help button home button Endocrine Society Molecular Endocrinology ENDO 08 Sessions Library
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Molecular Endocrinology, doi:10.1210/me.2006-0172
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Robins, T.
Right arrow Articles by Persson, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Robins, T.
Right arrow Articles by Persson, B.
Molecular Endocrinology 20 (11): 2946-2964
Copyright © 2006 by The Endocrine Society

Molecular Model of Human CYP21 Based on Mammalian CYP2C5: Structural Features Correlate with Clinical Severity of Mutations Causing Congenital Adrenal Hyperplasia

Tiina Robins1, Jonas Carlsson1, Maria Sunnerhagen, Anna Wedell and Bengt Persson

Department of Molecular Medicine and Surgery (T.R., A.W.), Center for Molecular Medicine, L8:02, Karolinska Institutet/Karolinska University Hospital, S-171 76 Stockholm, Sweden; The Department of Physics, Chemistry and Biology (IFM Bioinformatics) (J.C., B.P.) and Molecular Biotechnology (M.S.), Linköping University, S-581 83 Linköping, Sweden; Department of Cell and Molecular Biology (B.P.), Programme for Genomics and Bioinformatics, Karolinska Institutet, S-171 77 Stockholm, Sweden

Address all correspondence and requests for reprints to: Tiina Robins, Department of Molecular Medicine and Surgery, Center for Molecular Medicine (CMM) L8:02, Karolinska Institutet/Karolinska University Hospital, S-171 76 Stockholm, Sweden. E-mail: tiina.robins{at}ki.se.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Enhanced understanding of structure-function relationships of human 21-hydroxylase, CYP21, is required to better understand the molecular causes of congenital adrenal hyperplasia. To this end, a structural model of human CYP21 was calculated based on the crystal structure of rabbit CYP2C5. All but two known allelic variants of missense type, a total of 60 disease-causing mutations and six normal variants, were analyzed using this model. A structural explanation for the corresponding phenotype was found for all but two mutants for which available clinical data are also discrepant with in vitro enzyme activity. Calculations of protein stability of modeled mutants were found to correlate inversely with the corresponding clinical severity. Putative structurally important residues were identified to be involved in heme and substrate binding, redox partner interaction, and enzyme catalysis using docking calculations and analysis of structurally determined homologous cytochrome P450s (CYPs). Functional and structural consequences of seven novel mutations, V139E, C147R, R233G, T295N, L308F, R366C, and M473I, detected in Scandinavian patients with suspected congenital adrenal hyperplasia of different severity, were predicted using molecular modeling. Structural features deduced from the models are in good correlation with clinical severity of CYP21 mutants, which shows the applicability of a modeling approach in assessment of new CYP21 mutations.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
CYTOCHROME P450 (CYP) enzymes constitute a superfamily of more than 1300 heme-containing proteins that are physiologically important in numerous organisms ranging from protists to humans (1, 2, 3). Most CYPs are monooxygenases that catalyze hydroxylation reactions by incorporating a single oxygen atom from molecular dioxygen into their substrates, thus increasing their polarity, whereas the second oxygen atom is reduced to water. The CYPs are divided into families and subfamilies based on amino acid sequence similarity. In mammals, 22 families have been described that can be functionally subdivided into those that mainly detoxify xenobiotic compounds, and those that are essential in the metabolism of endogenous substrates such as steroid hormones, prostaglandins, leukotrienes, fatty acids, and bile acids (4). The CYP superfamily can also be subdivided into four classes based on the nature of their redox partner (5). Class I CYPs are found in most bacteria and in the inner mitochondrial membranes of eukaryotes. In contrast, class II CYPs are mainly found in the animal kingdom where they are bound to the cytosolic membrane of the smooth endoplasmic reticulum (ER). These ER enzymes are part of a two-component system and utilize electrons transferred from reduced nicotinamide adenine dinucleotide phosphate (NADPH) to the enzyme via cytochrome P450 NADPH oxidoreductase, a flavin adenine nucleotide- and flavin mononucleotide-containing flavoprotein (6). Certain drug-metabolizing CYPs in the ER receive electrons from NADH via NADH-cytochrome b5 reductase and cytochrome b5 (7, 8). Class III and IV CYPs do not require an exogenous source of electrons or the intervention of an electron carrier.

Congenital adrenal hyperplasia (CAH) comprises a group of recessively inherited syndromes characterized by impaired cortisol secretion, increased feedback secretion of ACTH from the pituitary gland, and subsequent hyperplasia of the adrenal cortex. In humans, CAH has been correlated with mutations in the genes encoding four different CYPs involved in the synthesis of the steroid hormone cortisol in the adrenal cortex (CYP11A1, CYP11B1, CYP17, and CYP21A2). CYP11A1 and CYP11B1 are located in the mitochondria, whereas CYP17 and CYP21A2 are microsomal, belonging to the class II enzymes. Depending on the enzyme affected, synthesis of the other adrenal steroid hormones, mineralocorticoids and sex steroids, are deranged in different ways. The vast majority of all cases of CAH are due to steroid 21-hydroxylase (CYP21A2) deficiency (21OHD), which causes reduced production of cortisol and aldosterone, together with an excessive secretion of adrenal androgens (9). CAH due to 21OHD has an incidence of around 1/10 000 (10) and presents with a wide spectrum of severity. The syndrome has traditionally been divided into three clinical forms, the most severe form being salt wasting (SW), with a life-threatening salt loss during the neonatal period together with prenatal virilization of external genitalia in affected females. The simple virilizing (SV) form does not include overt SW: affected girls present with prenatal genital virilization and boys are often diagnosed due to accelerated growth and precocious pseudopuberty during early childhood. The mildest form, nonclassic (NC) CAH, is usually not recognized until later in childhood or adulthood, with hyperandrogenic symptoms such as accelerated somatic growth, precocious pseudopuberty, hirsutism, and/or decreased fertility.

The CYP21 gene is part of a complicated structure, referred to as the RCCX-module located in the human leukocyte antigen class III locus on chromosome 6 (band 6p21.3) (11). Also present in this locus, approximately 30 kb upstream of CYP21 is a 98% identical pseudogene (CYP21A1/CYP21P), which does not encode a functionally active protein due to numerous deleterious mutations that have been gathered throughout the gene during evolution. The proximity of CYP21 and CYP21P predisposes to sequence exchanges between the two genes. Misalignment during meiosis followed by recombination can lead to a net deletion of CYP21, and deleterious sequences can be transferred from CYP21P to CYP21 in apparent gene-conversion events. CYP21 gene deletion together with nine pseudogene-derived mutations are responsible for around 95% of all alleles involved in CAH. The remaining 5% of affected alleles harbor rare mutations, the number of which has increased dramatically during the last decade (12). To date, more than 80 non-pseudogene-derived disease-causing mutations have been reported. Around 50 of these are missense mutations or combinations of such. In CAH due to 21OHD, there are clear genotype-phenotype relationships. Thus, the degree of disease presentation is largely dependent on each patient’s underlying combination of mutations, and CYP21 genotyping has therefore become a valuable diagnostic tool. Activities of mutant enzymes, when expressed in cultured cells, also generally correlate with disease severity. All reported disease-causing CYP21A2 mutations together with (when known) associated clinical disease severity and in vitro enzyme activity, are listed on the home page of The Human Cytochrome P450 (CYP) Allele Nomenclature at http://www.imm.ki.se/CYPalleles/cyp21.htm.

The substantial number of known missense mutations, for which corresponding clinical phenotypes as well as in vitro enzyme activities are known, provides a unique background enabling analysis of clinically relevant functional and structural relationships of the CYP21 protein. A number of different prokaryotic CYP enzymes were initially crystallized and their structures were determined. The bacterial CYP102 (formerly called P450BM-3) was originally proposed as the most accurate prototype for microsomal CYPs (13) and, in the absence of determined mammalian P450 cytochrome structures, it was used as the main template for modeling mammalian CYP structures (14) including human CYP21 (15, 16). A model of the CYP21 active site using the bacterial CYP101 as the template has also been reported (17). However, in the past few years, an increasing number of CYP structures have been resolved and deposited in the Protein Data Bank (PDB) (18) of which five are mammalian, including three human proteins (CYP2C5, CYP2B4 and CYP2C9, CYP2C8, CYP3A4, respectively) (19, 20, 21, 22, 23, 24). The first eukaryotic P450 structure solved, CYP2C5, is a progesterone 21-hydroxylase expressed in rabbit microsomes. Because CYP21 has progesterone as one of its substrates, it is intriguing to compare CYP21 with respect to CYP2C5. Sequence comparisons also reveal that CYP2C5 is among the closest homologs of CYP21. Furthermore, extensive structural evaluation of CYP2C5 has been undertaken (21, 25, 26, 27, 28), as it is the first membrane-bound CYP structure determined by x-ray diffraction. Thus, it represents an improved template structure for the modeling of human cytochrome P450 enzymes.

In the present study bioinformatic techniques have been used to analyze naturally occurring CYP21 mutants and normal variants to improve understanding of structure-function relationships of the protein and to better understand the molecular causes of CAH due to 21OHD. A structural model of human CYP21 was calculated based upon the known structure of rabbit CYP2C5. Furthermore, individual models were calculated to investigate molecular mechanisms of all known allelic variants of CYP21. Finally, seven novel CYP21 missense mutations detected in CAH patients, V139E, C147R, R233G, T295N, L308F, R366C, and M473I, are reported, and the present models have been used to predict functional and structural consequences of each mutation.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Molecular Model of Human CYP21
The homology model shows that the structure of CYP21 is compatible with the fold of CYP2C5 (Fig. 1Go). Most of the secondary structure elements are conserved, with the following exceptions. In the CYP21 model, the end of helix B does not fulfil the criteria of the Database of Secondary Structure Assignments for Proteins algorithm as implemented in Internal Coordinate Mechanics (ICM) used to define the regular secondary structure elements. Furthermore, the short two-residue ß-sheet 1–5 is not considered to form a ß-structure. Helix F is shorter in our model compared with the CYP2C5 structure, partly due to a Lys/Pro substitution at position 188. Finally, the region 215–219 of helix G is only forming a loop in our model. Based upon the bacterial CYP102 (13), differences from the former CYP21 model, regarding the ß-sheets, are that ß1 consists of only four strands, lacking the last ß1–5.


Figure 1
View larger version (69K):
[in this window]
[in a new window]
  		Fig. 1. Molecular Model of Human CYP21

Human CYP21 is shown in ribbon representation with notations of secondary structure elements. Residues affected by SW mutants are colored red, those affected by SV mutants are blue, and those affected by NC mutants are brown. Stick representations are used for the heme group in red and the steroid substrate in green.

 
Heme and Steroid Binding
The heme-binding site is sandwiched between helix I on the distal side (interior of the protein), and helix L on the proximal side to the protein surface, in homology to that of other cytochrome CYPs. In further support of this, most (~60%) of the heme-binding residues are conserved between CYP21 and CYP2C5. The steroid-binding site was defined by homology with CYP51, for which the complex with its steroid substrate (estriol) bound has been structurally determined (29). Totally 16 residues were found within a 5Å sphere around the steroid: S108 and L109 in the loop between helix B' and C, residues 203–207 located in the loop connecting helix F and G', D287 and residues 290–292 and 294–295 in the central part of helix I, and finally V359, L363 and P364 in the loop between helix K and ß1–4. All 16 substrate-binding residues are indicated in Fig. 3Go. The identified residues fall generally into the six different substrate recognition sites (SRS) defined by Gotoh (30) with only minor shifts of the secondary structure elements for SRS1–3. The majority of these residues are hydrophobic, contributing to hydrophobic interactions with the substrate. Docking calculations with progesterone showed that this steroid fits the substrate-binding site properly.


Figure 3
View larger version (31K):
[in this window]
[in a new window]
  		Fig. 3. The CYP21 Amino Acid Sequence with Important Properties Marked

Mutations are indicated in the sequence and shown below the sequence with color-coding according to their clinical severity. Regular secondary structure elements are given above the sequence. Conserved amino acid residues are in bold, and binding sites for heme and steroid are shown with red and blue circles below the sequence.

 
Redox Partner Interaction
CYP21 interacts with cytochrome P450 NADPH oxidoreductase for coenzyme regeneration (31). In CYP2B4, 12 positively charged amino acid residues are considered important for binding of the cytochrome P450 reductase (20). Of these 12, nine are structurally conserved in CYP21 with the same charge (Table 1Go). Six of the structurally conserved amino acid residues are closely located. These amino acid residues are K117 and R121 in helix C, R132 in helix D, R414 in a loop between K'' and K''', R426 in helix K''', and R435 in helix L (Fig. 2Go). Furthermore, for the homolog CYP17, structural studies have revealed that R347, corresponding to R339 in CYP21, is a candidate for redox partner interaction (32, 33) (Table 1Go). All these seven residues are superficially located and clustered spatially in a plane, forming a possible redox partner interaction site, located parallel to the heme group at the opposite side of the steroid (Fig. 2Go).


View this table:
[in this window]
[in a new window]
  		Table 1. Amino Acid Residues Believed to be Important in the Redox Partner Interaction in CYP2B4 and CYP17 (32 33 ) and the Corresponding Amino Acid Residues in CYP21

 

Figure 2
View larger version (60K):
[in this window]
[in a new window]
  		Fig. 2. Human CYP21 with the Putative Redox Partner Interacting Residues

The putative redox partner interacting residues are shown with blue side chains. The additional scattered amino acid residues that have been ascribed a possible redox partner binding role are indicated in cyan. The membrane-binding sites derived from the Optimal Docking Area calculations are shown in red. Stick representations are used for the heme group in red and the steroid substrate in green.

 
Large Hydrophobic Areas
The search for large hydrophobic areas using the Optimal Docking Area method (34) revealed three segments, corresponding to amino acid residues 30–42, 63–66, and 211–219. These three sequences together form a hydrophobic region that is exposed to the surface adjacent to the N-terminal transmembrane part (Fig. 2Go). Two of the segments have been postulated to be membrane interacting (21). The first loop extends from residue 31 to the {alpha}-helix A, the second loop extends between strand 1 and 2 of ß-sheet 1 (ß1–1 and ß1–2), and the third loop is positioned between a break in the {alpha}-helix G, dividing the helix into what we denote G' and G, and helix G'' (Fig. 3Go). Table 2Go shows the sequences and their hydrophobicity. It can be observed that this hydrophobic region contains 85% hydrophobic amino acids that are exposed to the surface, which is far more than what is seen for the entire protein, i.e. only 45% exposed hydrophobic amino acid residues. Such a highly hydrophobic surface would contribute negatively to the stability of the protein, unless it provides hydrophobic interactions, e.g. for membrane binding.


View this table:
[in this window]
[in a new window]
  		Table 2. The Amino Acid Sequences in CYP21 that form a Hydrophobic Patch, Possibly Binding to the ER Membrane

 
Analyses of Species Variants
By analyzing the multiple sequence alignment, information about amino acid residues that are conserved between species to a high degree were extracted (Table 3Go). Such residues are considered to be of critical importance for structure and function. It can be seen that most of the residues involved in the putative heme- and steroid-binding regions are highly conserved. Furthermore, at the steroid-binding site, all hydrophobic residues (nine of 16) have their hydrophobicities conserved. Six residues are identical whereas three are isoleucine/valine or leucine/isoleucine exchanges.


View this table:
[in this window]
[in a new window]
  		Table 3. Residue Conservation in CYP21

 
Human Mutants
There are many known allelic forms of CYP21 that give rise to protein variants that have been functionally characterized, both by clinical evaluations of patients and/or by measuring enzyme activities of recombinant proteins in cultured cells. Hitherto, 68 missense mutations (including six normal variants) have been reported, affecting residues distributed along the entire protein chain with no obvious hot spot region, although the majority (~70%) are located in the C-terminal half (Fig. 3Go). The location of all mutants, with the exception of M1I and A15T, that are in a transmembrane helix and not included in the three-dimensional model, are shown in Fig. 3Go, where the affected positions are marked with different colors according to resulting CAH severity. Information on secondary structure elements, heme- and steroid-binding sites, and residue conservation is also included. Table 3Go shows that the majority of mutations affecting conserved amino acid residues are correlated with impaired enzyme function and severe CAH, whereas alteration of nonconserved residues is correlated with less severe CAH. It can also be observed that all normal CYP21 variants affect nonconserved residues, as expected.

Molecular modeling has been carried out in an attempt to explain the structural effects for all reported mutations resulting in amino acid substitutions (Table 4GoGoGo). Predicted distances to the heme and steroid are also given together with hydrophobicity changes. Amino acid residues present in the protein core are usually of structural importance, and they are consequently often conserved among different species. More than 50% of all identical or closely homologous residues compared with CYP2C5 are found in the inner core of our CYP21 model (accessibility to surface ≤20%) indicating a conserved structural core. Furthermore, it can be seen that nearly all of the SW and SV mutants affect residues that are buried in the three-dimensional structure and strictly conserved between higher eukaryotes (Table 3Go). Many of these severe mutants are also located close to the heme- and steroid-binding sites (Table 3Go). The remaining severe mutants mainly cause large changes in residue hydrophobicity or size.


View this table:
[in this window]
[in a new window]
  		Table 4. List of Known Mutants of Human CYP21A2 Together with Data from the Molecular Models

 

View this table:
[in this window]
[in a new window]
  		Table 4A. Continued

 

View this table:
[in this window]
[in a new window]
  		Table 4B. Continued

 
Correlation between Mutation Class and Structural Stability
Investigations of whether there were any correlation between the mutation class and the stability of the protein were also performed. All mutations classified as N (normal variants, n = 6), SV (associated with SV CAH, n = 11), or SW (associated with SW CAH, n = 16), giving a total of 33 mutations were studied. For each mutant, three independent calculations were made, and the median energy value was obtained. A graph illustrating the cumulative proportion for each class of mutant plotted against the energy value is shown in Fig. 4AGo. There is a clear separation between the three curves, representing the different classes N, SV, and SW, suggesting an inverse correlation between the stability of the protein and clinical severity of the mutant. The cumulative proportion of class N mutants reaches 100% at much lower energy than the class SW mutants. The SV mutants are found on a curve in between those of the N and SW mutants. The curves are clearly separated (Fig. 4AGo), but from the viewpoint of the energy value, the curves partly overlap, and therefore the energy value alone cannot be used to classify an individual mutant. However, the curves can still be used to judge the likelihood for the mutant to belong to a separate class. In combination with other types of information, e.g. residue conservation, distances to functional sites, accessibility, hydrophobicity shifts, and size differences, the clinical severity of the mutant can be estimated. For example, 50% of all salt-wasting mutants have higher energy than all normal variants. Thus, if the structure corresponding to a new, uncharacterized mutant has an energy value in this range, there is a high probability that the mutant is associated with a severe phenotype.


Figure 4
View larger version (26K):
[in this window]
[in a new window]
  		Fig. 4. Protein Stability vs. Mutant Severity Classes for CYP21 and G6PD

The curves illustrate the distribution of energy values calculated for the molecular models corresponding to mutants of three different severity classes in CYP21 and G6PD. The cumulative fraction is plotted against the energy, i.e. on the y-axis are given the fraction of the mutants with an energy value equal to or lower than the energy value on the x-axis. Panel A shows the cumulative energy of the CYP21 mutants, whereas B shows that of the G6PD mutants. In both cases, the different curves are separated with the normal variants having the lowest energy and those corresponding to the severe mutants displaying the highest energy. For G6PD, the curves corresponding to severe and <10% activity are not clearly separated and they are also crossing each other, compatible with difficulties in classification of low enzyme activities. Furthermore, single mutants can affect the active site with only minor effects on the stability of the protein in general.

 
To investigate whether this inverse correlation between structural stability and functional impairment is a general phenomenon, searches were performed for similar cases by looking for proteins with known three-dimensional structures and with several documented allelic variants. These searches resulted in 44 proteins with more than 40 variants that were manually checked for notations that could be translated into classes of mutant severity, in a similar way as with CYP21 and CAH. The proteins found were: von Hippel-Lindau disease tumor suppressor (pVHL), glucose-6-phosphate 1-dehydrogenase (G6PD), coagulation factor VIII, coagulation factor IX, and antithrombin-III. Analysis of these proteins gave similar results as for CYP21, although for some proteins the curves were somewhat less separated, but there were no contradictions. As an example, the results for G6PD are shown in Fig. 4BGo. Thus, it seems to be a general phenomenon that the energy values reflect mutant severity.

Prediction of Severity of Novel Mutants
Seven novel mutations in CYP21 have also been investigated (Table 5Go). By combining information from modeling and residue properties, predictions of to which severity class each mutation belongs have been done. The information used for prediction was energy, surface accessibility, evolutionary conservation, changes in polarity, distances to heme and steroid, and structural information from visual inspection of the model.


View this table:
[in this window]
[in a new window]
  		Table 5. List of New Mutants of Human CYP21A2 not yet Functionally Characterized

 
In Table 5Go, predicted severities based on the suggested molecular mechanisms are listed. The V139E mutation is completely buried inside the structural core; it causes a change from a nonpolar to charged amino acid; and it affects a conserved residue. Thus, it is predicted that V139E correlates with a severe form of CAH, although it is hard to distinguish between the SV and SW forms. For C147R the predictions are similar to those for V139E; in addition, it could introduce repulsions to nearby charged amino acid residues, destabilizing the structure. R233G is presumably more harmless as it is exposed to the surface. However, it is a conserved residue, and introducing glycine inside an {alpha}-helix might cause instabilities. The conclusion is that it fits in the NC class. T295N is most likely deleterious (SW) because it is close to both the heme and the steroid. L308F is buried but leads to only a small polarity change, most likely making it moderately harmful (NC or SV). R366C causes a considerable polarity change that leads to lost electrostatic interactions. However, the residue is not conserved, and it has some access to the surface, which implies that it can be either SV or SW. A classical example of a harmless alteration (N) is M473I: it affects a fully exposed residue, which is only partly conserved, and causes a rather low polarity change, and the allelic form (isoleucine) is identical to the corresponding residue in the mouse and rat sequence.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
Human CYP21A2 is the cytochrome P450 protein with the largest number of known naturally occurring mutants (4), and associated functional information is available for the majority of these (12), offering valuable information about important residues. CAH due to CYP21 deficiency is unique among diseases in that it presents with a very wide spectrum of severity, ranging from a life-threatening neonatal disease to subtle hormone imbalances that present in adult life. Genetic diagnostics has become a valuable instrument in the management of CAH patients, because the established genotype-phenotype relationships enable prognostic evaluations to be made in each case. Especially in association with neonatal screening programs, when the diagnosis of CAH is established before the onset of symptoms, it is crucial to determine the degree of disease severity in each child to optimize treatment. In addition, CAH is recessively inherited, and disease severity reflects the function of the mildest mutation present in each patient. Thus, when a novel mutation occurs in trans with a mild allele, it is not possible to classify this mutant based on clinical evaluation of the patient.

Functional characterization of new mutants using site-directed mutagenesis followed by in vitro expression has been done on numerous mutants and has shown that there is generally a good agreement between degree of enzymatic impairment in vitro and clinical disease presentation [for overview see http://www.imm.ki.se/CYPalleles/cyp21.htm (12)]. However, these are laborious and time-consuming methods and cannot provide clinically useful information within a reasonable time frame when a novel mutation is encountered in a patient. Therefore, the use of bioinformatic techniques would be valuable to predict consequences of mutations.

Molecular Model of Human CYP21A2
The present extensive evaluation of the correlation between clinical and structural data shows that the molecular model of human CYP21A2 can be used to judge the physiological effects of patient-occurring mutations. To obtain an accurate structural model, a sequence identity of more than 30% between the known structure and the modeled sequence is usually required (35). However, even though the residue identity between the CYP21 sequence and the sequence of the template amounts to only 29%, it should be remembered that in the cytochrome P450 superfamily, low sequence identity in pair-wise comparisons is common. In spite of this, the three-dimensional structures experimentally determined so far are clearly similar, especially around the heme-binding region, where the structures are easily superimposable. Thus, it can be expected that the core region of our model is valid, even though some uncertainties might occur at the exterior loops. However, this would not change our conclusions regarding heme and steroid binding. The same has been shown for other enzyme families with distant relationships, e.g. the short-chain dehydrogenases/reductases, with pair-wise identity of typically 20% (36).

The N-terminal membrane-spanning part was not modeled, because no experimentally solved three-dimensional structure has any structural information for this part due to methodological difficulties. However, it is believed to form an {alpha}-helix that anchors the P450 molecule in the membrane and that this helix should not affect the overall structure of the globular part.

Structural Domains of CYP21 and Naturally Occurring Mutants in These Regions
Meander and the ERR-Triad.
The meander is a highly conserved region in all cytochrome P450s and consists of a stretch approximately 20 amino acid residues long between the K' helix and the Cys pocket (37). In the present model, this region consists entirely of residues exposed to the molecular surface, with the exception of the two buried prolines (P401 and P406) that are practically invariant in all CYPs. An additional notable position in the meander of CYP21 is position 408, which is a highly conserved arginine (varies between arginine, histidine, and asparagine in all other CYPs) and has been reported to clinically result in SW CAH when mutated to cysteine (R408C; CYP21A2*73) (38). This residue is the third amino acid residue in the so-called ERR-triad, also consisting of glutamate and arginine that are located in the K-helix (E351 and R354) and that are strictly conserved in all CYPs. The ERR-triad acts as a folding motif in the meander and assists in stabilizing the three-dimensional structure of this region, locking the Cys-pocket in a position that assures covalent binding of heme with the CYP (37). Several site-directed mutagenesis experiments of the ERR-triad residues in various CYPs have confirmed their significance for enzyme function (39, 40, 41, 42). Because the hydroxylation activity of CYP21 is dependent on heme binding, it is not surprising that an interruption of the ERR-motif leads to a severe form of CAH. Modeling of R408C made it evident that this surface residue completely changes the hydrophilicity when becoming nonpolar, and thereby loses its function as a structural stabilizer. Two naturally occurring CYP21 mutants have also been reported in the first arginine of the ERR-triad, (R354H; CYP21A*54) (43, 44) and (R354C; CYP21A2*59) (85), both of which are clinically associated with SW CAH. Our model shows that substitutions to either histidine or cysteine in this position result in loss of the naturally occurring interactions with E351 and E403, essential for maintaining the structure. Recently, a missense mutation in the most N-terminal residue of the ERR-triad (E351K; CYP21A2*96) was reported in a SV CAH patient, and functional studies confirmed the diagnosis, with severe, but not a complete, loss of in vitro activity. Structural changes induced by the glutamate to lysine exchange as well as other artificial mutants at position 351 of CYP21 have also been shown by modeling to interfere with correct folding, although direct heme binding with C428 was not believed to be disturbed (46). In our novel model, a substitution of the negatively charged glutamate to the positively charged lysine results in loss of the strong interaction with the first arginine in the ERR-triad (R354) and, moreover, induces electrostatic repulsion between the two interacting residues. Thus, the structural features shown in the model highly correlate with the clinical consequences of CYP21 mutations in the ERR-motif and support the importance of the meander being a fundamental structural domain.

The Cys-Pocket and the Heme-Binding Region.
The present investigation has identified the same residues central to heme binding as for the template CYP2C5. These heme-binding residues are identical to the predicted residues in the previous model of CYP21 (16), based on the prokaryotic CYP102 (13). Thus, this novel model demonstrates that the heme-binding region is maintained in CYP21, as suspected, because the three-dimensional structure of this site is highly conserved throughout all domains of life from prokaryotic to human CYPs, despite low sequence identity.

All amino acid residues in the Cys-pocket (amino acid residues 421–428) are conserved between CYP21 and the template. The heme iron is sulfide bonded by C428 in the loop connecting helix K to helix L, and the propionates of the heme interact with the side chains of W116 and K120 in helix C, H365 in ß1–4, and F421 and R426 in the Cys-pocket. C428 in CYP21 has also been shown experimentally to be heme iron binding, and experimental mutants (W428, M428, S428) have, in expression studies in mammalian and yeast cells, been shown to lose all enzymatic activity for conversion of 17OHP (47). CYP21 mutants involved in the Cys-pocket and heme binding are G424S (CYP21A2*55), R426H (CYP21A2*82), and H365Y (CYP21A2*99), and the reported clinical features (SV CAH) correlate well with the structural data presented here. The G424S mutant has been suggested to be a founder mutation with Mulatto origin, and several unrelated patients confirm its association to SV CAH (48). This glycine is important for structural maintenance of the Cys-pocket, because it enables a sharp turn where no other residue with a larger side chain can fit spatially. The R426H mutant was first reported in three hemizygous sisters, all with severe virilization of external genitalia but without any salt-wasting symptoms (49). However, recently we found this mutation in a young SV CAH girl with latent salt-wasting symptoms, and functional analyses revealed residual in vitro activity below 1% (50). The H365Y mutant (51) was recently detected in a CAH patient, but information of the clinical severity was not presented. Based on the present modeling studies of this mutant, it most probably results in classical CAH (SW or SV), because it is a residue directly involved in heme binding.

The Active Site; Substrate Binding and Catalysis.
The substrate-binding region is usually one of the most difficult regions to predict in modeling studies of CYPs, because it is the site of expected structural divergence depending on the substrate specificity of different proteins. Significant structural differences between substrate-binding sites have also been demonstrated (22, 23). Therefore, it is reasonable to believe that by using a template structure which catalyzes a related substrate, a more accurate model can be obtained regarding this specific matter. Rabbit CYP2C5, which was used as a template for the present modeling, utilizes several substrates including progesterone, which is also used by CYP21. Due to the lack of three-dimensional structures determined with bound CYP21 substrates (progesterone and 17OHP) in the active site, data from CYP51 were also used, which is the experimentally determined structure-ligand complex available with the most similar substrate, the steroid estriol (29). Residues identified as steroid binding in CYP51 correlate well with the putative substrate-binding site of docked progesterone in CYP21. Thus, we believe that our CYP21 model in this respect is an improvement over previous models based on prokaryotic CYP structures.

A conserved region among CYPs in the middle of helix I (corresponding to SRS 4) starts with a few hydrophobic residues followed by charged and polar amino acid residues that are believed to act as proton acceptors and donors in proton transfer to the ironbound oxygen. The present model shows that the conserved G291 and T295 are the corresponding residues in CYP21 that are connected by hydrogen bonding, shown in prokaryotic (13, 52) as well as mammalian (21) structures to form an oxygen-binding pocket. These helix I-interacting residues are furthermore involved in CAH-causing mutations. T295N is one of the novel mutants that are reported in this study, whereas three different mutations have been observed for G291. They are of missense type: G291S (CYP21A2*23) (53, 54), G291C (CYP21A2*53) (43, 55), and G291R (CYP21A2*77) (56), and all are correlated with severe, classical CAH. It has also previously been suspected that G291 may be part of catalysis, supported by experimental data from functional studies showing enzymatic activity in vitro of below 1% for both 17OHP and progesterone for the G291S mutant, whereas the half-life of the mutant was comparable to that of the normal protein (53). Among the identified substrate-binding residues, another disease-causing mutation is found between the K-helix and strand 4 of ß-sheet 1, corresponding to SRS5. The L363W mutant (CYP21A2*66) has been associated with SV CAH (57), which correlates well with the presented structural data suggesting this residue to be involved in steroid and heme binding. All these above-mentioned residues have been predicted to be involved with substrate binding in CYP2C5, which is also true for the substrate-binding residues S108, L109, D287, and V359 in our model.

Redox Partner Interactions
Electrostatic mechanisms are considered significant to orient the positively charged surface of the CYP-dipole toward the negative surface of the redox partner, whereas hydrophobic forces are likely to mediate the functional electron transfer between the two proteins. It is believed that hydrophobic interactions between the ER membrane and the CYP are significant to achieve proper alignment of the CYP and its redox partner (5).

Experimental evidence indicates that basic residues on the proximal face (with respect to heme) of various CYPs interact with acidic residues on the surface of the redox partner (5). A cluster of basic residues, mostly arginines, in CYP21 are prone to mutate resulting in CAH (58). The mutants consist of R339H, R341P/W, R354H, and R356Q/W/P and are part of the segment corresponding to residues 338–361 in CYP21 that is believed to be involved in redox partner interaction according to the crystal structure of CYP102 (13). Modeling of the related CYP17 as well as experimental work has shown that arginines R358 and R347 are important for interaction with the corresponding redox partner. R347, in particular, is evidently a crucial component of the redox partner interaction site (32, 33). The corresponding residue in CYP21 is the completely conserved R339. A naturally occurring missense mutation in this position, R339H (CYP21A2*24), is associated with NC CAH (59). It seems likely that R339 is crucial for redox partner interaction, such that even the minor change of the side chain properties from arginine to histidine leads to mild disease. The R358 in CYP17 does not correspond to a basic residue in CYP21, but to alanine (A350). Thus, it seems likely that some other residue in CYP21 fulfils the function of R358. One of the candidates might be R341, corresponding to R349 in CYP17, which is located close to the important R339 (R347 in CYP17) (cf. Fig. 2Go). All arginine residues in this region of CYP21 (R339, R341, R354, and R356) have been affected by disease-causing mutations, which most likely cause electrostatic disturbances as concluded from the present modeling studies (Table 3Go).

Large Hydrophobic Area
Two mutants, G64E and V211L, have so far been identified in the large hydrophobic surface of CYP21. The G64E mutant (CYP21A2*47) is positioned in the loop between the two first strands of ß-sheet 1 and has been associated with the most severe form of the disease (SW) and complete loss of activity in vitro (60). Modeling shows that this residue substitution may lead to structural hindrance due to very constricted space, because the small side chain of glycine is the only possible amino acid to make the sharp turn seen in this loop. The V211L variant (CYP21A2*21) (61) affects the first residue C-terminally of helix G' that is exposed to the surface. Although reported as a CYP21 sequence variant, it has not been claimed to be responsible for CAH. Data deduced from the present model supports the fact that this residue alteration is of no significance to disease, but rather a normal variant of CYP21, because it does not confer any change in hydrophobicity. However, because this position is part of the hydrophobic patch that seems to be important for membrane association, a different change might result in unfavorable interactions causing disease.

Protein Stability Correlation
The stability dependence can give useful information about the mutant in combination with other information, e.g. about surface accessibility. In our case all amino acid residues affected by the normal variations have access to the solvent. The deleterious mutations associated with residue exchanges on the surface, in most cases, affect residues close to the active site and, consequently, they do not need to significantly affect the stability of the protein. Thus, when we looked at the stability dependence for the mutants affecting residues in the core, a greater separation between SV and SW was observed (data not shown).

Prediction of Clinical Effect of Novel CYP21 Mutations
The V139E mutation was found in a compound heterozygous patient (V139E/I2 splice) diagnosed with SW CAH. The structural prediction of V139E resulting in classical (SV/SW) CAH thus correlates well with the patient’s diagnosis. The 7-yr-old boy with genotype C147R/Q318X was diagnosed with NC CAH or possibly SV CAH. The prediction based on modeling is that C147R is associated with classical CAH, and the boy most likely has SV rather than NC CAH. This is hence a good example where a structural prediction based on the present model is of value in classifying a novel mutation. Because this patient is a boy, it is difficult to distinguish between SV and NC CAH based on clinical evaluation, because the presence or absence of virilization of external genitalia cannot be used. The ability to distinguish between NC and SV CAH is important in these boys, because it has impact both on treatment and genetic counseling to the family. The female patient with genotype R233G/R233G or R233G/del was diagnosed with NC CAH, which matches the predictions for this novel mutation. The boy with genotype T295N/I172N was diagnosed with SV CAH. The structural prediction is that T295N is associated with the most severe form of the disease, SW CAH, and his slightly less severe clinical phenotype thus reflects I172N (CYP21A2*11), which is associated with SV CAH. Structural prediction of L308F revealed this mutation to be most likely associated with moderately severe CAH (NC/SV), again in agreement with the clinical diagnosis of this SV girl. The Finnish girl with genotype R366C/V281L was diagnosed with NC CAH. R366C cannot be classified based on clinical evaluation, because the mild V281L (CYP21A2*15) mutation is responsible for this form of the disease. The present modeling shows that the novel mutant R366C is most likely associated with SV/SW CAH. Thus, R366C, in combination with a severe allele, would most probably result in classical CAH. Regarding M473I, we suggest that it is not a disease-causing mutation but rather a normal CYP21 variant. Consequently, the girl with genotype M473I/V281L is most likely a heterozygous carrier rather than a girl affected by a mild form of CAH. In conclusion, the predicted mutant severities were in good agreement with degree of disease presentation and, in several cases, provided additional clinically useful information that could not be resolved by examining the patients.

Normal Variants of CYP21
The present models include six normal variants of CYP21, which are not associated with disease. These are K102R, D183E, V211L, M239K, S268T, and N493S (Table 4GoGoGo). As observed, they all affect nonconserved amino acid residues that are accessible to the surface to a higher degree than most disease-causing mutants. None of these alterations changes the electrostatic charge, except for M239K (CYP21A2*14), in which nonpolar methionine is exchanged for the positively charged lysine. It has recently been shown, by in vitro expression and functional analysis, that M239K is not disease causing on its own, but rather a normal CYP21 variant, with an activity of 95/98% for conversion of 17OHP and progesterone, respectively, compared with the normal enzyme. In combination with I236N and V237E, it forms the common Cluster E6 mutation, which is correlated to SW CAH and causes abolished enzyme function in vitro (62). M239K is also the only normal CYP21 variant in which modeling reveals some notable characteristics; the new mutant residue gains interaction with E238, as well as causes repulsion to R242 in the present model, which could be expected to result in structural disturbance. However, because this surface-exposed residue is found in the short loop between helix G and G'', which is not part of any domain of importance for enzyme function/structure, the introduced molecular interactions are unlikely to disturb overall protein structure.

Mutants of Unclear Severity
The present model has enabled structural explanations for all reported missense mutations, with the exception of G178A and E380D. The E380D mutant (CYP21A2*35) has been reported to be associated with SW CAH, although functional data have shown a reduction of normal enzyme activity to 30% (63, 64) that is usually associated with a NC phenotype. However, the modeling shows that there are only minor differences between a glutamate and an aspartate residue at this position. It has been speculated that this residue might be in electrostatic contact with residues in the membrane-binding helix (16), which is, however, not included in this model, and it thus remains speculative. G178A (CYP21A2*52) (43, 55), just as E380D, is also grouped among the unclear mutants in Table 4GoGoGo, because the reported clinical phenotypes are not in agreement with functional data.

In conclusion, the work presented here shows evidence as to how the joint evaluation of a clinical mutation data bank with structural and bioinformatical analysis can provide both enhanced knowledge of structure-function relationships of the native protein and better understanding of the molecular mechanisms of CAH due to 21-hydroxylase deficiency. The strong coherence between the degree of disturbed structural parameters with the physiological manifestations of the mutation is striking and could provide a new bioinformatic route for the prediction of clinical consequences of novel disease-causing mutations. Because biochemical and biophysical evaluation of a mutant protein in vitro is both time consuming and often impractical due to folding and/or solubility problems of mutant proteins, novel routes toward in silico analysis of the effects of patient mutations open up for high-throughput analysis of genetic data. To achieve efficient individualized therapies in mutation-linked diseases, such as CAH due to 21-hydroxylase deficiency, which was the case analyzed here, and in mutation-generating disorders, such as cancer and antibiotics resistance, such efforts will be crucial for success.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
Multiple Sequence Alignment
Sequences homologous to human CYP21 were found using FASTA (65) searches in the Uniprot database (66) and assembled in a multiple sequence alignment using ClustalW (67) with default parameters. The aligned sequences were from the following species: human (Uniprot accession no. P08686), pig (P15540), dog (Q9N2C7), cow (P00191), sheep (Q7M366), mouse (P03940), rat (Q64562), puffer fish (Q7SZG0), and eel (Q7ZZR9). Because the puffer fish and eel sequences were only distantly related and might have different catalytic properties, they were excluded in the conservation calculations (cf. below).

Homology Modeling
A molecular model was calculated based upon the structure of rabbit cytochrome P450 2C5 (27) at 2.3 Å resolution and with an R-value of 0.257 (PDB accession no. 1N6B) using the program ICM (version 3.2, Molsoft LLC, La Jolla, CA). The sequence identity is 31.1% on the aligned part of the chain. ICM uses the alignment to tether the CYP21 sequence to the template structure. A Monte-Carlo procedure is used to energy minimize the structure. The calculations are performed both globally over the entire protein and locally on loops between regular secondary structure elements. The Monte-Carlo method is optimized using biases toward regions with high energy and using common side chain conformations, so called rotamers (68). The quality of the model was evaluated using several methods. The widely used evaluation method Procheck (69, 70) gives an overall G value for the model of –0.01. This is well above the threshold –1.00 for a bad structure and –0.50 for a poor one. Procheck evaluates geometrical and stereochemical properties, i.e. close contacts, bond distances, bond angles, torsion angles, and chirality. Another evaluation method, considering energy strain (71), results in an average normalized residue energy for the model of only 0.74, which is in the better half of the theoretical models in PDB. It is also useful to look at the Ramachandran plot in which the model has only nine nonglycine outliers compared with six for the template crystal structure. Finally there are very few clashes, i.e. atoms too close to each other causing high van der Waals interaction energies. Thus, we consider the CYP21 model to be of appropriate quality for the investigations in this study.

For comparison, we also calculated a model based upon the CYP2B4 structure (PDB code 1PO5, 1.6 Å resolution, R-value 0.217) (20), which has 28.5% sequence identity to CYP21. However, this model was considered inferior to the one based upon CYP2C5, as judged by the quality evaluations mentioned above.

Human CYP21 is a membrane-bound protein with an N-terminal membrane-attaching segment of 26 amino acid residues that was excluded from the model. The first residue in the present homology model is thus leucine 27, with residue numbering according to the native full-length protein. The alignment was manually adjusted at a few places where gaps interrupting regular secondary structure elements could be shifted a limited number of residues.

Analyses of CYP21 Mutants and Normal Variants
Structural analyses were performed for all reported mutations affecting CYP21 except nonsense, frame shift, splice site, and promoter mutations. A total of 66 different missense mutations, including seven novel mutations and six normal variants, were studied. The analyses of two reported missense mutations, M1I (CYP21A2*83) (72) and A15T (CYP21A2*74) (73) are lacking, because the very N-terminal transmembrane part of the protein was not included in the present model. Allelic variants were grouped according to their reported clinical classification (12), dividing the phenotypes into the three groups of disease severity, SW (n = 16), SV (n = 11), and NC CAH (n =18), as well as N for normal CYP21 variants (n = 6). For mutants that have been functionally investigated, in vitro enzyme activities below 1% of normal were considered to correspond to a SW phenotype, activities between 1% and 15% were classified as SV, and residual activity greater than 20% was considered typical of mutations associated with NC CAH. Thirteen mutants are presented separately because seven of these are novel, and functional information (associated clinical symptoms and correlated in vitro enzyme activity) is lacking for the remaining six. The structure corresponding to each mutant was individually modeled in triplicate using ICM, as detailed above. The molecular free energy was calculated according to the ECEPP/3 forcefield (74) parameters for van der Waals interactions, hydrogen bonds, electrostatic interactions, and torsion energies using algorithms implemented in ICM. The molecular free energy was used in comparisons between the structures. Hydrophobicity changes were calculated using the Kyte and Doolittle parameters (75). Distances between the amino acid residues corresponding to the mutants and the heme- and steroid-binding sites were measured in ICM.

Patients with Novel CYP21 Mutations 2
The novel mutations (V139E, C147R, R233G, T295N, L308F, R366C, and M473I) were detected at the Clinical Genetics Unit, Karolinska University Hospital, Sweden, by sequencing of the entire coding region of CYP21 from Scandinavian patients investigated for CAH.

V139E (accession no. AM183945) was found in trans with the intron 2 splice mutation (CYP21*9, genotype V139E/I2 splice) in a Swedish boy who was diagnosed with CAH through neonatal screening. His sodium levels started to fall before treatment was begun, and he was diagnosed with SW CAH.

The C147R mutation (AM183946) was found in trans with Q318X (CYP21A2*17) in a Swedish boy who presented with accelerated growth and premature adrenarche at the age of 7 yr. He was diagnosed with NC CAH, although the SV form cannot be ruled out in this boy.

R233G (AM183947) was found in hemi- or homozygous form (genotype R233G/deletion or R233G/R233G) in a Swedish female born in 1960. She reported hyperandrogenic symptoms such as hirsutism and oligomenorrhea from the age of 20, and had been diagnosed with polycystic ovary syndrome at this age. At the age of 32, she was reinvestigated for infertility, and elevated 17-hydroxyprogesterone (17OHP) (293 nmol/liter) and slight clitoromegaly were found. She was diagnosed with NC CAH.

T295N (AM183948) was found in trans with I172N (CYP21A2*11) in a Swedish boy who was also found through the neonatal screening (17OHP was 127 nmol/liter, reference value 75 nmol/liter). A few days later, serum 17OHP was 500 nmol/liter. Treatment was promptly instituted during the first week of life, and the patient was diagnosed with SV CAH, although SW cannot be ruled out in this patient.

L308F (AM183949) was found in trans with Q318X (CYP21A2*17) in a Russian female immigrant who came to Sweden at the age of 16 yr. Medical records were incomplete, but she had been diagnosed with CAH at the age of 4 months although treatment had not been instituted until she was 2 yr of age. She had undergone genital surgery at the age of 5 yr. She was classified as having SV CAH.

R366C (AM183950) was found in trans with V281L (CYP21A2*15) in a Finnish girl who came to medical attention because of growth acceleration and pubic hair growth at the age of 7.5 yr. She was diagnosed with NC CAH.

M473I (AM183951) was found in trans with V218L (CYP21A2*15) in a Norwegian girl who was investigated due to pubic and axillary hair growth from 9 yr of age. She had slightly increased 17OHP (20 nmol/liter basally, 95.6 after ACTH stimulation), and was considered either to have a very mild form of NC CAH or to be a heterozygous carrier.

Redox Partner Interaction
The redox partner of CYP21 is known to be the cytochrome P450 NADPH oxidoreductase (31), which docks to the protein and transfers electrons from NADPH to the heme iron. For the homolog CYP2B4, showing 28.5% pair-wise identity with CYP21, the redox partner-binding site is well studied (76, 77, 78, 79, 80). The CYP2B4 structure was superimposed onto the CYP21 model using ICM, and, subsequently, the residues considered important for redox binding in CYP2B4 were mapped in the CYP21 model.

Hydrophobic Areas
Searches for large hydrophobic areas that might interact with other molecules using the Optimal Docking Area method (34), available as a web service at http://www.molsoft.com/oda, was also performed. The ODA algorithm generates protein surface of different sizes to analyze their docking surface energy. The energy calculations are based on atomic solvation parameters, for each type of atom exposed to the solvent, previously derived form octanol-water transfer experiments. Areas with low docking surface energy correspond to regions likely to be buried in the interaction with other proteins.

Steroid-Binding Region
In a search for a homologous protein structure containing bound progesterone at the active site, a steroid that is hydroxylated by CYP21, CYP51 with the steroid estriol bound was the closest match found. The CYP51-estriol complex has been structurally determined at 1.55Å resolution with an R-value of 0.206 (pdb-ID 1X8V) (29). The two steroids, estriol and progesterone, have similar structures and sizes, and the two proteins, CYP21 and CYP51, are homologous with 22% pair-wise sequence identity. Based on the structure of CYP51, the probable steroid-binding site in CYP21 was mapped by superimposing the two proteins using the heme groups as guidelines The heme groups superimposed with an root mean square deviation of 0.63 Å, thus showing good agreement. After superimposing in ICM, the estriol molecule was transferred to the CYP21 model and subsequently changed to progesterone. The complex structure was energy minimized to avoid clashes, predominantly between the steroid and the heme group. From the resulting model, the amino acid residues within a sphere of 5Å radius from the progesterone were considered as steroid binding. Similarly, the amino acid residues close to the heme group were classified as heme binding (cf. Fig. 1Go).


   ACKNOWLEDGMENTS
 
We thank Malin Fladvad for initial assistance with computer graphics.


   FOOTNOTES
 
This work was supported by grants from Stiftelsen Samariten and The Sven Jerring Foundation (to T.R.); the Swedish Research Council (Grant 6475) and the Carl Trygger Foundation (to M.S.); the Swedish Research Council (Grant 12198), the Novo Nordisk Foundation, The Centre of Gender Related Medicine, Karolinska Institutet, the Stockholm County Council, and the Stiftelsen Frimurare Barnhuset (to A.W.); the Linköping University (to M.S. and B.P.).

Disclosure statement: The authors have nothing to disclose.

First Published Online June 20, 2006

1 T.R. and J.C. contributed equally to this work. Back

Abbreviations: CAH, Congenital adrenal hyperplasia; CYP, cytochrome P450; CYP21A2, CYP21, steroid 21-hydroxylase; CYP21A2, CYP21, the gene encoding steroid 21-hydroxylase; CYP21A1, CYP21P, the CYP21 pseudogene; ER, endoplasmic reticulum; G6PD, glucose-6 phosphate 1-dehydrogenase; NADPH, reduced nicotinamide adenine dinucleotide phosphate; NC, non-classic; 21OHD, steroid 21-hydroxylase deficiency; 17OHP, 17-hydroxyprogesterone; PDB, Protein Data Bank; SRS, steroid recognition site; SV, simple virilizing; SW, salt wasting.

2 Novel nucleotide sequence data reported are available in the DDBJ/EMBL/GenBank databases under the GenBank accession numbers: AM183945 (CYP21-V139E), AM183946 (CYP21-C147R), AM183947 (CYP21-R233G), AM183948 (CYP21-T295N), AM183949 (CYP21-L308F), AM183950 (CYP21-R366C), and 183951 (CYP21-M473I). Back

Received for publication April 24, 2006. Accepted for publication June 12, 2006.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  1. Degtyarenko KN and Archakov AI 1993 Molecular evolution of P450 superfamily and P450-containing monooxygenase systems. FEBS Lett 332:1–8[CrossRef][Medline]
  2. Nelson DR, Kamataki T, Waxman DJ, Guengerich FP, Estabrook RW, Feyereisen R, Gonzalez FJ, Coon MJ, Gunsalus IC, Gotoh O, Okuda K, Nebert DW 1993 The P450 superfamily: update on new sequences, gene mapping, accession numbers, early trivial names of enzymes, and nomenclature. DNA Cell Biol 12:1–51[Medline]
  3. Omura T, Sato R 1964 The carbon monoxide-binding pigment of liver microsomes. II. Solubilization, purification, and properties. J Biol Chem 239:2379–2385[Free Full Text]
  4. Homepage of the Human Cytochrome P450 (CYP) Allele Nomenclature Committee: http://www.imm.ki.se/CYPalleles
  5. Hlavica P, Schulze J, Lewis DF 2003 Functional interaction of cytochrome P450 with its redox partners: a critical assessment and update of the topology of predicted contact regions. J Inorg Biochem 96:279–297[CrossRef][Medline]
  6. Degtyarenko KN 1995 Structural domains of P450-containing monooxygenase systems. Protein Eng 8:737–747[Abstract/Free Full Text]
  7. Ingelman-Sundberg M, Johansson I 1980 Cytochrome b5 as electron donor to rabbit liver cytochrome P-450LM2 in reconstituted phospholipid vesicles. Biochem Biophys Res Commun 97:582–586[CrossRef][Medline]
  8. Truan G, Cullin C, Reisdorf P, Urban P, Pompon D 1993 Enhanced in vivo monooxygenase activities of mammalian P450s in engineered yeast cells producing high levels of NADPH-P450 reductase and human cytochrome b5. Gene 125:49–55[CrossRef][Medline]
  9. White PC, Speiser PW 2000 Congenital adrenal hyperplasia due to 21-hydroxylase deficiency. Endocr Rev 21:245–291[Abstract/Free Full Text]
  10. Thilén A, Nordenström A, Hagenfeldt L, von Döbeln U, Guthenberg C, Larsson A 4 April 1998 Benefits of neonatal screening for congenital adrenal hyperplasia (21-hydroxylase deficiency) in Sweden. Pediatrics 101:E11
  11. Yang Z, Mendoza AR, Welch TR, Zipf WB, Yu CY 1999 Modular variations of the human major histocompatibility complex class III genes for serine/threonine kinase RP, complement component C4, steroid 21-hydroxylase CYP21, and tenascin TNX (the RCCX module). A mechanism for gene deletions and disease associations. J Biol Chem 274:12147–12156[Abstract/Free Full Text]
  12. Database of CY21A2 by the Human Cytochrome P450 (CYP) Allele Nomenclature Committee: http://www.imm.ki.se/CYPalleles/cyp21.htm
  13. Ravichandran KG, Boddupalli SS, Hasemann CA, Peterson JA, Deisenhofer J 1993 Crystal structure of hemoprotein domain of P450BM-3, a prototype for microsomal P450’s. Science 261:731–736[Abstract/Free Full Text]
  14. Szklarz GD, Halpert JR 1997 Use of homology modeling in conjunction with site-directed mutagenesis for analysis of structure-function relationships of mammalian cytochromes P450. Life Sci 61:2507–2520[CrossRef][Medline]
  15. Lewis DF, Lee-Robichaud P 1998 Molecular modelling of steroidogenic cytochromes P450 from families CYP11, CYP17, CYP19 and CYP21 based on the CYP102 crystal structure. J Steroid Biochem Mol Biol 66:217–233[CrossRef][Medline]
  16. Mornet E, Gibrat J-F 2000 A 3D model of human P450c21: study of the putative effects of steroid 21-hydroxylase gene mutations. Hum Genet 106:3309–3339
  17. Lin D, Zhang LH, Chiao E, Miller WL 1994 Modeling and mutagenesis of the active site of human P450c17. Mol Endocrinol 8:392–402[Abstract/Free Full Text]
  18. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, Shindyalov IN, Bourne PE 2000 The protein data bank. Nucleic Acids Res 28:235–242[Abstract/Free Full Text]
  19. Schoch GA, Yano JK, Wester MR, Griffin KJ, Stout CD, Johnson EF 2004 Structure of human microsomal cytochrome P450 2C8. Evidence for a peripheral fatty acid binding site. J Biol Chem 279:9497–9503[Abstract/Free Full Text]
  20. Scott EE, He YA, Wester MR, White MA, Chin CC, Halpert JR, Johnson EF, Stout CD 2003 An open conformation of mammalian cytochrome P450 2B4 at 1.6-A resolution. Proc Natl Acad Sci USA 100:13196–13201[Abstract/Free Full Text]
  21. Williams PA, Cosme J, Sridhar V, Johnson EF, McRee DE 2000 Microsomal cytochrome P450 2C5: comparison to microbial P450s and unique features. J Inorg Biochem 81:183–190[CrossRef][Medline]
  22. Williams PA, Cosme J, Ward A, Angove HC, Matak Vinkovic D, Jhoti H 2003 Crystal structure of human cytochrome P450 2C9 with bound warfarin. Nature 424:464–468[CrossRef][Medline]
  23. Williams PA, Cosme J, Vinkovic DM, Ward A, Angove HC, Day PJ, Vonrhein C, Tickle IJ, Jhoti H 2004 Crystal structures of human cytochrome P450 3A4 bound to metyrapone and progesterone. Science 305:683–686[Abstract/Free Full Text]
  24. Yano JK, Wester MR, Schoch GA, Griffin KJ, Stout CD, Johnson EF 2004 The structure of human microsomal cytochrome P450 3A4 determined by X-ray crystallography to 2.05-A resolution. J Biol Chem 279:38091–38094[Abstract/Free Full Text]
  25. Cosme J, Johnson EF 2000 Engineering microsomal cytochrome P450 2C5 to be a soluble, monomeric enzyme. Mutations that alter aggregation, phospholipid dependence of catalysis, and membrane binding. J Biol Chem 275:2545–2553[Abstract/Free Full Text]
  26. Johnson EF, Wester MR, Stout CD 2002 The structure of microsomal cytochrome P450 2C5: a steroid and drug metabolizing enzyme. Endocr Res 28:435–441[CrossRef][Medline]
  27. Wester MR, Johnson EF, Marques-Soares C, Dansette PM, Mansuy D, Stout CD 2003 Structure of a substrate complex of mammalian cytochrome P450 2C5 at 2.3 A resolution: evidence for multiple substrate binding modes. Biochemistry 42:6370–6379[CrossRef][Medline]
  28. Wester MR, Johnson EF, Marques-Soares C, Dijols S, Dansette PM, Mansuy D, Stout CD 2003 Structure of mammalian cytochrome P450 2C5 complexed with diclofenac at 2.1 A resolution: evidence for an induced fit model of substrate binding. Biochemistry 42:9335–9345[CrossRef][Medline]
  29. Podust LM, Yermalitskaya LV, Lepesheva GI, Podust VN, Dalmasso EA, Waterman MR 2004 Estriol bound and ligand-free structures of sterol 14{alpha}-demethylase. Structure (Camb) 12:1937–1945[Medline]
  30. 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[Abstract/Free Full Text]
  31. Kominami S, Hara H, Ogishima T, Takemori S 1984 Interaction between cytochrome P-450 (P-450C21) and NADPH-cytochrome P-450 reductase from adrenocortical microsomes in a reconstituted system. J Biol Chem 259:2991–2999[Abstract/Free Full Text]
  32. Auchus RJ, Miller WL 1999 Molecular modeling of human P450c17 (17{alpha}-hydroxylase/17,20-lyase): insights into reaction mechanisms and effects of mutations. Mol Endocrinol 13:1169–1182[Abstract/Free Full Text]
  33. Geller DH, Auchus RJ and Miller WL 1999 P450c17 mutations R347H and R358Q selectively disrupt 17,20-lyase activity by disrupting interactions with P450 oxidoreductase and cytochrome b5. Mol Endocrinol 13:167–175[Abstract/Free Full Text]
  34. Fernandez-Recio J, Totrov M, Skorodumov C, Abagyan R 2005 Optimal docking area: a new method for predicting protein-protein interaction sites. Proteins 58:134–143[CrossRef][Medline]
  35. Sanchez R, Pieper U, Melo F, Eswar N, Marti-Renom MA, Madhusudhan MS, Mirkovic N, Sali A 2000 Protein structure modeling for structural genomics. Nat Struct Biol 7(Suppl):986–990
  36. Nordling E, Oppermann UC, Jornvall H, Persson B 2001 Human type 10 17 ß-hydroxysteroid dehydrogenase: molecular modelling and substrate docking. J Mol Graph Model 19:514–520, 591–593[CrossRef][Medline]
  37. Hasemann CA, Kurumbail RG, Boddupalli SS, Peterson JA, Deisenhofer J 1995 Structure and function of cytochromes P450: a comparative analysis of three crystal structures. Structure 3:41–62[Medline]
  38. Billerbeck AE, Mendonca BB, Pinto EM, Madureira G, Arnhold IJ, Bachega TA 2002 Three novel mutations in CYP21 gene in Brazilian patients with the classical form of 21-hydroxylase deficiency due to a founder effect. J Clin Endocrinol Metab 87:4314–4317[Abstract/Free Full Text]
  39. Chen S, Zhou D 1992 Functional domains of aromatase cytochrome P450 inferred from comparative analyses of amino acid sequences and substantiated by site-directed mutagenesis experiments. J Biol Chem 267:22587–22594[Abstract/Free Full Text]
  40. Furuya H, Shimizu T, Hirano K, Hatano M, Fujii-Kuriyama Y, Raag R, Poulos TL 1989 Site-directed mutageneses of rat liver cytochrome P-450d: catalytic activities toward benzphetamine and 7-ethoxycoumarin. Biochemistry 28:6848–6857[CrossRef][Medline]
  41. Kitamura M, Buczko E, Dufau ML 1991 Dissociation of hydroxylase and lyase activities by site-directed mutagenesis of the rat P45017 {alpha}. Mol Endocrinol 5:1373–1380[Abstract/Free Full Text]
  42. Shimizu T, Tateishi T, Hatano M, Fujii-Kuriyama Y 1991 Probing the role of lysines and arginines in the catalytic function of cytochrome P450d by site-directed mutagenesis. Interaction with NADPH-cytochrome P450 reductase. J Biol Chem 266:3372–3375[Abstract/Free Full Text]
  43. Lobato MN, Ordonez-Sanchez ML, Tusie-Luna MT, Meseguer A 1999 Mutation analysis in patients with congenital adrenal hyperplasia in the Spanish population: identification of putative novel steroid 21-hydroxylase deficiency alleles associated with the classic form of the disease. Hum Hered 49:169–175[CrossRef][Medline]
  44. Nunez BC, Lobato MN, White PC, Meseguer A 1999 Functional analysis of four CYP21 mutations from Spanish patients with congenital adrenal hyperplasia. Biochem Biophys Res Commun 262:635–637[CrossRef][Medline]
  45. Krone N, Riepe FG, Grotzinger J, Partsch CJ, Sippell WG 2005 Functional characterization of two novel point mutations in the CYP21 gene causing simple virilizing forms of congenital adrenal hyperplasia due to 21-hydroxylase deficiency. J Clin Endocrinol Metab 90:445–454[Abstract/Free Full Text]
  46. Krone N, Riepe, FG, Grotzinger J, Partsch CJ, Bramswig J, Sippell WG 2005 The residue E351 is essential for the activity of human 21-hydroxylase: evidence from a naturally occurring novel point mutation compared with artificial mutants generated by single amino acid substitutions. J Mol Med 83:561–568[CrossRef][Medline]
  47. Wu DA, Chung BC 1991 Mutations of P450c21 (steroid 21-hydroxylase) at Cys428, Val281, and Ser268 result in complete, partial, or no loss of enzymatic activity, respectively. J Clin Invest 88:519–523[Medline]
  48. Billerbeck ANC, Bachega TASS, Frazzatto ET, Nishi MY, Goldberg AC, Marin MLC, Madureira G, Monte O, Arnhold IJP, Mendonca BB 1999 A novel missense mutation, G424S, in Brazilian patients with 21-hydroxylase deficiency. J Clin Endocrinol Metab 84:2870–2872[Abstract/Free Full Text]
  49. Baumgartner-Parzer SM, Schulze E, Waldhausl W, Pauschenwein S, Rondot S, Nowotny P, Meyer K, Frisch H, Waldhauser F, Vierhapper H 2001 Mutational spectrum of the steroid 21-hydroxylase gene in Austria: identification of a novel missense mutation. J Clin Endocrinol Metab 86:4771–4775[Abstract/Free Full Text]
  50. Barbaro M, Baldazzi L, Balsamo A, Lajic S, Robins T, Barp L, Pirazzoli P, Cacciari E, Cicognani A, Wedell A 2006 Functional studies of two novel and two rare mutations in the 21-hydroxylase gene. J Mol Med 84:521–528[CrossRef][Medline]
  51. Zeng X, Witchel SF, Dobrowolski SF, Moulder PV, Jarvik JW, Telmer CA 2004 Detection and assignment of CYP21 mutations using peptide mass signature genotyping. Mol Genet Metab 82:38–47[CrossRef][Medline]
  52. Poulos TL, Finzel BC, Howard AJ 1987 High-resolution crystal structure of cytochrome P450cam. J Mol Biol 195:687–700[CrossRef][Medline]
  53. Nikoshkov A, Lajic S, Vlamis-Gardikas A, Tranebjærg L, Holst M, Wedell A, Luthman H 1998 Naturally occurring mutants of human steroid 21-hydroxylase (P450c21) pinpoint residues important for enzyme activity and stability. J Biol Chem 273:6163–6165[Abstract/Free Full Text]
  54. Wedell A, Ritzén ME, Haglund-Stengler B, Luthman H 1992 Steroid 21-hydroxylase deficiency: three additional mutated alleles and establishment of phenotype-genotype relationships of common mutations. Proc Natl Acad Sci USA 89:7232–7236[Abstract/Free Full Text]
  55. Robins T, Bellanne-Chantelot C, Barbaro M, Cabrol S, Wedell A, Lajic S, Characterization of novel missense mutations in CYP21 causing congenital adrenal hyperplasia. J Mol Med, in press
  56. Stikkelbroeck NMML, Hoefsloot LH, De Wijs IJ, Otten BJ, Hermus ARMM, Sistermans EA 2003 CYP21 gene mutation analysis in 198 patients with 21-hydroxylase deficiency in The Netherlands: six novel mutations and a specific cluster of four mutations. J Clin Endocrinol Metab 88:3852–3859[Abstract/Free Full Text]
  57. Levo A and Partanen J 2001 Novel mutations in the human CYP21 gene. Prenat Diagn 21:885–889[CrossRef][Medline]
  58. Lajic S, Levo A, Nikoshkov A, Lundberg Y, Partanen J, Wedell A 1997 A cluster of missense mutations at Arg356 of human steroid 21-hydroxylase may impair redox partner interaction. Hum Genet 99:704–709[CrossRef][Medline]
  59. Helmberg A, Tusie-Luna M-T, Tabarelli M, Kofler R, White PC 1992 R339H and P453S: CYP21 mutations associated with nonclassic steroid 21-hydroxylase deficiency that are not apparent gene conversions. Mol Endocrinol 6:1318–1322[Abstract/Free Full Text]
  60. Ohlsson G, Müller J, Skakkebæk NE, Schwartz M 1999 Steroid 21-hydroxylase deficiency: mutational spectrum in Denmark, three novel mutations, and in vitro expression analysis. Hum Mutat 13:482–486[CrossRef][Medline]
  61. Speiser PW, New MI and White PC 1988 Molecular genetic analysis of nonclassical steroid 21-hydroxylase deficiency associated with HLA-B14, DR1. N Engl J Med 319:19–23[Abstract]
  62. Robins T, Barbaro M, Lajic S, Wedell A 2005 Not all amino acid substitutions of the common cluster E6 Mutation in CYP21 cause congenital adrenal hyperplasia. J Clin Endocrinol Metab 90:2148–2153[Abstract/Free Full Text]
  63. Hsu NC, Guzov VM, Hsu LC, Chung BC 1999 Characterization of the consequence of a novel Glu-380 to Asp mutation by expression of functional P450c21 in Escherichia coli. Biochim Biophys Acta 1430:95–102[CrossRef][Medline]
  64. Kirby-Keyser L, Porter CC, Donohoue PA 1997 E380D: a novel point mutation of CYP21 in an HLA-homozygous patient with salt-losing congenital adrenal hyperplasia due to 21-hydroxylase deficiency. Hum Mutat 9:181–182[CrossRef][Medline]
  65. Pearson WR, Lipman DJ 1988 Improved tools for biological sequence comparison. Proc Natl Acad Sci USA 85:2444–2448[Abstract/Free Full Text]
  66. Bairoch A, Apweiler R, Wu CH, Barker WC, Boeckmann B, Ferro S, Gasteiger E, Huang H, Lopez R, Magrane M, Martin MJ, Natale DA, O’Donovan C, Redaschi N, Yeh LS 2005 The Universal Protein Resource (UniProt). Nucleic Acids Res 33 (Database issue):D154–D159
  67. Thompson JD, Higgins DG, Gibson TJ 1994 CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22:4673–4680[Abstract/Free Full Text]
  68. Abagyan R, Totrov M 1994 Biased probability Monte-Carlo conformational searches and electrostatic calculations for peptides and proteins. J Mol Biol 235:983–1002[CrossRef][Medline]
  69. Laskowski RA, MacArthur MW, Moss DS, Thornton JM 1993 PROCHECK: a program to check the stereochemical quality of protein structures. J Appl Cryst 26:283–291[CrossRef]
  70. Morris AL, MacArthur MW, Hutchinson EG, Thornton JM 1992 Stereochemical quality of protein structure coordinates. Proteins 12:345–364[CrossRef][Medline]
  71. Maiorov V, Abagyan R 1998 Energy strain in three-dimensional protein structures. Fold Des 3:259–269[CrossRef][Medline]
  72. Usui T NK, Kaji M, Ikuno N, Yorifuji T, Yasudo T, Kuzuya H, Shimatsu A 2004 Three novel mutations in Japanese patients with 21-hydroxylase deficiency. Horm Res 61:126–132[CrossRef][Medline]
  73. Dolzan V, Stopar-Obreza M, Zerjav-Tansek M, Breskvar K, Krzisnik C, Battelino T 2003 Mutational spectrum of congenital adrenal hyperplasia in Slovenian patients: a novel Ala15Thr mutation and Pro30Leu within a larger gene conversion associated with a severe form of the disease. Eur J Endocrinol 149:137–144[Abstract]
  74. Nemethy G, Gibson KD, Palmer KA, Yoon CN, Paterlini G, Zagari A, Rumsey S, Scherega HA 1992 Energy parameters in polypeptides. 10. Improved geometric parameters and nonbonded interactions for use in the ECEPP/3 algorithm, with application to proline-containing peptides. J Phys Chem 96:6472–6484[CrossRef]
  75. Kyte J, Doolittle RF 1982 A simple method for displaying the hydropathic character of a protein. J Mol Biol 157:105–132[CrossRef][Medline]
  76. Bernhardt R, Kraft R, Otto A, Ruckpaul K 1988 Electrostatic interactions between cytochrome P-450 LM2 and NADPH-cytochrome P-450 reductase. Biomed Biochim Acta 47:581–592[Medline]
  77. Bernhardt R, Makower A, Janig GR, Ruckpaul K 1984 Selective chemical modification of a functionally linked lysine in cytochrome P-450 LM2. Biochim Biophys Acta 785:186–190[CrossRef][Medline]
  78. Bridges A, Gruenke L, Chang YT, Vakser IA, Loew G, Waskell L 1998 Identification of the binding site on cytochrome P450 2B4 for cytochrome b5 and cytochrome P450 reductase. J Biol Chem 273:17036–17049[Abstract/Free Full Text]
  79. Lehnerer M, Schulze J, Achterhold K, Lewis DF, Hlavica P 2000 Identification of key residues in rabbit liver microsomal cytochrome P450 2B4: importance in interactions with NADPH-cytochrome P450 reductase. J Biochem (Tokyo) 127:163–169[Abstract/Free Full Text]
  80. Shen S and Strobel HW 1993 Role of lysine and arginine residues of cytochrome P450 in the interaction between cytochrome P4502B1 and NADPH-cytochrome P450 reductase. Arch Biochem Biophys 304:257–265[CrossRef][Medline]
  81. Lajic S, Nikoshkov A, Holst M, Wedell A 1999 Effects of missense mutations and deletions on membrane anchoring and enzyme function of human steroid 21-hydroxylase (P450c21). Biochem Biophys Res Commun 257:384–390[CrossRef][Medline]
  82. Pinto G, Tardy V, Trivin C, Thalassinos C, Lortat-Jacob S, Nihoul-Fekete C, Morel Y, Brauner R 2003 Follow-up of 68 children with congenital adrenal hyperplasia due to 21-hydroxylase deficiency: relevance of genotype for management. J Clin Endocrinol Metab 88:2624–2633[Abstract/Free Full Text]
  83. Loke KY, Lee YS, Lee WW, Poh LK 2001 Molecular analysis of CYP-21 mutations for congenital adrenal hyperplasia in Singapore. Horm Res 55:179–184[CrossRef][Medline]
  84. Nikoshkov A, Lajic S, Holst M, Wedell A, Luthman H 1997 Synergistic effect of partially inactivating mutations in steroid 21-hydroxylase deficiency. J Clin Endocrinol Metab 82:194–199[Abstract/Free Full Text]
  85. Krone N, Braun A, Roscher AA, Knorr D, Schwarz HP 2000 Predicting phenotype in steroid 21-hydroxylase deficiency? Comprehensive genotyping in 155 unrelated, well defined patients from Southern Germany. J Clin Endocrinol Metab 85:1059–1065[Abstract/Free Full Text]
  86. Chiou SH, Hu MC, Chung BC 1990 A missense mutation at Ile172—Asn or Arg356—Trp causes steroid 21-hydroxylase deficiency. J Biol Chem 265:3549–3552[Abstract/Free Full Text]
  87. Lajic S, Clauin S, Robins T, Vexiau P, Blanché H, Bellanne-Chantelot C, Wedell A 2002 Novel mutations in CYP21 detected in individuals with hyperandrogenism. J Clin Endocrinol Metab 87:2824–2829[Abstract/Free Full Text]
  88. Kharrat M, Tardy V, M’Rad R, Maazoul F, Jemaa LB, Refai M, Morel Y, Chaabouni H 2004 Molecular genetic analysis of Tunisian patients with a classic form of 21-hydroxylase deficiency: identification of four novel mutations and high prevalence of Q318X mutation. J Clin Endocrinol Metab 89:368–374[Abstract/Free Full Text]
  89. Wedell A, Luthman H 1993 Steroid 21-hydroxylase (P450c21): a new allele and spread of mutations through the pseudogene. Hum Genet 91:236–240[Medline]
  90. Amor M, Parker KL, Globerman H, New MI, White PC 1988 Mutation in the CYP21B gene (Ile-172->Asn) causes steroid 21-hydroxylase deficiency. Proc Natl Acad Sci USA 85:1600–1604[Abstract/Free Full Text]
  91. Tusie-Luna M-T, Traktman P, White PC 1990 Determination of functional effects of mutations in the steroid 21-hydroxylase gene (CYP21) using recombinant vaccinia virus. J Biol Chem 265:20916–20922[Abstract/Free Full Text]
  92. Lajic S, Robins T, Krone N, Schwarz HP, Wedell A 2001 CYP21 mutations in simple virilizing congenital adrenal hyperplasia. J Mol Med 79:581–586[CrossRef][Medline]
  93. Tusie-Luna M-T, Speiser PW, Dumic M, New MI, White PC 1991 A mutation (Pro-30 to Leu) in CYP21 represents a potential nonclassic steroid 21-hydroxylase deficiency allele. Mol Endocrinol 5:685–692[Abstract/Free Full Text]
  94. Ezquieta B, Cueva E, Varela J, Oliver A, Fernandez J, Jariego C 2002 Non-classical 21-hydroxylase deficiency in children: association of adrenocorticotropic hormone-stimulated 17-hydroxyprogesterone with the risk of compound heterozygosity with severe mutations. Acta Paediatr 91:892–898[CrossRef][Medline]
  95. Deneux C, Tardy V, Dib A, Mornet E, Billaud L, Charron D, Morel Y, Kuttenn F 2001 Phenotype-genotype correlation in 56 women with nonclassical congenital adrenal hyperplasia due to 21-hydroxylase deficiency. J Clin Endocrinol Metab 86:207–213[Abstract/Free Full Text]
  96. Bojunga J, Welsch C, Antes I, Albrecht M, Lengauer T, Zeuzem S 2005 Structural and functional analysis of a novel mutation of CYP21B in a heterozygote carrier of 21-hydroxylase deficiency. Hum Genet 117:558–564[CrossRef][Medline]
  97. Gunn S K, SLD, Therrel B L, Owebach DI 1993 Molecular genetics of 21-hydroxylase deficient late-onset adrenal hyperplasia. Semin Reprod Endocrinol 11:347–352[CrossRef]
  98. Owerbach D, Sherman L, Ballard AL, Azziz R 1992 Pro-453 to Ser mutation in CYP21 is associated with nonclassic steroid 21-hydroxylase deficiency. Mol Endocrinol 6:1211–1215[Abstract/Free Full Text]
  99. Balsamo A, Cacciari E, Baldazzi L, Tartaglia L, Cassio A, Mantovani V, Piazzi S, Cicognani A, Pirazzoli P, Mainetti B,Zappulla F 2000 CYP21 analysis and phenotype/genotype relationship in the screened population of the Italian Emilia-Romagna region. Clin Endocrinol (Oxf) 53:117–125[CrossRef][Medline]
  100. Barbaro M, Lajic S, Baldazzi L, Balsamo A, Pirazzoli P, Cicognani A, Wedell A, Cacciari E 2004 Functional analysis of two recurrent amino acid substitutions in the CYP21 gene from Italian patients with congenital adrenal hyperplasia. J Clin Endocrinol Metab 89:2402–2407[Abstract/Free Full Text]
  101. Rodrigues NR, Dunham I, Yu CY, Carroll MC, Porter RR, Campbell RD 1987 Molecular characterization of the HLA-linked steroid 21-hydroxylase B gene from an individual with congenital adrenal hyperplasia. EMBO J 6:1653–1661[Medline]
  102. Higashi Y, Hiromasa T, Tanae A, Miki T, Nakura J, Kondo T, Ohura T, Ogawa E, Nakayama K, Fujii-Kuriyama Y 1991 Effects of individual mutations in the P-450(C21) pseudogene on the P-450(C21) activity and their distribution in the patient genomes of congenital steroid 21-hydroxylase deficiency. J Biochem (Tokyo) 109:638–644[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 J. Clin. Endocrinol. Metab.Home page
F. G. Riepe, O. Hiort, J. Grotzinger, W. G. Sippell, N. Krone, and P.-M. Holterhus
Functional and Structural Consequences of a Novel Point Mutation in the CYP21A2 Gene Causing Congenital Adrenal Hyperplasia: Potential Relevance of Helix C for P450 Oxidoreductase-21-Hydroxylase Interaction
J. Clin. Endocrinol. Metab., July 1, 2008; 93(7): 2891 - 2895.
[Abstract] [Full Text] [PDF]

Home page
 J. Clin. Endocrinol. Metab.Home page
F. C. Soardi, M. Barbaro, I. F. Lau, S. H. V. Lemos-Marini, M. T. M. Baptista, G. Guerra-Junior, A. Wedell, S. Lajic, and M. P. de Mello
Inhibition of CYP21A2 Enzyme Activity Caused by Novel Missense Mutations Identified in Brazilian and Scandinavian Patients
J. Clin. Endocrinol. Metab., June 1, 2008; 93(6): 2416 - 2420.
[Abstract] [Full Text] [PDF]

Home page
 J. Clin. Endocrinol. Metab.Home page
R. Menassa, V. Tardy, F. Despert, C. Bouvattier-Morel, J. P. Brossier, M. Cartigny, and Y. Morel
p.H62L, a Rare Mutation of the CYP21 Gene Identified in Two Forms of 21-Hydroxylase Deficiency
J. Clin. Endocrinol. Metab., May 1, 2008; 93(5): 1901 - 1908.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Robins, T.
Right arrow Articles by Persson, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Robins, T.
Right arrow Articles by Persson, B.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals