Chimeric studies of the extracellular domain of the rat thyrotropin (TSH) receptor: amino acids (268-304) in the TSH receptor are involved in ligand high affinity binding, but not in TSH receptor-specific signal transduction

A series of chimeric TSH-LH/CG receptors were constructed by substituting homologous segments of the extracellular domain of the rat TSH receptor with corresponding segments of rat LH/CG receptor: C1 (amino acids 37-123 substituted), C2 (91-112), C3 (173-234), C4 (233-266), C5 (268-304), C6 (112-305) and C7 (36-404). After transfection in Cos-7 cell, TSH- and LH/CG-receptor activities of these chimeras were evaluated and compared with those of deletion mutants involving the same residues [Kosugi et al. Thyroid 1:321 (1991)]. Western blot analyses revealed that most of the chimeric receptor proteins were normally synthesized and integrated in the membrane of transfected Cos-7 cells: an antibody to a TSH receptor specific synthetic peptide (residues 352-366) identified 170-190kDa and 90-100kDa TSH receptor structures in the plasma membrane fractions of Cos-7 cells transfected with wild-type TSH receptor cDNA and the C1 to C6 chimeras, but not C7 or wild LH/CG receptor cDNA. Despite this, no receptor except C5 exhibited any significant TSH receptor activities either in [12I]TSH binding or in cAMP responses to TSH and thyroid-stimulating antibodies (TSAbs) from Graves' patients. The chimeric receptor C5 exhibited only low affinity TSH binding (Kd = 3.5 x 10(-8) M), as did its counterpart the M2C mutant with residues 268-304 deleted. However, unlike M2C, C5 demonstrated a significant cAMP response to TSH as well as to TSAbs. The cAMP increase in response to TSH in the wild type receptor was observed at 10(-11) M TSH. In C5 the response was first evident at 10(-10) M TSH, but the maximum cAMP stimulation by TSH and TSAbs in C5 (EC50 = 6.7 x 10(-10) M) was approximately the same as the wild type receptor (EC50 = 1.5 x 10(-10) M). Inhibition of either TSH- or TSAb- stimulated cAMP increase by thyroid-stimulating blocking antibodies (TSBAbs) was also preserved in C5. These results suggest that amino acids 268-304 do not include an important determinant required for signal transduction, since a significant cAMP response to TSH and TSAbs was observed in the C5 receptor with these residues substituted. Additionally, these residues appear to be involved in ligand high affinity binding because high affinity TSH binding was lost in the chimeric receptor C5.

Characterization of epidermal growth factor receptor in a rat thyroid cell line, FRTL-5

A rat thyroid cell line FRTL-5 showed no significant binding of epidermal growth factor(EGF) either in the presence or absence of thyrotropin(TSH). Conditioned medium from FRTL-5 did not inhibit the specific binding of [125I]EGF to HeLa cell which has EGF receptors (EGF-R). Two bands of 170KD and 80KD obtained by immunoprecipitation with solubilized FRTL-5 and anti-EGF-R monoclonal antibody indicated autophosphorylation activity. The activity was enhanced by several thyroid cell growth factors including TSH. dibutyryl-cAMP(db-cAMP), insulin like growth factor I(IGF-I). Interleukin 1 beta (IL-1 beta) and phorbol ester. EGF alone enhanced the autophosphorylation activity. These data indicate that FRTL-5 possesses unique EGF-R with low affinity to EGF and that the tyrosine kinase of the receptor is activated during cell proliferation.

Pituitary resistance to thyroid hormone associated with a base mutation in the hormone-binding domain of the human 3,5,3'-triiodothyronine receptor-beta

Point mutations in the human T3 receptor-beta (TR beta) gene causing single amino acid substitutions have been identified in several different kindreds with generalized resistance to thyroid hormone. Until now, no study has been reported on the TR gene in cases of pituitary resistance (PRTH). In the present study, we analyzed the TR beta gene in a 30-yr-old Japanese female with PRTH. She exhibited clinical features of hyperthyroidism, elevated serum thyroid hormone levels accompanied by inappropriately increased secretion of TSH, mildly elevated basal metabolic rate, and increased urinary excretion of hydroxyproline. No pituitary tumor was detected. DNA fragments of exons 3-8 of the genomic TR beta gene were generated by the polymerase chain reaction and analyzed by a single stranded conformation polymorphism method. Exon 7 of the patient's TR beta gene showed an abnormal band, suggesting the existence of mutation(s). By subcloning and sequencing the DNA, a point mutation was identified in one allele at nucleotide 1297 (C to T), which altered the 333rd amino acid, arginine, to tryptophan. Neither of her apparently normal parents had any mutations of the TR beta gene. In vitro translation products of the mutant TR beta gene showed remarkably decreased T3-binding activity (Ka, 2.1 x 10(8) M-1; normal TR beta Ka, 1.1 x 10(10) M-1). Since the molecular defect detected in a patient with PRTH is similar to that seen in subjects with generalized resistance to thyroid hormone, both types of the syndrome may represent a continuous spectrum of the same etiological defect with variable tissue resistance to thyroid hormone.

Low doses of endothelin-1 inhibit atrial natriuretic peptide secretion

To clarify the interaction between endothelin-1 (ET-1) and atrial natriuretic peptide (ANP), the effects of ET-1 on ANP secretion were investigated in isolated perfused rat hearts and in conscious unrestrained rats. Perfusion with 10(-9) M ET-1 stimulated ANP secretion from the isolated perfused rat heart. However, 10(-10) M ET-1 significantly decreased ANP secretion for the initial 15 min of the perfusion period. The perfusion with 10(-11) M ET-1, which is near the plasma level of ET-1 (2 x 10(-12) M), inhibited ANP secretion throughout the perfusion period. The perfusion of 10(-12) M ET-1 slightly decreased ANP secretion. After the ET-1 perfusion period, ANP secretion increased in proportion to ET-1 dose. The inhibitory action of ET-1 on ANP secretion was almost abolished by simultaneous perfusion of indomethacin, a cyclooxygenase inhibitor, but not by methylene blue, an inhibitor of soluble guanylate cyclase. In conscious unrestrained rats the iv infusion of 1 pmol/kg.min ET-1, which doubled the plasma ET-1 level, slightly but significantly decreased the plasma ANP level 5 and 10 min after the initiation of the infusion. The infusion of 10 and 30 pmol/kg.min ET-1 increased the plasma ANP level. These results demonstrate that low doses of ET-1 exert an inhibitory and short-acting action on ANP secretion from the heart, although high doses of ET-1 exert stimulating and long-lasting action, and suggest that prostanoids are involved in this inhibitory action.

Enhancing effect of 17 beta-estradiol on human NK cell activity

The in vitro effect of 17 beta-estradiol on NK activity was studied. The proliferation and NK activity of YT-N17 (a human NK-like cell line) were enhanced by 17 beta-estradiol (E2), and the enhancement was blocked by tamoxifen (Tx), an antagonist of E2. On the contrary, other steroid hormones such as Tx, progesterone, and testosterone had no effect. YT-N17 contained 11.8 fmol/mg protein of estrogen receptor (mean of two independent assays), a value which was 5-10-fold higher than that of other hematopoietic cell lines. An enhancement of NK activity by E2 was also seen in large granular lymphocytes obtained from normal subjects, and it was again suppressed by Tx. These data suggest that E2 is one of the activating factors for NK/LGL cells.

Galanin-induced growth hormone secretion in conscious rats: evidence for a possible involvement of somatostatin

Intracerebroventricular administration of galanin (GAL) potently evoked growth hormone (GH) secretion in conscious male rats. Pretreatments with neostigmine and cysteamine blunted the GAL-induced GH secretion. Pretreatment of animals with a specific anti-somatostatin serum significantly inhibited the GAL-induced GH secretion. On the contrary, GH-releasing hormone-induced GH secretion was significantly enhanced with cysteamine and anti-somatostatin serum. These results suggest that somatostatin is involved in GAL-induced GH secretion in rats.

Inhibition of PMA-induced, LFA-1-dependent lymphocyte aggregation by ADP ribosylation of the small molecular weight GTP binding protein, rho

Botulinum C3 exoenzyme specifically ADP-ribosylates a group of ras-related small molecular weight GTP-binding proteins, rho, and inhibits their biological activity. Using this enzyme, we examined the function of rho in PMA-induced activation of lymphocyte function-associated antigen-1 (LFA-1) in a B lymphoblastoid cell line, JY. Northern blot analysis revealed that among the three rho genes, rhoA mRNA was predominantly expressed in JY cells. Consistently, only one [32P]ADP-ribosylated band was found when the lysate of the cells was subjected to ADP ribosylation by C3 exoenzyme. When the cells were cultured with C3 exoenzyme, this substrate was ADP-ribosylated in situ in a time- and concentration-dependent manner. Concomitant with this ADP ribosylation, PMA-induced LFA-1/intercellular adhesion molecule (ICAM)-1-dependent aggregation of JY cells was inhibited. This inhibition was blocked by prior treatment of the enzyme with an anti-C3 monoclonal antibody, and overcome by stimulation with higher concentrations of PMA. The C3 exoenzyme-induced inhibition was not affected by shaking of the cell suspension, while inhibition of aggregation by cytochalasin B was abolished by this procedure, suggesting that the inhibitory effect of the C3 exoenzyme treatment was not due to decrease in cell motility. The C3 exoenzyme treatment affected neither protein phosphorylation in JY cells before and after PMA stimulation, nor affected surface expression of LFA-1 and ICAM-1. These results suggest that rhoA protein works downstream of protein kinase C activation linking PMA stimulation to LFA-1 activation and aggregation in JY cells.

Iodothyronine 5'-deiodinase is present in mouse sublingual gland

Homogenates from male ddy mouse sublingual glands were incubated with 100 nM [125I]rT3 in the presence of 2 mM dithiothreitol (DTT). Metabolites were analyzed by HPLC or Dowex-50 minicolumn assay. 3,3'-Diiodothyronine and I- were the only appreciable products. Iodide release from rT3 was compatible with the enzymatic nature. The Km and maximum velocity from three separate determinations were 326, 356, and 629 nM and 29.8, 42.1, and 31.3 pmol I- release/mg protein.min, respectively. The deiodinase activity was DTT dependent and had higher affinity for rT3 than for T4 or T3. 6-Propyl-2-thiouracil inhibited the deiodination, which was competitively overcome by DTT. Sublingual 5'-deiodinase activity was approximately 80% of that in the liver, while the submandibular gland showed no deiodination. Our results show the presence of 5'-deiodinase (type I) in mouse sublingual gland for the first time. Selective localization and abundance of the enzyme suggest a previously unrecognized role of the sublingual gland in thyroid hormone physiology.

Specific activity of phosphatidylinositol 3-kinase is increased by insulin stimulation

We investigated whether phosphatidylinositol 3-kinase (PI3K) is phosphorylated and whether its specific activity is increased by insulin stimulation in vivo using Fao cells and antibodies raised against the 85 kDa subunit of PI3K, insulin-receptor substrate-1 (IRS-1), and phosphotyrosine (pTyr). PI3K activity was detected in the immunoprecipitate produced with anti-PI3K at a basal state. The activity was increased 2-3-fold by insulin stimulation, although the protein concentration of kinase in the anti-PI3K immunoprecipitates was the same before and after insulin stimulation. Both anti-pTyr and anti-IRS-1 antibodies immunoprecipitated the kinase activity only after insulin stimulation. After the first immunoprecipitation with anti-pTyr, the supernatant was immunoprecipitated once more with anti-PI3K. PI3K activity in the second immunoprecipitate revealed little difference between the basal and insulin-stimulated states, suggesting that most of the insulin-activated portion of PI3K was precipitated by anti-pTyr. Both IRS-1 and the insulin-receptor beta-subunit (95 kDa) were phosphorylated on tyrosine residues by insulin stimulation and immunoprecipitated with anti-pTyr. However, phosphorylation of neither subunit of PI3K (85 kDa or 110 kDa) was detectable in the immunoprecipitate produced with anti-pTyr. The 185 kDa pTyr-containing protein was immunoprecipitated with anti-PI3K after insulin stimulation, although there was little phosphorylation of the 85 kDa protein. pTyr in the 110 kDa protein immunoprecipitated with anti-PI3K was below detectable levels. These results indicate that the specific activity of PI3K is increased by insulin stimulation without detectable tyrosine phosphorylation of PI3K itself in Fao cells. The majority of the insulin-activated portion of PI3K is associated with pTyr-containing proteins including IRS-1, which suggests that this is important for activation of PI3K by insulin.

The human endothelin-B receptor gene. Structural organization and chromosomal assignment

The gene encoding the human endothelin-B receptor (hET-BR) has been isolated, and its structural organization and chromosomal assignment have been determined. Southern blot analysis demonstrated a single copy of the hET-BR gene in the human genome. The hET-BR gene spans 24 kilobases and consists of seven exons and six introns. The size range for exons is 109-2855 base pairs, although that for introns is 0.2-14.5 kilobases. Every intron occurs near the border of the putative transmembrane domain in the coding region. The major transcription initiation sites were mapped to the positions 258 and 229 base pairs upstream of the ATG initiation codon by primer extension and nuclease S1 protection experiments. The 5'-flanking region of the hET-BR gene lacks conventional TATA and CCAAT boxes but contains a sequence of potential Sp1 binding sites upstream of the transcription initiation sites. There are some canonical consensus sequences of cis-elements including GATA motif, acute phase reactant regulatory element, and E box. Using human-rodent somatic hybrid cell lines, the hET-BR gene was assigned to human chromosome 13. The present study will lead to a better understanding of the mechanism for the transcriptional regulation of the hET-BR gene and will give a clue as to how to search for possible genetic disorders of hET-BR.

Liver damage in patients with colony-stimulating factor-producing tumors

PURPOSE

We have demonstrated that colony-stimulating factor (CSF)-producing tumor cell lines produce not only CSF but also interleukin-1 (IL-1) and interleukin-6 (IL-6). Clinically, we have observed that patients bearing such tumors present with liver dysfunction and fever in addition to marked leukocytosis. The purpose of this study was to determine whether or not the liver damage was specifically related to CSF-producing tumors.

PATIENTS AND METHODS

Clinicopathologic examinations were performed in six autopsied patients with CSF-producing tumors. We also transplanted two tumor cell lines (KHC287 and CHU-2), which produce granulocyte (G)-CSF, IL-1, and IL-6, to nude mice.

RESULTS

Of the six patients, five had G-CSF- and one had granulocyte/macrophage (GM)-CSF-producing tumors. IL-1 and IL-6 concentrations in plasma or culture supernatant were elevated in these patients. Biochemical examinations revealed high serum enzyme levels of the biliary system in contrast to normal or slight increases in transaminase levels in all patients studied. Serum direct bilirubin was elevated in five of the six patients. Three common pathologic changes of the liver were found: (1) focal necrosis associated with neutrophil infiltration in the centrilobular zones, (2) fibrous change and enlargement of the portal area associated with neutrophil infiltration, and (3) intrahepatic cholestasis. The same pathologic changes, except for cholestasis, were reproduced in the liver of mice transplanted with KHC287 or CHU-2.

CONCLUSION

These results indicate that patients with CSF-producing tumors have characteristic liver damage, and suggest a new paraneoplastic syndrome of leukocytosis and liver damage.

Different secretion patterns of atrial natriuretic peptide and brain natriuretic peptide in patients with congestive heart failure

BACKGROUND

The plasma levels of atrial natriuretic peptide (ANP) and brain natriuretic peptide (BNP) are increased in relation to the severity of congestive heart failure (CHF). This study was designed to examine whether the secretion patterns of ANP and BNP vary with underlying cardiac disorders of CHF with different degrees of overload in atria and ventricles.

METHODS AND RESULTS

We measured plasma levels of ANP and BNP in the aorta in 20 patients with mitral stenosis (MS) in whom atria are mainly overloaded, 30 patients with dilated cardiomyopathy (DCM) in whom both atria and ventricles are overloaded, and 20 control subjects during cardiac catheterization. Pulmonary capillary wedge pressure (PCWP) was significantly higher in the MS and DCM groups (16.7 +/- 4.7 mm Hg and 15.1 +/- 7.7 mm Hg, respectively) than in the control group (7.2 +/- 1.1 mm Hg, p < 0.01), whereas there was no significant difference between the MS and DCM groups. Left ventricular end-diastolic pressure (LVEDP) was significantly higher in the DCM group than in the MS group (16.4 +/- 7.8 mm Hg versus 7.6 +/- 2.0 mm Hg, p < 0.01), and the level was comparable between the MS and control groups (7.6 +/- 2.0 mm Hg versus 6.8 +/- 1.2 mm Hg, p = NS). The plasma ANP level was significantly higher in the MS and DCM groups (356 +/- 169 pg/ml and 331 +/- 323 pg/ml, respectively) than in the control group (98 +/- 41 pg/ml, p < 0.01), whereas there was no significant difference between the MS and DCM groups. The plasma BNP level was significantly higher in the DCM group than in the MS group (333 +/- 405 pg/ml versus 147 +/- 54 pg/ml, p < 0.01), and the level was significantly higher in the MS group than in the control group (147 +/- 54 pg/ml versus < 10 pg/ml, p < 0.01). The plasma levels of ANP and BNP had a highly positive correlation with PCWP in the DCM group (p < 0.01). On the other hand, in the MS group, the plasma ANP level had a highly significant correlation with PCWP (p < 0.01) but the plasma BNP level did not.

CONCLUSIONS

We conclude that plasma levels of BNP mainly reflect the degree of ventricular overload and that the secretion patterns of ANP and BNP vary with underlying cardiac disorders of CHF with different degrees of overload in atria and ventricles.

Preparation of anti-phosphoserine and anti-phosphothreonine antibodies and their application in the study of insulin- and EGF-induced phosphorylation

We prepared antibodies against phosphoserine (P-Ser) and phosphothreonine (P-Thr) by immunizing rabbits with P-Ser or P-Thr conjugated to bovine serum albumin. The antibodies (anti-P-Ser and anti-P-Thr) were purified using P-Ser or P-Thr affinity columns. Anti-P-Thr was highly specific for P-Thr, while anti-P-Ser showed weak cross-reactivity with P-Thr. We showed that these antibodies can immunodetect serine/threonine phosphorylated insulin and epidermal growth factor (EGF) receptors and several proteins which are phosphorylated on serine/threonine residues in response to insulin or EGF stimulation. The antibodies will certainly provide a good tool for discovering novel kinases and substrates involved in signal transduction.

Role and mutational heterogeneity of the p53 gene in hepatocellular carcinoma

The mutational spectrum of the p53 gene was analyzed in 53 hepatocellular carcinomas. Somatic mutations of the p53 gene were detected in 17 cases (32%). Among these 17 mutations, 9 were missense mutations; the mutations in the other 8 cases were nonsense mutations, deletions, or mutations at the intron-exon junctions. These mutations were found in a wide region stretching from exon 4 to exon 10 without any single mutational hot spot. G:C to T:A transversions were predominant, suggesting the involvement of environmental mutagens in the mutagenesis of the p53 gene in a subset of the hepatocellular carcinoma cases. Mutations of the p53 gene occurred frequently in advanced tumors, although several tumors in the early stages also showed mutations. A deletion map of chromosome 17 was constructed by using 10 polymorphic probes and was compared with the p53 gene mutation in each case. Loss of heterozygosity (LOH) on chromosome 17p was observed in 49% of the cases (24 of 49), and two commonly deleted regions were detected (around the p53 locus and at 17p13.3 to the telomere). Sixteen of the 17 cases with p53 gene mutations showed LOH around the p53 locus, and mutations were rare in hepatocellular carcinomas without LOH. However, no mutations were detected in 8 cases with LOH on 17p, suggesting the possibility that an unidentified tumor suppressor gene(s) located on 17p may have also been involved in hepatocarcinogenesis.

Estimation of the protein content of thyroid hormone receptor alpha 1 and beta 1 in rat tissues by western blotting

Recent studies of the expression of c-erbA/thyroid hormone receptor (TR) mRNAs have revealed a dissociation between T3-binding activity and the behavior of the mRNAs that code the functional TRs in some tissues. Compared with T3-binding activity, TR(alpha 1 + beta 1) mRNA is disproportionally high in the brain and low in the liver. Using anti-TR antiserum, 4BII, which recognizes TR alpha 1 and beta 1, but not the alpha 2-variant, we measured TR protein content in rat tissues by Western blotting. Two protein bands of 47 and 55 kilodaltons (kDa) were specifically identified as TR proteins. The positions of the in vitro transcription/translation products of c-erbA/TR alpha 1 and beta 1 cDNA on the gel were consistent with those of the 47- and 55-kDa bands, respectively. The 47- and 55-kDa proteins in nuclear proteins extracted with 0.4 M KCl from rat tissues were analyzed by Western blotting, and the intensity of TR protein bands in each tissue was measured by a densitometer. The relative TR protein concentration was highest in liver, followed by brain, kidney, and testis. We compared the TR protein level measured by Western blotting with the maximal T3-binding capacity (Cmax) in the same aliquot of samples from liver and brain. Both the TR protein level and the Cmax in the brain were about 40% of those in the liver, suggesting that the Cmax per receptor molecule is constant in these two tissues, and an abundant amount of functional TR proteins exists in the liver, corresponding to the high level of T3-binding activity.

The natriuretic peptide family

The natriuretic peptide system is a complicated system comprising at least three endogenous peptides, atrial natriuretic peptide, brain natriuretic peptide, and C-type natriuretic peptide and three receptors, the atrial natriuretic peptide-A receptor (guanylyl cyclase A), the atrial natriuretic peptide-B receptor (guanylyl cyclase B), and the clearance receptor. The accumulated evidence indicates that this system is implicated in the control of blood pressure, body fluid homeostasis, and vascular remodeling as both cardiac hormone and local regulator.

Abnormal facilitative glucose transporter gene expression in human islet cell tumors

Our previous studies have shown that increased expression of GLUT1/erythrocyte and GLUT3/brain type glucose transporter genes in human tumors is associated with cellular transformation. We have now determined the levels of messenger RNAs (mRNAs) encoding these two glucose transporter isoforms as well as that of GLUT2/liver isoform in insulin-, glucagon-, and gastrin-secreting islet cell tumors. Northern blot analysis and reverse transcriptase-polymerase chain reaction revealed the presence of GLUT1 and GLUT3 mRNA in all human islet cell tumors and normal islets examined. In contrast, GLUT2 mRNA, which is present at high levels in normal islets, was not detected in insulinomas or other types of islet cell tumors. These results imply that GLUT1 and GLUT3 are primarily responsible for glucose uptake by these tumors.

Ventricular expression of atrial and brain natriuretic peptides in dilated cardiomyopathy. An immunohistocytochemical study of the endomyocardial biopsy specimens using specific monoclonal antibodies

Although brain natriuretic peptide is expressed in ventricles of failing hearts including dilated cardiomyopathy, its morphological localization is still unclear. In this study, we analyzed the immunohistocytochemical localization of atrial and brain natriuretic peptides in ventricles of dilated cardiomyopathy at both light and electron microscopic levels. Ventricular specimens were obtained by endomyocardial biopsy in 31 patients (26 with dilated cardiomyopathy and 5 controls without any specific cardiac disease). By light microscopic immunohistochemistry using specific monoclonal antibodies, all (26 of 26) of the left ventricular endomyocardial biopsy specimens and 31% (8 of 26) of the right ventricular specimens showed immunoreactivity for both of these natriuretic peptides in dilated cardiomyopathy. In contrast, none of the normal controls showed immunoreactivity for either of these peptides. The percentage of atrial natriuretic peptide-containing or brain natriuretic peptide-containing myocytes in the left ventricular specimens showed an inverse correlation with the left ventricular ejection fraction (r = -0.72 and r = -0.69, respectively). By electron microscopy, we identified specific secretory granules in ventricular myocytes from patients with dilated cardiomyopathy, but not in those from normal controls. Double immunocytochemistry using a two-face immunogold staining method revealed brain natriuretic peptide colocalized with atrial natriuretic peptide in the same ventricular granules. These findings suggest that brain natriuretic peptide is expressed in ventricular myocytes in response to hemodynamic stress in dilated cardiomyopathy. Brain natriuretic peptide may be, at least in part, synthesized simultaneously and secreted together with atrial natriuretic peptide by granules from failing ventricles, although the secretory turnover is different between these two peptides.

The chimeric gene linked to glucocorticoid-suppressible hyperaldosteronism encodes a fused P-450 protein possessing aldosterone synthase activity

Glucocorticoid-suppressible hyperaldosteronism (GSH) is one variety of primary aldosteronism with hypertension and is inherited in an autosomal dominant mode. A recent report has indicated that GSH is caused by a gene duplication arising from unequal crossing over between the two genes, CYP11B1 and CYP11B2, encoding P-450(11 beta) and P-450C18, respectively (Lifton et al. Nature (1992) 355, 262-265). The nucleotide sequence analysis in the present study has demonstrated that unequal crossing over in the chimeric gene formed by the gene duplication occurs within the region from the 3'-portion of exon 4 through the 5'-portion of intron 4 in Australian GSH patients. Namely, the chimeric gene encodes a fused P-450 protein consisting of the amino-terminal side of P-450(11 beta) (encoded by exons 1-4 of CYP11B1) and the carboxyl-terminal side of P-450C18 (encoded by exons 5-9 of CYP11B2). When a cDNA corresponding to the chimeric gene is transfected into COS-7 cells, the fused P-450 protein expressed in the mitochondria exhibits steroid 18-hydroxylase or aldosterone synthase activity. These results provide the molecular genetic basis for the characteristic biochemical phenotype of GSH patients.