WCK1090
`Wockhardt Bio AG v. Janssen Oncology, Inc.
`IPR2016-01582
`
`1
`
`

`

`
`
`
`
`
`
`
`
`
`
`HmmmmtSwamimama:BASESmmwmmfiEmmaS
`
`
`
`
`
`6memtmouodto230$mmm._.939m
`
`onohfimfimouoémfinA||_
`HmavHomflouummlém
`
`
`3:5.meA|onohowmowmorwAJIEgonoumewaa
`Sengnmmm
`
`
`2.8380330m
`
`90:59:%EsomsfiooflomHA.........
`
`
`
`Enema$2.3m
`
`
`
`
`
`.5235cam:wm
`
`nofinouom
`
`qunmawm
`
`uu.uuuo-ooo...o..ou..-
`
`oomcmvm
`
`3.138150
`
`
`
`enouoamoEQAllDOEA|||||onohowmomemAll]mafia—Hunwam
`£28meRuemvm593%
`
`102
`
`:3va
`
`
`
`
`
`
`
`cm‘8-Mfl04%EmFSQOSoQ3AllenououwowouméwozAlfimlmmmoncfiononwohmémou
`
`afichwmmA|oficwfiufioH—mOHfiH—dwAJI
`
`DwmmSagacmwm
`
`
`
`AweAna
`
`
`
`bHocmvmhHucmvm
`
`
`
`
`
`2
`
`
`
`
`
`
`
`
`

`

`MANAGEMENT OF HUMAN DEFICIENCIES OF P450c17
`
`103
`
`lone—to—dehydroepiandrosterone (DHEA) reaction than for the 170L-hy-
`droxyprogesterone~to-androstenedione reaction.3/ 35 Although the rate of
`the lyase reaction can be increased more than 10~fold by the addition of
`cytochrome b5,3/ 31' 35 the A5 preference persists, and the lyase rate never
`quite achieves the rate of the hydroxylase reactions. In addition, human
`P450c17 160L—hydroxylates progesterone but not pregnenolonefi' 37' 62 In
`the presence of cytochrome b5, P450c17 converts approximately 10% of
`pregnenolone substrate to a A16 andiene product}5 which is also formed
`by porcine P450c17 and acts as a pheromone precursor in pigs.48 Al—
`though experiments to study the chemistry of P450c17 often require
`certain conditions, such as detergent solubilization that could be consid-
`ered nonphysiologic, the remarkable consistency of substrate preferences
`and kinetic constants observed for the modified solubilized P450c17
`
`expressed in Escherichia COZi,31’ 35 the native P450c17 expressed in yeast
`microsomes5 or intact COS-1 cells/"7'38 and that obtained from human
`tissues and cells3 62 strengthens these conclusions.
`One consequence of this A5 preference of human P450c17 for the
`17,20~lyase reaction is that the vast majority of sex steroids in humans
`derive from DHEA as an intermediate. This A5 preference also allows
`the phenomenon of adrenarche to occur in humans, an event that is
`characterized by a dramatic rise in adrenal DHEA production that occurs
`at about age 8 to 10 years}; 60 whereas cortisol production remains
`relatively constant. Adrenarche is an exemplary manifestation of the
`biochemistry of P450c17, in which the 17cc—hydroxylase and 17,20—lyase
`activities are differentially regulated. In fact, this dichotomy between
`adrenal 17cc-hydroxylase activity, reflected by relatively constant cortisol
`production, and 17,20-lyase activity, reflected by drastically age-depen-
`dent changes in DHEA production, previously suggested that distinct
`enzymes performed the two transformations; however, later copurifica-
`tion of the 170c—hydroxylase and 17,20—1yase activities of neonatal pig
`testes suggests otherwise/‘7 This controversy was settled when the cDNA
`for bovine P450c17 was expressed in COS-1 cells, conferring 170L—hydrox-
`ylase and 17,20—lyase activities to these nonsteroidogenic cells77 and
`proving genetically that the 170t~hydroxylase and 17,20-lyase enzymes
`
`
`
`Figure 1. Major steroidogenesis pathways in humans and feedback loops controlling
`glucocorticoid and mineralocorticoid production. Ordinarily, cortisol is the major glucocorti-
`coid produced by the adrenal zona fasciculata/reticularis, and cortisol exerts negative
`feedback inhibition (double vertical bars) to regulate pituitary adrenocorticotropic hormone
`(ACTH) production. Aldosterone is the principal mineralocorticoid of the adrenal zona
`glomerulosa, and aldosterone synthase (P450c11AS) expression is stimulated by volume
`depletion, which activates the renin-angiotensin (All) system, and to a lesser extent, by
`ACTH. Aldosterone acts to stimulate kaluresis and salt and water retention, which feeds
`back on the kidney to suppress renin production. The production of corticosterone, a weak
`glucocorticoid, and of 11-deoxycorticosterone (DOC), a potent mineralocorticoid, is relatively
`low and unimportant
`in healthy individuals with intact feedback systems. Note that
`P4500115 in the zona fasciculata also 18—hydroxylates (180H) D00 and corticosterone as
`minor products.
`
`3
`
`

`

`104
`
`AUCHUS
`
`were, in fact, both embodied in a single enzyme, P450c17. Differential
`regulation of the two principal activities of P450c17 is possible because
`the abundance of P450—oxidoreductase37 and the addition3lr35r 50 or coex—
`
`pression3 of cytochrome b5 preferentially augments the 17,20—lyase activ-
`ity, and phosphorylationz 76 also selectively enhances 17,20-lyase activity
`Recent data showing high expression of b5 in the zona reticularis of
`monkeys40 and humans71 suggest that the developmentally regulated
`expression of b5 might be a key event in the genesis of adrenarche in
`higher primates.
`
`PATHOPHYSIOLOGY
`
`P450c17 deficiencies are a form of congenital adrenal hyperplasia in
`which not only adrenal but also gonadal steroidogenesis is impaired. In
`humans, one gene for P450c17 is expressed in the adrenals and gonads11
`instead of two tissue-specific isozymes. A single 2.1—kb mRNA species
`yields a 57-kd protein in these tissues, and mutations in this gene
`produce a spectrum of deficiencies in 17—hydroxysteroids and C19 ste—
`roids. Loss of P450c17 in the adrenal gland impairs cortisol and DHEA
`production, whereas gonadal deficiency of P450c17 abrogates sex steroid
`production. The initial description of 17—hydroxylase deficiency was a
`case in which both 170t-hydroxylase and 17,20-lyase products were ab-
`sent.10 When the gene for human P450c17 was cloned,54 patients with
`17-hydroxylase deficiency were found to harbor mutations in the CYP17
`gene/1' 67 but molecular techniques and subsequent clinical evaluations
`failed to implicate CYP17 mutations as the cause of isolated 17,20—lyase
`deficiency.73 Recently, three cases of isolated 17,20-lyase deficiency have
`been confirmed by molecular genetics)"20 demonstrating that amino acid
`substitution mutations in P450c17 can cause an isolated loss of 17,20-
`lyase activity.
`
`Combined 17a-Hydroxylase/17, 20-Lyase Deficiency
`
`Loss of P450c17 in the human adrenal gland prohibits the biosynthe-
`sis of cortisol and C19 steroids. Curiously, the adrenal glands of patients
`with 17~hydroxylase deficiency are similar to those of rodents, which do
`not express P450c17f’3 such that rodents rely on corticosterone as their
`principal glucocorticoid, and their adrenal glands cannot make C19 ste—
`roids. Patients with 17~hydroxylase deficiency rarely26 manifest symp—
`toms of adrenal insufficiency owing to sustained corticosterone produc—
`tion. Because corticosterone is a weaker glucocorticoid than cortisol,
`abnormally high corticosterone production is necessary before feedback
`inhibition on pituitary corticotropin (ACTH) secretion occurs,45 establish—
`ing a new steady state (Fig. 2). To produce sufficient corticosterone to
`make up for the absence of cortisol, dramatically elevated quantities of
`intermediate steroids, such as progesterone and 11~deoxycorticosterone
`
`4
`
`

`

`MANAGEMENT OF HUMAN DEFICIENCIES OF P450c17
`
`105
`
`(DOC), must accumulate, as well as unusual metabolites, such as 18-
`hydroxycorticosterone33 and 19—nor—deoxycorticosterone.23 This ACTH-
`driven overproduction of mineralocorticoids leads to hypertension, a
`characteristic presenting feature of this disease. The hypertension usually
`develops in early adulthood9 but can present in infancy15 and can be
`severe.46 As is true in other hypertensive disorders caused by mineralo~
`corticoid excess,39 the hypertension can become fixed if the disease is
`not treated for many years.52
`Although the general description given herein is true for most
`patients with this disorder, considerable variation in phenotype and
`laboratory findings has been described. These variables include the
`degree of genital virilization in 46,XY subjects and the capacity for
`menstruation in 46,XX subjects; the severity of the hypertension and
`hypokalemia; the aldosterone secretion rate; the type and amount of
`adrenocortical hyperplasia; the gonadal morphology and histology; and
`the coexistence of additional disorders, such as 21-hydroxylase defi—
`ciency53 or maternal androgen excess.14 This heterogeneity has not been
`completely explained, but many factors, including the severity of the
`P450c17 deficiency, variations in genes regulating hormone respon-
`siveness, diet (sodium consumption), and environment, undoubtedly
`contribute. The reader is referred to a detailed discussion of case re-
`
`ports,72 which is beyond the scope of this article.
`
`Isolated 17, 20-Lyase Deficiency
`
`This disoder is extremely rare because mutations that cause this
`phenotype must not only destroy most 17,20-lyase activity but preserve
`most 170i-hydroxylase activity. Patients who are 46,XY present with
`ambiguous genitalia at birth or with inguinal hernias with or without
`pubertal delay as adolescents72 (Table 1). Patients do not show the
`consequences of mineralocorticoid excess because preserved cortisol pro-
`duction prevents excessive DOC and corticosterone accumulation. Clini—
`cal laboratory findings vary considerably owing to the age of diagnosis,
`the severity of the disease, and the discrepancy between the Nor—hydrox—
`ylase and 17,20-1yase activities in a given individual. Nonetheless, C19
`steroid production is
`severely, although not completely,
`impaired,
`whereas 17—hydroxylated steroid production is nearly or completely nor-
`mal.
`
`DIAGNOSIS
`
`Unlike forms of congenital adrenal hyperplasia, such as the lipoid
`type and 21-hydroxylase deficiency, in which glucocorticoid and miner~
`alocorticoid production are impaired, patients with 17—hydroxylase defi—
`ciency do not have an adrenal crisis in the postnatal period. Conse—
`quently,
`the diagnosis is often not entertained until hypertension,
`hypol<alemia, or pubertal delay is evaluated during adolescence or early
`
`5
`
`

`

`“,un-nuluqu.
`
`
`
`WOO—HAHme»%enououmcquUAHHOwH
`
`ongowmcoicU333mg
`
`
`
`
`
`dqumtmoQQotoUthQmmm
`
`
`
`.N0.59".
`
`
`
`
`
`
`
`ofiPfificfiEA..........:
`
`
`Vgmmocmvm“Nucmvm
`..fl0:95mewcaAquariuflumocacao»
`
`
`QOQwmQmmamHm
`
`53>?maxflaw
`
`mew—58m
`
`uumomvm
`
`106
`
`MOHwaquHMU
` hfioawmm
`
`
`6
`
`
`
`

`

`MANAGEMENT OF HUMAN DEFICIENCIES OF P450c17
`
`107
`
`adulthood. Patients with a 46,XY l<aryotype and incomplete deficiency
`may be misdiagnosed with androgen insensitivity or defects in later
`steps of dihydrotestosterone biosynthesis. As is true for all steroidogenic
`enzyme deficiencies, the diagnosis is most convincingly established by
`measuring precursor-to—product ratios during ACTH stimulation testing.
`In particular, circulating concentrations of the 17—deoxysteroids proges-
`terone, corticosterone, and DOC rise to 5 to 10 times normal after ACTH
`administration.15 In addition, 17—hydroxylase deficiency, in distinct con—
`trast to 11—hydroxylase and 21-hydroxylase deficiencies, is characterized
`by elevated production of 18-hydroxycorticosterone and 18—hydroxy-
`DOC (Table 2).33 The ratio of corticosterone to DOC (or of their 18-
`hydroxy—derivatives) distinguishes 17— from ll—hydroxylase deficiency.
`Table 3 compares the clinical, laboratory, and genetic characteristics of
`the various mineralocorticoid excess states that may arise in children
`and young adults.
`Although production of the precursors corticosterone and DOC is
`markedly elevated in 17—hydroxy1ase deficiency, DOC production can be
`much greater in 11-hydroxylase deficiency, whereas plasma 18~hydroxy-
`DOC concentrations are not elevated.33 The reason for this apparent
`discrepancy is that P450c118 (the product of the CYPIIBI gene) is
`not exclusively an llB-hydroxylase but exhibits weak 18—hydroxylase
`activity”,64 (Fig. 2) and trace amounts of aldosterone synthase activity?5
`The low 18-hydroxy-DOC production in 11—hydroxylase deficiency, de-
`spite enormous DOC concentrations, is compelling genetic evidence that
`P450c11l3 is responsible for elevated 18-hydroxy~DOC and 18—hydroxy—
`corticosterone production in 17-hydroxylase deficiency. Analogously, in
`glucocorticoid-remediable aldosteronism, abundant 18—oxygenase activi-
`ties in the zona fasciculata owing to the presence of a chimeric CYPIIBZ/
`11B1 gene36 lead to excessive 18-oxygenated steroid production.16 Pa—
`tients with 17—hydroxylase deficiency with paradoxically measurable, if
`
`
`
`Figure 2. Physiologic disturbances in glucocorticoid, mineralocorticoid, and sex steroid
`homeostasis in complete 17-hydroxylase deficiency. The inability to 17a-hydroxylate C21
`steroids in the adrenal gland eliminates all steroids within shaded region and shunts
`pregnenolone flux to progesterone, 11-deoxycorticosterone (DOC), corticosterone, and
`possibly aldosterone (large open arrows). Absence of negative feedback by cortisol (dashed
`line) causes overproduction of adrenocorticotropic hormone (ACTH) (top-most large open
`arrow), and the resultant abundance of the weak glucocorticoid corticosterone provides
`adequate systemic glucocorticoid action and feedback on ACTH secretion (solid line). The
`hypothalamic-pituitary-adrenal axis then reaches a steady state at a higher set-point;
`however, the drive to overproduce corticosterone allows the accumulation of intermediates
`such as the potent mineralocorticoid DOC, and high DOC production stimulates salt and
`water retention, which suppresses renin secretion (dashed arrow). Thus, aldosterone pro—
`duction is low (dashed arrow), but hypertension and hypokalemia develop because of DOC
`excess.
`In addition, the unusually high concentrations of DOC and corticosterone in the
`presence of robust P450011B expression leads to excessive production of ordinarily minor
`metabolites 18—hydroxy (180H)-DOC and 18-0H—corticosterone. Because 17-hydroxy
`(17OH)-pregnenolone and dehydroepiandrosterone (DHEA) synthesis is nil (shaded region)
`in the fetus and at puberty, no androgen or estrogen synthesis is possible, and sexual
`infantilism results.
`
`7
`
`

`

`
`
`
`
`mEonwmumOuhHQEHOZ
`
`
`
`\CQBVEmmfifimwmdodmfigmlenHawmmfiémfiwmuflofl
`
`
`
`
`
`
`
`
`
`
`
`.EoukxoywkmaunfiHmUEOIBH“mucufimohvfimimofikfimwH<mED«EmobovmfiomquEofiFEES;HUUQacoEmfimumOHEHmHD<uwfiowymwmngoHPESHHEDIE“
`
`
`
`.wqobwmofluLonxowUAHHUODwmwwo‘aymou‘mxlhfiHWMthwmwfionmumou
`
`GU;8%SAVofinfioéeEmwohvfim4min;
`
`
`
`\mgmfimvfio&£\qgamtm&m
`
` uooéafibfl9odoHoEmfimqwmummlenH\mUEOSHa$39.05BEE:
`
`
`
`
`EmmanuthESXmm
`
`
`
`ecoHBmoUEHou\mofionfima
`
`UOQGOD\mfioHfimoonn:
`
`
`
`mammohmm\mamwOHEHm
`
`$3185
`
`
`
`
`
`mEmmE memfl\mEonymIEOIBHawthme
`
`9Emmfl?OHUN€T§HflmEDrEOU
`
`
`
`
`
`cozflcmwmi325.05:22th
`
`
`
`
`
`
`
`>OZm__0_u_m_Dmm<>lTONKFamp/.309DZ<>OZm:O_n_m_DmmSflémNF\m_m<:_>XOmD>ILoNFDMZEEOOu_OZOw_m<n__>_OO.Fmin...
`
`
`
`
`
`
`
`
`
`108
`
`8
`
`

`

`MANAGEMENT OF HUMAN DEFICIENCIES OF P450c17
`
`109
`
`Table 2. COMPARISON OF STEROID PROFILES IN ADULTS WiTH 17-, 11-, AND 21-
`HYDROXYLASE DEFICIENCES
`
`Type of
`DOC
`18-0H-DOC Corticosterone
`18-OH-Corticosterone Aldosterone
`
`Deficiency
`ng/dL
`ng/dL
`ng/dL
`ng/dL
`ng/dL
`17-OH
`25—500
`100—600
`4000—40,000
`60—1000
`<10
`11-OH
`50 to >1000
`<10
`<200
`<10
`<3
`21—OH
`10—100
`3—20
`100—500
`10~200
`10—60
`Normal
`2—20
`1—20
`100—500
`10—40
`10—30
`
`DOC = 11-deoxycorticosterone;18-OH—DOC = 18-hyd1'0xy~DOC; OH = hydroxylase.
`Data adapted from Kater CE, Biglieri EG: Disorders of steroid 17 alpha-hydroxylase deficiency. Endocrinol Metab
`Clin North Am 23:341—357, 1994.
`
`not elevated, aldosterone production have been described. It is possible
`that, in these instances, artifacts owing to laboratory methods or intercur-
`rent glucocorticoid therapy confound the data.32 It is equally likely that
`other genetic and environmental modifiers contribute to these variations
`in disease manifestations, such as polymorphisms that alter the aldoste-
`rone synthase activity of P450c11[3. The latter hypothesis is consistent
`with the finding that most 17-hydroxylase deficiency cases with measur-
`able aldosterone production are from Iapan.52' 72 Until a large series
`of patients with 17-hydroxylase deficiency is compiled with uniform
`evaluation, these conundrums will persist.
`Although heterozygous family members of patients with 17-hydrox—
`ylase deficiency without other endocrine abnormalities usually have
`clinically normal adrenal and gonadal physiology, it is sometimes possi-
`ble to detect heterozygosity using biochemical testing. Elevated cortico-
`sterone and 18-hydroxycorticosterone concentrations, as well as the 18~
`hydroxycorticosterone~to—aldosterone ratio, after ACTH stimulation are
`perhaps the most readily available means to detect heterozygotes if an
`index case has been identified.65 More precisely, the ratio of total urinary
`metabolites of corticosterone to those of cortisol is elevated (reflecting
`low 17a—hydroxylation), and the ratio of total urinary metabolites of C19
`steroids to those of C21 steroids is low (reflecting low 17,20—lyase activ-
`ity).13 If a compelling reason for ascertainment of an individual’s zygos-
`ity exists, molecular genetics provides a highly sensitive, although te—
`dious, method that must be performed in a research laboratory.20
`
`MOLECULAR GENETICS
`
`Deletions, Premature Truncations, Frameshifts, and
`Splicing Errors
`
`Among the genetic abnormalities described in the CYP17 gene, the
`largest deletion reported involves the substitution of 518 bp (most of
`exon 2 and part of exon 3) with 469 bp of unknown DNA, disrupting
`the protein near its beginning and causing complete 170L-hydroxylase
`
`9
`
`

`

`ESCSED
`
`555?:
`
`8505?:
`
`
`
`@888:mwobwnm
`
`8mmuumfiuwBENEEC
`
`
`
`Hwfio\mdoflfidfiMU
`
`883:8mmnamm
`883:8KEC
`mmmuxmEQUEOUBSQE203E?
`
`
`
`SufimfimwsmfiHemuuouMomfioafikm
`
`
`moSofimmmhasmmcommfiwfidxwfl
`modofimmmfifidmmnemwfimamxmfl
`
`bamsamgmfififiQ33
`\mmmuxmHOwEOUmomfioaam
`
`mamasdmx\qowmzwfimg:
`
`mfimmufimvfidowmfimtmnfi;
`
`mgHemwuouWEOZ
`
`mgHOmEOU€8.82
`
`EOIwHwad\mfioHBonEoU\UOD<
`
`[$0139Qmfio~mfi0wwfi0§<<
`
`
`mfimfigdowusvomm3380<
`23055mfiU\mEonymIEOIBH>
`
`\HOmwuouoxoiHflaw-EO|wH<
`
`$58386?EmbedKUOQ<
`885%$50mfifibfiw
`
`$509.?mewfiOmfiuoU>
`
`mnemmevamzmfig
`
`£253533%.on<
`
`EmfiqoymwmowfimmzmwwmfimHIEOUEHouSEQn
`
`EOUwHouofimHmEE398%?wemaobwawn
`
`
`
`Emmm$58.05.
`
`3.33".>3.
`
`
`
`mmEuEu.€39.33
`
`wwmmwfi.
`
`
`
`
`
`558vammflkxofirmmunF
`
`
`
`
`
`.8383UmouuwououflmHMUhoney?HE0«mnemgmoumfivumxomwéfiHUOD
`
`
`
`
`
`.mmmUXmEouflwgofiumfifiwfimvfiwmmflugvab
`
`mm>wm>flow
`
`onHBmofiZ>
`
`mzohflmousug
`
`mfiofimovz>
`
`$30.83$50
`
`mEOHBm2U<
`
`EmEonymoEmEavfl
`
`m2.8.08
`
`mmmuxw
`
`
`
`85.83338330839“
`
`
`
`mEonSnmwEEmSU*
`
`
`
`
`
`
`
`
`
`>._._>_._.O<2.2mm<_>.m<n_n_mewmmnEDmI._._>>mmbfimwwmoxmQ_OO_._.EOOOI_<N_MZ=>_“.0ZOwE<n=ZOO.w053.
`
`
`
`
`
`
`
`
`
`110
`
`10
`
`10
`
`

`

`MANAGEMENT OF HUMAN DEFICIENCIES OF P450c17
`
`111
`
`deficiency.6 A 4—bp duplication of the sequence CATC following Ile47930
`was originally observed in Canadian Mennonites29 and has been subse—
`quently found in at least six Dutch Frieslander families.28 This duplica—
`tion leaves 95% of the protein unaffected and creates a mutant P450c17
`that has an altered sequence in its last 25 residues, that is truncated
`three residues prematurely, and that is wholly devoid of enzymatic
`activity. The crucial nature of the carboxy terminus of P450c17 is also
`shown by the complete absence of activity in the 9-bp, in—frame deletion
`of residues Asp487, Ser488, and Phe48919 and in a Gln461—>stop muta-
`tion.73 Although these mutants retain the heme-binding region, these
`ostensibly minor alterations in the extreme carboxy terminus are cata-
`strophic for enzymatic activity.
`A computer model of human P450c17 suggests why the enzyme is
`so sensitive to alterations in its carboxy terminus.5 The last 48 residues
`of P450c17 are involved in an extended B—sheet structure that folds down
`from the protein surface to form the “roof” of the active site, which is
`critical for proper substrate binding and subsequent catalysis. The CATC
`duplication after Ile479,30 the deletion of residues 487 to 489,19 and the
`mutant Gin461—>stop,73 which all retain the heme—binding site, disrupt
`or lack this critical stretch of residues required for activity.
`The mutation delTG300,301 shifts the reading frame and alters the
`codon use beginning within exon 5.43 Mutation 7bp dup 120 changes the
`reading frame from exon 2 onward.7o The premature truncation Trp
`17—>stop has been found in a homozygous68 and a compound heterozy—
`gous61 patient, and mutations Glu194—>stop and Arg239estop each com-
`prise separate alleles in a single patient with complete 170L-hydroxylase
`deficiency.56 These three early truncations are not
`informative for
`structure/ function studies because they delete the heme-binding region
`as well as residues important for substrate and redox partner binding.
`Two deleterious intronic mutations have been described, a G to T
`substitution at nucleotide +5 in intron 261 and an analogous G to A
`substitution at position +5 of intron 7.66 These splice junction mutations
`delete exons 2 or 7, respectively, during RNA processing (“exon skip-
`ping”). The excision of these exons introduces early premature stop
`codons well before the heme—binding region. The deletion of a G within
`codon 438 has been found in a homozygous patient.51 This mutant gene
`encodes a protein in which the Gly—Pro—Arg-Ser-Qfi-Ile motif at residues
`438—443 (the underlined Cys ordinarily donates the axial sulfhydryl to
`the heme iron) is converted to Asp-Leu-Ala-Pro—Val—Stop, which destroys
`all enzymatic activity. An ATGeATC substitution in the intiating methi—
`onine codon has been described in a patient with complete 170L—hydroxy-
`lase deficiency and hypokalemic myopathy.58
`
`Amino Acid Substitutions—Combined 17a-
`
`Hydroxylase/17,20-Lyase Deficiency
`
`Careful biochemical and computational analyses of mutant enzymes
`from patients with unusual phenotypes can provide insight into the
`
`11
`
`11
`
`

`

`112
`
`AUCHUS
`
`functional roles of specific amino acids in P450c17. For example, the
`mutation His373Leu, when expressed in E. coli, lacks the classical P450
`difference spectrum,44 strong evidence that this protein does not bind
`heme properly. Modeling studies5 predict that His373 lies distant from
`the heme moiety, suggesting that structural changes elsewhere in the
`His373Leu mutant secondarily abolish heme binding. In contrast, the
`mutation iAirgllAOI-lis18 lies two residues away from the heme-liganding
`Cys442, and the reason for loss of activity in this mutant is more
`straightforward. In most P450 enzymes, an analogous arginine residue
`in this position is critical for neutralization of a negative charge on a
`heme proprionate and stabilization of heme incorporation”; hence, this
`mutation also interferes with heme binding.
`The mutation Ser106Pro, found in two apparently unrelated Gua-
`manian patients,38 introduces a helix-breaking proline into what is pre-
`dicted to be the B’-helix, near residues that form a lateral boundary of
`the substrate-binding pocket. P450c17 is sensitive to perturbations in
`this region, such that even the conservative replacement of Ser106 with
`Thr (the corresponding residue found in rainbow trout P450c1757) abol-
`ishes most enzymatic activity.37 Specifically, Ile112 is predicted to interact
`directly with substrate, suggesting why mutation insIle112 is devoid of
`measurable activity.27 Nearby, mutations Gly9OAsp67 and Arg96Trp34 are
`predicted to reposition the second strand of B-sheet 1, containing the
`key residue Gly95. Computer simulations predict that 38—hydroxyl and
`3-l<eto groups of A5 and A4 substrates, respectively, form hydrogen bonds
`to the carbonyl group or the amide hydrogen of Gly95.5' 41 The four
`mutants insIle112, Ser106Pro, Arg96Trp, and Gly90Asp may all primarily
`impair substrate binding.
`Three mutations that retain partial enzymatic activity have also
`been described. Mutations Tyr64=Ser27 and Pr0342Thr1 retain approxi-
`mately 15% and 20% of wild—type activity, respectively. The loss of one
`of two contiguous Phe residues in the APhe53/54 mutation69 destroys
`all but a trace of enzymatic activity,72 and this mutation has been found
`in other cases of 17—hydr0xylase deficiency in Japan,42 suggesting a
`founder effect. The structural alterations responsible for the loss of
`activity in these mutants are not entirely clear, but these regions of the
`protein must be somewhat more tolerant of such structural changes
`than, for example, the active site and the heme-binding region.
`
`Mutations Causing Isolated 17,20-Lyase Deficiency
`
`The first patient with isolated 17,20-lyase deficiency in whom the
`CYP17 gene was sequenced proved to be a compound heterozygote for
`the Gln46lestop and Arg496Cys mutations.73 When studied in trans-
`fected cells, the Gln46lastop mutant was inactive, but the Arg496Cys
`mutant retained a small amount of 170L-hydroxylase and 17,20—lyase
`activities.” When restudied as an adult,74 the patient’s steroid hormone
`profile reflected nearly complete deficiencies of 170L-hydroxylase and
`17,20-lyase activities, consistent with the molecular genetics and bio-
`
`12
`
`12
`
`

`

`MANAGEMENT OF HUMAN DEFICIENCIES OF P450c17
`
`113
`
`chemistry of the mutant proteins. This case illustrates many of the
`pitfalls in the diagnosis of isolated 17,20—lyase deficiency and emphasizes
`that the clinical features, the molecular genetics, and the biochemistry
`of the mutant P450c17 protein(s) must all be congruent to ensure an
`accurate diagnosis.
`Recently, two 46,XY Brazilian patients presented with convincing
`clinical evidence of isolated 17,20—lyase deficiency, that is, genital ambi-
`guity and diminished C19 steroid production yet normal 17-hydroxycorti—
`costeroid production. One patient was homozygous for mutation
`Arg347His and the other for Arg358Gln, whereas each parent was het-
`erozygous for the respective mutant allele.20 When expressed in COS-1
`cells, the mutants hydroxylated progesterone and pregnenolone,20 but
`only a trace of 17,20—lyase activity could be reconstituted by coexpressing
`an excess of oxidoreductase and b5.21 Although 170L—hydroxypregneno—
`lone is a poor substrate for the mutant enzymes, competition experi—
`ments unequivocally show that the affinity of the mutant proteins for
`170L-hydroxypregnenolone is equivalent
`to that of
`the wild—type
`enzyme,” 21 suggesting that arginines 347 and 358 do not lie in or near
`the active site.
`
`Computer modeling studies demonstrate that R347H and R358Q
`neutralize positive charges in the redox partner binding site.5' 20 Biochem-
`ical studies confirm that mutations R347H and R358Q impair interactions
`of P450c17 with its electron donor P450—oxidoreductase and with cyto-
`chrome b521; therefore, isolated 17,20—lyase deficiency is not caused by an
`inability of the mutant enzymes to bind the intermediate 170L-hydroxy-
`pregnenolone but rather by subtle disturbances in interactions with
`redox partners?” 20' 21 Another patient subsequently shown to have iso—
`lated 17,20—lyase deficiency was found to harbor mutation F417C,8 which
`is predicted to lie on the edge of this redox partner binding surface.5
`The biochemistry of the F417C mutant has not been studied in detail, so
`it is not known if the same mechanisms as for the R347H and R358Q
`mutants apply to F417C.
`A male pseudohermaphrodite with congenital methemoglobinemia
`owing to a mutation in the gene for cytochrome b5 has been described.22
`It is possible that this patient was incompletely virilized because of low
`(but not absent) testicular 17,20-lyase activity and testosterone deficiency
`in utero owing not to a P450c17 mutation but rather to the loss of b5, the
`cofactor protein that stimulates 17,20-lyase activity. Neither circulating
`steroid hormone concentrations nor a genetic analysis of the CYP17 gene
`were reported for this subject. If this patient has isolated 17,20—lyase
`deficiency owing to the loss of b5, the physiologic importance of b5 in
`P450c17 chemistry would be proved.
`
`MANAGEMENT
`
`The child with 17—hydroxylase deficiency is chronically exposed to
`elevated circulating mineralocorticoid (DOC) concentrations but roughly
`normal amounts of glucocorticoids (as corticosterone). Mineralocorticoid
`
`13
`
`13
`
`

`

`114
`
`AUCHUS
`
`excess in the neonatal period is of no consequence because mineralocorti—
`coid (aldosterone) production is normally high in infants”; however, as
`the child ages and begins to consume solid foods, sodium intake rises,
`and mineralocorticoid excess can lead to sodium retention, hypertension,
`and hypokalemia. The hypertension can become fixed if not treated for
`many years”; hence, some control of DOC production is desirable.
`Moderation of dietary sodium content is prudent as an adjunct to phar-
`macologic therapy, which consists of glucocorticoid supplementation to
`reduce aberrant DOC production. Special considerations in the child
`with 17-hydroxylase deficiency include the avoidance of highly potent
`fluorinated glucocorticoids, such as dexamethasone, that have dispro-
`portionately large detrimental effects on linear growth and bone mineral
`accrual. Hydrocortisone administered in two or three divided doses will
`generally suffice, although direct comparison of steroid regimens in this
`uncommon disease are lacking. The glucocorticoid dose should be ti~
`trated to normalization of blood pressure and plasma potassium concen-
`trations, as well as restoring plasma renin activity to the measurable
`range as endpoints. The frank normalization of plasma DOC and cortico-
`sterone concentrations may require overtreatment with glucocorticoids.52
`It is preferable to err on the side of undertreatment because the dire
`consequences of glucocorticoid excess throughout childhood are less
`desirable than modest mineralocorticoid excess.
`
`As is true for patients with Turner’s syndrome, gonodal dysgenesis,
`androgen insensitivity, or some other steroid biosynthetic defects, pa—
`tients with 17—hydroxylase deficiency fail to exhibit pubertal develop—
`ment, and fetal testosterone deficiency causes all but the most mildly
`affected patients to present phenotypically as prepubertal females. In
`addition, the testosterone surge that occurs during the first year of life 7
`in 46,XY children is absent in 17—hydroxylase deficiency, which could
`theoretically impair responsiveness to testosterone later in life for mildly
`affected individuals. In most cases, estrogen replacement therapy is
`initiated at the time of expected puberty or on diagnosis if that time has
`already passed. Estrogen replacement not only allows the development
`of female secondary sexual characteristics but stimulates the increase in
`bone mass that normally occurs during puberty.24 In a few cases, testos—
`terone supplementation has been given to mildly affected 46,XY patients
`to stimulate penile development“; however, as is true for patients with
`partial androgen insensitivity, the rearing of these individuals as males
`and the choice of appropriate therapy are complex decisions that unfor—
`tunately may yield less than satisfactory results.
`The treatment of 17—hydroxy1ase deficiency in the adult patient
`strives to achieve four goals: (1) reduction of the production or action of
`mineralocorticoids; (2) avoidance of the untoward effects of glucocorti-
`coid excess; (3) replacement of sex steroids; and (4) prevention of the
`long-term consequences of the abnormal physiology. Although the cave-
`ats and special considerations are somewhat different in the two age
`groups, the cornerstone of the therapeutic plan remains sodium restric-
`tion plus glucocorticoid supplementation, traditionally consisting of a
`
`14
`
`14
`
`

`

`MANAGEMENT OF HUMAN DEFICIENCIES OF P450c17
`
`115
`
`daily dose of dexamethasone. Patients with 17~hydroxylase deficiency
`given dexamethasone demonstrate a prompt reduction in DOC and
`corticosterone production, with naturesis and resolution of kaluresisfi'3
`The hypertension usually resolves with glucocorticoid therapy,52 but if
`the diagnosis has been delayed for many years, the hypertension can
`become fixed.39 The goal of glucocorticoid therapy is to restore the
`blood pressure and plasma potassium concentration to normal using the
`minimal amount of drug possible, usually 0.25 to 1 mg/ d of dexametha—
`sone or 2 mg/d to 5 mg / d of prednisone. Circulating concentrations of
`DOC and corticosterone may not completely normalize on this regi-
`men,52 but a rise in renin and aldosterone during glucocorticoid adminis-
`tration indicates that the therapeutic goal of eliminating ACTH—depen-
`dent mineralocorticoid excess has been achieved.53 As is true in other
`states of ACTH—driven mineralocorticoid excess, such as glucocorticoid—
`remediable aldosteronism and apparent mineralocorticoid excess, care
`must be taken not to suppress the hypothalamic—pituitary—adrenal axis
`overzealously, which can lead to complications of glucocorticoid excess.16
`Instead, small doses of mineralocorticoid antagonists, such as spironolac—
`tone or potassium canrenoate, can be added to the regimen, allowing
`modest glucocorticoid doses during long—term therapy in the adult as
`well.39 Particularly with the development of newer mineralocorticoid
`antagonists lacking the side effects of spironolactone,