Crystal structures of the multispecific 17beta-hydroxysteroid dehydrogenase type 5: critical androgen regulation in human peripheral tissues

Human type 5 17beta-hydroxysteroid dehydrogenase (17beta-HSD5;AKR1C3) plays a major role in the metabolism of androgens in peripheral tissues. In prostate basal cells, this enzyme is involved in the transformation of dehydroepiandrosterone into dihydrotestosterone, the most potent androgen. It is thus a potential target for prostate cancer therapy because it is understood that the testosterone formation by this enzyme is an important factor, particularly in patients who have undergone surgical or medical castration. Here we report the first structure of a human type 5 17beta-HSD in two ternary complexes, in which we found that the androstenedione molecule has a different binding position from that of testosterone. The two testosterone-binding orientations in the substrate-binding site demonstrate the structural basis of the alternative binding and multispecificity of the enzyme. Phe306 and Trp227 are the key residues involved in ligand recognition as well as product release. A safety belt in the cofactor-binding site enhances nicotinamide adenine dinucleotide phosphate binding and accounts for its high affinity as demonstrated by kinetic studies. These structures have provided a dynamic view of the enzyme reaction converting androstenedione to testosterone as well as valuable information for the development of potent enzyme inhibitors.     

Intracellular-specific colocalization of prostaglandin E2 synthases and cyclooxygenases in the brain

Prostaglandin E2 (PGE2) is the major primary prostaglandin generated by brain cells. However, the coordination and intracellular localization of the cyclooxygenases (COXs) and prostaglandin E synthases (PGESs) that convert arachidonic acid to PGE2 in brain tissue are not known. We aimed to determine whether microsomal and cytosolic PGES (mPGES-1 and cPGES) colocalize and coordinate activity with either COX-1 or COX-2 in brain tissue, particularly during development. Importantly, we found that cytosolic PGES also associates with microsomes (cPGES-m) from the cerebrum and cerebral vasculature of the pig and rat as well as microsomes from various cell lines; this seemed dependent on the carboxyl terminal 35-amino acid domain and a cysteine residue (C58) of cPGES. In microsomal membranes from the postnatal brain and cerebral microvessels of mature animals, cPGES-m colocalized with both COX-1 and COX-2, whereas mPGES-1 was undetectable in these microsomes. Accordingly, in this cell compartment, cPGES could coordinate its activity with COX-2 and COX-1 (partly inhibited by NS398); albeit in microsomes of the brain microvasculature from newborns, mPGES-1 was also present. In contrast, in nuclei of brain parenchymal and endothelial cells, mPGES-1 and cPGES colocalized exclusively with COX-2 (determined by immunoblotting and immunohistochemistry); these PGESs contributed to conversion of PGH2 into PGE2. Hence, contrary to a previously proposed model of exclusive COX-2/mPGES-1 coordination, COX-2 can coordinate with mPGES-1 and/or cPGES in the brain, depending on the cell compartment and the age group.     

Interchromatid and interhomolog recombination in Arabidopsis thaliana

Intermolecular recombination events were monitored in Arabidopsis thaliana lines using specially designed recombination traps consisting of tandem disrupted beta-glucuronidase or luciferase reporter genes in direct repeat orientation. Recombination frequencies (RFs) varied between the different lines, indicating possible position effects influencing intermolecular recombination processes. The RFs between sister chromatids and between homologous chromosomes were measured in plants either hemizygous or homozygous for a transgene locus. The RFs in homozygous plants exceeded those of hemizygous plants by a factor of >2, implying that in somatic plant cells both sister chromatid recombination and recombination between homologous chromosomes exist for recombinational DNA repair. In addition, different DNA-damaging agents stimulated recombination in homozygous and hemizygous plants to different extents in a manner dependent on the type of DNA damage and on the genomic region. The genetic and molecular analysis of recombination events showed that most of the somatic recombination events result from gene conversion, although a pop-out event has also been characterized.     

Competitive inhibition by Rauber's sickle of the primitive streak and/or (pre)neural plate inducing effects of sickle endoblast in avian blastoderms

When in unincubated chicken blastoderms the Rauber's sickle is (sub)totally mechanically removed by selective scraping, the further evolution of the blastoderm in culture is often profoundly disturbed, going from only expansion of the upper layer and preneural plate formation to the development of a slowly growing miniature embryo. Our results suggest that the developmental potencies of the embryo are related to the presence or absence of Rauber's sickle material left after its removal. This can be checked after culture by the presence or nonpresence of junctional endoblast (derived from Rauber's sickle) and the concomitant induction of blood islands in the immediate neighborhood. Our study thus indicates that without Rauber's sickle (in the cases of successful total selective removal), an avian blastoderm cannot develop normally, even in the presence of an intact caudal marginal zone. After placing a fragment of quail sickle endoblast on the anti-sickle region of unincubated chicken blastoderms from which the Rauber's sickle was (sub)totally removed, different developmental scenarios were seen, according to the degree of removal, both in the anti-sickle as in the sickle regions. 1) If Rauber's sickle activity is strongly reduced, then besides a centripetally directed miniature embryo, induced by the remnants of the autochthonous Rauber's sickle, an additional centripetally directed embryo or preneural plate (without accompanying blood islands) develops in the anti-sickle region under inductory influence of the apposed quail sickle endoblast. We make a distinction between a neural plate and a preneural plate. The latter consists of a thickening of the upper layer (with the same initial aspect as a neural plate) adjacent to endophyll or sickle endoblast in the absence of chordomesoblast and gastrulation phenomena. 2) If Rauber's sickle activity is totally absent, then the inducing power of the sickle endoblast fragment becomes maximal and, starting from the anti-sickle region, one single embryo (without blood islands) extending over the whole area centralis appears. 3) If much of the Rauber's sickle material has been left in the blastoderm, then the inducing activity of the sickle endoblast, placed on the anti-sickle region, will be totally suppressed (although the sickle endoblast remains intact) and neither a preneural plate nor a primitive streak was induced. After placing a fragment of quail sickle endoblast on the anti-sickle region of an unincubated chicken blastoderm from which the Rauber's sickle and surrounding tissues were completely excised, an embryo was always induced by the sickle endoblast in the adjacent upper layer of this anti-sickle region. In the absence of sickle endoblast, this never occurred. Thus, our experiments demonstrate that in the absence of the Rauber's sickle, a parent tissue (sickle endoblast) induces both gastrulation and neurulation phenomena, while in the full presence of Rauber's sickle these functions are totally suppressed. Moreover, Rauber's sickle not only organizes gastrulation and blood island formation by itself but also influences neurulation at a distance (in space and time) by part of its cell lineage (i.e., sickle endoblast). Our study suggests that the inhibitory effect of Rauber's sickle on its parent tissue (sickle endoblast) represents an early mechanism impairing polyembryony, so that only a single primary major organizer (Rauber's sickle) remains active in the young avian germinal disc.     

Increased platelet-activating factor-induced periventricular brain microvascular constriction associated with immaturity

Oxidant stress contributes to the pathogenesis of hypoxic-ischemic encephalopathies. Platelet-activating factor (PAF) is generated during oxidant stress. We studied the vasomotor mode of actions of PAF on periventricular (PV) microvessels of fetal ( approximately 75% of term), newborn (1-3 days), and adult pigs. PAF constricted PV microvessels from fetal (29.27 +/- 2.6%) and newborn (22.14 +/- 3.2%) pigs but was ineffective in adults (<2.5%). Specific [(3)H]PAF binding was greater in fetus and newborn than in adults; a concordant developmental PAF-induced inositol phosphate formation was observed. PAF-induced vasoconstriction was abrogated by thromboxane A(2) (TXA(2)) synthase and receptor inhibitors, calcium channel blockers, and by removal of endothelium; vasoconstriction to TXA(2) mimetic U-46619 did not differ with age. Immunoreactive TXA(2) synthase expression and PAF-evoked TXA(2) formation revealed a fetus> newborn>adult profile. Thus the greater PAF-induced PV microvascular constriction in younger subjects seems attributable to greater PAF receptor density and mostly secondary to TXA(2) formation from endothelium. The resulting decrease in blood flow may contribute to the increased vulnerability of the PV brain regions to oxidant stress-induced injury in immature subjects.     

Human 20alpha-hydroxysteroid dehydrogenase: crystallographic and site-directed mutagenesis studies lead to the identification of an alternative binding site for C21-steroids

Human 20alpha-hydroxysteroid dehydrogenase (h20alpha-HSD; AKR1C1) catalyzes the transformation of progesterone (Prog) into 20alpha-hydroxy-progesterone (20alpha-OHProg). Although h20alpha-HSD shares 98% sequence identity with human type 3 3alpha-HSD (h3alpha-HSD3, AKR1C2), these two enzymes differ greatly in their activities. In order to explain these differences, we have solved the crystal structure of h20alpha-HSD in a ternary complex with NADP(+) and 20alpha-OHProg at 1.59A resolution. The steroid is stabilized by numerous hydrophobic interactions and a hydrogen bond between its O20 and the N(epsilon ) atom of His222. This new interaction prevents the formation of a hydrogen bond with the cofactor, as seen in h3alpha-HSD3 ternary complexes. By combining structural, direct mutagenesis and kinetic studies, we found that the H(222)I substitution decreases the K(m) value for the cofactor 95-fold. With these results, we hypothesize that the rotation of the lateral chain of His222 could be a mediating step between the transformation of Prog and the release of the cofactor. Moreover, crystal structure analysis and direct mutagenesis experiments lead us to identify a new residue involved in the binding of Prog. Indeed, the R(304)L substitution leads to a 65-fold decrease in the K(m) value for Prog reduction. We thus propose that Prog is maintained in a new steroid-binding site composed mainly of residues found in the carboxy-terminal region of the protein.     

In the absence of Rauber's sickle material,no blood islands are formed in the avian blastoderm

Using the quail-chick chimera technique, we followed the fate of Rauber's sickle cells in older whole blastoderms (cultured for approximately 2 days): after removal of the autochthonous Rauber's sickle from an unincubated chicken blastoderm, a quail Rauber's sickle was grafted isotopically and isochronically in its place. In transverse sections through these chimeras, the grafted quail Rauber's sickle cells were seen to have transformed into a broad row or ridge of quail junctional endoblast cells extending at the inner border of the area containing blood islands. After unilateral removal of the junctional endoblast from an intermediate streak chicken blastoderm (Stage 3; Hamburger and Hamilton [1951] J Morphol 88:49-92), we observed during further in vitro culture that at the operated side, in the area previously occupied by this junctional endoblast, blood islands no longer developed. If after such a unilateral removal of the chicken junctional endoblast quail junctional endoblast was apposed in its place, then blood islands reappeared in the operated area. The intimate contact between the apposed quail junctional endoblast and the recently formed blood islands, derived from peripherally migrating mesoderm, was very obvious on sections through such chimeras. We further demonstrate that Rauber's sickle vs. junctional endoblast is indispensable for the anlage of blood islands in avian blastoderms. Indeed, in the absence of Rauber's sickle material no blood islands develop (even when mesoderm is present after ingression of the upper layer via a primitive streak) in the isolated central region of the area centralis of unincubated chicken blastoderms after culture in vitro. Also, no junctional endoblast and no sickle canal appear in these explants. By contrast, if a Rauber's sickle fragment is placed on such an isolated central blastoderm region, then blood islands develop. These blood islands start to develop from peripherally migrating mesoderm in the neighborhood of the Rauber's sickle-derived junctional endoblast.     

Studies on the angiotensin-converting enzyme and the kinin B2 receptor in the rabbit jugular vein: modulation of contractile response to bradykinin

The rabbit jugular vein (rbJV) was used as a bioassay system to validate some early and new hypothetical interactions between the angiotensin-converting enzyme (ACE) and the B2 receptor, which may be influenced by ACE inhibitors (ACE-I). These involve the potentiation of the contractile effect of bradykinin (BK) and BK analogues, which are inactivated by ACE (e.g., [Hyp3, Tyr(Me8)]-BK (R556)), the prevention of BK-induced B2 receptor desensitisation, and the restoration of receptor sensitivity in tissues desensitised with B2 receptor agonists. Enzymatic degradation studies performed in vitro and in vivo revealed that BK and R556 are readily degraded by rabbit ACE whereas [Phe8psi(CH2-NH)Arg9]-BK (R379) is totally resistant. BK, R556, and R379 contracted endothelium-denuded veins with similar potencies (pEC50 range 8.10-8.50). Tissues pretreated with ACE-I showed an increase in pEC50 values for BK and R556 but not for R379. ACE-I (captopril, enalaprilat) were unable to prevent B2 receptor desensitisation induced by BK (1 microM). ACE-I partially restored B2 receptor-mediated contraction in tissues initially exposed to BK but not to R379. These effects were antagonised by HOE 140 (0.1 microM) but were unaffected by AcLys[Dbeta-Nal7, Ile8]-desArg9BK (R715) (1 microM) or by Losartan (1 microM). In conclusion, the potentiation of BK and its analogues relates exclusively on prevention of their metabolism, B2 receptor desensitisation is not affected by ACE-I, and restoration of tissue responsiveness to BK by ACE-I may be attributed to changes in BK concentrations in the vicinity of the B2 receptor.     

Kinin B1 receptor antagonists with multi-enzymatic resistance properties

The kinin B, receptor has been implicated in a variety of pathological states; therefore, potent, selective, and specific antagonists with prolonged duration of action in vivo are needed. Using R-715 (AcLys[D-beta-Nal(7),Ile(8)] desArg9BK) as a template, new peptides containing alpha-MePhe in position 5, Oic in position 2, and AcOrn instead of AcLys at the N-terminal were prepared and tested for their antagonist potency, their selectivity, and their specificity for the kinin B1 receptor. In vitro metabolic stabilities toward aminopeptidase M (from human plasma), aminopeptidase P (from human platelets), and angiotensin-converting enzyme (purified from rabbit lung) were also investigated. The results of this study indicate that the three modifications applied separately are as well tolerated as they are when present conjointly in the template R-715. Indeed, pA2 values of R-715 (ranging from 8.40 to 8.5) do not differ significantly from the analogues R-954 and R-955 (both ranging from 8.4 to 8.6) when measured at kinin B1 receptors from rabbit aortas and human umbilical veins. Moreover, the chemical modifications utilized in the peptides R-954 and R-955 have provided resistance against aminopeptidases M and P, as well as the angiotensin-converting enzyme, unlike the early (e.g., Lys[Leu8]desArg9BK) and more recent (e.g., R-715, B-9858) generations of B, receptor antagonists. Ongoing in vivo assays will validate the assumption that the analogues R-954 and R-955 have a prolonged duration of action.     

Structural basis for the high-affinity interaction of nidogen-1 with immunoglobulin-like domain 3 of perlecan

Nidogen and perlecan are large multifunctional basement membrane (BM) proteins conserved in all metazoa. Their high-affinity interaction, which is likely to contribute to BM assembly and function, is mediated by the central G2 domain in nidogen and the third immunoglobulin (IG)-like domain in perlecan, IG3. We have solved the crystal structure at 2.0 A resolution of the mouse nidogen-1 G2-perlecan IG3 complex. Perlecan IG3 belongs to the I-set of the IG superfamily and binds to the wall of the nidogen-1 G2 beta-barrel using beta-strands C, D and F. Nidogen-1 residues participating in the extensive interface are highly conserved, whereas the corresponding binding site on perlecan is more variable. We hypothesize that a second, as yet unidentified, activity of nidogen overlaps with perlecan binding and accounts for the unusually high degree of surface conservation in the G2 domain.