`Printed in U.S.A.
`
`The Journal of Clinical Endocrinology & Metabolism 87(12):5714 –5721
`Copyright © 2002 by The Endocrine Society
`doi: 10.1210/jc.2001-011880
`
`Differential Inhibition of 17␣-Hydroxylase and 17,20-
`Lyase Activities by Three Novel Missense CYP17
`Mutations Identified in Patients with P450c17 Deficiency
`
`ERICA L. T. VAN DEN AKKER, JAN W. KOPER, ANNEMIE L. M. BOEHMER, AXEL P. N. THEMMEN,
`MIRIAM VERHOEF-POST, MARIANNA A. TIMMERMAN, BARTO J. OTTEN, STENVERT L. S. DROP,
`AND FRANK H. DE JONG
`
`Department of Pediatrics, Division of Endocrinology (E.L.T.v.d.A., A.L.M.B., S.L.S.D.), and Department of Internal
`Medicine, Section of Endocrinology (J.W.K., A.P.N.T., M.V.-P., M.A.T., F.H.D.J.), Erasmus Medical Center, 3000 DR
`Rotterdam, The Netherlands; and Department of Pediatrics, Division of Endocrinology, Radboud University Hospital
`(B.J.O.), 6500 HB Nijmegen, The Netherlands
`
`The microsomal enzyme cytochrome P450c17 is an important
`regulator of steroidogenesis. The enzyme has two functions:
`17␣-hydroxylase and 17,20-lyase activities. These functions
`determine the ability of adrenal glands and gonads to syn-
`thesize 17␣-hydroxylated glucocorticoids (17␣-hydroxylase
`activity) and/or sex steroids (17,20-lyase activity). Both en-
`zyme functions depend on correct steroid binding, but it was
`recently shown that isolated lyase deficiency can also be
`caused by mutations located in the redox partner interaction
`domain. In this article we present the clinical history and
`molecular analysis of two patients with combined 17␣-hydrox-
`ylase/17,20-lyase deficiency and four patients with isolated
`17,20-lyase deficiency. In these six patients, four missense
`CYP17 mutations were identified. Two mutations were lo-
`cated in the steroid-binding domain (F114V and D116V), and
`the other two mutations were found in the redox partner
`
`interaction domain (R347C and R347H). We investigated the
`activity of these mutated proteins by transfection experi-
`ments in COS-1 cells using pregnenolone, progesterone, or
`their hydroxylated products as a substrate and measuring
`17␣-hydroxylase- and 17,20-lyase-dependent metabolites in
`the medium. The mutations in the steroid-binding domain
`(F114V and D116V) of P450c17 caused combined, complete
`(F114V), or partial (D116V) 17␣-hydroxylase and 17,20-lyase
`deficiencies, whereas mutations in the redox partner inter-
`action domain (R347C and R347H) displayed less severe 17␣-
`hydroxylase deficiency, but complete 17,20-lyase deficiency.
`These findings are consistent with the clinical data and sup-
`port the observation that the redox partner interaction do-
`main is essential for normal 17,20-lyase function of P450c17.
`(J Clin Endocrinol Metab 87: 5714 –5721, 2002)
`
`IN THE STEROIDOGENIC pathway, cholesterol is con-
`
`verted into pregnenolone, which subsequently can be
`processed to either mineralocorticoids (no 17␣-hydroxyla-
`in adrenal
`tion) or glucocorticoids (17␣-hydroxylation)
`glands or to sex steroids in adrenals and gonads (17␣-
`hydroxylation and 17,20-lyase activity). The microsomal en-
`zyme cytochrome P450c17 is an important switchpoint in this
`steroidogenic pathway because it has both 17␣-hydroxylase
`and 17,20-lyase activities (Fig. 1). The first step necessary for
`P450c17 enzyme activity is steroid binding; then electron
`transfer occurs, enhanced by oxidoreductase to catalyze the
`hydroxylase reaction. The lyase activity is dependent on
`facilitation of the interaction of oxidoreductase with the re-
`dox partner-binding site (1). This interaction is enhanced by
`cytochrome b5 (2) or phosphorylation of phosphoserine res-
`idues (3). Optimal functioning of the redox partner-binding
`site is especially essential for the lyase reaction.
`The P450c17 enzyme is encoded by the CYP17 gene, which
`is located on chromosome 10q24.3 (4). CYP17 gene mutations
`are known to cause either complete or partial, combined, or
`isolated 17␣-hydroxylase/17,20-lyase enzyme deficiencies.
`The study of these mutant enzymes found in patients with
`17␣-hydroxylase/17,20-lyase enzyme deficiencies can help
`
`Abbreviations: AIS, Androgen insensitivity syndrome; DHEA, de-
`hydroepiandrosterone; hCG, human chorionic gonadotropin; Vmax,
`maximum velocity.
`
`us to understand the factors involved in P450c17 enzyme
`function. Until now 15 single base pair CYP17 gene muta-
`tions have been found, causing combined 17␣-hydroxylase
`and 17,20-lyase deficiencies (Table 1) (5–19). Only 2 muta-
`tions identified in patients with isolated complete 17,20-lyase
`deficiency were examined in vitro (Table 1) (14). These 2
`mutations (R347H and R358Q) are located in the redox part-
`ner interaction domain. The observation of differential re-
`sidual enzyme activity in naturally occurring mutations in
`various regions of the CYP17 gene supports the hypothesis
`that 17,20-lyase activity depends on normal
`function
`of the redox partner interaction site of the P450c17 enzyme
`(Fig. 1) (20).
`In this article we present the clinical and molecular data of
`six patients with 17,20-lyase deficiency. In the CYP17 gene of
`these patients four different missense mutations were iden-
`tified. Two mutations were located in the steroid-binding
`domain and two in the redox partner interaction domain. The
`effects of these four mutations on the enzymatic activity of
`the protein were examined by in vitro expression studies.
`
`Subjects and Methods
`
`Patients
`
`The clinical picture and hormone levels of patient 1 have been de-
`scribed previously (21). This 17-yr-old female patient was referred be-
`Amerigen Exhibit 1155
`Amerigen v. Janssen IPR2016-00286
`
`5714
`
`The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 11 January 2017. at 21:06 For personal use only. No other uses without permission. . All rights reserved.
`
`
`
`van den Akker et al. • 17,20-Lyase Deficiency
`
`J Clin Endocrinol Metab, December 2002, 87(12):5714 –5721 5715
`
`FIG. 1. Model of the mechanism of action of micro-
`somal P450c17 and its differential regulation of 17-
`hydroxylase and 17,20-lyase activities. OR, Oxi-
`doreductase; b5, cytochrome b5; e⫺, electron
`transfer [adapted from Auchus et al. (2)].
`
`TABLE 1. All single base pair missense mutations in the CYP17 gene described until now, their location, and their residual enzyme
`activity studied in vitro
`
`Mutation
`
`R35L
`Y64S
`G90N
`F93C
`R96W
`S106P
`F114V
`D116V
`N177D
`P342T
`R347H
`
`R347C
`R358Q
`H373L
`P409R
`R415C
`F417C
`R440H
`R496C
`R496H
`
`17␣-Hydroxylase
`(%)
`
`17,20-Lyase
`(%)
`
`Site of mutationa
`
`Reference
`
`38
`15
`⬍1
`11
`25
`⬍1
`2.2
`37.7
`10
`20
`65
`44.1
`13.6
`65
`⬍1
`⬍1
`8
`⬍1
`⬍1
`⬍10
`38
`
`33
`
`⬍1
`10
`25
`⬍1
`⬍1
`10.7
`10
`20
`⬍5%
`⬍1
`⬍1
`⬍5
`⬍1
`⬍1
`10
`⬍1
`⬍1
`⬍10
`33
`
`Membrane
`Membrane
`
`Steroid
`
`Steroid
`Steroid
`Steroid
`Steroid
`
`Redox
`
`Redox
`Redox
`Heme
`Steroid
`Heme(?)
`Heme
`Heme
`Heme
`Heme
`
`15
`8
`11
`19
`12
`5
`This article
`This article
`15
`6
`14
`This article
`This article
`14
`9
`16
`18
`13 and 14
`10
`7
`15
`
`a Membrane, Membrane attachment domain; Steroid, steroid binding domain; Redox, redox partner interaction domain; Heme, heme binding
`domain; and Heme(?), uncertainty of affected domain (28).
`
`cause of primary amenorrhea and lack of secondary sexual development
`(Tanner stage M1P1). She had one sister with normal secondary sexual
`development. The patient had female external genitalia, but no uterus
`and a 46,XY karyotype. Based on these results, the presumptive diag-
`nosis of androgen insensitivity syndrome (AIS) was made. Secondary
`sexual development began after the start of estrogen substitution ther-
`apy. The patient underwent a hormonal evaluation before undergoing
`bilateral gonadectomy at the age of 20 yr. With the exception of fatigue,
`she had no complaints. Her blood pressure was elevated (150/110 mm
`Hg) despite low renin levels. Basal levels of androgens were low, and
`progesterone levels were high; basal cortisol was within the normal
`range, but rose insufficiently after ACTH (Table 2). These findings led
`to the diagnosis of combined 17␣-hydroxylase/17,20-lyase deficiency.
`Results of in vitro studies of testis tissue of this patient, using preg-
`nenolone,
`17-hydroxypregnenolone,
`and
`dehydroepiandrosterone
`(DHEA) as substrate, followed by measurement of metabolites confirmed
`the absence of 17␣-hydroxylase and 17,20-lyase activities (21).
`Patient 2 is a 46,XY individual, who was born with an enlarged clitoris
`and no uterus. Her gonads were removed when she was 3 yr old. Based
`
`on the combination of ambiguous genitalia, lack of pubic hair at puberty,
`low-normal levels of androgens, and cortisol (Table 2), combined partial
`17␣-hydroxylase/17,20-lyase deficiency was diagnosed.
`Patient 3 was born with complete female external genitalia and raised
`as a girl. She was the first child of consanguineous parents. The family
`history did not reveal any sexual differentiation disorders. At the age of
`2 months, she had bilateral inguinal hernias, which contained testes, and
`subsequently, her karyotype was shown to be 46,XY. The presumptive
`diagnosis of AIS was made, and gonadectomy was performed. She was
`reevaluated at the age of 10 yr. Her blood pressure was 140/80 mm Hg
`despite normal renin levels. She underwent uneventful surgery twice.
`Basal levels of androgens were low, and progesterone was elevated. An
`ACTH test showed a low-normal basal level of cortisol that did not
`respond to ACTH (Table 3), suggesting partial 17␣-hydroxylase defi-
`ciency with complete 17,20-lyase deficiency.
`Patient 4, raised as a girl, was evaluated at age 14 yr because of
`delayed puberty. She had a 46,XY karyotype and complete female ex-
`ternal genitalia with an absent uterus. The presumptive diagnosis of AIS
`was made, and she underwent gonadectomy. At age 28 yr she was
`
`The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 11 January 2017. at 21:06 For personal use only. No other uses without permission. . All rights reserved.
`
`
`
`5716 J Clin Endocrinol Metab, December 2002, 87(12):5714 –5721
`
`van den Akker et al. • 17,20-Lyase Deficiency
`
`TABLE 2. Serum hormone concentrations of patients 1 and 2 with CYP17 mutations in the steroid binding domain
`
`Hormone
`
`LH (U/liter)
`FSH (U/liter)
`Progesterone (nmol/liter)
`17OH-progesterone (nmol/liter)
`Cortisol (nmol/liter)
`DHEA (nmol/liter)
`Androstenedione (nmol/liter)
`Testosterone (nmol/liter)
`Renin (pg/ml)
`
`Patient 1 (F114V)a
`
`Basal
`
`107.4 (1.5–9.4)
`38.4 (2.6 –7.4)
`14.5 (0.5–2)
`0.29 (⬍10)
`220 (200 – 800)
`⬍0.1 (3.5–25)
`⬍0.1 (2–10)
`⬍0.2 (0.5–3)
`⬍0.1 (0 –2.5)
`
`ACTH
`
`14.9
`0.31
`240 (⬎500)
`⬍0.1
`⬍0.1
`⬍0.2
`⬍0.1
`
`Patient 2 (D116V)a
`
`Basal
`
`3.2 (0.5–2)
`0.1 (⬍10)
`248 (200 – 800)
`3.6 (3.5–25)
`1.67 (2–10)
`0.3 (0.5–3)
`
`Normal age-adjusted reference values are shown in parentheses. In the ACTH test, venous catheters were inserted, and serum levels of several
`hormones were determined in blood samples taken before and at 60 min after the iv administration of 0.25 mg ACTH.
`a Age: patient 1, 17 yr; patient 2, 27 yr.
`
`TABLE 3. Serum hormone concentrations of patients 3 and 4 with CYP17 mutation R347C in the redox partner interaction domain
`
`Hormone
`
`LH (U/liter)
`FSH (U/liter)
`Progesterone (nmol/liter)
`17OH-progesterone (nmol/liter)
`Cortisol (nmol/liter)
`DHEA (nmol/liter)
`Androstenedione (nmol/liter)
`Testosterone (nmol/liter)
`Renin (pg/ml)
`
`Patient 3 (R347C)a
`
`Patient 4 (R347C)a
`
`Basal
`
`ACTH
`
`Basal
`
`ACTH
`
`6.4 (⬍1)
`41.5 (⬍1)
`7.4 (0.5–2.0)
`2.4 (0.4 –2.1)
`216 (200 – 800)
`0.3 (0.4 – 4.9)
`0.11 (0.24 – 0.8)
`0.1 (⬍1)
`12.5 (60 –300)
`
`5.4
`46.5
`8.4
`2.6 (3.5– 6.0)
`228 (⬎500)
`0.5 (1.2–9.4)
`0.13 (0.4 –1.6)
`0.1
`
`10.6 (1.5– 8)
`37.3 (2–7)
`6.1 (0.5–2)
`1.9 (⬍10)
`232 (200 – 800)
`0.0 (3.5–25)
`0.57 (2–10)
`0.1 (0.5–3.0)
`10.2 (60 –300)
`
`12.2
`2.4
`285 (⬎500)
`0.0
`0.51
`
`Normal age-adjusted reference values are shown in parentheses. The ACTH test is described in the legend to Table 2.
`a Age: patient 3, 10 yr; patient 4, 28 yr.
`
`reevaluated. Her blood pressure was 170/90 mm Hg despite normal
`renin levels. She underwent surgery several times without complica-
`tions. She developed breasts on estrogen substitution. Her main com-
`plaint was the complete absence of pubic hair, which developed with
`testosterone propionate ointment therapy. Basal levels of androgens
`were extremely low with elevated progesterone. She had low-normal
`basal levels of cortisol, which did not rise after ACTH stimulation (Table
`3). These data are consistent with partial 17␣-hydroxylase deficiency and
`complete 17,20-lyase deficiency.
`Patients 5 and 6 are siblings from consanguineous parents. Both 46,XY
`siblings were born with ambiguous external genitalia (Prader stage III).
`Patient 5 was assigned the male sex, and patient 6 was assigned the
`female sex on the basis of the sex designated by the parents at birth.
`There were no clinical signs of insufficient cortisol secretion. Both sib-
`lings underwent surgery several times without complications. Their
`basal serum levels of androgens were low and rose insufficiently after
`human chorionic gonadotropin (hCG) stimulation. Basal plasma renin
`activity was normal. Progesterone and 17-hydroxyprogesterone levels
`were high after hCG stimulation, and basal serum levels of cortisol were
`normal with some, but insufficient, rise during ACTH treatment (Table
`4). On the basis of these observations the diagnosis of isolated 17,20-lyase
`deficiency was made.
`
`Mutation analysis of the CYP17 gene
`
`Genomic DNA was isolated from leukocytes according to standard
`procedures (22). Exons 1– 8 and their flanking intron sequences of the
`CYP17 gene were amplified individually by PCR using the primers and
`PCR conditions described by Monno et al. (5), followed by single strand
`conformation polymorphism analysis (23) and sequencing of the frag-
`ments that showed abnormal single strand conformation polymorphism
`patterns. To determine whether two mutations identified in the same
`patient were on separate alleles, allele-specific amplification was carried
`out (patient 3), or DNA of the parents was sequenced (patients 1, 2,
`and 4).
`
`Construction of mutant expression plasmids
`
`Mutant CYP17 expression plasmids were constructed using the con-
`ditions described previously for the LH receptor (24). pcDNA3 was used
`as the expression vector. For the exchange of fragments containing the
`mutation in the wild-type CYP17 expression vector, flanking primers
`were used: T7 forward, AATACGACTCACTATAG; and 638 reverse,
`CTGTATGACATTCAACTC for the F114V and D116V mutants; and 670
`forward, GCAAAGACAGCCTGGTGGACC; and SP6 reverse, CTAT-
`AGTGTCACCTAAAT for the R347C and R347H mutations. The frag-
`ments were digested with the restriction enzymes BamHI and BstEII for
`the F114V and D116V mutants and BspEI and XhoI for the R347C and
`R347H mutations, respectively, and subsequently ligated into the ex-
`pression vector that had been digested with the same enzymes. Primers
`that carry the mutation were as follows: F114V, GGGTATCGC-
`CGTCGCTGACTCTG; D116V, CGCCTTCGCTGTCTCCGGAGCA-
`CACTGG; R347C, CAGTGACTGTAATCGATTGCTCCTGCTG; and
`R347C, CCAACTATCAGTGATCATAACCGTCTC and their reverse
`complements.
`
`Culture and transfection of cells
`
`COS-1 cells were grown in 24-well plates to 50% confluence and
`transfected with 0.4 g/well (four wells per plasmid) of pcDNA ex-
`pression plasmid containing wild-type or mutant CYP17. Transfection
`efficiency was monitored by cotransfection with a -galactosidase ex-
`pression plasmid. The transfected cells were washed and incubated in
`fresh medium. After 40 h when the COS-1 cells were 80 –90% confluent,
`1 m pregnenolone, progesterone, 17-hydroxypregenenolone, or 17-
`hydroxyprogesterone was added to the medium. After 8 h, i.e. during
`the period of linear steroid production against time (data not shown), the
`medium was removed and assayed for products using RIAs for 17-
`hydroxypregnenolone (DRG Diagnostics, Marburg, Germany) and
`DHEA or 17-hydroxyprogesterone and androstenedione (Diagnostic
`Products, Los Angeles, CA). All transfections were performed at least
`twice.
`
`The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 11 January 2017. at 21:06 For personal use only. No other uses without permission. . All rights reserved.
`
`
`
`van den Akker et al. • 17,20-Lyase Deficiency
`
`J Clin Endocrinol Metab, December 2002, 87(12):5714 –5721 5717
`
`In separate experiments, Km and maximum velocity (Vmax) were
`measured for the 17␣-hydroxylase and 17,20-lyase activities of the en-
`zyme by culturing the cells in the presence of 0.1, 0.2, 0.5, 1, or 2 m
`pregnenolone or 17-hydroxypregnenolone, respectively, and measuring
`the concentration of the above-mentioned products using the same
`assays. Results were obtained with three or four wells of transfected cells
`per dose and were corrected for the concentrations measured in the
`medium of cells transfected with the empty vector.
`
`Mutations
`
`Results
`
`The mutations identified in the CYP17 gene of our six
`patients are shown in Table 5. Mutation analysis of patients
`1, 2, and 4 revealed compound heterozygosity for three novel
`mutations: F114V (TTC3 GTC), D116V (GAC3 GTC), and
`R347C (CGT3 TGT), combined with a frameshift mutation
`on the other allele: a 4-base duplication near codon 480. This
`duplication has previously been observed in several families
`of Dutch and German Mennonite descent and is known to
`completely abolish all P450c17 enzyme activity (25). The
`novel mutations identified in patients 1 and 2 (F114V and
`D116V) are in the steroid-binding domain. Patient 3 was a
`compound heterozygote for two new mutations: R347C
`(CGT3 TGT) and a 25-bp deletion in exon 1 (nucleotides
`204 –228 of the coding sequence, counting from the A of the
`start codon). This deletion renders the message out of frame,
`resulting in a premature stop codon located at 28 residues
`after the deletion. We therefore expect that this deletion in
`allele 2 does not lead to the production of any functional
`protein. Patient 4 showed the same R347C mutation in one
`allele, whereas the 4-bp dup exon 8 mutation that was also
`present in patients 1 and 2 was present in the other. The
`mutation R347C is located in the redox partner interaction
`domain. Localization of two different mutations identified in
`the CYP17 gene in patients 1– 4 on separate alleles was con-
`firmed by determining that each of their parents carried only
`one of these mutations (patients 1, 2, and 4) or by allele-
`specific amplification (patient 3). Finally, patients 5 and 6
`(siblings) were homozygous for the R347H (CGT3 CAT)
`mutation that has been described previously (14).
`
`Expression of mutant proteins in COS-1 cells
`
`The 17␣-hydroxylase and 17,20-lyase activities of the mu-
`tated proteins were estimated using transient transfection in
`COS-1 cells and compared with those of the wild-type en-
`zyme. The conversion of various concentrations of preg-
`nenolone to 17-hydroxypregnenolone and DHEA was used
`as a measure for 17␣-hydroxylase activity. Results are shown
`in Fig. 2A. Lineweaver-Burk plots calculated from these data
`are shown in Fig. 3. The r values for the regression lines were
`
`TABLE 5. Mutations identified in the CYP17 gene of six patients
`
`Patient
`
`Allele 1
`
`Site of mutationa
`
`Allele 2
`
`F114V
`1
`D116V
`2
`R347C
`3
`R347C
`4
`5 and 6 R347H
`
`Steroid
`Steroid
`Redox partner
`Redox partner
`Redox partner
`
`4-bp duplication exon 8
`4-bp duplication exon 8
`25-bp deletion exon 1
`4-bp duplication exon 8
`R347H
`
`a Steroid indicates steroid binding domain; redox partner indicates
`redox partner interaction domain (28).
`
`Normalage-adjustedreferencevaluesareshowninparentheses.TheACTHtestisdescribedinthelegendtoTable2.InthehCGtest,serumhormonelevelsweredetermined
`
`inbloodsamplestakenbeforeand72haftertheimadministrationof1500IUhCG.
`
`0.09(0.07–0.21)
`0.05(0.4–2.4)
`0.3(0.7–3.3)
`
`427(⬎500)
`
`2.1(2.0–8.1)
`
`11.0
`
`0.07(0.2–1.7)
`0.1(0.3–1.5)
`243(157–690)
`1.3(0.1–3.5)
`2.6(0.5–2)
`
`2.6(6.8–26)0.01(⬍0.5)
`0.4
`1.4
`
`150
`
`19.6
`
`ACTH
`
`Basal
`
`hCG
`
`1.2(0.03–26)
`0.3(0.9–15)
`2.0(0.9–8.2)
`270(120–1158)
`13.4(0.33–5.2)
`
`4.7(0–3.9)
`6.0(2–26)
`
`Basal
`
`0.05
`0.09(0.4–1.6)
`0.44(1.2–9.4)
`
`385(⬎500)
`
`1.5(3.5–6)
`6.1
`
`4.1(6.8–26)
`0.15(1.0–4.0)0.10(0.7–3.8)
`0.7(0.4–4.9)
`0.35
`190(157–414)
`1.3(0.4–2.1)
`2.7(0.5–2.0)
`
`187
`
`14.9
`
`0.02(⬍0.5)
`
`0.7(0.03–26)
`0.14(0.9–1.8)
`0.25(0.9–8.2)
`194(120–1158)
`4.5(0.03–0.2)
`9.6(0.5–2)
`2.9(0–3.9)
`4.6(2–12)
`
`Testosterone(nmol/liter)
`Androstenedione(nmol/liter)
`DHEA(nmol/liter)
`Cortisol(nmol/liter)
`17OH-progesterone(nmol/liter)
`Progesterone(nmol/liter)
`FSH(U/liter)
`LH(U/liter)
`
`ACTH
`
`Basal
`
`hCG
`
`Basal
`
`2.5yrold
`
`2wkold
`
`7.5yrold
`
`2monthsold
`
`Hormone
`
`Patient6(R347H)
`
`Patient5(R347H)
`
`TABLE4.Serumhormoneconcentrationsofpatients5and6withCYP17mutationR347Hintheredoxpartnerinteractiondomain
`
`The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 11 January 2017. at 21:06 For personal use only. No other uses without permission. . All rights reserved.
`
`
`
`5718 J Clin Endocrinol Metab, December 2002, 87(12):5714 –5721
`
`van den Akker et al. • 17,20-Lyase Deficiency
`
`FIG. 2. Production of 17-hydroxypregnenolone (17OH-Preg) and
`DHEA after 8-h culture of transfected COS-1 cells in the presence of
`various concentrations of steroid precursors as a measure of 17␣-
`hydroxylase activity (A; culture with pregnenolone) and of 17,20-lyase
`activity (B; culture with 17-hydroxypregnenolone). The cells were
`transiently transfected with wild-type CYP17 and the CYP17 mu-
`tants. Substrate concentrations were 0.1, 0.2, 0.5, 1, or 2 M. Data are
`the mean ⫾ SEM (n ⫽ 3 in A and n ⫽ 4 in B).
`
`all above 0.95. Resulting apparent values for Km and Vmax
`have been summarized in Table 6. All Vmax values for the
`mutated proteins were lower than that for the wild-type
`protein, whereas Km values of the proteins with mutations in
`the steroid-binding domain were comparable to that of the
`wild-type protein, and the Km values for the other two pro-
`teins were lower. As an example for the separate amounts of
`17-hydroxypregnenolone and DHEA produced, the data ob-
`tained with 1 m pregnenolone are plotted in Fig. 4. The use of
`a logarithmic axis allows a more clear indication of the amounts
`of DHEA. The wild-type enzyme and the F114V and D116V
`mutants converted between 10–20% of 17-hydroxypreg-
`nenolone to DHEA, whereas the R347C and R347H mutants did
`not produce measurable concentrations of DHEA despite the
`production of 17-hydroxypregnenolone.
`Similar results were obtained using 17-hydroxypreg-
`nenolone as the substrate (Fig. 2B). Significant amounts of
`DHEA were only produced by the wild-type enzyme and the
`D116V mutant. The apparent Vmax and Km values for the
`wild-type enzyme were 1.55 min⫺1 and 0.35 m, respectively.
`The production of DHEA by the mutated enzymes was too
`low to calculate Vmax and Km for the 17-hydroxypreg-
`nenolone to DHEA conversion by these enzymes. Conver-
`sion percentages at
`the highest concentration of 17-
`hydroxypregnenolone (2 m) were 31.3%, 0.15%, 4.8%,
`0.03%, and 0.12% for wild-type, F114V, D116V, R347C, and
`R347H, respectively.
`
`FIG. 3. Lineweaver-Burk plot for the conversion of pregnenolone to
`17-hydroxypregnenolone plus DHEA, shown in Fig. 2A, by wild-type
`CYP17 (F), and the D116V (‚), R347C (E), and R347H (Œ) mutants.
`Results for the F114V mutant are not shown, because all measured
`values are outside the axes used.
`
`TABLE 6. Enzyme kinetic data on the 17␣-hydroxylase activity of
`the CYP17 mutants investigated
`
`Wild type
`
`F114V
`
`D116V
`
`R347C
`
`R347H
`
`Km (M)
`Vmax (min⫺1)
`
`0.97
`3.0
`
`0.76
`0.04
`
`0.95
`0.56
`
`0.08
`0.10
`
`0.27
`0.50
`
`FIG. 4. Separate productions of 17-hydroxypregnenolone (f) and
`DHEA (䡺) from 1 M pregnenolone by the COS-1 cells described in
`Fig. 2A. Note the logarithmic y-axis.
`
`When progesterone was used as a substrate for testing the
`⌬4-steroid biosynthetic pathway,
`the production of 17-
`hydroxyprogesterone,
`indicative of hydroxylase activity,
`showed a similar pattern as
`the production of 17-
`hydroxypregnenolone from pregnenolone, but at lower lev-
`els (Fig. 5). With neither progesterone nor 17-hydroxypro-
`gesterone (results not shown) as substrate did we observe
`significant production of androstenedione by the mutant
`
`The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 11 January 2017. at 21:06 For personal use only. No other uses without permission. . All rights reserved.
`
`
`
`van den Akker et al. • 17,20-Lyase Deficiency
`
`J Clin Endocrinol Metab, December 2002, 87(12):5714 –5721 5719
`
`experiments. Neither of these mutations has been described
`before. Like these two mutations, all other known mutations
`in the steroid-binding domain cause defects that affect 17␣-
`hydroxylase and 17,20-lyase activities to a similar extent.
`With regard to the patients with mutations in the redox
`partner interaction domain, special attention should be paid
`to the genotypic and phenotypic differences. Complete 17,20-
`lyase deficiency results in a complete female phenotype in
`46,XY individuals (patients 3 and 4), but is theoretically not
`expected to result in ambiguous genitalia (patients 5 and 6).
`Patients 5 and 6 are, just like one of the previously described
`patients (10), homozygous for the R347H mutation and sim-
`ilarly had ambiguous genitalia at birth, indicating slight re-
`sidual capacity for androgen synthesis. In contrast to this,
`patients 3 and 4 were heterozygous for another mutation in
`the same codon (R347C), the other allele being completely
`inactive due to a frameshift mutation. These two patients
`(both with an 46,XY genotype) showed female genitalia at
`birth and had barely detectable androgen levels. The differ-
`ences between patients 5 and 6, on the one hand, and patients
`3 and 4, on the other, indicate that the R347C mutation is
`more deleterious for the 17␣-hydroxylase and/or 17,20-lyase
`reaction than the R347H mutation, as is also shown by the
`results of the transfection experiments. To explain the large
`difference in 17␣-hydroxylase activity of mutations R347C
`and R347H, we hypothesize that the change of arginine to
`cysteine disrupts the function of the whole protein more
`seriously than a change to histidine, because of the possibility
`of the formation of abnormal cysteine dimers, causing not
`only disruption of 17,20-lyase activity, but also of 17␣-
`hydroxylase. In addition, the functional allele in patients 3
`and 4 may be haplo-insufficient and result in residual en-
`zyme activity that is too low to stimulate male genital dif-
`ferentiation, whereas the homozygous presence of the R347H
`mutation may cause the residual presence of sufficient 17,20-
`lyase activity.
`The partially virilized genitalia of patients 5 and 6 and the
`ability to synthesize some testosterone in the hCG test are not
`in accordance with our finding of absence of lyase activity in
`vitro. As described previously, R347 and R358, which are
`located in the redox partner interaction domain and contrib-
`ute to the positive charges on the proximal surface of
`P450c17, are known to be key residues involved in the in-
`teraction with redox partner proteins. Geller et al. (20) re-
`ported that the absence of these charged amino acids selec-
`tively impairs 17,20-lyase activity without substantial
`reductions in 17␣-hydroxylase activity or 17-hydroxy-
`pregnenolone binding. Coexpression of the R347 and R358
`mutants with P450 oxidoreductase did not result in a sig-
`nificant increase in 17,20-lyase activity, but addition of excess
`cytochrome b5 partially restored 17,20-lyase activity (20).
`Thus, the ambiguous genitalia in R347H homozygous pa-
`tients might be explained by a partial rescue of the R347H
`mutation through in vivo accumulation of cytochrome b5 in
`these patients.
`To obtain a more detailed view of the P450c17 enzyme
`function in our patients, we compared the clinical data with
`in vitro expression studies. The Km values obtained for the
`wild-type enzyme are in line with reported values (18, 27),
`and the R347H mutation yields a lower value as described
`
`FIG. 5. Relative production of 17-hydroxyprogesterone as a measure
`of 17␣-hydroxylase activity (䡺) and androstenedione as a measure of
`17,20-lyase activity (o) in COS-1 cells that were transiently trans-
`fected with wild-type (wt) CYP17 or the CYP17 mutants and cultured
`with 1 M progesterone as substrate. For CYP17 wild-type transfected
`cells, the 17-hydroxyprogesterone concentration was 760 nmol/liter,
`and the androstenedione concentration was 33 nmol/liter. Data are
`the mean ⫾ SEM (n ⫽ 4).
`
`proteins, indicating the absence of 17,20-lyase activity under
`these conditions.
`
`Discussion
`
`Combined 17␣-hydroxylase/17,20-lyase deficiency (pa-
`tients 1 and 2) is a well defined disorder (26). Isolated 17,20-
`lyase deficiency (patients 3– 6) with subnormal 17␣-hydrox-
`ylase function leading to a female phenotype or ambiguous
`genitalia, absence of sexual development, and normal basal
`cortisol levels without hypertension is extremely rare, with
`few clinical data. Isolated 17,20-lyase deficiency can be dif-
`ferentiated from combined 17␣-hydroxylase/17,20-lyase
`deficiency by an ACTH test, showing elevated progesterone
`in 17␣-hydroxylase-deficient patients and elevated 17-
`hydroxyprogesterone with subnormal rise of cortisol in pa-
`tients with isolated 17,20-lyase deficiency. All of our patients
`had normal basal cortisol serum levels, but insufficient re-
`sponse to stimulation by ACTH. No guidelines are available
`for the hydrocortisone substitution treatment in these pa-
`tients. In our view, daily hydrocortisone substitution treat-
`ment is not needed when there are no complaints, but in case
`of stress, a hydrocortisone stress dose is advised. Neverthe-
`less, some of our patients underwent uneventful surgery
`without a hydrocortisone stress scheme.
`In four of our six patients (patients 1– 4), one allele of the
`CYP17 gene was completely inactive due to insertions or
`deletions resulting in a frameshift. Patient 1 had a missense
`mutation in the steroid-binding domain in the other allele:
`F114V. This resulted in a combined complete 17␣-hydrox-
`ylase/17,20-lyase deficiency, with a complete female phe-
`notype. Patient 2 also had a mutation in the steroid-binding
`domain in the second allele (D116V), but had milder com-
`bined 17␣-hydroxylase/17,20-lyase deficiency with low, but
`measurable, levels of androgens, explaining her ambiguous
`genitalia. The F114V mutation (patient 1) appears to affect
`steroid binding more seriously than the D116V mutation
`(patient 2) both in vivo as well as in the in vitro transfection
`
`The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 11 January 2017. at 21:06 For personal use only. No other uses without permission. . All rights reserved.
`
`
`
`5720 J Clin Endocrinol Metab, December 2002, 87(12):5714 –5721
`
`van den Akker et al. • 17,20-Lyase Deficiency
`
`previously (20); these researchers also described an approx-
`imately 3-fold reduction of the Vmax for this mutant enzyme.
`The enzyme activities of the mutated proteins in vitro were
`consistent with the clinical data and the hypothesis that
`mutations in the steroid-binding domain result in combined
`complete or partial 17␣-hydroxylase and 17,20-lyase defi-
`ciency, whereas mutations in the redox partner interaction
`domain result in isolated 17,20-lyase deficiency. The latter
`point becomes especially clear from the results of the con-
`version of pregnenolone to DHEA; although the production
`of 17-hydroxypregnenolone by the cells transfected with the
`R347C and R347H mutated CYP17 was larger than or com-
`parable to that of the F114V and D116V cells, the amount of
`DHEA produced by the former two cell types was much
`lower than that secreted by the latter.
`The in vitro results of the present study were obtained
`using the conversion of nonradioactive precursors to prod-
`ucts, which were measured by RIA. Using this type of de-
`tection, Biason-Lauber et al. (13) showed that the F417C mu-
`tation in P450c17 can lead to isolated 17,20-lyase deficiency;
`these researchers used one dose of pregnenolone or proges-
`terone as substrate. In contrast, using the conversion of tri-
`tiated precursors, Gupta et al. (17) indicated that this muta-
`tion affects both the 17␣-hydroxylase and 17,20-lyase
`activities of the enzyme and argued that accurate enzymic
`studies of the mutant proteins should be performed to be able
`to conclude that isolated lyase deficiency is present. The
`latter researchers explained the discrepancies between the
`two studies on the basis of differences in the methods used
`to measure the products of the enzymic conversions; RIAs
`might lack the specificity needed to obtain reliable results. As
`we used RIAs of nonradioactive steroids for the detection of
`enzyme deficiencies of the mutated proteins, similar objec-
`tions might be raised against our conclusions. However, it is
`highly unlikely that insufficient specificity of our methods
`plays a role, because we obtained straight lines in the
`Lin