The effect of amino-acid substitutions I112P, D147E and K152Nin CYP11B2 on the catalytic activities of the enzyme Stephanie Bechtel1, Natalya Belkina2and Rita Bernhardt1 1 Universita¨t d
Trang 1The effect of amino-acid substitutions I112P, D147E and K152N
in CYP11B2 on the catalytic activities of the enzyme
Stephanie Bechtel1, Natalya Belkina2and Rita Bernhardt1
1
Universita¨t des Saarlandes, Saarbru¨cken, Germany;2Insitute of Biomedical Chemistry RAMS, Moscow, Russia
By replacing specific amino acids at positions 112, 147 and
152 of the human aldosterone synthase (CYP11B2) with the
corresponding residues from human, mouse or rat
11b-hydroxylase (CYP11B1), we have been able to
investi-gate whether these residues belong to structural
determi-nants of individual enzymatic activities When incubated
with 11-deoxycorticosterone (DOC), the 11b-hydroxylation
activity of the mutants was most effectively increased by
combining D147E and I112P (sixfold increase) The two
substitutions displayed an additive effect The same tendency
can be observed when using 11-deoxycortisol as a substrate,
although the effect is less pronounced The second step of the
CYP11B2-dependent DOC conversion, the
18-hydroxyla-tion activity, was not as strongly increased as the
11b-hydroxylation potential Activity was unaffected by
D147E, whereas the single mutant I112P displayed the most
pronounced activation (70% enhancement), thus causing
different increasing effects on the first two enzymatic reaction
steps A slightly enhanced aldosterone synthesis from DOC
could be measured due to increased levels of the
intermedi-ates However, the 18-oxidation activity of all the mutants,
except for I112S and D147E, was slightly reduced The strongly enhanced 18-hydroxycorticosterone and aldoster-one formation observed in the mutants provides important information on a possible role of such amino-acid replace-ments in the development of essential hypertension Furthermore, the results indicate the possibility of a differ-ential as well as independent modification of CYP11B2 reaction steps The combination of functional data and computer modelling of CYP11B2 suggests an indirect involvement of residue 147 in the regulation of CYP11B isoform specific substrate conversion due to its location on the protein surface In addition, the results indicate the functional significance of amino-acid 112 in the putative substrate access channel of human CYP11B2 Thus, we present the first example of substrate recognition and conversion being attributed to the N-terminal part of human CYP11B2
Keywords: cytochrome P450; 11b-hydroxylase, aldosterone synthase; N-terminal protein region; engineering substrate specificity
Cytochromes P450 are key enzymes in the
biotransforma-tion of drugs, xenobiotics and steroids (reviewed in [1])
The synthesis of the most important glucocorticoid and
mineralocorticoid hormones in humans (cortisol and
aldosterone, respectively), take place in the adrenal gland
It has been shown that in pig [2] and frog [3] this synthesis
is performed by a single P450 enzyme (CYP11B1) In
contrast, bovine has two closely related isoenzymes
encoded by different genes [4,5] that synthesize both
cortisol and aldosterone In several other species, including human [6,7], mouse [8] and rat [9,10], two distinct isoforms
of the CYP11B subfamily, namely CYP11B1 and CYP11B2, have been characterized, which are specialized
to synthesize cortisol or aldosterone In human, the terminal three steps in the biogenesis of aldosterone are catalyzed by the aldosterone synthase (CYP11B2) exclu-sively in the zona glomerulosa [11] The 11b- and 18-hydroxylation of the substrate 11-deoxycorticosterone (DOC) leads to corticosterone (B) and 18-hydroxycorticos-terone (18-OH-B), whose 18-oxidation yields aldos18-hydroxycorticos-terone
In the zona fasciculata/reticularis, the 11b-hydroxylase (CYP11B1) catalyzes the 11b-hydroxylation of 11-deoxy-cortisol to produce 11-deoxy-cortisol which is normally secreted 100- to 1000-fold in excess over aldosterone [12] CYP11B1
is also able to produce corticosterone from 11-deoxycorti-costerone but it cannot convert corti11-deoxycorti-costerone into aldosterone [7,13] The translated proteins of the two human isoenzymes of CYP11B contain 503 amino acids, including a 24-residue N-terminal mitochondrial targeting sequence, and share 93% sequence identity [6] There are only 32 amino-acid differences in the mature forms of the two cytochrome P450 proteins The apparent molecular masses of the aldosterone synthase and 11b-hydroxylase have been determined to be 48.5 and 50 kDa, respectively [13] Both enzymes are localized in the inner mitochondrial membrane and function alongside the flavoprotein adreno-doxin reductase (AdR) [14], and adrenoadreno-doxin (Adx) [15]
Correspondence to R Bernhardt, Universita¨t des Saarlandes, FR 8.8
Biochemie, PO Box151150, D-66041 Saarbru¨cken, Germany.
Fax: + 49 681302 4739, Tel.: + 49 681302 4241,
E-mail: ritabern@mx.uni-saarland.de
Abbreviations: CYP11B1, cytochrome P450 11b , 11b-hydroxylase;
CYP11B2, cytochrome P450 aldo , aldosterone synthase; Adx,
adreno-doxin; AdR, adrenodoxin reductase; SS, Dahl salt-sensitive rat; SR,
Dahl salt-resistant rat; SRS, substrate recognition site; DOC,
11-deoxycorticosterone; B, corticosterone; 18-OH-B,
18-hydroxy-corticosterone; Aldo, aldosterone; HPTLC, high performance thin
layer chromatography; DMEM, Dulbecco’s modified Eagle’s medium.
Enzymes: steroid 11b-hydroxylase and aldosterone synthase
(EC 1.14.15.4); adrenodoxin reductase (EC 1.18.1.2).
Note: a website is available at http://www.uni-saarland.de/fak8/
bernhardt/
(Received 29 August 2001, revised 30 November 2001, accepted 7
December 2001)
Trang 2Lifton et al [16] described a patient carrying a chimeric
gene consisting of a 5¢-CYP11B1 regulatory sequence fused
to a 3¢-CYP11B2 portion, causing
glucocorticoid-remedi-able aldosteronism The encoded chimeric protein, which is
a result of an unequal meiotic cross-over upstream of
intron 5, possessed efficient aldosterone synthase activity
Previous studies have primarily concentrated on the
C-terminal amino acids, emphasizing their importance for
the individual activities of CYP11B1 and CYP11B2 For
instance, by substituting the positions 301, 302 and 320 in
CYP11B2 by CYP11B1-specific residues, a switch in the
regio- and stereospecificity of the enzymatic activity can be
observed [17] Moreover, an aldosterone synthase activity
could be converted from CYP11B2 to the 11b-hydroxylase,
when creating a CYP11B1 double mutant containing the
aldosterone synthase specific amino acids glycine and
alanine at positions 288 and 320, respectively [18] Bo¨ttner
et al [19] have shown that the mutant A320V of CYP11B1
displays only 20% aldosterone synthase wild-type activity
when expressed in COS-1 cells in the presence of DOC,
indicating that other amino acids, including some at the
N-terminus, contribute to efficient CYP11B1 and CYP11B2
wild-type activity In addition, it is known from the crystal
structures of CYP101, CYP108 and CYP102 that the
N-terminal region encodes an amino-acid sequence that is
involved in substrate recognition and binding as well as
redox partner binding [20] This finding was also supported
by results obtained with microsomal P450 proteins
Ridderstro¨m et al [21] have shown the functional
impor-tance of Arg97 and Arg108 in the activity of CYP2C9,
especially for substrate binding, by site-directed mutagenesis
and homology modelling
The phenotypical abnormality of hypertension was
examined using the model system of Dahl salt-sensitive
(SS) and salt-resistant (SR) rats demonstrating the essential
role of exons 3 and 4 of aldosterone synthase [22], which
also implicates the significance of the N-terminal region of
CYP11B2 in enzymatic activity These studies prompted us
to perform protein sequence- and structure-based
align-ments of human CYP11B family members with mouse and
rat CYP11B1 and CYP11B2, human CYP2C9 and P450s
with known three-dimensional structures We concentrated
our efforts on the N-terminal amino acids, which differ
between the human CYP11B1 and CYP11B2 enzyme, and
are candidate residues for influencing the enzymatic activity
of human aldosterone synthase As the two helices, B and C,
of the structurally known cytochromes P450 located in the
N-terminal protein regions play an essential role in the high
substrate selectivity and redox partner interaction [23,24],
we investigated whether the amino acids of human
CYP11B2 located in regions aligned with these helices are
of functional importance They were replaced by the
corresponding amino acids of human, mouse and rat
CYP11B1 using site-directed mutagenesis and the mutants
were characterized with respect to their hydroxylation
selectivity
M A T E R I A L S A N D M E T H O D S
Materials
Expression vector pSVL was purchased from Pharmacia
Biotech Inc Oligonucleotides were synthesized on an
Applied Biosystems model 380A DNA synthesizer at BioTez (Berlin) COS-1 cells were obtained from the American Type Culture Collection Cell culture media, pyruvate, glutamine, antibiotics and Hepes were from Sigma Fetal bovine serum and DEAE-dextran were obtained from GibcoBRL and Pharmacia Biotech Inc., respectively Chloroquine, Hank’s balanced salt solution, dimethylsulfoxide, 11-deoxycorticosterone, corticosterone, 18-hydroxycorticosterone, aldosterone, 11-deoxycortisol, cortisol, 4-chlor-1-naphthol and secondary horseradish conjugated anti-(rabbit IgG) Ig were all from Sigma [14C]11-deoxycorticosterone and [3H]11-deoxycortisol were purchased from DuPont NEN HPTLC plates silica gel 60
F254and solvents were from Merck The BCA assay kit for quantitation of total protein was purchased from Pierce
Site-directed mutagenesis and expression vectors Mutations were inserted into human CYP11B2 cDNA by site-directed mutagenesis using the Quick Change Kit from Stratagene Ltd (Cambridge, UK), according to manufac-turer’s instructions and using mutagenic primers listed in Table 1 The cell culture expression construct pSVL/ CYP11B2 was used as a template This construct contains the cDNA encoding human aldosterone synthase The sequence corresponds to that published by Kawamoto et al [7] with one variation at position 249, where we found Ser instead of Arg, as described by Mornet et al [6] All exchanges were confirmed by automatic sequencing using a LiCor-4000 DNA sequencer (MWG Biotech, Ebersberg, Germany), thus excluding undesired mutations
To express the human 11b-hydroxylase enzyme, we used the cDNA sequence corresponding to that described by Mornet et al [6], except for three modifications These modifications led to the following variations in the encoded protein: Leu at position 52 is replaced by Met, Ile 78 is replaced by Val, and at position 494 we found Phe instead of Cys, as published by Kawamoto et al [25] The cDNA was cloned into the mammalian cell expression vector pSVL All standard procedures were carried out as described by Sambrook et al [26]
Cell culture COS-1 cells were grown at 37°C and 6% CO2 in Dulbecco’s modified Eagle’s medium (DMEM) supple-mented with 5% fetal bovine serum, 0.1 mgÆmL)1 strepto-mycin, 100 UÆmL)1 penicillin, 1 mM pyruvate and 4 mM
L-glutamine
Table 1 Sequences of forward oligonucleotides employed for the mutagenesis of the human aldosterone synthase and the corresponding amino-acid exchanges Nucleotides represented in bold characters indicate mismatched bases in CYP11B2 Codons for the changed amino acids are underlined.
Mutation Oligonucleotide sequences
D147E/K152N ACCCAGAAGTGCTGTCGCCCAACGCCG
TGC
Trang 3Transient transfections and enzymatic assays
Transfections were performed using the DEAE-dextran
method as described previously [27], modified as follows:
COS-1 cells were plated at a density of 6· 105cells per 6-cm
dish and grown overnight Next day, the medium was
aspirated and the cells were subjected to starvation by
incubating in 2 mL fetal bovine serum-free medium
containing Hepes to a final concentration of 50 mM The
incubation time was fixed to 2 h After removing the
medium, the COS-1 cells were cotransfected with 5 lg of
CYP11B2 or CYP11B1 expression plasmid and 3 lg of
pBAdx4 (a generous gift from M Waterman, Department
of Biochemistry, Vanderbilt University School of Medicine,
Nashville, USA) mixed with 1 mL starvation Medium
supplemented with 250 lg DEAE-dextran After 1 h, 2 mL
of complete medium containing chloroquine to a final
concentration of 100 lMwere added, and the incubation of
the cells was continued for 2 h For the subsequent
dimethylsulfoxide treatment, the medium was replaced by
2 mL of Hank’s balanced salt solution supplemented with
10% dimethylsulfoxide for exactly 2 min Afterwards, the
cells were washed twice with Hank’s balanced salt solution
and cultured with 3 mL of complete medium To assay for
CYP11B1- and CYP11B2-dependent activities, the cells
were incubated 24 h after transfection with 2 mL complete
medium containing either 30 lM DOC and 6 nCi
14C-labelled DOC or 30 lM11-deoxycortisol and 0.6 lCi
3H-labelled 11-deoxycortisol Following a 48-h incubation
period, steroids were extracted twice from the cell culture
supernatant with methylene chloride and the organic phase
was dried The residues were dissolved in 10 lL methanol
and spotted onto glass-baked silica-coated high
perfor-mance thin layer chromatography (HPTLC) plates The
HPTLC plates were developed twice in methylene chloride/
methanol/water (300 : 20 : 1, v/v/v) The reaction products
were identified by comigration of unlabeled steroid
refer-ences and quantified after 2 days exposure on a bioimaging
analyser (BAS-2500, Fuji Photo Film Co., Ltd) After
substrate incubation, the transfected COS-1 cells were lysed,
as described previously [19], and subjected to
immunolog-ical detection of cytochrome P450 expression according to
standard procedures [26,28] using an anti-(human CYP11B)
serum (a kind gift from H Takemori, Department of
Physiological Chemistry, Osaka University Medical School
Osaka, Japan) The total amounts of protein were
quanti-fied using a BCA assay kit, according to the manufacturer’s
protocol
Alignment of P450 sequences and protein modelling
Multiple sequence alignment was carried out using
CLUSTALW1.8 [29] The secondary structure predictions
were produced by the network method usingPHDSEC[30]
The modelling of the three-dimensional structure of
CYP11B1 was carried out by homology modelling with
bacterial cytochromes with known three-dimensional
struc-ture from the Protein Data Bank [31], using theSYBYL6.6
subroutineCOMPOSER (Tripos Inc., St Louis, MO, USA)
The standard procedure of protein modelling using
COMPOSER includes the following steps: (a) determination
of an initial set of topologically equivalent residues by using
the multiple sequence alignment method, which is then used
to produce an optimal structural alignment of the cyto-chromes P450 with known structure; (b) determination of structurally conserved regions (SCRs) of the proteins based
on this structural alignment; (c) building of the backbone of each SCR in the model by fitting a most appropriate fragment from one of the cytochromes P450 with known three-dimensional structure and determination of the side-chain conformations based on information about the backbone secondary structure and the side chains of the corresponding residues in each of the protein templates; (d) searching for protein loops in order to design the backbone conformations of the structurally variable regions (SVRs) with visual inspection to avoid poor steric interaction with surrounding parts of the protein model
The models of the three-dimensional structure of CYP11B2 and the mutants were made by using point mutations and protein loop search for regions which are different for CYP11B1 and CYP11B2 by means of the SYBYL programme suite, as described previously [32] Energy minimization was performed for the structures of the models in the presence of water; the Tripos Force Field was used The optimum was reached when the energy gradient was lower than 0.05 kcalÆmol)1ÆA˚)1 However, no more than 500 minimization steps were used The Powell Conjugate Gradient method was used for energy minimi-zation in both cases Verification of the obtained models was carried out usingPROCHECK[33] andPROSA[34] and all the models showed appropriate quality
R E S U L T S
Alignment of human, mouse and rat CYP11B1 and CYP11B2 with crystallized cytochromes P450 and human microsomal CYP2C9
Although the sequence identities between the multitude of P450 enzymes, identified to date, are frequently less than 20%, there is a Ôstructural coreÕ common to all P450s [23], indicating high conservation of secondary structure Based
on this fact, we performed amino-acid sequence and structure alignments of human 11b-hydroxylase and aldo-sterone synthase with structurally known P450s and the human CYP2C9 (Fig 1) We focused on the distribution of
32 amino acids that differ in the mature forms of CYP11B1 and CYP11B2, in order to identify candidates residues for determining the efficient catalytic functions of the two enzymes We discovered that residue 112 is located in a region aligned with the substrate recognition site (SRS) 1 of CYP2 family members [35] and the B helix of the crystal-lized P450s (Fig 1) As the helices A, B, B¢, F and G of the crystallized P450 proteins may contribute to the high substrate specificity to cytochrome P450 [23], and as the conversion of a multitude of compounds might be due to the high variability in the SRS of the family 2 P450s [35], amino acid 112 of CYP11B1 and CYP11B2 may therefore
be involved in specific substrate recognition of human 11b-hydroxylase and aldosterone synthase Residues 147 and
152 are encoded by exon 3 Its functional relevance was demonstrated by the use of the model system of Dahl SS and SR rats [22] encoding the amino-acid substitution E136D The double mutant E136D/Q251R in Dahl SR rats resulted in a 1000-fold enhanced enzymatic activity in Dahl
SR rats Furthermore, amino acids 147 and 152 are placed
Trang 4Fig 1 Multiple sequence alignment between several cytochromes P450 The alignment was done using CLUSTALW 1.8 (31) The regions corre-sponding to helices and the heme-binding area of the structurally known P450s are indicated and examplified by the underlines in the CYP101 sequence The shaded positions in the human CYP11B sequences represent the residues selected for investigation, whereas the shaded part in the CYP2C9 sequence indicates its putative SRS1, belonging to the substrate recognition sites in CYP2 family members identified by Gotoh (37).
Trang 5in an area aligned with the C helix of the so far crystallized
P450 enzymes (Fig 1) These amino acids could play an
important functional role, especially with regard to the
interaction with Adx, in accordance with the observation
that the helices B, C, J, J¢, K, L of several known bacterial
P450s seem to be involved in redox partner binding [24]
Site-directed mutagenesis and expression
of CYP11B2 mutants
Three single mutants, two double mutants and one triple
mutant of CYP11B2 were created by site-directed
muta-genesis using the oligonucleotides listed in Table 1, in
addition to the complementary oligonucleotides Thus, the
human aldosterone synthase wild-type amino acids were
replaced with the corresponding residues of human, mouse
and rat CYP11B1, respectively, as summarized in Table 2
The successful insertion of the intended mutations was
verified by sequence analysis
By performing three independent transfection
experi-ments, we found no substantial deviations in expression
levels between the wild-type and mutant proteins This result
suggests that the amino-acid exchanges had no influence on
protein stability or expression level (data not shown)
Enzymatic activity of aldosterone synthase mutants
To analyse the enzymatic specificities of the CYP11B2
mutants, as compared to the wild-type proteins, we
contransfected the resultant plasmids together with pBAdx4
into COS-1 cells The coexpression of bovine adrenodoxin
has been demonstrated to be a useful approach to increase
the activity of the human steroidogenic enzymes, as well as
the sensitivity of the test system [17,36–38] To estimate the
aldosterone-producing or cortisol-synthesizing potential,
the cells were incubated with either DOC or
11-deoxycor-tisol, respectively Different concentrations of DOC or
11-deoxycortisol (ranging from 10 to 80 lM) and different
incubation times were used to optimize the incubation
conditions; the optimal conditions were found to be 30 lM
DOC or 30 lM 11-deoxycortisol and 48 h incubation
Under the conditions tested, comparable relative activities
between the respective constructs were detected without
affecting the viability of COS-1 cells during substrate
incubation (data not shown) Using the optimized
condi-tions, the different mutants and the wild-type enzymes were
characterized with respect to all three catalytic activities
11b-hydroxylation, 18-hydroxylation and 18-oxidation
The mutated CYP11B2 enzymes were analysed by
incubating them with DOC as substrate (Fig 2) No
significant alteration in substrate conversion was detectable
for mutant I112S, as compared to CYP11B2 wild-type, indicating that this amino-acid exchange had no effect on the enzymatic activity The same observation was made for the single mutant K152N (M Hampf, Max-Delbru¨ck-Centre, Berlin, Germany, personal communication) In contrast, all other mutants induced markedly different steroid profiles relative to the wild-type of CYP11B2, as shown in Fig 2 It is obvious that more intermediates (B and 18-OH-B) were produced from DOC due to a substantial increase in the activities of the mutants How-ever, the three enzymatic steps were affected to different extents, represented by the relative activities as shown in Fig 2B As evident from the comparison of the 11b-hydroxylation activities of all constructs (Fig 2B), the introduction of Pro at position 112 enhanced the activity of the first enzymatic reaction step more than threefold,
Table 2 Corresponding amino acids of human CYP11B2 and
CYP11B1 as well as mouse and rat CYP11B1 at the positions selected
for mutagenesis.
Position
Human
CYP11B2
Human CYP11B1
Mouse and rat CYP11B1
Fig 2 Enzymatic activities of aldosterone synthase and hydroxy-lase (A) Enzymatic activities of aldosterone synthase and 11b-hydroxylase wild-type enzymes and different CYP11B2 site-directed mutants expressed in COS-1 cells towards 11-deoxycorticosterone (30 l M DOC and 6 nCi of [ 14 C]DOC) Mock represents the transfec-tion with the empty vector pSVL Steroid patterns of DOC conversion are given as means ± SEM of four similar independent experiments performed in duplicate The amounts of the substrate, the intermedi-ates corticosterone (‘B’) and 18-hydroxycorticosterone (18-OH-B) and the final product aldosterone (Aldo) are presented as percentages of total activity (B) Relative aldosterone synthase activities The effects
of the aldosterone synthase mutants on the 11b-hydroxylation (ratio of
RB + 18-OH-B + Aldo/DOC), 18-hydroxylation [ratio of
R18-OH-B plus Aldo)/R18-OH-B] and 18-oxidation (ratio of Aldo/18-OH-R18-OH-B) activity of CYP11B2 are presented The activities are shown as means ± SEM (n ¼ 8).
Trang 6whereas a fourfold increase was observed for the D147E
mutant, representing the strongest effect on the
11b-hydroxylation activity of all single mutants investigated
here When both mutations were introduced into CYP11B2,
the 11b-hydroxylation capacity was additionally activated,
obtaining a sixfold enhancement in relation to the wild-type
enzyme (Fig 2B) In contrast, the introduction of another
amino-acid exchange (I112P/D147E/K152N) led to a 26%
reduction in 11b-hydroxylation activity, as compared to the
double mutant I112P/D147E, which demonstrated slightly
increased activity of the first enzymatic reaction step, as did
the single mutant D147E (Fig 2B) The same observation
was made for mutant D147E/K152N (exhibiting a 20%
reduction), as compared to the single mutant D147E The
11b-hydroxylation activity of the double replacement
mu-tant, D147E/K152N, equalled almost that of mutant I112P
Obviously, K152N in combination with the mutations
D147E and I112P/D147E minimized the activating
charac-ter of the corresponding mutants (Fig 2B) The second
catalytic step performed by human CYP11B2 was not as
strongly enhanced as the first enzymatic modification in all
mutants studied (Fig 2B) The construct containing the
I112P substitution could be clearly identified as the single
mutant displaying the strongest activation of the
18-hydroxylation; 1.7-fold compared to the CYP11B2
wild-type, suggesting a critical role of this residue in the second
enzymatic reaction step of CYP11B2 (Fig 2B) In contrast,
this reaction step seems to be unaffected by the single
replacement D147E The same observation was made for
the double replacement mutant I112P/D147E showing
18-hydroxylation activity comparable to CYP11B2 wild-type
(Fig 2B), thus suggesting a slightly negative influence of
D147E on the second hydroxylation step when combined
with I112P
Interestingly, insertion of one more human
CYP11B1-specific residue at position 152 (I112P/D147E/K152N) leads
to an increase (13%) in hydroxylation at position 18
(Fig 2B), compared to the corresponding double mutant
without K152N This data indicates that K152N positively
affected the 18-hydroxylation potential when combined
with I112P and D147E Investigation of aldosterone
synthesizing abilities demonstrated that all mutants
pro-duced slightly higher amounts of this steroid than CYP11B2
wild-type (Fig 2A) Comparing the relative amounts of
aldosterone and 18-OH-B formation (Fig 2A), it becomes
clear that 18-oxidation activity displays a slightly decreased
efficiency in all investigated mutants, except for I112S and
D147E, when compared to the CYP11B2 wild-type emzyme
(Fig 2B)
In the second set of experiments, we investigated the
activity of wild-type and mutant proteins towards the
11b-hydroxylase-specific substrate, 11-deoxycortisol As seen for
DOC, we observed an overall tendency of all mutants,
except I112S, to strongly improve the substrate conversion
in relation to the CYP11B2 wild-type protein (Fig 3) By
replacing the amino acids in positions 112 and 147 of
CYP11B2 with those found in mouse, rat and human
CYP11B1, the two single substituted proteins I112P and
D147E were obtained These mutants displayed increases of
80% (1.8-fold) and 90% (1.9-fold) in cortisol-synthesizing
activities, respectively, as compared to the CYP11B2
wild-type enzyme (Fig 3A,B) As shown in Fig 3A, the product
formation for the double mutant I112P/D147E was
enhanced by more than 200%, which represents a 2.7-fold increase on CYP11B2 wild-type activity (Fig 3B), when incubated with 11-deoxycortisol The data from the I112P/ D147E mutant indicate an additive effect of the two single mutants The combined substitutions at positions 147 and
152 (double mutant D147E/K152N), and 112, 147 and 152 (triple mutant I112P/D147E/K152N) gave rise to cortisol-producing activity increases of 1.6-fold and 2.5-fold, respectively, compared to the CYP11B2 wild-type These results show that the replacement of lysine 152 by gluta-mine did not further enhance the cortisol production of the corresponding single or double mutant (Fig 3B), demonstrating that the 11b-hydroxylase activity of CYP11B2 seems to be unaffected by an amino-acid change
at position 152
D I S C U S S I O N
In humans, certain phenotypical abnormalities, such as essential hypertension, cardiovascular or endocrine diseases,
Fig 3 Assessment of 11b-hydroxylase activity and determination of 11b-hydroxylase capacity (A) Assessment of 11b-hydroxylase activity
of CYP11B2 variants expressed in COS-1 cells Cells were cotrans-fected with the indicated wild-type proteins, mutants or the empty vector pSVL as a negative control (Mock) and the cDNA of bovine Adx Data shown are means ± SEM of four separate transfections, each done in duplicate (B) Determination of 11b-hydroxylase capacity
of CYP11B2 mutants in relation to the wild-type enzyme, when incubated with 11-deoxycortisol The 11b-hydroxylation of 11-deoxycortisol catalysed by the mutated proteins is shown as percentage of CYP11B2 wild-type activity, fixed to 100% The values given are means ± SEM of four separate transfections, each performed in duplicate.
Trang 7are partially caused by genetic variations of CYP11B1 and/
or CYP11B2 [39,40] Due to this fact, it is of great interest to
obtain a deeper insight into the structural features
under-lying the determination of individual activities of these
enzymes Several structural determinants of human
11b-hydroxylase and aldosterone synthase have already been
elucidated in previous studies [17–19,41,42] These stuctures
are mainly located in the C-terminal regions of CYP11B1
and CYP11B2 So far, the role of distinct amino acids of the
N-terminal regions of human CYP11B isozymes has not
been studied extensively, although it is known that the
N-terminal domains of CYP11B1 and CYP11B2 differ
more from each other than the C-terminal ones, as also seen
in CYP11B isoforms of other mammals such as rat, hamster
or mouse (Fig 1) Therefore, our studies were focussed on
the residues at positions 112, 147 and 152 due to their
location in protein regions aligned with functionally
important areas of crystallized P450 enzymes [20,43]
(Fig 1) In this way, we intended to identify key
amino-acid residues of CYP11B2 implicated in the regulation of
individual reaction steps Swapping the amino acid at
position 147 from CYP11B2 to CYP11B1, led to a stronger
increase in the hydroxylation at the 11b-position of the
substrates than mutant I112P, with a smaller effect in case of
11-deoxycortisol compared with DOC The results obtained
with the single substitution (D147E) are in contrast to those
presented by Fisher et al [44] They reported no effect of
D147E on the 11b-hydroxylation of 11-deoxycortisol The
observed difference might be due to polymorphisms in the
CYP11B locus, different experimental conditions or
differ-ent steroid detection methods used by either group The
double mutant I112P/D147E exerted the most pronounced
enhancement of the 11b-hydroxylation of both substrates
(sixfold and 2.7-fold increases, as compared to the
CYP11B2 wild-type activity, in the case of DOC and
11-deoxycortisol, respectively), indicating an almost
addi-tive effect, but not a synergistic effect, of the two
substi-tutions The conversion of 11-deoxycortisol was not altered
by the replacement K152N, while the substitution slightly
influenced the enzymatic reaction steps of aldosterone
synthesis from DOC, suggesting only minor functional
relevance of lysine 152 in human CYP11B2
In contrast to the insertion of glutamic acid in position
147, the replacement I112P also increased the
18-hydroxy-lation activity (1.7-fold increase compared to the CYP11B2
wild-type enzyme; Fig 2B), in addition to significantly
enhancing the 11b-hydroxylation potential The absolute
amount of aldosterone formation was slightly enhanced for
all mutants (Fig 2A) However, the 18-oxidation activity
(Fig 2B) was either equal to the wild-type (D147E only), or
even slightly decreased (all other mutants) Although the
enzymatic activity remained unchanged by the intraspecies
replacement I112S (Table 2), the essential role of residue
112 of human aldosterone synthase was clearly shown by
mutant I112P This demonstrated the importance of the
correct residue at position 112 to ensure the species-specific
selectivity of substrate hydroxylation Thus, mutant I112P
produced an increased amount of 18-OH-B compared to
the wild-type This is in accordance with the observation
that rat CYP11B2, which contains proline instead of
isoleucine in position 112, produces higher levels of
18-OH-B than human CYP11B2 [45,46] I112 and S112 seem
to be conserved in the human enzymes to prevent the
strongly increased 18-OH-B production as seen when proline is inserted The position of residue 112 in the recently developed computer model of human CYP11B2 [32] (Fig 4) suggests structural modifications in the sub-strate access channel induced by its replacement Therefore, the observed significantly higher hydroxylation activities of the resulting mutants might be attributed to a faster and easier passage of the substrate, possibly caused by a substrate access channel enlargement (Fig 5) Also, the slightly reduced oxidation activity of these constructs suggests a facilitated intermediate dissociation from the active site before being oxidized at the 18th position Thus, the amino-acid replacement I112P located in the B-helix of the CYP11B2 model (Fig 4), might lead to a partial loss of enzymatic specificity This suggestion is in agreement with the observed contribution of helices A, B, B¢, F and G to the high specificity of other cytochromes P450 [47]
Our finding of an exclusive increase in the 11b-hydroxy-lation capacity of CYP11B2 by the replacement D147E indicates that residue 147 in the CYP11B isoform is involved in specific substrate conversion This conclusion agrees with earlier observations made by Bo¨ttner et al [36] who, while evaluating the functional relevance of the region flanked by amino acids 296 and 339 in human CYP11B1, found out that residues other than those investigated, appeared to be required for efficient 11b-hydroxylation The position of D147 on the protein surface of the CYP11B2 model (Fig 4) suggests, however, that an indirect influence exists, possibly via the mediation of structural modifications induced by redox partner binding or by the interaction with other proteins of the mitochondrial membrane, such as CYP11A1 It was previously shown that CYP11B1 and CYP11B2 were able to interact not only with the redox partner but also with CYP11A1 [37,48] As a consequence,
in the bovine system an enhancement of the 11b-hydroxy-lase activity was observed, whereas the aldosterone
synthe-Fig 4 Computer model of the three-dimensional structure of the human CYP11B2 The view is focused on the investigated amino acids and the heme-group of the P450 enzyme which are marked The arrow indicates the putative substrate access channel The putative I-helix, running through the molecule like a tunnel, is shown in the center.
Trang 8sizing activity was suppressed [49] However, this effect
seems to be species-specific, as in the human system no effect
of CYP11A1 on the product pattern has been found [37] As
the observed effects of mutant D147E investigated here can
be attributed to a conservative amino-acid exchange, the
side-chain size variations at position 147 seem to be
important A similarly crucial effect on the enzymatic
activity was demonstrated for mutant E198D of human
CYP11B2, leading to a reduction in aldosterone synthase
activity [50]
Taken together, our data clearly demonstrate for the first
time the functional relevance of N-terminal amino acids in
human CYP11B2 for substrate recognition In addition,
they provide evidence that amino acids that are placed
outside the active center (Fig 4) are essential for efficient
catalytic activity of human aldosterone synthase Our
observations are supported by data obtained with other
cytochrome P450 family members Amino-acid 4 of Gunn rat CYP2C11 has been shown to play an important role in testosterone hydroxylation, possibly in modulating sub-strate channel conformation [51], whereas Arg112 of CYP101, located on the protein surface, is essential for electron transfer from putidaredoxin to this cytochrome P450 enzyme [52]
However, it becomes apparent by our data that in contrast to studies on Dahl SR rats [22], the examined amino-acid replacements between the two human CYP11B isoenzymes in exon 3 exerted a more modulating effect than
a dramatically increasing effect on the enzymatic activity Nevertheless, it is conceivable that pathological abnormal-ities observed in patients with essential hypertension could
be caused by similar mutations as the analysed ones, due to their strongly increased 18-OH-B and increased aldosterone formation Our hypothesis is in accordance with the report
of Fardella et al [53], suggesting essential hypertension for the mutant K251R of CYP11B2 This mutation caused a 400% and 50–80% enhancement in the formation of 18-OH-B and aldosterone, respectively
In conclusion, the studies presented here are the first example of conferring CYP11B1 specific cortisol-producing function to the aldosterone synthase, thereby simultan-eously increasing the CYP11B2 specific catalytic activity Furthermore, we were able to demonstrate that the three enzymatic reaction steps of aldosterone synthesis could not only be modified independently, as evident with mutant D147E (where only the first reaction step was increased), but also differentially, as seen by mutant I112P (where the three hydroxylation steps were affected to a different amount) This indicates the possibility of dissecting the single reactions in aldosterone synthase activity by mutating defined positions in the primary structure, supporting the idea of divergent structural determinants of each reaction step
A C K N O W L E D G E M E N T S
This work was supported by a Grant from the Deutsche Forschungs-gemeinschaft to R B., Be 1343/2-6, and a visitor Grant from the Deutsche Forschungsgemeinschaft to N B We thank Michael Lisurek for assistance with computer modelling and Katharina Bompais for expert DNA sequencing We also express our gratitude to Achim Heinz for helpful discussion.
R E F E R E N C E S
1 Bernhardt, R (1996) Cytochrome P450 structure, function, and generation of reactive oxygen species Rev Physiol Biochem Pharmacol 127, 137–221.
2 Yanagibashi, K., Haniu, M., Shively, J.E., Shen, W.H & Hall, P (1986) The synthesis of aldosterone by the adrenal cortex Two zones (fasciculata and glomerulosa) possess one enzyme for 11 beta-, 18-hydroxylation, and aldehyde synthesis J Biol Chem.
261, 3556–3562.
3 Nonaka, Y., Takemori, H., Halder, S.K., Sun, T., Ohta, M., Hatano, O., Takakusu, A & Okamoto, M (1995) Frog cytochrome P-450 (11 beta, aldo), a single enzyme involved in the final steps of glucocorticoid and mineralocorticoid biosynthesis Eur J Biochem 229, 249–256.
4 Morohashi, K., Yoshioka, H., Gothoh, O., Okada, Y., Yamam-oto, K., Miyata, T., Sogawa, K., Fujii-Kuriyama, Y & Omura, T (1987) Molecular cloning and nucleotide sequence of DNA of
Fig 5 Putative structures of the substrate access channel of human
CYP11B2 wild-type enzyme (A) and the two mutants I112S (B) and
I112P (C) The heme-groups and the analysed amino acids in position
112 are displayed in capped sticks.
Trang 9mitochondrial cytochrome P-450 (11 beta) of bovine adrenal
cortex J Biochem 102, 559–568.
5 Mitani, F., Shimizu, T., Ueno, R., Ishimura, Y., Izumi, S.,
Komatsu, N & Watanabe, K (1982) Cytochrome P-45011 beta
and P450scc in adrenal cortex: zonal distribution and
intrami-tochondrial localization by the horseradish peroxidase-labeled
antibody method J Histochem Cytochem 30, 1066–1074.
6 Mornet, E., Dupont, J., Vitek, A & White, P.C (1989)
Charac-terization of two genes encoding human steroid 11
beta-hydrox-ylase (P-450 11b ) J Biol Chem 264, 20961–20967.
7 Kawamoto, T., Mitsuuchi, Y., Ohnishi, T., Ichikawa, Y.,
Yokoyama, Y., Sumimoto, H., Toda, K., Miyahara, K.,
Kuribayashi, I & Nakao, K., et al (1990) Cloning and expression
of a cDNA for human cytochrome P-450 aldo as related to primary
aldosteronism Biochem Biophys Res Commun 173, 309–316.
8 Domalik, L.J., Chaplin, D.D., Kirkman, M.S., Wu, R.C., Liu, W.,
Howard, T.A., Seldin, M.F & Parker, K.L (1991) Different
iso-zymes of mouse 11 beta-hydroxylase produce mineralocorticoids
and glucocorticoids Mol Endocrinol 5, 1853–1861.
9 Nonaka, Y., Matsukawa, N., Morohashi, K., Omura, T.,
Ogihara, T., Teraoka, H & Okamoto, M (1989) Molecular
cloning and sequence analysis of cDNA encoding rat adrenal
cytochrome P-450 11b FEBS Lett 255, 21–26.
10 Mukai, K., Imai, M., Shimada, H & Ishimara, Y (1993) Isolation
and characterization of rat CYP11B genes involved in late steps
of mineralo- and glucocorticoid syntheses J Biol Chem 268,
9130–9137.
11 Mitani, F (1979) Cytochrome P450 in adrenocortical
mitochon-dria Molec Cell Biochem 24, 21–43.
12 Pascoe, L., Curnow, K.M., Slutsker, L., Connell, J.M.C., Speiser,
P.W., New, M.I & White, P.C (1992)
Glucocorticoid-suppress-ible hyperaldosteronism results from hybrid genes created by
unequal crossovers between CYP11B1 and CYP11B2 Proc Natl
Acad Sci USA 89, 8327–8331.
13 Ogishima, T., Shibata, H., Mitani, F., Suzuki, H., Saruta, T &
Ishimura, Y (1991) Aldosterone synthase cytochrome P-450
expressed in the adrenals of patients with primary aldosteronism.
J Biol Chem 266, 10731–10734.
14 Sagara, Y., Takata, Y., Miyata, T., Hara, T & Horiuchi, T (1987)
Cloning and sequence analysis of adrenodoxin reductase cDNA
from bovine adrenal cortex J Biochem (Tokyo) 102, 1333–1336.
15 Grinberg, A.V., Hannemann, F., Schiffler, B., Mu¨ller, J.,
Heinemann, U & Bernhardt, R (2000) Adrenodoxin: structure,
stability, and electron transfer properties Proteins 40, 590–612.
16 Lifton, R.P., Dluhy, R.G & Powers, M (1992) Chimaeric 11
beta-hydroxylase/aldosterone synthase gene causes
glucocorticoid-remediable aldosteronism and human hypertension Nature 355,
262–265.
17 Bo¨ttner, B., Schrauber, H & Bernhardt, R (1996) Engineering a
mineralocorticoid- to a glucocorticoid-synthesizing cytochrome
P450 J Biol Chem 271, 8028–8033.
18 Curnow, K.M., Mulatero, P., Emeric-Blanchouin, N.,
Aupetit-Faisant, B., Corvol, P & Pascoe, L (1997) The amino acid
substitutions Ser288Gly and Val320Ala convert the cortisol
producing enzyme, CYP11B1, into an aldosterone producing
enzyme Nat Struct Biol 4, 32–35.
19 Bo¨ttner, B & Bernhardt, R (1996) Changed ratios of
glucocor-ticoids/mineralocorticoids caused by point mutations in the
putative I-helix regions of CYP11B1 and CYP11B2 Endocr Res.
22, 455–461.
20 Graham-Lorence, S.E & Peterson, J.A (1996) Structural
align-ments of P450s and extrapolations to the unknown Methods
Enzymol 272, 315–325.
21 Ridderstro¨m, M., Masimirembwa, C., Trump-Kallmeyer, S.,
Ahlefelt, M., Otter, C & Andersson, T.B (2000) Arginines 97 and
108 in CYP2C9 are important determinants of the catalytic
function Biochem Biophys Res Commun 270, 983–987.
22 Cover, C.M., Wang, J.M., St-Lezin, E., Kurtz, T.W & Mellon, S.H (1995) Molecular variants in the P450c11AS gene as deter-minants of aldosterone synthase activity in the Dahl rat model of hypertension J Biol Chem 270, 16555–16560.
23 Peterson, J.A & Graham, S.E (1998) A close family resemblance: the importance of structure in understanding cytochromes P450 Structure 6, 1079–1085.
24 Graham-Lorence, S & Peterson, J.A (1996) P450: structural similarities and functional differences FASEB J 10, 206–214.
25 Kawamoto, T., Mitsuuchi Y., Toda, K., Miyahara, K., Yokoyama, Y., Nakao, K., Hosoda, K., Yamamoto, Y., Imura,
H & Shizuta, Y (1990) Cloning of cDNA and genomic DNA for human cytochrome P-45011 beta FEBS Lett 269, 345–349.
26 Sambrook, J., Fritsch, E.F & Maniatis, T (1989) Molecular Cloning: a Laboratory Manual, 2nd edn Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York.
27 Zuber, M.X., Mason, J.I., Simpson, E.R & Waterman, M.R (1988) Simultaneous transfection of COS-1 cells with mitochon-drial and microsomal steroid hydroxylases: incorporation of a steroidogenic pathway into nonsteroidogenic cells Proc Natl Acad Sci USA 85, 699–703.
28 Laemmli, U.K (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4 Nature 227, 680–685.
29 Thompson, J.D., Higgins, D.G & Gibson, T.J (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap pen-alties and weight matrix choice Nucleic Acids Res 22, 4673–4680.
30 Rost, B & Sander, C (1993) Prediction of protein structure at better than 70% accuracy J Mol Biol 232, 584–599.
31 Berman, H.M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T.N., Weissig, H., Shindyalov, I.N & Bourne, P.E (2000) The protein data bank Nucleic Acids Res 28, 235–242.
32 Belkina, N.V., Lisurek, M., Ivanov, A.S & Bernhardt, R (2001) Modelling of 3D-structures of cytochromes P450 11B1 and 11B2.
J Inorg Biochem 87, 197–207.
33 Laskowski, R.A., MacArthur, M.W., Moss, D.S & Thornton, J.M (1993) PROCHECK: a program to check the stereochemical quality of protein structures J Appl Cryst 26, 283–291.
34 Sippl, M.J (1993) Recognition of errors in three-dimensional structures of proteins Proteins 17, 355–362.
35 Gotoh, O (1992) Substrate recognition sites in cytochrome P450 family 2 (CYP2) proteins inferred from comparative analyses of amino acid and coding nucleotide sequences J Biol Chem 267, 83–90.
36 Bo¨ttner, B., Denner, K & Bernhardt, R (1998) Confer-ring aldosterone synthesis to human CYP11B1 by replacing key amino acid residues with CYP11B2-specific ones Eur J Biochem.
252, 458–466.
37 Cao, P.-R & Bernhardt, R (1999) Interaction of CYP11B1 (cytochrome P-450 11b ) with CYP11A1 (cytochrome P-450 scc ) in COS-1 cells Eur J Biochem 262, 720–726.
38 Cao, P.-R & Bernhardt, R (1999) Modulation of aldosterone biosynthesis by adrenodoxin mutants with different electron transport efficiencies Eur J Biochem 265, 152–159.
39 White, P.C., Curnow, K.M & Pascoe, L (1994) Disorders of steroid 11b-hydroxylase isozymes Endocr Rev 15, 421–438.
40 Peter, M., Dubuis, J.-M & Sippell, W.G (1999) Disorders of the aldosterone synthase and steroid 11b-hydroxylase deficiencies Horm Res 41, 211–222.
41 White, P.C., Dupont, J., New, M.I., Leiberman, E., Hochberg, Z.
& Rosler, A (1991) A mutation in CYP11B1 (Arg448 fi His) associated with steroid 11b-hydroxylase deficiency in Jews of Moroccan origin J Clin Invest 87, 1664–1667.
42 Geley;, S., Jo¨hrer, K., Peter, M., Denner, K., Bernhardt, R., Sippell, W.G & Kofler, R (1995) Amino acid substitution R384P
in aldosterone synthase causes corticosterone methyloxidase type I deficiency J Clin Endocrinol Metab 80, 424–429.
Trang 1043 Hasemann, C.A., Kurumbail, R.G., Boddupalli, S.S., Peterson,
J.A & Deisenhofer, J (1995) Structure and function of
cytochromes P450: a comparative analysis of three crystal
struc-tures Structure 3, 41–62.
44 Fisher, A., Fraser, R., Mc-Connell, J & Davies, E (2000) Amino
acid residue 147 of human aldosterone synthase and
11beta-hydroxylase plays a key role in 11beta-hydroxylation J Clin.
Endocrinol Metab 85, 1261–1266.
45 Nonaka, Y., Fujii, T., Kagawa, N., Waterman, M.R., Takemori,
H & Okamoto, M (1998) Structure/function relationship of
CYP11B1 associated with Dahl’s salt-resistant rats – expression of
rat CYP11B1 and CYP11B2 in Escherichia coli Eur J Biochem.
258, 869–878.
46 Nonaka, Y., Fujii, T., Bernhardt, R & Okamoto, M (1998)
Amino acid residues in I- and K-helices of rat CYP11B1 and
CYP11B2 are important in expression of 18-hydroxylation
activ-ity Endocr Res 24, 615–618.
47 Graham, S.E & Peterson, J.A (1999) How similar are P450s and
what can their differences teach us? Arch Biochem Biophys 369,
24–29.
48 Schwarz, D., Chernogolov, A & Kisselev, P (1999) Complex
formation in vesicle-reconstituted mitochondrial cytochrome P450
systems (CYP11A1 and CYP11B1) as evidenced by rotational
diffusion experiments using EPR and ST-EPR Biochemistry 38,
9456–9464.
49 Ikushiro, S., Kominami, S & Takemori, S (1992) Adrenal P-450scc modulates activity of P-45011 beta in liposomal and mitochondrial membranes Implication of P-450scc in zone specificity of aldosterone biosynthesis in bovine adrenal J Biol Chem 267, 1464–1469.
50 Portrat-Doyen, S., Tourniaire, J., Richard, O., Mulatero, P., Aupetit-Faisant, B., Curnow, K.M., Pascoe, L & Morel, Y (1998) Isolated aldosterone synthase deficiency caused by simul-taneous E198D and V386A mutations in the CYP11B2 gene.
J Clin Endocrinol Metab 83, 4156–4161.
51 Biagini, C.P., Philpot, R.M & Celier, C.M (1999) Nonsubstrate recognition site residues are involved in testosterone hydroxylation
by cytochrome P450 CYP 2C11 Arch Biochem Biophys 361, 309–314.
52 Koga, H., Sagara, Y., Yaoi, T., Tsujimura, M., Nakamura, K., Sekimizu, K., Makino, R., Shimada, H., Ishimura, Y & Yura, K.,
et al (1993) Essential role of the Arg112 residue of cytochrome P450cam for electron transfer from reduced putidaredoxin FEBS Lett 331, 109–113.
53 Fardella, C.E., Rodriguez, H., Hum, D.W., Mellon, S.H & Miller, W.L (1995) Artificial mutations in P450c11AS (aldosterone synthase) can increase enzymatic activity: a model for low-renin hypertension? J Clin Endocrinol Metab 80, 1040– 1043.