres. We have 1987). GLC-1 and GLC-1-M13 were a gift from Dr M. Brouwer, Netherlands Cancer Institute, Amsterdam, the Netherlands, and Dr L. de Leij, University of Groningen, Correspondence: A. White. the Netherlands (de Leij et al., 1985). All the cell lines were derived from patienls with pathologically confirmed SCLC conccntralions between I and i, 110 pmol i 1 (4.5— with the exception of COR L32 which was obtained from a 5.000 n g l"'). The assay sensitivity (2.5 x standard deviation patient with a poorly differentiated squamous carcinoma of (s.d.) at zero ACTH) is 0.8 pm oll -1 (3.5ngl~') and the the lung. The cell line, however, exhibited the characteristics within and between batch coefficients of variation (c.v.) are of SCLC (Baillie-Johnson et al., 1985). The majority of the <10% at 5-1,1 lOpmol I -1 (22-5,000 ngl-1) and 6- cell lines grow as floating aggregates of cells in suspension. 1,110 pmol l -1 (27-5,000 n g l-1) respectively. The assay measures ACTH 1-39 and there is no interference from Culture media fragments of ACTH such as a-MSH, ACTH 18-39 and ACTH 1-24. Using this combination of ACTH MAbs, the Cell lines were routinely cultured in a growth medium, assay retains some cross-reactivity with the ACTH RS(10), consisting of RPMI 1640 (Gibco, Paisley, Scotland) precursors (POMC <1% and pro-ACTH <10%). supplemented with 10% fetal calf serum (FCS, Gibco), 4mM glutamine (Flow Laboratories, Irvine, Ayrshire, UK), 1 m M Immunoradiometric assay for ACTH precursors sodium pyruvate (Flow) and 10 m M iV-2-hydroxycthyl- pipcrazine-iV'-2-ethancsulphonic acid (HEPES) buffer The development of the precursor assay is described in detail (Flow). Antibiotics were not used. To determine whether elsewhere (Crosby et al., 1988). MAb 1A12 (specific for different culture media influenced ACTH secretion, RS(10) ACTH 10-18) was radio-iodinatcd and MAb 1C11 (specific was compared with two hormone supplemented media, in for gammaj MSH) was coupled to Sephacryl S300 as solid which the concentration of FCS was reduced to 2.5%. phase. A partially purified POMC standard was prepared RHS(2.5) contained HITES supplements (Carney et al., from growth medium of a cultured human pituitary tumour 1981), i.e. 10- 8 m hydrocortisone (Sigma Chemical Co., by Sephadex G-75 chromatography under acid dissociating Poole, Dorset, UK, H4001), 5jigm l -1 bovine insulin conditions. The POMC fraction was initially assigned an (Sigma, 15500), 10/zgml -1 human transferrin arbitrary potency of 10,000 precursor units per litre (Sigma, T2252), 10- 8 M oestradiol (Sigma, E8875) and corresponding to 26 nmol 1- 1 as determined by the fluorometric 3 x lO -8 M sodium selenite (Sigma, S1382). RTISS(2.5) assay for //-terminal tryptophan (Hakanson & Sundler, contained transferrin, insulin and selenite but steroid 1971). Standards were prepared at concentrations of 2.6- hormones were omitted. 2,600pm oll-1. The assay sensitivity (2.5 x s.d. at zero POMC) is 2.6pmoll -1 and the within and between assay Cell culture c.v.s are <10% between 20-2,600 pmol I -1 and 37- 2,600pmoll-1 respectively. Using the POMC standard, pro- All incubations were carried out at 37°C in a 5% C 0 2 ACTH cross-reacts 100% in this assay, thus it fully atmosphere. Cells were routinely passaged by diluting the quantitates both ACTH precursor peptides but does not cultures 2-fold. For growth experiments, cells were pipetted distinguish between them. Other POMC derived peptides e.g. to reduce the size of the aggregates to 5-10 cells and ACTH, /Mipotrophin and //-proopiocortin do not cross- 100 passaged by diluting in % fresh medium to give a seeding react at levels up to 1,000 pmol l-1. density of approximately 0.5 x 10s cells ml-1 . Before each experiment, cells were cultured for at least three passages in Sephadex G-75 chromatography the appropriate growth medium. Cultures were judged to be ‘confluent’ when crowded with cell aggregates causing the Growth medium (2 ml) from the SCLC cells was acidified medium to change to an acid pH. This corresponded with an with formic acid (1% final concentration). The acidified approximate cell density of 1 x 106 cells ml-1. medium was chromatographed on a Sephadex G-75 superfine column (1.5 x 90 cm; Pharmacia, Uppsala, Sweden) Cell counts and eluted with 1% formic acid containing Polypep Single cell suspensions were prepared by trituration and (lm gm l-1) (Sigma P5115) (Ratter et al., 1980). Fractions viable counts obtained using trypan blue exclusion. (4 ml) were immediately neutralised with 5 m NaOH However, for many of the cell lines this provides only an (approximately 85/d) and 0.5m sodium phosphate buffer approximate guide, since it is difficult to disaggregate the cell (400 pi) pH 7.5 and flash frozen before assay. The column clumps without a significant effect on their viability. was calibrated with glyceraldehyde-3-phosphate dehydrogenase (36 kD) 12SI-prolactin (25 kD), alpha- Estimation o f cellular DNA lactalbumin (14.2 kD) and 12SI-ACTH 1-39 (4.5 kD). Gel chromatography markers for POMC, pro-ACTH and ACTH DNA was assayed according to a fluorometric method (West 1-39 were obtained from the culture medium of human et al., 1985), the fluorochrome dye being Hoechst 33258 pituitary adenoma cells (Crosby et al., 1988). (Uniscience, London, UK). The standard was salmon testes DNA type III (Sigma D1626). Results Extraction o f cell pellets for ACTH Cell pellets were immediately frozen on dry ice and stored at Detection o f ACTH and ACTH precursors —20°C until extraction. The pellets were thawed, weighed Using a two-site IRMA for ACTH based on monoclonal and 1ml of 0.01 m HCI added. They were sonicated for antibodies (as shown in Figure 1) low levels of ACTH 2 min, centrifuged at 3,000£ for 10 min at 4°C, and immunoreactivity (range 1.4-16.7 pmol l-1) were initially supernatants were frozen and stored at — 70°C before detected in culture medium from the SCLC cell lines ACTH assay. These procedures were optimised for COR L24, COR L27, COR L31, COR L42 and COR L103. preservation of ACTH by extraction of exogenous ACTH Markedly higher levels of the precursors of ACTH were added to non-secreting cell lines, and by extraction of detected in these cells lines (range 62-3,640 pmol l-1) with endogenous ACTH from rat pituitary cells. the precursor assay which directly quantitates POMC and pro-ACTH without detecting ACTH 1-39. Since POMC and Immunoradiometric assay for ACTH pro-ACTH cross-react in the ACTH IRMA (<1% and The ACTH IRMA was developed and optimised as <10% respectively), secretion of these high levels of ACTH previously described (White et al., 1987) and employs two precursors would account for the ACTH immunoreactivity. MAbs: MAb 1A12 (specific for ACTH 10-18) was radio­ To determine the optimal stage of growth to detect ACTH iodinated and MAb 2A3 (specific for ACTH 25-39) was precursor peptides, levels were measured throughout the coupled to Sephacryl S300 as solid phase. Human ACTH growth of these five SCLC cell lines over a 14-day period as standards (NIBSC Code 74/555) were prepared at shown in Figure 2. ACTH precursor levels increased 22 M.F. STEWART et al.

Table I Comparison of peptide levels obtained in three different 120 -i growth media ACTII precursors pmol mg 1 DNA < 100- Cell tine RS(U)) RHS(2.5) RTISSU__ 5) COR L24 365 62 8 0 - COR L27 ' 77 66 43 COR L3I 246 31 64 COR L42 213 175 245 6 0 - COR LI03 119 67 34 o. M ean(±s.d.) 204 (113) 80 (55) 124 (106) 4 0 -

20- o. Table II Levels of ACTH and precursor peptides in supernatant medium ACTH 0 4 7 11 14 Precursors ACTH No. o f (pm oll~ l) (pmoll~') Molar Days in culture Cell line observations mean{±s.d.) mean(±s.d.)ratio Figure 2 Secretion of ACTH precursors throughout cell growth COR L24 5 426 (172) 2.6 (1.4) 164:1 in five SCLC cell lines. Growth medium = RTISS( 2.5). The cell COR L27 7 596 (290) 3.4 (1.4) 175:1 lines are COR L24 (X X), COR L27 ( • ------•), COR L31 COR L31 3 1200 (630) 12.4 (1) 97:1 (O O), COR L42 (A A) and COR L103 (A A). COR L42 7 1017 (980) 5.2 (2.5) 196:1 COR L88 3 135 (60) <0.8 >135:1 COR L99 1 44 <0.8 >44:1 progressively in the medium as the cells proliferated. This COR L103 27 1405 (770) 9.0 (2.8) 156:1 was accompanied by a parallel increase in the peptide levels NCI H I28 3 37 (10) 1.5 (0.2) 25:1 measured in the ACTH IRMA (data not shown). In four of GLC-1-M13 3 198 (190) 1.4 (0.4) 141:1 the five cell lines ACTH precursor levels began to decline by HC12 3 97 (20) <0.8 >97:1 14 days. The increase in precursors reflected accumulation of Mean = 516 Mean = 5.0 peptides in the medium since endogenously secreted ACTH ACTH precursors <2.6 pmol I-1 and ACTH <0.8 pmol 1 _l were precursors were relatively stable in culture medium. This was observed in the following cell lines on at least three occasions: assessed by separating peptide-containing medium from cells COR L32, COR L47, COR L51, NCI H82, NCI H209, NCI N417, and re-incubating at 37°C. ACTH precursor concentrations GLC-1, HX149. The cell lines were sampled at a peak viable density were 62% and 74% of the initial values in RS10 and of 1 x 106 cells ml-1 . RTISS 2.5 respectively after 72 h incubation. The five COR cell lines were cultured in three growth media, RS(10), RHS(2.5) and RTISS(2.5) (Table I). Mean Table in Levels of ACTH and precursor peptides in cell pellet ACTH precursor levels in confluent cultures were extracts 204 pmol mg -1 DNA, 80 pmol mg -1 DNA and Precursors ACTH 124 pmol mg -1 DNA respectively. Although RHS(2.5) Cell line (pmol mg~1 DNA) (pmolmg~ \ DNA) (containing 10 “ 8 M hydrocortisone) gave the lowest mean COR L24 5.40 n.d. level of secretion, this did not hold for all individual cell COR L27 3.30 0.06 lines. RTISS(2.5) was chosen for subsequent experiments as COR L31 0.40 n.d. this had the advantage of reduced serum concentrations. COR L42 4.40 0.018 Further studies are in progress to establish whether the COR L103 0.10 n.d. hydrocortisone in RHS(2.5) was having a specific effect on GLC-1-M13 0.40 0.003 ACTH precursor production. n.d., not detected. Cell pellet extracts in the 12 cell lines not listed were negative for both ACTH and precursors. Prevalence of ACTH precursor and ACTH secretion The prevalence of secretion of both ACTH precursors and but a minor peak was detected in the 22 kD region, ACTH was formally assessed in all the 18 cell lines consistent with the known cross-reactivities of POMC and (Table II). Cells were seeded at approximately 0.5 x 10s pro-ACTH in the ACTH IRMA. This chromatographic cells m l-1 and allowed to proliferate to high density profile is representative in that immunoreactivity (approximately 1 x 10® cells ml-1) without medium change corresponding to authentic ACTH 1-39 could not be as this was most likely to yield maximal peptide levels identified in other SCLC cells lines studied (data not shown). (Figure 2). Ten of the cell lines (56%) secreted ACTH precursor peptides giving mean levels of 516 pmol I-1. Low levels of ACTH immunoreactivity (mean 5pm oll_I) were Discussion detected in medium from seven (39%) of the cell lines using the ACTH IRMA. In cell pellet extracts, ACTH precursors Secretion of ACTH by SCLC cell lines in culture has been and ACTH were measurable at only low levels from six reported previously (Ellison et al., 1976; Sorenson et al., (33%) and three (17%) of the cell lines respectively 1981; Luster et al., 1985a). However, characterisation of (Table III). ACTH precursors and estimates of their prevalence in cell lines as well as in plasma and tumour extracts, has been Chromatography o f ACTH-related peptides limited by technical problems. Previously the precursor Sephadex G-75 chromatography of medium from confluent forms could only be detected by chromatographic separation cultures of COR L103 (Figure 3) showed that the precursor and analysis of fractions by radioimmunoassay. This is a IRMA detected two significant peaks of immunoreactivity, relatively insensitive technique and quantitation is critically corresponding to 31 kD POMC and 22 kD pro-ACTH. dependent on how well the antiserum recognises the HMW Measurement of fractions with the ACTH IRMA confirmed forms, which cannot usually be determined because of the that no significant ACTH could be detected. The peak in the lack of availability of POMC and pro-ACTH standards. region of 31 kD was not recognised with the ACTH assay Thus while HMW precursor forms of ACTH have been ACTH PRECURSORS IN SMALL CELL LUNG CANCER 23

ACTH Precursor IRMA became exhausted and possibly with increased release of 200— 1 Vo 34k 24k ACTH proteolytic enzymes. We have shown that in this panel of SCLC cell lines, 180 ACTH precursors are secreted but ACTH 1-39 cannot be 160 — detected to any great extent. ACTH precursor peptides were synthesised and secreted at significant levels by 10 of the 18 _ 140 — cell lines (56%). The low levels of ACTH detected in the £ 120- seven SCLC cell lines were measured in the presence of high Q. levels of ACTH precursors and reflect the cross-reactivity of « 100 — precursors in the ACTH IRMA rather than the presence of ACTH 1-39 itself. Thus the two-site IRMA for the ACTH precursors provides a more sensitive approach for the detection of POMC expression in SCLC cell lines and 4 0 - because it is a direct quantitative method has demonstrated that the overwhelming majority of the ACTH secreted is in 20- HMW precursor forms. Classical Sephadex G-75 chromalo- graphy with measurement of fractions in both the ACTH IRMA and the precursor IRMA confirmed that ACTH 0 5 10 15 20 25 30 35 40 45 50 55 50 precursors predominated in culture medium and that 10. ACTH IRMA COR LI03 cells secreted almost equal amounts of both Vo 34k 24k ACTH POMC and pro-ACTH (Figure 3). No significant levels of 9 I ACTH were detected and while the presence of very small 8 I I i amounts of ACTH 1-39 cannot be ruled out from this data alone, it is clear that processing of POMC beyond pro- 7 ■ ACTH occurs to a negligible extent, if at all, in these tumour o 6 cells. E CL The frequency of expression of POMC together with the 5 — high levels of peptides secreted, invite speculation on the HX O 4 — significance of ACTH-related peptides in SCLC. Although < no definitive functional role has been ascribed either to 3 POMC or pro-ACTH, it has been suggested that pro-ACTH 2 may be bioactive and could give rise to the hypokalaemic alkalosis, characteristic of the ectopic ACTH syndrome 1 A. (Hale et al., 1984). There are also limited and contradictory reports of the ability of ACTH 1-39 to stimulate the in vitro i i I i i i i i i r \ i growth of SCLC cells (Luster et al., 19856, Bepler et al., 0 5 10 15 20 25 30 35 40 45 50 55 60 1987). It is thought that A-terminal peptides derived from Fraction number POMC may have a role in stimulating adrenal growth Figure 3 Sephadex G-75 chromatography of COR L103 culture (Estivariz et al., 1982), but to date there is no convincing medium. The 34 K, 24 K and ACTH markers correspond to the evidence that ACTH-related peptides act as autocrine or elution positions of POMC, pro-ACTH and ACTH 1-39 from paracrine growth factors in SCLC. It is of note that the human pituitary adenoma cell culture medium. human POMC gene maps to chromosome 2, in close proximity to the N -myc oncogene (Hozier et al., 1987) raising the possibility that POMC expression in SCLC might demonstrated in SCLC, both in vivo (Hale et al., 1986) and occur because abnormal deregulation of DNA in malignant in vitro (Bertagna et al., 1978), the quantitaton is inexact. cells produces activation not only of oncogenes but also of Our approach to the measurement of ACTH precursors closely linked marker genes. and ACTH using two-site immunoradiometric assays based In summary, we have applied a direct approach to the on monoclonal antibodies offers many advantages over quantitation of ACTH precursors and established that these earlier methodology. Both assays are simple, robust and peptides are secreted at significant levels by 56% of SCLC suitable for large numbers of samples, requiring no cell lines studied. POMC is processed to pro-ACTH but very extraction procedure. Using these novel assays to detect little, if any, ACTH 1-39 is synthesised or secreted by these ACTH and precursors it was noted that peptide levels were tumour cells. influenced by culture conditions and an attempt was made to estimate prevalence under optimal conditions. The rise in ACTH precursor peptides in medium which occurred as the cells proliferated reflected accumulation of a stable peptide The support of the North West Regional Health Authority for A.W. and the decline in precursor levels observed at 14 days and S.R.C. and Boots Celltech Diagnostics for S.G. is gratefully coincided with increasing cell death as the nutrient medium acknowledged.

References

BAILLIE-JOHNSON, H., T W E N T Y M A N , P.R., FOX, N. and 6 others CARNEY, D.N., BUNN, P.A., j r . , GAZDAR, A.F., PAGAN, J.A. & (1985). Establishment and characterisation of cell lines from MINNA, J.D. (1981). Selective growth in serum-free hormone- patients with lung cancer (predominantly small cell carcinoma). supplemented medium of tumor cells obtained by biopsy from Br. J. Cancer, 52, 495. patients with small cell carcinoma of the lung. Proc. Natl Acad. BEPLER, G., CARNEY, D.N., GA Z D A R A.F. & MINNA, J.D. (1987). In Sci. USA. 78, 3185. vitro growth inhibition of human small cell lung cancer by CARNEY, D.N., GAZDAR, A.F., BEPLER, G. and 5 others (1985). physalaemin. Cancer Res., 47, 2371. Establishment and identification of small cell lung cancer cell BERTAGNA, X.Y., NICHOLSON. W.E., SORENSON. G.D., lines having classic and variant features. Cancer Res., 45, 2913. PETTENGILL, O.S., MOUNT, C.D. & ORTH, D.N. (1978). Cortico- CROSBY, S.R., STEWART, M.F., RATCLIFFE. J.G. & WHITE, A. trophin lipotrophin and ^-endorphin production by a human (1988). Direct measurement of the precursors of adrenocorti- non-pituitary tumor in culture: evidence for a common cotropin in human plasma by two-site immunoradiometric precursor. Proc. Natl Acad. Sci. USA, 75, 5160. assay. J. Clin. Endocrinol. Metab.. 67, 1272. 24 M.F. STEWART et al.

DE LEIJ. L.. POSTMUS. P.O.. BUYS. C.H.C.M. and 7 others (1985). LUSTER. W.. GROPP. KERN. H.F.. WAIIL. R.. KOHER, If.D. & Characterisation of three new variant type cell lines derived from HAVEMANN. K. (1985 />). Peptide hormone production in lung small cell carcinoma of the lung. Cancer Res., 45, 6024. cancer cell lines of different histological types. Rec. Results DUCHESNE. G.M., EADY. J.J., PEACOCK, J.H. & PERA, M.F. (1987). Cuncer Res., 99, 117. A panel of human lung carcinoma cell lines: establishment, RATCLIFFE. J.G.. PODMORE, J.. STACK, B.H.R., SPILG. W.G.S. Sl properties and common characteristics. Br. J. Cancer, 56, 287. GROPP, C. (1982). Circulating ACTH and related peptides in ELLISON, M.L., HILLYARD, C.J., BLOOMFIELD, G.A., REES, L.H., lung cancer. Br. J. Cancer. 45, 230. COOMBES, R.C. & NEVILLE, A.M. (1976). Ectopic hormone RATTER, S.J., LOWRY, P.J., BESSER, G.M. &. REES, L.H. (1980). production by bronchial carcinomas in culture. Clin. Endocrinol., Chromatographic characterisation of adrenocorticotrophin in 5, 397S. human plasma. J. Endocrinol., 85, 359. ESTIVARIZ, J.E., ITURRIZA, F„ McLEAN, C., HOPE. J. & LOWRY. P.J. SORENSON, G.D.. PETTENGILL, O S., BRINCK-JOHNSEN, T„ CATE. (1982). Stimulation of adrenal mitogenesis by JV-terminal pro- C.C. & MAURER. L.H. (1981). Hormone production by cultures opiocortin peptides. Nature, 297, 419. of small cell carcinoma of the lung. Cancer, 47, 1289. HAKANSON. R. & SUNDLER, F. (1971). Fluorometric determination WEST. D.C., SATTAR. A. & K U M A R . S. (1985). A simplified in situ of iV-terminal tryptophan-peptides after formaldehyde solubilisation procedure for the determination of DNA and cell condensation. Biochem. Pharmacol., 20, 3223. number in tissue cultured mammalian cells. Anal. Biochem., 147, HALE. A.C., RATTER, S.J., BESSER. G.M. & REES, L.H. (1984). 289. Characterisation and bioactivity of vV-pro-opiomelanocortin WHITE, A.. SMITH. H., HOADLEY, M.. DOBSON, S.H. & RATCLIFFE, fragments from tumours associated with Cushing’s syndrome. J.G. (1987). Clinical evaluation of a two-site immunoradiometric Proceedings of the 7th International Congress of Endocrinology assay for adrenocorticotrophin in unextracted human plasma (abstract 561), Quebec, July 1984. using monoclonal antibodies. Clin. Endocrinol., 26, 41. HALE, A.C., BESSER, G.M. & REES, L.H. (1986). Characterisation of YAMAGUCHI, K., ABE, K., ADACHI, I. and 6 others (1985). Peptide pro-opiomelanocortin-derived peptides in pituitary and ectopic hormone production in primary lung tumours. Rec. Results adrenocorticotrophin-secreting tumours. J. Endocrinol., 108, 49. Cancer Res., 99, 107 HOZIER, J.C., MASS, M.J. & SIEGFRIED, J.M. (1987). Genes for tumor markers are clustered with cellular proto-oncogenes on human chromosomes. Cancer Lett., 36, 235. LUSTER, W„ GROPP, C., KERN, H.F. & HAVEMANN, K. (1985a). Lung tumour cell lines synthesising peptide hormones established from tumours of four histological types: characterisation of the cell lines and analysis of their peptide hormone production. Br. J. Cancer, 51, 865. Journal of Clinical Endocrinology and Metabolism Vol. 67, No. 6 Copyright© 1988 by The Endocrine Society Printed in U.S.A.

Direct Measurement of the Precursors of Adrenocorticotropm in Human Plasma by Two™Site Immunoradiometric Assay* S. R. CROSBY, M. F. STEWART, J. G. RATCLIFFE, And A. WHITE University of Manchester, Department of Clinical Biochemistry, Hope Hospital, Clinical Sciences Building, Salford M6 8HD, United Kingdom and Wolfson Research Laboratories (J.G.R.), Queen Elizabeth Medical Centre, Edgbaston, Birmingham B15 2TH, United Kingdom •

ABSTRACT. An immunoradiometric assay (IRMA) for the normal subjects sampled at 0930 h ranged from 5-34 pmol/L. direct measurement of the precursors of ACTH in unextracted The concentrations in the patient groups studied were: 260- human plasma has been developed and evaluated clinically in 2300 pmol/L in 5 patients with the ectopic ACTH syndrome normal subjects and patients with disorders of the hypothalamic- associated with small cell lung cancer, less than 2.0-104 pmol/ pituitary-adrenal axis. The IRMA is based on an iodinated L in 10 patients with pituitary-dependent Cushing’s disease, 23 monoclonal antibody to ACTH and a monoclonal antibody to pmol/L in a patient with Nelson’s syndrome, and 3.0-230 pmol/ -yMSH coupled to Sephacryl S300. The assay detects only pep­ L in 5 patients with Addison’s disease. tides containing both epitopes, Le. POMC (31K) and pro-ACTH We conclude that this IRMA offers a simple and reliable (22K). The reference standard wa 9 partially purified POMC method for measuring ACTH precursors in unextracted plasma. from culture medium of human corticotroph adenoma cells. The The proportionately greater elevation of ACTH precursors com­ detection limit (>+2.5sd of the 0 standard) was 2.0 pmol/L and pared to ACTH in patients with the ectopic ACTH syndrome the within-assay coefficient of variation was less than 10% associated with small cell lung cancer but not in pituitary- between 29 and 2600 pmol/L. dependent Cushing’s syndrome, suggests that this assay may be Plasma concentrations of ACTH precursor peptides in 11 clinically useful. (J Clin Endocrinol Metab 6 7 : 1272,1988)

I N COMMON with other hormonal peptides, ACTH not in the plasma of normal subjects or patients with is synthesized as a high molecular weight precursor, disorders of the hypothalamic-pituitary-adrenal (HPA) POMC, which has a molecular size of approximately 31 axis. In contrast, Hale et al (3, 4) detected a high K and yields a cleavage product of approximately 2 2 K molecular weight peptide in plasma from patients with (pro-ACTH). These precursor peptides circulate in the either pituitary or ectopic ACTH hypersecretion and plasma, particularly in certain pathological conditions from normal subjects after insulin-induced hypoglyce­ such as the ectopic ACTH syndrome, but their functional mia. However, the proportion of the precursors was role, if any, is uncertain ( 1). Specific measurement of the significantly higher in patients with the ectopic ACTH precursors currently involves chromatographic separa- " syndrome. tion followed by RIA using antibodies raised to compo­ We have adopted a more direct approach to the meas­ nent peptides such as ACTH, ^MftH, or jff-lipotropin urement of ACTH precursors by using monoclonal an- 1 (0 LPH). Such methods are impractical for most clinical tibodies (MAbs) to component peptides of POMC in a studies. Furthermore, quantification is variable, depend­ two-site immunoradiometric assay (IRMA) (Fig. 1 ). We ing upon antiserum specificity and the peptide used as previously described MAbs which bind to the mid-mole­ standard. Consequently, reports on the prevalence of cule (10-18) and C-terminal (25-39) sequences of ACTH ACTH precursors in plasma are limited and contradic­ (5). We now report the production and characterization tory. Ratter et al (2) detected a high molecular weight of MAbs to 7iMSH and their use in combination with form of ACTH corresponding to pro-ACTH in the the ACTH MAbs in the development of a sensitive IRMA plasma of patients with the ectopic ACTH syndrome but for ACTH precursors in human plasma. A positive signal Received February 2,1988 in the assay requires that both peptide sequences be Address correspondence and requests for reprints to: Dr. A. White, present in the molecule(s) detected. The IRMA elimi­ University of Manchester, Department of Clinical Biochemistry, Hope nates the requirement for chromatography and is tech­ 6 Hospital, Clinical Sciences Building, Eccles Old Road, Salford M nically simple, rapid, and precise. Plasma levels of pre­ 8HD, United Kingdom. •Financial support for this project was given by the Northwest cursor peptides in normal subjects and patients with Regional Health Authority and from Boots-Celltech Diagnostics. disorders of the HPA axis are described.

1272 SPECIFIC ASSAY OF A C T H PRECURSOR PEPTIDES 1273

-*--- POMC 131K)------► Characterization of MAbs -«------Pro-ACTH-(22K)------►- Antisera, hybrid supernatant media, and ascites fluid were stored at -20 C and diluted for use in the appropriate buffer. -« N-POC -4— ACTH— >—4---/JLPH ► For the RIA, the diluent was 0.05 mol/L sodium phosphate buffer, pH 7.5, containing 0.25% BSA and 0 .0 1 % sodium azide. Antibody (100 /xL) was incubated in duplicate with [, 2!VI]7iMSH (10,000 cpm in 100 /zL) and peptide standards or diluent (100 /zL) for 18 h. Bound and free radioligand were separated with T T Y Sac Cel for 30 min at 22 C. Cross-reactivity was calculated ICII IAI2 2A3 according to the method of Abraham ( 10). Fig. 1. Binding sites of the monoclonal antibodies to the POMC The MAbs to the ACTH mid-molecule region [ 1A12, specific molecule. MAb 1C11 recognizes the 71MSH sequence (H) MAb 1A12 for ACTH-(10-18)] and the C-terminus [2A3, specific for recognizes the ACTH-(10-18) sequence and MAb 2A3 recognizes the . ,ACTH-(25-39)] were produced as described by White et al. (5). ACTH-(25-39) sequence. • Preparation of RIA and IRMA reagents Materials and Methods 7iNtSH was iodinated by the chloramine-T method ( 11) and Hybridoma cell culture media were obtained from Flow Lab­ purified by adsorption to 10 mg Vycor glass followed by elution oratories (Irvine, Ayrshire, UK) dnd tumor cell culture media with 50% acetone, 1% acetic acid. The mean specific activity from Gibco (Paisley, Scotland, UK). Costar culture dishes (24- was 149 tiCi/ng. well no. 3424 and 96-well no. 3596) were purchased from The procedures used for the purification, iodination, and Northumbria Biologicals Ltd. (Crainlington, Northumbria, solid-phase coupling of antibodies were described previously UK), polyethylene glycol (mol wt, 4000) for the fusion experi­ (12). ment from Merck (Darmstadt, W. Germany), and pristane (2, 6, 10, 14 - tetramethylpentadecane) from Aldrich Chemical Preparation of ACTH precursors for standardization Company (Gillingham, UK). All IRMA and RIA reagents were of Analar grade from Sigma Corticotroph adenoma cell culture. A pituitary tumor was ob­ Chemical Co. (Poole, Dorset, UK) unless otherwise specified. tained at hypophysectomy from a 42-yr-old man with Cushing’s Polypep was also obtained from Sigma. Donkey anti-mouse disease. Histologically, it was an invasive corticotroph ade­ immunoglobulin G (IgG) bound to cellulose (Sac Cel) and horse noma. The tissue was dispersed for 15 min at 37 C using a serum was obtained from Wellcome Diagnostics, (Dartford, Teflon paddle. The cells were cultured in RPM I1640 with 10% fetal calf serum, 4 mmol/L L-glutamine, 1 mmol/L sodium UK). Polystyrene LP3 (63 X 10 mm) and LP5 (75 X 13 mm) tubes used for the RIAs and IRMAs, respectively, were obtained pyruvate, 10 mmol/L HEPES buffer, 80 kallikrein inhibiting from LIP Ltd. (Shipley, Yorkshire UK). units/mL aprotinin and 0.2 mg/mL gentamicin for 7 weeks, Sephadex G-75 superfine, Sephacryl S300, and Sepharose changing the medium weekly. Supernatant medium removed CL4B-Protein A used in the purification of MAbs were pur­ from the cells was immediately frozen and stored at -70 C chased from Pharmacia (Uppsala, Sweden). before assay. Chromatography of corticotroph adenoma cell culture medium. Peptides Growth medium (2 mL) from the tumor cells was acidified with formic acid ( 1% final concentration) and chromatographed on Synthetic human y,MSH [N-terminal proopiocortin (N- a Sephadex G-75 superfine column (1.5 X 90 cm), eluted with POC) 51-62] was purchased from Bachem UK (Saffron Wal­ den, Essex, UK), and purified human 0LPH from UBL (Lon­ l"% formic acid containing Polypep (1 mg/mL) by the method of Ratter et al. (2). Fractions (4 mL) were immediately neu­ don, UK). Purified human ACTH [code 74/555, content 6.2 IU, tralized with approximately 85 nL 5 mol/L NaOH and 400 nL ~25 /rg/ampoule] was a gift from the National Institute for Biological Standards and Control (Holly Hill, London, UK). 0.5 mol/L sodium phosphate buffer, pH 7.5, and flash frozen ynMSH, corresponding to human N-POC 51-76, was kindly before assay. ACTH precursors in the fractions were detected provided by Dr. N. Ling, and highly purified human N-POC by using the ACTH IRMA described previously (13). Professor P. Lowry ( 6) and hNT (N-POC 1-76) (7) by Professor POMC standard. The reference standard was prepared from M. Chretien. the POMC peak eluting from the Sephadex G-75 column. The peak fraction contained 26 nmol POMC/L as determined by Immunization and MAb production measurement of the N-terminal tryptophan using N-POC as standard (14). Aliquots were diluted in assay diluent or dexa- YiMSH was conjugated to BSA by the carbodiimide method methasone-suppressed plasma. based on the procedure described by Orth ( 8). Balb/c mice were hyperimmunized with the71 MSH conjugate as described pre­ Plasma IRMAs for ACTH and ACTH precursors. The ACTH viously (5). Hybridomas secreting MAbs to 71MSH were pro­ IRMA has been described previously (13). The IRMA for duced by methods described previously for the production of detection of ACTH precursors in plasma used labeled anti- MAb to ACTH (5). Cloning and antibody production in vivo ACTH-(10-18) antibody 1A12 and 71MSH MAb 1C11 coupled were carried out as described by Coding (9). to Sephacryl S300 as the solid phase. The IRMA diluent was 1274 CROSBY E T A L JCE4M-1988 Vol 67* No 6

0.05 mol/L sodium phosphate buffer, pH 7.5, containing 0.25% Data analysis BSA, 0.1% Triton X-100, and 0.01% sodium azide. POMC All standards and unknowns were assayed in duplicate. Pre­ standards (100 fiL containing 2.6-2600 pmol/L) diluted in cision profiles were calculated as described by Edwards and ACTH-free plasma or plasma samples were incubated in dupli­ Ekins (16). cate with 2,r’I-labeled MAb (100 /xL 150,000 cpm) overnight at 4 C. MAb linked to Sephacryl S300 (100 pL, 5% slurry) then Results was added, and the mixture was incubated with constant shak­ ing for 2 h at22 C. The antibody-POMC complex was separated Characterization of MAbs to y^MSH by sucrose layering using a two-pass system (15), and samples From the fusion, six hybrids secreted antibodies to were counted for 3 min in a 7-counter. 7 1MSH and were stabilized as secondary clones. All six The optimal incubation time was assessed in a one-stage MAbs recognized yjMSH containing peptides (Table 1) assay by incubating POMC standard for 2 h at 22 C with both and cross-reacted variably with ACTH (1-100%) and l2SI-MAb and Sephacryl-linked MAb with constant shaking. In the two-stage protocol, the length of the first incubation Was. 0 LPH (1- 1 0 0 %). MAb 1C11 was highly specific for varied from 0.5 to 10 h at 4 C. 7MSH-related peptides, and its cross-reactivity with Possible interference due to cross-reaction with POMC- ACTH and jSLPH was only 1%. Therefore, this MAb derived peptides was assessed by incubating standard POMC . was chosen for further evaluation. in concentrations from 0 to 2600 pmol/L with ACTH-(l-39), Selection of MAbs for a precursor IRMA N-POC-(l-76), and 0LPH at concentrations from 22 to l l l i pmol/L. Two precursor IRMAs were established using MAb Stability of ACTH precursors 1C11 to txMSH and either MAb 1A1 2 to ACTH-(10-18) The stability of ACTH precursors was assessed by allowing- (precursor IRMA 1) or MAb 2A3 to ACTH-(25-39) aliquots of whole blood or plasma to stand at 22 C for various (precursor IRMA 2 ) (Table 2 ). These assays were com­ periods, after which plasma aliquots from each sample were pared for their ability to detect POMC peptides isolated frozen and stored at —70 C. In addition, a known amount of by Sephadex G-75 chromatography (Fig. 2 , upper panel). POMC was added to whole blood or plasma from normal Analysis of the fractions with precursor IRMA 1 revealed subjects to give a final concentration of 114 pmol/L and treated two peaks corresponding to POMC and pro-ACTH. Us­ as described above to assess the stability of exogenous peptide. ing precursor IRMA 2 , much lower concentrations were Clinical studies detected and the POMC peak was not detected. An ACTH and ACTH precursor peptides were measured in ACTH IRMA based on MAbs 1A1 2 and 2A3 also de­ plasma of 11 normal subjects (5 women, 6 men, aged 22 to 39 tected peaks of immunoactivity corresponding to POMC yr). We also studied: 1) 6 patients with the ectopic ACTH and pro-ACTH as well as ACTH-(l-/39) (Fig. 2 , lower syndrome, 5 of whom had small cell lung cancer (SCLC), panel). 1 confirmed at autopsy, and who had a lung carcinoid tumor, With precursor IRMA 1 , the relative proportion of diagnosed by bronchoscopic biopsy, 2) 10 patients with pitui­ POMC to pro-ACTH in corticotroph adenoma cell cul­ tary-dependent Cushing’s disease, in each of whom the diag­ ture medium was 2.3:1 (Table 3). This ratio was con­ nosis was confirmed at hypophysectomy, 3) 1 patient with Nelson’s syndrome, and 4) 5 patients with Addison’s disease, firmed by fluorometric estimation of the number of N- in whom glucocorticoid therapy had been withdrawn for 1-5 terminal tryptophan residues, which gave a 2 .6:1 ratio of days. The blood samples were collected between 0900-1000 h POMC to pro-ACTH. Assessment by RIA using 71MSH in lithium-heparin tubes, and the plasma was separated im­ MAb 1C 11 also revealed an excess of POMC over pro- mediately, flash-frozen, and stored at -70 C. ACTH. It would appear that MAb 1A12 (precursor

T able 1. Characterization of 71MSH MAbs

Displacement % Cross-reactivity relative to 71MSH MAb Titer" (pmol/L)* 73MSH N-POC hNT ACTH jSLPH 1C11 1:600,000 156 50 23 33 1 1 2B3 1:34,000 3,750 100 100 100 13 17 3E11 1:5,600 50,000 100 100 100 0" 0 3F1 1:2,600 81,250 100 0 0 0 0 5C11 1:650 187,500 100 100 100 100 100 6H5 1:25,000 27,500 100 100 100 40 10 N-POC and hNT are preparations of highly purified human N-terminal proopiocortin from Professor P. Lowry and Professor M. Chretien, respectively. " Titer defined as dilution of MAb which bound 50% of ['“ IJyiMSH (see section on Characterization of MAbs in Materials and Methods). 6 Concentration of 71MSH causing 50% inhibition of binding of [‘“IJyiMSH. ' No inhibition with 100,000 pmol/L. SPECIFIC ASSAY OF AC T H PRECURSOR PEPTIDES 1275

Table 2. Components of ACTH and ACTH Precursor IRMAs ACTH-(10-18) was radioiodinated. The optimal slurry of Sephacryl-linked MAb was 5%, chosen from a range MAbs Assay of 5-50%. Labeled MAb 1A12 between 1 2 0 ,0 0 0 and 1C11 1A12 2A3 240,000 cpm/tube gave low detection limits with accept­ Specificity 7.MSH ACTH-UO-18) ACTH-(25-39) able count rates, so 150,000 cpm/tube was adopted for Precursor IRMA 1 SP I2AJ subsequent experiments. Optimal sensitivity was ob­ m i Precursor IRMA 2 SP tained with a two-stage protocol involving at least a 4-h ACTH IRMA I2SJ SP incubation of 125I-MAb 1A12 followed by a 2 -h incubation SP, Solid-phase was Sephacryl S300. with Sephacryl-linked MAb 1C 11. For convenience, an overnight initial incubation was chosen for routine use. ACTH PRECURSOR IRMA* The effects of POMC standards diluted in assay buffer, Vo 34k 24k 4.5k Vt plasma from normal subjects given dexamethasone (ACTH-free plasma), and horse serum were virtually

cpm identical (data not shown). Dilutions of endogenous x 10* ACTH precursors from three patients with the ectopic 2 0 - ACTH syndrome diluted 1:10 in ACTH-free plasma were parallel to the standard curve (data not shown). ACTH- . 1 5 - free plasma was chosen as the matrix for standardization

10 - of the assay for clinical studies.

8- Sensitivity, specificity, and precision , The limit of detection of the assay was 2.0 pmol/L (± i— r 2.5 SD of 0 dose). The assay recognized both POMC and pro-ACTH but did not distinguish between them (Fig. 3) ACTH IRMA and did not cross-react with up to 2 2 2 pmol/L other 14.3- POMC-derived peptides. Since ACTH and N-POC both 1 4 - ng/L bound to one or another of the MAbs, they could poten­ 6- tially interfere with binding of POMC. However, in the presence of N-POC or 0 LPH at concentrations up to 430 pmol/L, no reduction in binding of POMC was found 4 — over the whole standard curve (data not shown). Binding of POMC was not affected by ACTH up to 2 2 2 pmol/L, 2- but at 1111 pmol/L the binding of high levels of POMC r-J (1300 pmol/L) was reduced by half. The within-assay coefficient of variation was less than 1 0 % over the range Friction no. Fig. 2. Chromatographic profiles of POMC peptides from the growth medium of corticotroph adenoma cells in culture. Upper panel, ACTH TABLE 3. Quantification of ACTH precursors purified from cortico­ precursors measured with precursor IRMA 1 using MAbs to YjMSH troph adenoma cell culture medium and ACTH 10-18 ( • -----• ) and precursor IRMA 2 using MAbs to 7,MSH and ACTH 25-39 (O --- O). Lower panel, ACTH measured by Assay POMC peak pro-ACTH peak Ratio an IRMA using MAbs to ACTH-( 10-18) and ACTH-(25-39). ACTH IRMA (MAbs 158 pmol/L 1,200 pmol/L 0.1:1 1A12 + 2A3) IRMA 1) recognized the precursors completely while Signal generated in 23,800 cpm 10,300 cpm 2.3:1 precursor IRMA 1 precursor IRMA 2 and the ACTH IRMA did not. There­ (MAbs 1C11 + fore precursor IRMA 1 with MAbs 1C11 and 1A12 was 1A12) selected for optimization. Since no other source of stand­ Signal generated in 990 cpm 3,100 cpm 0.3:1 ard was available, POMC from fraction 2 0 (Fig. 2 , upper precursor IRMA 2 panel) was chosen as standard and was assigned a con­ (MAbs 1C11 + 2A3) 2 6:1 centration of 26 nmol/L based on the N-tryptophan N-Tryptophan assay 26 nmol/L 10 nmol/L . Concentration meas­ 12 nmol/L 3 nmol/L 4.0:1 assay. ured by N-POC RIA Optimization of the plasma IRMA The specificity of each MAb is shown in Fig. 1. The precursors were purified by Sephadex G-75 chromatography (Fig. 2) and the TOMC Monoclonal antibody 1C11 to 71MSH was coupled to and pro-ACTH peak fractions were 20 and 25, respectively. The total Sephacryl S300 as the solid-phase and MAb 1A1 2 to radioactivity added to each precursor IRMA was 150,000 cpm. 1276 CROSBY ET AL JCE & M * 1988 Vol 67 • No 6

29-2600 pmol/L (Fig. 4). The between-assay coefficient Plasma levels of ACTH and precursors of variation was less than 10% between 37-2600 pmol/ L. In normal subjects at 0930 h, low but measurable concentrations of precursors were found (mean, 13 pmol/ Stability of ACTH precursors L; range, 5.0 to 34 pmol/L), with comparable levels of ACTH of 1.5 to 12.2 pmol/L (Fig. 5). There was no Partially purified POMC was stable in whole blood or obvious sex difference in the concentration of either plasma for at least 4 h at 22 C, the recovery being 95% peptide. and 91%, respectively. After 16 h, the recovery fell to Patients with the ectopic ACTH syndrome associated 86% in whole blood and 73% in plasma. Endogenous with SCLC had grossly elevated precursor levels, ranging plasma ACTH precursor concentrations were 83% and from 260-2340 pmol/L. The plasma ACTH levels in 105% of the initial concentrations in whole blood and these patients were also elevated (18-171 pmol/L). One plasma, respectively, after 4 h at 22 C, from a patient patient with the ectopic ACTH syndrome associated with with the ectopic ACTH syndrome and 106% and 87%’, a lung carcinoid tumor had a low normal plasma precur­ respectively, in blood and plasma from a patient with sor peptide concentration (4.0 pmol/L), compared to a Nelson’s syndrome. plasnia ACTH level of 32 pmol/L. In patients with Cushing’s disease precursor levels ranged from less than 2.0 to 104 pmol/L compared to ACTH levels of 6.6 to 43 10000- pmol/L. The one patient with Nelson’s syndrome had a cpm precursor level within the normal range (23 pmol/L) and a highly elevated ACTH level (137 pmol/L). In Five

1000- patients with Addison’s disease, ACTH precursor con­ centrations ranged from undetectable to 234 pmol/L, whereas ACTH levels ranged between 13 and 333 pmol/ L. 100—H The molar concentrations of ACTH precursor peptides in patients with the ectopic ACTH syndrome due to SCLC were at least 10-fold greater than those of normal T l subjects and at least 2-fold greater than those of patients 1 10 100 1000 with pituitary-dependent Cushing’s syndrome. In con­ POMC & Derived Peptides (pmol/L) trast, the ACTH levels in the patients with the ectopic Fig . 3. Specificity of the precursor IRMA: POMC ( • -----• ) , pro- ACTH syndrome due to SCLC overlapped with those in ACTH (V 7), and POMC*derived peptides; N-POC (▼ -----▼), the patients with Cushing’s disease. ACTH (O O), and (3LPH (□-----D), were each incubated separately with the MAbs. Eachpoint represents the mean of duplicate determi­ nations. Discussion The practical problems associated with the measure­

10.000- ment of ACTH precursor peptides in plasma have se­ verely limited clinical studies so far. The IRMA described cpm here overcomes these difficulties and provides a simple, -SO -45 1000 - C V { * | 2000-1 -40 ACTH pmol/L Prieurwil pmol/L -35 1000- -30 -26 600- 100- -20 -15 -10 0-1 10 100 1000 -200 ACTH Prtcuriori (pmol/L)

Fig . 4. Standard curve ( • -----• ) and precision profiles for the pre­ 100- cursor IRMA using POMC purified from corticotroph adenoma cell culture medium and diluted in ACTH-free plasma as standard. The 1 within-assay precision profile (i— ■) is based on 42 samples and the Cvthipfi EctopH Nthom Addinni Cirtinoid between-assay precision profile (A A) is based on three separate Fig . 5. Comparison of plasma ACTH (O) and ACTH precursor (•) assays and 102 samples. levels in normal subjects and patients with disorders of the HPA axis. SPECIFIC ASSAY OF ACTH PRECURSOR PEPTIDES 1277

rapid, and reliable method for the combined direct meas­ significant concentrations of immunoreactive ACTH, urement of both precursors, POMC and pro-ACTH. This predominantly in a large molecular weight form (mol wt has been achieved because the measured signal requires 20,000). More recent reports (2, 3) also suggest that a binding of two MAbs, one to 7MSH and the other to relatively high proportion of circulating ACTH-like pep­ ACTH. Both sequences are present together only in tides are present as high molecular weight forms in these POMC and pro-ACTH. patients. Although the number of patients we studied Standardization of ACTH precursor assays is difficult was small, ACTH precursor levels clearly differentiated because of the lack of readily available reference prepa­ those with the ectopic ACTH-syndrome due to SCLC rations. We used POMC partially purified from human from those with pituitary-dependent Cushing’s disease, corticotroph adenoma cell culture medium. Analysis of whereas the ACTH levels for these groups overlapped. the fractionated medium with two precursor IRMAs Acknowledgments showed that detection of POMC was related to the speo ificity of the ACTH MAb used. When C-terminal ACTH * We are grateful to Professor L. H. Rees and Dr. A. Clark for provision of the pituitary tumor and Mrs. S. Gibson for preparation of reagents MAb 2A3 was used with 7MSH MAb 1C11 (precursor • forthe ACTH IRMA. IRMA 2 ), the major form of precursor detected corre­ References sponded to pro-ACTH. In contrast, when mid-ACTH 1. Hale AC, Ratter SJ, Tomlin SJ, Lytras N, Besser GM, Rees LH MAb 1A12 was used with 7MSH MAb 1C11 (precursor 1984 Measurement of immunoreactive yMSH in human plasma. IRMA 1), POMC was the major precursor detected. Clin Endocrinol (Oxf) 21:139 Thus, MAb 2A3 appears to be specific for the extreme 2. Ratter SJ, Lowry PJ, Besser GM, Rees LH 1980 Chromatographic characterisation of adrenocorticotropin in human plasma. J En­ C-terminus of ACTH and reacts poorly with peptides in docrinol 85:359 which this sequence *is obscured, as in POMC. This 3. Hale AC, Besser GM, Rees LH 1985 Characterisation of proopi­ interpretation is supported by the poor cross-reactivity omelanocortin-derived peptides in pituitary and ectopic adrenocor- ticotrophin-secreting tumours. J Endocrinol 108:49 of POMC in the ACTH IRMA which uses this antibody. 4. Hale AC, Millar JG, Ratter SJ, Pickard JD, Doniach I, Rees LH The choice of mid-ACTH MAb 1A12 in combination 1985 A case of pituitary dependent Cushing’s disease with biochem­ with 7MSH MAb 1C1 1 appears to overcome this steric ical features of the ectopic ACTH syndrome. Clin Endocrinol (0x0 22:478 problem. 5. White A, Gray C, Ratcliffe JG 1985 Characterisation of monoclonal A sensitive IRMA for ACTH precursors was obtained antibodies to adrenocorticotrophin. J Immunol Methods 79:185 with solid-phase MAb 1C 11 and labeled MAb 1A1 2 , with 6. Estivariz FE, Hope J, McLean C, Lowry PJ 1980 Purification and characterisation of a y-melanotropin precursor from frozen human a formal detection limit of 2 .0 pmol/L. The IRMA has pituitary glands. Biochem J 191:125 similar precision within and between assays, which sug­ 7. Seidah NG, Chretien M 1981 Complete amino acid sequence of a gests that minor variations in reagents or technique do human pituitary glycopeptide: an important Saturation product of pro-opiomelanocortin. Proc Natl Acad Sci USA 78:4236 not affect the results. Plasma samples generated dilution 8. Orth DN 1979 Adrenocorticotropic hormone (ACTH) In: Jaffe M, curves parallel to that of the POMC standard. An im­ Behrman HR (eds): Methods in Hormone Radioimmunoassay. portant practical point is the stability of POMC immu- Academic Press, New York, pp 245-283 9. Goding JW 1980 Antibody production by hybridomas. J Immunol noactivity in whole blood or plasma for at least 4 h at 22 Methods 39:285 C, obviating the need for rigorous care in handling spec­ 10. Abraham GE 1969 Solid phase radioimmunoassay of oestradiol- imens. This is similar to the stability of ACTH immu- 17,0. J Clin Endocrinol Metab 29:866 11. Hunter WM, Greenwood FC 1962 Preparation of Iodine-131 la- noactivity measured by IRMA (13). * belled human growth hormone of high specific activity. Nature There are no reports of unstimulated plasma levels of 194:495 ACTH precursors in normal subjects. Circulating high 12. Dobson SH, Gray C, Smith H, Baker T, Ratcliffe JG, White A 1986 Selection and optimisation of monoclonal antibodies for a molecular weight forms of ACTH and 7MSH have pre­ two-site immunoradiometric assay for ACTH. J Immunol Methods viously been detected in normal subjects only after stim­ 88:83 ulation by insulin-induced hypoglycemia (3). In patients 13. White A, Smith H, Hoadley M, Dobson SH, Ratcliffe JG 1987 Clinical evaluation of a two-site immunoradiometric assay for with Addison’s disease, the levels of ACTH precursors adrenocorticotrophin in unextracted human plasma using mono­ were elevated to a similar degree as were ACTH levels. clonal antibodies. Clin Endocrinol (0x0 26:41 The lower concentrations of ACTH precursors than 14. Hakanson R, Sundler F 1971 Fluorometric determination of N- terminal tryptophan-peptides after formaldhyde condensation. ACTH in one patient with Nelson’s syndrome and an­ Biochem Pharmacol 20:3223 other with ectopic ACTH syndrome due to a lung carci­ 15. Hunter WM, Budd PS 1981 Immunoradiometric versus radio­ noid tumor require confirmation. immunoassay: a comparison using alpha fetoprotein as model analyte. J Immunol Methods 45:255 Of particular note were the markedly elevated plasma 16. Edwards PR, Ekins RP 1982 Development of a microcomputer ACTH precursor levels in five patients with the ectopic radioimmunoassay data processing program for distribution by the ACTH syndrome due to SCLC. This finding was not World Health Organisation. In: Radioimmunoassay and Related Procedures in Medicine. I.A.E.A., Vienna, pp 423-442 unexpected, since earlier studies by Gerwitz and Yalow 17. Gerwitz G, Yalow RS1974 Ectopic ACTH production in carcinoma (17) revealed that extracts of lung carcinomas contained of the lung. J Clin Invest 53:1022