Recent letters
Displaying 1-10 letters out of 13 published
-
Author Response to Zak letter
Submit responseResponse to the letter of F. Zak containing his comments to the paper Kasparkova, J., Novakova, O., Vrana, O., Intini, F., Natile, G. and Brabec, V. (2006) Molecular aspects of antitumor effects of a new platinum(IV) drug. Mol Pharmacol, 70, 1708-1719.
In the following text we summarize our disposition of the Zak’s comments.
Zak mentions in the version of his letter which we had available that “I work in the field of platinum anticancer drugs for long time. In this article, are shown results of research with platinum complexes adamplatin(II) and adamplatin(IV). According to citations in this article, these compounds are identical with compounds LA-9, CAS No. [250613-98-0] and LA-12, CAS No. [250611-20-2], developed by our team in PLIVA-Lachema a.s., Brno, Czech Republic. Unfortunately, in "Materials and Methods" is probably described synthesis and characterisation of different compounds.“
Zak mentiones that he works in the field of platinum drugs for a long time. However, we found in the database “Web of Science” that he published only 5 papers in the field of platinum antitumor drugs and the first appeared in 2004. On the other hand, for instance the corresponding author (Viktor Brabec) has published in the field of platinum drugs about 130 papers in international refereed journals and the first appeared in 1983. Prof. Giovanni Natile from University of Bari, co-author of the above- mentioned paper, who synthesized the compounds adamplatin(II) and adamplatin(IV), has the competence in the synthesis, characterization, stereochemistry, and reactivity of platinum compounds documented by almost 200 papers published in international refereed journals since 1980. His competence and achievements in the field of platinum chemistry have been also appreciated by electing him the president of the European Association for Chemical and Molecular Sciences (former Federation of European Chemical Societies).
We do not understand why Zak mentions that adamplatin(II) and adamplatin(IV) are identical to those synthesized by them earlier. This information is given in the paper and few references related to this fact are also included. On the other hand, the idea of platinum compounds containing the adamantylamine ligands was described first by other authors (e.g. Inorg. Chem. 32 (1993) 2717-2723) and not by Zak at al. in 2004. And finally we do not agree with his view that synthesis and characterization of different compounds is described in our paper. The reasons are contained below in the letter written by professor Natile:
The statement of professor Giovanni Natile:
I am a bit surprised by the letter sent by F. Zak and I will briefly answer his questions in the same order.
“Preparation of platinum(II) species”. The procedure used by us is very general and in fact similar to that used by F. Zak (and briefly described in his paper in J. Med. Chem. 2004, 47, 761-763), i.e. the reaction of trichlorido(ammine)platinum(II) anion with adamantylamine. The only differences between our procedure and that of Zak concern the counter cation (which does not take part in the reaction) and the solvent.
“Preparation of platinum(IV) species”. In our case, as in the case of Zak ( J. Med. Chem. 2004, 47, 761-763), the platinum(IV) species was obtained from the platinum(II) species by oxidation with hydrogen peroxide followed by the treatment with acetic anhydride. Also this latter procedure is very general and was used by many other authors before Zak. This is the reason why we only quoted the paper of Giandomenico et al. (Inorg. Chem. 1995, 34, 1015) and did not give any detailed description of the preparation procedure. In addition, the journal Molecular Pharmacology as well as our article is not focused to details of chemical synthesis so that the synthesis of the compounds was described in the extent usual for the articles published in Molecular Pharmacology.
“Characterization of the compounds”. Since the prepared compounds are not new and have already been reported in the literature, which is obvious from the article, we have given in our paper only the analytical and spectroscopic data which are, according to our opinion, necessary and sufficient to prove the identity of the compounds.
A final comment of Zak concerns the elemental analysis given for the platinum(IV) species. It is clearly indicated in our paper (and I am surprised that Zak has overlooked the chemical formula following the elemental composition C14H26Cl3KN2Pt) that the analyzed sample of the platinum(IV) species contained one mole of crystallized KCl per mole of the complex. In the specific preparation (not described in detail) the starting platinum(II) species already contained a cocrystallized KCl which was found also in the crystallized final platinum(IV) species. The presence of cocrystallized KCl cannot affect the type of investigations we had planned and therefore we did not proceed to its separation. In addition, Zak wrote in his letter that the analysis was incomplete, however over 99% of papers reporting elemental analysis give figures only for C, H, and N (and sulphur when present), the reason being that these are the elements which are analyzed by standard elemental analysis equipments (in fact it can be clearly seen in the Zak’s paper in J. Med. Chem. 2004 that Pt was not analyzed).
In conclusion, we believe that we have provided a sufficient information and explanation of our view that the compounds used in our work were characterized in a sufficient way. Thus, we strongly reject any doubts of their identity raised by Zak.
January 3, 2007
Viktor Brabec
-
Method of preparation and identity of compounds tested
Submit responseTo the Editor,
I have worked in the field of platinum anticancer drugs for long time. The article by Kasparkova, et al. , “Molecular Aspects of Antitumor Effects of a New Platinum(IV) Drug,” shows the results of research with platinum complexes adamplatin(II) and adamplatin(IV). According to citations in this article, these compounds are identical to compounds LA-9, CAS No.[250613-98-0] and LA-12, CAS No. [250611-20-2], developed by our team in PLIVA - Lachema a.s., Brno, Czech Republic.
Unfortunately, I believe that the "Materials and Methods" probably describes the synthesis and characterization of different compounds.
During the preparation of adamplatin(II) by the method described in the article, ion, and not coordination, compounds of adamantylamine and platinum are created or a mixture of both. The type of compounds developed by this procedure depends on many parameters that are not sufficiently described in the article.
For preparation of adamplatin(IV) from adamplatin(II), two synthetic steps are necessary. However, this is described only as a citation to Giandomenico et al. Inorg. Chem. 34, 1015 (1995). This article does not describe complexes of platinum with adamantylamine, so the process was probably modified, but no modification is mentioned. Preparation of Pt complexes with adamantylamine (LA-9 and LA-12) was first described in patents (WO99/61450, WO99/61451) and in the article [Zak et al.; J.Med. Chem. 47, 761 (2004)].
The characterization of the identity of the prepared compounds is inadequate to determine which compounds were prepared and subsequently tested. Incomplete results of elemental analysie are shown (content of Pt and Cl is missing). Only 1H NMR spectra are described and in the case of IR spectrometry, only Pt-Cl bound is evaluated, which corresponds to many types of platinum complexes. The summary formula of adamplatin(IV) - C14H26Cl3KN2Pt neither corresponds to LA-12 nor to the structural formula, shown in the article.
Due to the lack of information about the exact method of preparation and the incomplete results of identification, we have some misgivings about the identity of the compounds used for biological tests. It would be helpful to provide this information which is necessary to the elucidation of the structure of compounds in the article.
Frantisek Zak PLIVA-Lachema, Brno, Czech Republic
-
Authors' Response
Submit responseDear Editor,
In our recent article (Kohlstedt et al., 2006. Angiotensin-converting enzyme (ACE) dimerization is the initial step in the ACE inhibitor-induced ACE signaling cascade in endothelial cells. Mol Pharmacol 69:1725-1732) we reported that human somatic ACE can be detected in the monomeric as well as the dimeric form in human endothelial cells that endogenously express ACE and in porcine aortic endothelial cells stably overexpressing the human enzyme. Moreover, our study demonstrated that ACE inhibitors enhance dimmer formation and that this step is essential for the subsequent activation of the c-Jun N terminal kinase (JNK), a step we have previously shown to be involved in the “ACE signaling cascade” (Kohlstedt et al., 2004; Kohlstedt et al., 2005).
To address the mechanisms involved in the regulation of ACE dimerization and the link to ACE inhibitor-induced JNK activation we assessed the effects of different antibodies and carbohydrates on ACE dimerization as previous studies, performed mainly in CHO cells and in a biomembrane model system, had led to the identification of a carbohydrate- recognizing domain in the N terminus of ACE and had linked enzyme dimerization with its cleavage/secretion (Kost et al., 2000; Kost et al., 2003; Balyasnikova et al., 2005). We found that treating ACE expressing endothelial cells with antibodies to specific epitopes in the N domain (Danilov et al., 1994) exerted relatively discrete effects on enzyme dimerization that were not significantly different from that detected in solvent-treated cells. The ACE inhibitor, ramiprilat on the other hand, induced a pronounced ACE dimerization that was unaffected by any of the antibodies tested, including one previously reported to prevent ACE dimerization in reverse micelles (Kost et al., 2003). The ACE antibodies tested did however result in the cleavage/secretion of human somatic ACE from endothelial cells, a finding which is consistent with the findings of Balyasnikova et al., (2002). It seems that the link between carbohydrate- induced dimerization and ACE secretion is likely to be a complex one as an antibody that promoted ACE secretion (9B9) as well as an antibody that prevented this process (3G8) were both reported to prevent enzyme dimerization in reverse micelles (Kost et al., 2003).
To address the role of the putative carbohydrate-recognizing domain on ACE dimerization in endothelial cells we performed experiments with sugars including galactose but were only able to use concentrations up to 10 µmol/L as higher concentrations significantly compromised endothelial cell viability. Thus, we may have missed an effect of this carbohydrate on ACE dimerization reported in reverse micelles (Kost et al., 2000) and which these authors now point out can only be assessed in intact ACE- expressing cells using 100-fold higher concentrations of galactose. However, we also addressed which domain of the ACE enzyme is implicated in ACE inhibitor-induced enzyme dimerization using a series of ACE mutants. We found that specific mutation of the C but not the N domain, attenuated both basal ACE dimerization and that induced by the application of an ACE inhibitor (Kohlstedt et al., 2006). These data tend to rule out a role for the N terminal domain in regulating ACE inhibitor-induced ACE dimerization but say nothing about enzyme dimerization elicited by high concentrations of carbohydrates, a response we did not assess.
In summary, we respectfully disagree that there are “fundamental flaws” in our reasoning and while we acknowledge the significant input of Dr’s Danilov, Kost and Sturrock to this topic, we feel it is necessary to exert caution when extrapolating the results of studies obtained in experimental systems in which the human enzyme is overexpressed (e.g. CHO cells) or studies based on the use of a biomembrane model system. This is because the extent and sites of protein glycosylation can vary markedly from cell type to cell type as well as between native and in cultured cells (Bevilacqua et al., 1996). Indeed, somatic ACE contains 17 potential sites for N-glycosylation, mainly of the complex type (Das and Soffer, 1975) and deglycosylated ACE as well as sequentially desialylated and degalactosylated ACE fail to dimerize (Kost et al., 2000). We however wholeheartedly agree with the comment that there is no reason to assume that carbohydrate-controlled ACE dimers and ramiprilat-induced ACE dimers are the same. Indeed, it is clear that human ACE can dimerize via a mechanism that is independent of the carbohydrate-recognizing domain (Kost et al. 2003), indicating that distinct mechanisms may be involved in the regulation of ACE dimerization in response to specific stimuli. Finally, while the physiological/pathophysiological role of ACE dimerization via the carbohydrate-recognizing domain remains to be determined, our data (Kohlstedt et al., 2006) highlight the importance of the C terminal domain of ACE in enzyme dimerization in response to a specific clinically- relevant stimulus i.e., an ACE inhibitor, and link this phenomenon to functional initiation of ACE signaling.
Karin Kohlstedt and Ingrid Fleming Johann Wolfgang Goethe-Universität, Frankfurt, Germany.
References
Balyasnikova IV, Karran E H, Albrecht R F and Danilov S M (2002) Epitope-specific antibody-induced cleavage of angiotensin-converting enzyme from the cell surface. Biochem J 362:585-595.
Balyasnikova IV, Woodman Z L, Albrecht R F, Natesh R, Acharya K R, Sturrock E D and Danilov S M (2005) Localization of an N-domain region of angiotensin-converting enzyme involved in the regulation of ectodomain shedding using monoclonal antibodies. J Proteome Res 4:258-267.
Bevilacqua M, Vago T, Rogolino A, Conci F, Santoli E and Norbiato G (1996) Affinity of angiotensin I-converting enzyme (ACE) inhibitors for N- and C-binding sites of human ACE is different in heart, lung, arteries, and veins. J Cardiovasc Pharmacol 28:494-499.
Danilov S, Jaspard E, Churakova T, Towbin H, Savoie F, Wei L and Alhenc-Gelas F (1994) Structure-function analysis of angiotensin I- converting enzyme using monoclonal antibodies. Selective inhibition of the amino-terminal active site. J Biol Chem 269:26806-26814.
Das M and Soffer R L (1975) Pulmonary angiotensin-converting enzyme. Structural and catalytic properties. J Biol Chem 250:6762-6768.
Kohlstedt K, Brandes R P, Muller-Esterl W, Busse R and Fleming I (2004) Angiotensin-converting enzyme is involved in outside-in signaling in endothelial cells. Circ Res 94:60-67.
Kohlstedt K, Busse R and Fleming I (2005) Signaling via the angiotensin-converting enzyme enhances the expression of cyclooxygenase-2 in endothelial cells. Hypertension 45:126-132.
Kohlstedt K, Gershome C, Friedrich M, Muller-Esterl W, Alhenc-Gelas F, Busse R and Fleming I (2006) Angiotensin-converting enzyme (ACE) dimerization is the initial step in the ACE inhibitor-induced ACE signaling cascade in endothelial cells. Mol Pharmacol 69:1725-1732.
Kost OA, Balyasnikova I V, Chemodanova E E, Nikolskaya I I, Albrecht R F and Danilov S M (2003) Epitope-dependent blocking of the angiotensin- converting enzyme dimerization by monoclonal antibodies to the N-terminal domain of ACE: possible link of ACE dimerization and shedding from the cell surface. Biochemistry 42:6965-6976.
Kost OA, Bovin N V, Chemodanova E E, Nasonov V V and Orth T A (2000) New feature of angiotensin-converting enzyme: carbohydrate-recognizing domain. J Mol Recognit 13:360-369.
-
The Missing Link: ACE Dimerization and Shedding
Submit responseDear Editor,
The article entitled “Angiotensin converting enzyme (ACE) dimerization is the initial step in the ACE inhibitor-induced ACE signaling cascade in endothelial cells” by Kohlstedt et al. (2006) provides evidence that ACE exists as both a monomer and a dimer, not only on the surface of transfected CHO cells, as demonstrated previously (Kost et al., 2003), but also on the surface of cultured human umbilical vein endothelial cells. Moreover, the amount of ACE dimers on the surface of human endothelial cells increased dramatically in the presence of the ACE inhibitor ramiprilat and this ligand binding initiates a signaling cascade involving phosphorylation of the ACE cytoplasmic region. This is the first evidence for cell signaling being mediated via ectodomain interaction(s) between the ACE monomeric forms and the study has important implications regarding our understanding of the unique roles of the ACE protein’s N- and C- domains. Kohlstedt et al. suggest that their experiments do not support our hypothesis of a link between ACE dimerization and shedding from the cell surface (Kost et al., 2003, Balyasnikova et al., 2005). We believe that there are three fundamental flaws in this reasoning, based on previous work in our laboratories. First, the authors were unable to prevent ramiprilat-induced ACE dimerization on the surface of human endothelial cells using different oligosaccharides (glucose, mannitol and galactose). We have shown (Kost et al., 1998, 2000) that ACE forms dimers in reverse micelles, whereas deglycosylated ACE failed to dimerize. Carbohydrates, especially Neu5Ac- and Gal-terminated saccharides, competitively inhibited ACE-ACE interaction. Moreover, the most effective glycan inhibitors of ACE dimerization were the total pool of ACE oligosaccharides and biantennary complex oligosaccharides from other glycoproteins. These findings allowed us to conclude that ACE possesses a specific carbohydrate-recognizing domain (CRD, lectin-like center).
The putative region of this CRD was further localized on the N-domain by using truncated N- and C-domains of ACE and monoclonal antibodies to different epitopes of the N-domain (Kost et al. 2003). Furthermore, galactose not only prevented ACE dimerization in a biomembrane model system, but also affected antibody-induced ACE shedding from the surface of ACE-expressing CHO cells (Kost et al. 2003). Reduced glycosylation of somatic ACE also significantly increased the basal rate of ACE shedding. Two monoclonal antibodies (9B9 and 3G8) with overlapping epitopes that blocked ACE dimerization in the reverse micelles, also affected ACE shedding from the surface of CHO cells. Kohlstedt et al. did not observe any effect of galactose, glucose or mannitol on ramiprilat-induced ACE dimerization. These data are not surprising as glucose and mannitol have a four- and two-fold order of magnitude lower affinity for the CRD than galactose, respectively. The only monosaccharide used in Kohlstedt’s study that was likely to affect ACE dimerization was galactose. However, the galactose concentration (10 ìM) that was needed to block ACE dimerization in a biomembrane model system (Kost et al. 2000) is inappropriate for cell membrane experiments where galactose can be involved in numerous other glycan-glycan and glycan-protein interactions. Thus, only 100-fold higher concentrations of galactose affected basal and antibody-induced ACE shedding from ACE-expressing cells (Kost et al., 2003).
Secondly, there is no reason to accept a priori that carbohydrate- controlled ACE dimers and ramiprilat-induced ACE dimers are the same. In fact, bovine ACE was able to form three types of carbohydrate-mediated dimers in reverse micelles and even formed a tetramer that was not controlled by carbohydrates (Kost et al. 2003, Grinshtein et al.1999). Human ACE was able to form only one compact carbohydrate-controlled dimer and larger dimer that was not formed via carbohydrate-recognizing domain (Kost et al. 2003).
Finally, Kohlstedt et al exclude the possibility of ACE dimerization affecting ectodomain shedding based on their data showing that the antibody mAb 9B9 did not prevent ramiprilat–induced ACE dimerization on the cell surface. Once again, these data are not surprising and the conclusions not entirely appropriate since this antibody only inhibited dimerization in the reverse micelles. On the other hand, the mAb 3G8, which has an overlapping epitope with that of mAb 9B9 blocks both dimerization in reverse micelles and shedding from the surface of ACE- expressing cells (Kost et al. 2003). In contrast, mAb 9B9 dramatically increased ACE shedding from the surface of these cells (Balyasnikova et al. 2002, Balyasnikova et al. 2005, Kohlstedt et al. 2006).
Important findings have emerged from our studies on the effects of mAbs on different functions of the ACE enzyme that are pertinent to this discussion: 1) the binding of mAbs to a region of the N domain, defined as the overlapping surface of epitopes 9B9 and 3A5, significantly induced ACE shedding; 2) the binding of mAbs to another region of the N domain, defined as the overlapping surface of epitopes 9B9 and 3G8, inhibited dimerization of the somatic ACE in the reverse micelles. Therefore, using a set of mAbs, we have identified a region on the N domain of ACE, which is very sensitive to the binding of monoclonal antibodies. mAb 3G8 results in inhibition of ACE shedding on the cell surface and prevention of dimerization in reverse micelles. Binding of mAb 3A5 to the N-domain resulted in conformational changes that prevented binding of other mAbs to the N domain, inhibition of ACE catalytic activity (Danilov et al., 1994), and induction of ACE shedding (Balyasnikova et al., 2002). However, binding of mAb 9B9 to the N domain prevented dimerization in reverse micelles (as with mAb 3G8) and increased ACE shedding (as with mAb 3A5) - Balyasnikova et al. (2005). Thus, the inability of mAb 9B9 to prevent ramiprilat-induced dimerization of ACE on the cell surface should not preclude the proposed influence of ACE dimerization on ACE ectodomain shedding. Undoubtedly, the crosstalk between proteins is a complex and exquisitely controlled mechanism for triggering and controlling cell function. Further work in this field is needed to reveal the full impact of ACE and various types of signaling on the renin angiotensin system and cardiovascular health.
Sergei M. Danilov (University of Illinois at Chicago, USA), Olga A. Kost (Moscow State University, Russia ), Edward E. Sturrock (University of Cape Town, South Africa).
REFERENCES 1. Kohlstedt K, Gershome C, Friedrich M, Muller-Esterl W, Alhenc-Gelas F, Busse R, Fleming I. (2006) Angiotensin converting enzyme (ACE) dimerization is the initial step in the ACE inhibitor-induced ACE signaling cascade in endothelial cells. Mol. Pharm 69, 1725-1732. 2. Kost OA, Balyasnikova IV, Chemodanova EE, Nikolskaya II, Albrecht RF II, Danilov SM (2003). Epitope-dependent blocking of the angiotensin- converting enzyme dimerization by monoclonal antibodies to N-terminal domain of ACE: Possible Link of ACE dimerization and shedding from the cell surface. Biochemistry 42, 6965-6976. 3. Balyasnikova IV, Woodman ZL, Albrecht RFII, Natesh R, Acharya KR, Sturrock ED, Danilov SM. (2005) Localization of an N domain region of angiotensin-converting enzyme involved in the regulation of ectodomain shedding using monoclonal antibodies. J Proteome Res. 4, 258-267. 4. Kost OA, Orth TA, Nikolskaya II, Nametkin SN, Levashov AV. (1998) Carbohydrates regulate the dimerization of angiotensin-converting enzyme, Biochem. Mol. Biol. Int. 44, 535-542. 5. Kost OA, Bovin NV, Chemodanova EE, Nasonov VV, Orth TA. (2000) New feature of angiotensin-converting enzyme: carbohydrate-recognizing domain, J. Mol. Recognit. 13, 360-369. 6. Grinshtein SV, Nikolskaya II, Klyachko NL, Levashov AV, Kost OA (1999) Structural organization of membrane and soluble forms of somatic angiotensin-converting enzyme. Biochemistry (Moscow), 64, 571-580. 7. Danilov S, Jaspard E, Churakova T, Towbin H, Savoie, F, Lei W, Alhenc Gelas F. (1994) Structure-function analysis of angiotensin I-converting enzyme using monoclonal antibodies. J Biol Chem. 269, 26806-26814. 8. Balyasnikova IV, Karran EH, Albrecht RFII, Danilov, SM. (2002) Epitope- specific antibody-induced cleavage of angiotensin-converting enzyme from the cell surface. Biochem. J. 362, 585-595.
-
Response to Savolainen Letter
Submit responseIt goes naturally and conveniently to ascribe biological changes induced by hydrogen sulfide to toxic phenomena. But the real issues raised in Dr. Savolainen’s letter are two-fold, distinguishing a “toxic effect” from a physiological action of H2S being one and understanding the mechanisms for H2S-induced modulation of ion channel functions being the other.
Dr. Savolainen assumed that functional changes in “synaptosomes from hydrogen sulfide-treated rat brain samples” were toxic outcomes. Endogenous production of H2S is a well-documented fact. Without knowing in vivo concentration of hydrogen sulfide after sulfide injection into animals one cannot tell whether H2S in vivo has reached a toxic level or remained in a physiological range. Furthermore, the inhibition of mitochondrial cytochrome oxidase by hydrogen sulfide is not necessarily a consequence of toxic damage. Would it also be important if H2S at physiological range can modulate the activity of this enzyme? More detailed discussion on and comparison between physiological and toxicological effects of H2S have been described before (1, 2).
Activation of KATP channels in smooth muscle cells by H2S (3) was not only observed with the whole-cell recording but also with the cell-free single channel recording configurations of the patch-clamp technique. In the absence of intracellular milieu, the effect of H2S on single KATP channel proteins on a membrane patch would not be affected by cellular metabolism or integrity of respiratory machinery in mitochondria. As such, different mechanisms may be involved in cellular and molecular effects of H2S, depending on metabolic status of the systems under investigation.
1. Wang R. The Evolvement of Gasotransmitter Biology and Medicine: From atmospheric toxic gases to endogenous gaseous signaling molecules. In "Signal Transduction and the Gasotransmitters: NO, CO, and H2S in Biology and Medicine", (Wang R, ed.), pp. 3-32, Totowa, Humana Press, 2004.
2. Wang R. Two’s company, three’s a crowd – Can H2S be the third endogenous gaseous transmitter? FASEB J. 16: 1792-1798, 2002.
3. Tang G, Wu L, Liang W, Wang R. Direct stimulation of KATP channels by exogenous and endogenous hydrogen sulfide in vascular smooth muscle. Mol. Pharmacol. 68: 1757-1764, 2005.
-
Toxicity of hydrogen sulfide
Submit responseHydrogen sulfide is highly toxic because it is an inhibitor of mitochondrial cytochrome oxidase at minute concentrations. Histotoxic hypoxia associated with the inhibition initiates lipid peroxidation which also affects function of ion channel. For example, we found a decrease in veratridine-dependent transmembrane potential in synaptosomes from hydrogen sulfide-treated rat brain samples (1). Such findings lead us to wonder whether similar changes in other organs might result from this toxic effect. 1 Rafalowska U, Zitting A, Savolainen H. Metabolic changes in rat brain synaptosomes after exposure to sulfide in vivo. Toxicol Lett 34:193-200 (1986)
-
Author's Response
Submit responseIn answer to the letter by Marceau et al., we wish to provide a brief background. Members of this laboratory discovered two human enzymes, carboxypeptidase N of plasma (Levin et al., 1982; Skidgel and Erdös, 2004) and carboxypeptidase M on plasma membranes (Skidgel, 2004; Skidgel et al., 1989), which cleave off the C-terminal Arg of bradykinin and kallidin. Although this was initially considered to be an inactivation step of B2 kinin receptor agonists, Regoli and associates subsequently found that des -Arg kinins are ligands for a different receptor they named B1 (Regoli and Barabe, 1980). Thus, carboxypeptidase-mediated cleavage of kinins is an obligatory activation step to generate B1 agonists. In addition to the studies mentioned above, we have published extensively on the biochemistry and functions of the angiotensin converting enzyme (ACE) or kininase II and appreciate the complexities of this important molecule (Erdös, 1990; Erdös et al., 1999). Because ACE inhibitors are such widely used drugs, their direct effects on the B1 receptor to release nitric oxide (NO) is of obvious importance and the subject of our publications (Ignjatovic et al., 2004; Ignjatovic et al., 2002a; Ignjatovic et al., 2002b).
The letter by Marceau et al., purports to address our results published in a recent issue of Molecular Pharmacology (Ignjatovic et al., 2004) by stating in their first sentence that we report “Angiotensin converting enzyme (ACE) inhibitors activate the kinin B1 receptor in a direct manner, presumably by interacting with an extracellular Zn2+ binding domain present in the extracellular loop…”. In fact, those results were reported in our initial paper in the Journal of Biological Chemistry in 2002 (Ignjatovic et al., 2002b). The actual subject of the Molecular Pharmacology report was to investigate how the B1 signaling pathways are stimulated by peptide ligands and ACE inhibitors in endothelial cells. One of the novel findings is that receptor activation by the two types of ligands initiates different signaling pathways, leading to the possibility that for some responses, peptide ligands may be effective, but ACE inhibitors may not (and possibly vice versa). Marceau et al., did not quite address the results of the Molecular Pharmacology paper, but rather primarily attacked our results in the previous publication. Nevertheless, they have already been afforded a forum to present their point of view in a previous publication (Fortin et al., 2003), much of which they repeat in this letter. The letter by Marceau et al., contains an element of surprise as it gives the impression that our results reported in the Journal of Biological Chemistry(Ignjatovic et al., 2002b) and extended in Molecular Pharmacology (Ignjatovic et al., 2004) cannot be repeated. However, it is to be noted that Marceau et al., have never tried to replicate our findings by employing the same cells, species or model systems we used, but have used different species, cell types and transfected tagged receptors and still dispute our results. As we showed (Ignjatovic et al., 2004), in bovine endothelial cells the ACE inhibitor response depends the influx of extracellular calcium into the cells and not on initial intracellular calcium release in contrast to results with the peptide ligand. Consequently, ACE inhibitors did not have the same direct effects as peptide ligands in signaling pathways in these cells. So far, many of Marceau et al’s experiments with endogenously expressed receptors (Fortin et al., 2003) have relied on responses resulting in smooth muscle contraction (rabbit aorta, isolated mouse stomach, human umbilical vein) or ERK1/2 phosphorylation in isolated smooth muscle cells. In addition, their vessel preparations were incubated for 6 hours in a Krebs solution to induce the B1 receptor response. Under these conditions it is expected that the endothelial layer would be damaged or not responsive as evidenced by contractions they obtained in response to B1 agonist. In contrast, our studies have focused on endothelial responses in which B1 receptor stimulation results in significant and profound NO production as established with real time direct measurements of NO. This would lead to generation of cGMP and smooth muscle relaxation in vivo. Besides these differences, variation in experimental protocols could also play a role. For example, one major difference in our cell culture protocols is Fortin et al’s routine use of prolonged serum starvation (overnight for transfected HEK cells and 36 h for smooth muscle cells (Fortin et al., 2003)), for the ERK1/2 phosphorylation studies, a pretreatment that we chose not to use. In preliminary experiments, we have found that serum starvation can selectively inhibit or abolish the response to ACE inhibitor without affecting peptide ligand mediated responses.
Despite the very different experimental approaches and protocols, the conclusion of Fortin et al., as stated in the first sentence of the discussion was “We failed to reproduce the claim of Ignjatovic et al.” (Fortin et al., 2003). The main reason they failed to reproduce our findings (which have been repeated many times by different laboratory members through the years) is because they didn’t try to reproduce them, as stated above. A more scientific approach would be to investigate the reasons for the differences. As different signaling pathways are undoubtedly responsible for smooth muscle contraction or ERK1/2 phosphorylation versus endothelial nitric oxide production, the most likely explanation at this point is that ACE inhibitors do not stimulate all the same signal transduction pathways as peptide ligands of the B1 receptor. Had Marceau, carefully read and considered the findings in our paper (Ignjatovic et al., 2004), they would have found that this is indeed one of the major conclusions of the studies.
With regard to the lack of response of the yellow fluorescent protein (YFP)-tagged rabbit B1 receptor (which they refer to as “a form of rabbit recombinant B1 receptor” in their letter) to ACE inhibitors, we are not sure of the reasons for the difference from our results using transfected untagged human B1 receptors, but there are several possibilities. First of all, they do not use the same assay systems for the rabbit receptors that we have used for the human B1 receptor. Instead of measuring an increase in intracellular calcium as we did, they did either phospholipase A2 assays or ERK1/2 phosphorylation assays. One potential reason for the lack of response in their studies, as mentioned above, is the triggering of different signal transduction pathways for B1 peptide agonists versus ACE inhibitors. A second possibility is that the YFP tag alters the receptor so that it no longer responds to ACE inhibitors, but maintains its responsiveness to peptide ligands. We have tried tagging the B1 receptor with a variety of markers at both the N- and C-terminus and have found that it can alter the signaling properties of the receptor in significant ways that are unpredictable. Thus, caution must always be used when one is unable to reproduce a response with a tagged receptor. In this regard, one of the assays used by Fortin et al., to assess the stimulation of the YFP-labeled B1 receptor was movement into caveolae- related rafts after addition of B1 agonist, which they had previously reported in this journal (Sabourin et al., 2002). In contrast to these findings, Lamb et al., using untagged human B1 receptors transfected in HEK cells, showed that B1 receptor stimulation did not move the B1 receptor into caveolae-related rafts whereas under the same conditions, B2 receptor stimulation transiently increased its localization in caveolae- related rafts (Lamb et al., 2002). Whether this discrepancy was due to species differences or artifacts induced by tagging of the rabbit receptor has not been addressed, but raises concerns that the YFP-labeled rabbit B1 receptor may not recapitulate all of the responses of the native human receptor.
Marceau et al., misrepresent our statements regarding the Zn2+ binding motif by stating “The Zn2+ binding motif HEXXH is postulated to be conserved in mammalian species, which is not true (table 1, compiled from GenBank)”. In the introduction to our paper in Molecular Pharmacology, we stated (par. 2, line 6): “This sequence is conserved in B1 receptors in several species,…” (emphasis added). Their table actually supports our statement, where they show the sequence is strictly conserved in 7 out of 11 species, with 2 of the 3 critical residues present in 3 of the 4 remaining species. In dog, only one of the 3 critical residues is conserved. We actually already reported the alignment of the dog sequence in a short review published in a symposium volume (Ignjatovic et al., 2002a), to which Marceau et al., also contributed (Marceau et al., 2002). We would also like to point out that the HEAWH sequence is strictly conserved in the chimpanzee genomic sequence as well.
The question of whether the sequence of the bovine B1 receptor is consistent with our hypothesis regarding the importance of the HEXXH consensus Zn-binding sequence for the ACE inhibitor-mediated responses would require further experimentation. This recently released partial genomic sequence containing HDWAP instead of HEAWH clearly has not been shown to encode a functional bovine B1 receptor and the corresponding protein has not been expressed. Caution must be used in interpreting initial reports of partial genomic sequences which may contain pseudogenes, polymorphisms or sequencing errors. So far, our results are consistent with the hypothesis that the HEXXH sequence on the B1 receptor is necessary for stimulation by ACE inhibitors in cells of various origins. Supporting data include the following: 1) The B1 receptor response to ACE inhibitors, but not peptide ligands, was abolished when the first His in this sequence was mutated to Ala in cells transfected with the mutant construct. 2) The ACE inhibitor response, but not peptide ligand response, was blocked by a synthetic peptide containing this sequence. 3) The esterified form of enalaprilat, enalapril, which does not bind Zn2+, did not activate the B1 receptor. If it is confirmed that a functional bovine B1 receptor is encoded by this sequence, in which the second zinc binding His is replaced by Pro, it could mean another residue fulfills this function in the bovine receptor. In any case, the evidence presented so far falls short of disproving our hypothesis.
Marceau et al., provide two alternative explanations for our data; the possibility that ACE inhibitors potentiate the local level of kinins endogenously generated by the cells or that ACE itself acts as the signaling molecule in response to ACE inhibitors and that this signaling can be blocked by a B1 antagonist interacting with the ACE active site. However, there are several lines of evidence that negate these interpretations.
First, the hypothetical level of kinin production that could be expected, as reported by Houle et al. (Houle et al., 2003) is quite small, orders of magnitude less than concentration used to stimulate the B1 responses in our system and were also generated over 10 minutes. Our letter in Hypertension (Deddish et al., 2003) already pointed out the various errors and misrepresentations in this publication. Second, this possibility was disproven in our first publication (Ignjatovic et al., 2002b), where we showed that cells transfected with the B2 receptor alone did not respond to enalaprilat, but responded to exogenous addition of bradykinin. Had kinins been generated or their levels increased by ACE inhibitor, the B2 receptor should have responded. Third, cells with the mutant B1 receptor (HEAWH to AEAWH) did not respond to ACE inhibitor but did to des-Arg-kallidin. If ACE inhibitors were activating B1 via generation of endogenous kinins, then the mutant receptor should have responded as well. Finally, for endogenous kinins to be involved, they would first have to be converted to the des-Arg kinin B1 agonists by a cellular kininase I – type enzyme (carboxypeptidases) as mentioned above. For example, we have found that in cytokine-stimulated human lung microvascular endothelial cells (HLMVEC), the B1- mediated NO production stimulated by addition of B2 agonist kinin was blocked to the same degree by either a B1 receptor antagonist or the carboxypeptidase inhibitor 2- mercaptomethyl-3-guanidinoethylthiopropanoic acid (MGTA, the specific kininase-I inhibitor) added to block conversion of the B2 agonist to B1 agonist (Sangsree et al., 2003). To directly address whether endogenous generation of des-Arg kinin could be involved in the ACE inhibitor response, we carried out experiments in which cytokine-stimulated HLMVEC were preincubated without or with 20 µM MGTA for 15 - 20 min and then enalaprilat (100 nM) was added to stimulate NO production via the B1 receptor. Under these conditions NO production in response to enalaprilat was the same in the control cells and in the cells preincubated with MGTA, showing endogenous kinins were not involved.
With regard to the possibility that ACE inhibitors may act directly through ACE as a signaling molecule this is unlikely as the only signaling cascade shown to be activated by ACE inhibitor binding to ACE is the c-Jun N-terminal kinase (JNK) pathway that leads to increased protein expression of ACE or COX-2 (Kohlstedt et al., 2004a; Kohlstedt et al., 2004b). In addition, our published data (Ignjatovic et al., 2004; Ignjatovic et al., 2002b) provide compelling evidence against this possibility. First, B1- mediated responses are stimulated by ACE inhibitors in CHO and HEK cells transfected with the native untagged human B1 receptor, which are blocked by B1 receptor antagonists. These cells lack any detectable ACE expression and untransfected cells did not respond to ACE inhibitors. Second, cells transfected with ACE alone did not respond to ACE inhibitors. Third, lisinopril, which is also a highly potent ACE inhibitor, but has a different structure than the other ACE inhibitors used, did not stimulate the B1 receptor in our system. Finally, control human lung microvascular endothelial cells, which express ACE, gave only rather negligible responses to ACE inhibitor compared to cells in which B1 receptor was induced with cytokines. Although we have considered and explored many possible alternate explanations for our data over the years, the only conclusion consistent with our findings so far is that ACE inhibitors directly activate the B1 receptor and that the region containing the HEAWH sequence is important for this effect.
References
Deddish PA, Hecquet C, Erdos EG, Marceau F, Houle S, Molinaro G and Adam A (2003) B2R of Bradykinin Activated by Proteases * Response: Does the Bradykinin B2 Receptor Function as a Protease-Activated Receptor? Hypertension 42:1e-2.
Erdös EG (1990) Angiotensin I converting enzyme and the changes in our concepts through the years. Lewis K. Dahl memorial lecture. Hypertension 16:363-70.
Erdös EG, Deddish PA and Marcic BM (1999) Potentiation of Bradykinin Actions by ACE Inhibitors. Trends Endocrinol Metab 10:223-229.
Fortin J-P, Gobeil F, Jr, Adam A, Regoli D and Marceau F (2003) Do angiotensin-converting enzyme inhibitors directly stimulate the kinin B1 receptor? Am J Physiol Heart Circ Physiol 285:H277-282.
Houle S, Molinaro G, Adam A and Marceau F (2003) Tissue kallikrein actions at the rabbit natural or recombinant kinin B2 receptors. Hypertension 41:611-7.
Ignjatovic T, Stanisavljevic S, Brovkovych V, Skidgel RA and Erdos EG (2004) Kinin B1 Receptors Stimulate Nitric Oxide Production in Endothelial Cells: Signaling Pathways Activated by Angiotensin I-Converting Enzyme Inhibitors and Peptide Ligands. Mol Pharmacol 66:1310-1316.
Ignjatovic T, Tan F, Brovkovych V, Skidgel RA and Erdos EG (2002a) Activation of bradykinin B1 receptor by ACE inhibitors. International Immunopharmacology 2:1787-1793.
Ignjatovic T, Tan F, Brovkovych V, Skidgel RA and Erdos EG (2002b) Novel mode of action of angiotensin I converting enzyme inhibitors: direct activation of bradykinin B1 receptor. J Biol Chem 277:16847-52.
Kohlstedt K, Brandes RP, Muller-Esterl W, Busse R and Fleming I (2004a) Angiotensin-converting enzyme is involved in outside-in signaling in endothelial cells. Circ Res 94:60-7.
Kohlstedt K, Busse R and Fleming I (2004b) Signaling via the Angiotensin- Converting Enzyme Enhances the Expression of Cyclooxygenase-2 in Endothelial Cells. Hypertension.
Lamb ME, Zhang C, Shea T, Kyle DJ and Leeb-Lundberg LM (2002) Human B1 and B2 Bradykinin Receptors and Their Agonists Target Caveolae-Related Lipid Rafts to Different Degrees in HEK293 Cells. Biochemistry 41:14340-7.
Levin Y, Skidgel RA and Erdös EG (1982) Isolation and characterization of the subunits of human plasma carboxypeptidase N (kininase I). Proc Natl Acad Sci U S A 79:4618-22.
Marceau F, Sabourin T, Houle S, Fortin JP, Petitclerc E, Molinaro G and Adam A (2002) Kinin receptors: functional aspects. Int Immunopharmacol 2:1729-39.
Regoli D and Barabe J (1980) Pharmacology of bradykinin and related kinins. Pharmacol Rev 32:1-46.
Sabourin T, Bastien L, Bachvarov DR and Marceau F (2002) Agonist-induced translocation of the kinin B(1) receptor to caveolae-related rafts. Mol Pharmacol 61:546-53.
Sangsree S, Brovkovych V, Minshall RD and Skidgel RA (2003) Kininase I- type carboxypeptidases enhance nitric oxide production in endothelial cells by generating bradykinin B1 receptor agonists. Am J Physiol Heart Circ Physiol 284:H1959-1968.
Skidgel RA (2004) Carboxypeptidase M, in Handbook of Proteolytic Enzymes (Barrett AJ, Rawlings ND and Woessner JF eds) pp 851-854, Academic Press, London.
Skidgel RA, Davis RM and Tan F (1989) Human carboxypeptidase M. Purification and characterization of a membrane-bound carboxypeptidase that cleaves peptide hormones. J Biol Chem 264:2236-2241.
Skidgel RA and Erdös EG (2004) Lysine carboxypeptidase, in Handbook of Proteolytic Enzymes (Barrett AJ, Rawlings ND and Woessner JF eds) pp 837- 840, Academic Press, London.
-
Reply to Ignjatovic et al.
Submit responseA recent article published in Molecular Pharmacology reports that angiotensin converting enzyme (ACE) inhibitors activate the kinin B1 receptor in a direct manner, presumably by interacting with an extracellular Zn2+-binding domain present in the second extracellular loop of this G protein coupled receptor (Ignjatovic et al., 2004). We would like to draw the attention of the readers to a certain number of inconsistencies in this hypothesis. Firstly, the proposed activation of the B1 receptor by ACE inhibitors has little predictive value of a stimulant response in established bioassays for this receptor, including the ones based on rabbit receptors that possess the HEAWH motif (Fortin et al., 2003). Thus, the contractility and phosphorylation of ERK1/2 in fresh or cultured human or rabbit vascular smooth muscle cells that express the B1 receptors in a regulated manner were unaffected by ACE inhibitors; the conventional B1 receptor agonist Lys-des-Arg9-bradykinin was active in all these systems. Enalaprilat did not modify the contractile response to the peptide agonist if combined with it (rabbit isolated aorta). Further, HEK 293 expressing a form of rabbit recombinant B1 receptor was also unresponsive to enalaprilat (phospholipase A2 assay, ERK1/2 phosphorylation assay, agonist-induced receptor translocation based on confocal microscopy), whereas Lys-des-Arg9-bradykinin exerts documented effects on all these end points (Fortin et al., 2003).
There is a further important inconsistency in the interpretation of results from Ignjatovic et al. (2004). The Zn2+-binding motif HEXXH is postulated to be conserved in mammalian species, which is not true (table 1, compiled from the GenBank). B1 receptor aminoacid sequences are available from several species; in table 1, we have grouped them according to a recent evolutionary classification of placental mammals (Madsen et al., 2001; Murphy et al., 2001). With one exception (Tupaia), the HEXXH sequence is generally conserved in the group III of placental mammals, to which the human species belongs. However, the pentameric consensus sequence is not conserved in the clade IV, in which the bovine species is found. A large part of the present and past experimental results of Ignjatovic et al. (2004) was based on bovine endothelial cells. The bovine genome is not available under an annotated or complete form, but the partial sequence identified in table 1 is from a contig complementary to a nucleotide sequence that codes for most of a G protein coupled receptor 80% identical to other mammalian B1 receptors at the level of nucleotides, and 77% identical to the human B1 receptor at the aminoacid level for the available bovine sequence (from the start codon to within TM7). The 3 predicted N-glycosylation sites common to all the sequences are conserved in the bovine receptor. The identity of the intronless bovine coding nucleotide sequence is further supported by the absence of homology upstream a previously identified intron-exon boundary located 10 bases 5’ from the start codon in the human B1 receptor gene (Bachvarov et al., 1996).
It can be argued that ACE inhibitors stimulate B1 receptors only in endothelial cells via a cell-type specific signaling pathway, but these cells are those naturally expressing ACE, which may be a source of confounding factors if some kinins are formed at the surface of cells that have taken up kininogen from fetal bovine serum present in the culture medium; there is some experimental support for the artefactual production of small kinin quantities in cultured cells systems (Houle et al., 2003). Further confounding factors are that the B1 receptor antagonist used by Ignjatovic et al. (2004), Lys-[Leu8]des-Arg9-bradykinin, is a substrate of purified ACE (Drapeau et al., 1993), and therefore may compete for ACE inhibitor binding at this level, and that ACE may mediate some forms of signaling upon binding an ACE inhibitor (Kohlstedt et al., 2004). Thus, without the evidence for direct binding of zinc and of ACE inhibitors to B1 receptors, it cannot be safely concluded that there is a direct molecular interaction between ACE inhibitors and the kinin B1 receptors.
Table 1. Hypothetical Zn2+ binding consensus sequence HEXXH
organism
sequence
GenBank accession
Clade III of placental mammals
Homo sapiens
HEAWH
NM_000710
Macacamulatta
HEAWH
AF540785
Cercopithecuspygerythrus
HEAWH
AF540784
Cercopithecusaethiops
HEAWH
AY045569
Tupaia minor
HQAWH
AF540786
Oryctolaguscuniculus
HEAWH
U20507
Musmusculus
HEAWH
NM_007539
Rattusnorvegicus
HEAWH
NM_030851
Clade IV of placental mammals (Laurasiatheria)
Susscrofa
HEAWA
AF540788
Bostaurus
HDAWP
AAFC01139146
Canisfamiliaris
PGAWH
AF334947
References
Drapeau G, Audet R, Levesque L, Godin D, and Marceau F (1993) Development and in vivo evaluation of metabolically resistant antagonists of B1 receptors for kinins.J Pharmacol Exp Ther266:192-199.
Fortin JP, Gobeil F, Adam A, Regoli D, and Marceau F (2003) Do angiotensin-converting enzyme inhibitors directly stimulate the kinin B1receptor ?Am J Physiol285:H277-H282.
Houle S, Molinaro G, Adam A, and Marceau F (2003) Tissue kallikrein actions at the rabbit natural or recombinant kinin B2 receptors. Hypertension 41:611-617.
Ignjatovic T, Stanisavljevic S, Brovkovych V, Skidgel RA, and Erdös EG (2004) Kinin B1 receptors stimulate nitric oxide production in endothelial cells: signaling pathways activated by angiotensin I-converting enzyme inhibitors and peptide ligands. Mol Pharmacol66:1310-1316.
Kohlstedt K, Brandes RP, Müller-Esterl W, Busse R, and Fleming I (2004) Angiotensin-converting enzyme is involved in outside-in signaling in endothelial cells. Circ Res94:60-67.
Madsen O, Scally M, Douady CJ, Kao DJ, DeBry RW, Adkins R, Amrine HM, Stanhope MJ, de Jong WW, and Springer MS (2001) Parallel adaptive radiations in two major clades of placental mammals. Nature 409:610-614.
Murphy WJ, Eizirik E, Johnson WE, Zhang YP, Ryder OA, and O'Brien SJ (2001) Molecular phylogenetics and the origins of placental mammals. Nature409:614-618.
-
Author's Response
Submit responseThe authors wish to indicate that upon the request of the reviewers, we evaluated the levels of the 3a-HSD gene product in the 2 cell lines (LA-20, and LnCAP) we believe reflect the genotropic effect of estren. However, using up to 35 cycles of PCR, we were unable to detect the gene product for 3a-HSD. In contrast, liver mRNA, used as positive control, showed the appropriately amplified product after 20-25 cycles of PCR.
In addition, the authors have clearly stated (last paragraph of discussion) "in addition to the conversion to 19-nortestosterone in certain cells, estren has a direct genotropic action via AR". This statement is consistent with the fundamental similarity in the take home message made by these 2 independent studies, namely, the potential risk for genotropic action mediated by AR, when estren is used in a clinical setting.
Gary Krishnan
Lilly Research Laboratories
Indianapolis, IN -
Primary versus indirect effects of estren
Submit responseThe article by Krishnan and colleagues, pre-published online on November 22, 2004 in Molecular Pharmacology (1), carefully shows that 4- estren-3alpha,17beta-diol, commonly termed estren, has potent androgenic effects in vitro and in vivo. The authors state that their work reveals a previously unidentified genotropic action of estren by androgen receptor. We feel that this somewhat inadequately acknowledges the results that we reported earlier (2). Indeed, using direct and indirect biochemical and biological assays, and estrogen and androgen receptor antagonists, we established the same androgen receptor dependency in primary cultures of osteoblasts, a model similar to one where Kousteni and colleagues first identified the skeletal effects of this compound (3,4). We were gratified to note that the binding affinities by estren for estrogen and androgen receptors, as determined by Krishnan and colleagues, were essentially identical to those that we measured, although we each used different receptor sources. The authors stated simply our finding that the conversion of estren to 19-nortestosterone was responsible for some of the androgenic effects of estren. Indeed, we showed that estren is rapidly and facilely converted to the potent androgen, 19-nortestosterone, by 3alpha- hydroxysteroid dehydrogenase, an activity that is expressed ubiquitously by multiple enzyme families (5,6). It is clear from both of our efforts that estren has a 5- to 10-fold higher affinity for androgen receptor relative to estrogen receptor. However, we feel it is important to note again that its androgen receptor affinity is nearly 200-fold less than dihydrotestosterone, and its estrogen receptor affinity is nearly 300-fold less than estradiol. In contrast, 19-nortestosterone is an avid ligand for the androgen receptor, with an affinity 100-fold greater than estren. To us, this predicts that high, pharmacological levels of estren would be required for androgenic activity relative to its potent metabolite, 19- nortestosterone. Even if some androgenic effects by estren could occur through direct sex steroid receptor activation, the current report does not eliminate the likelihood or contribution of its metabolism to 19- nortestosterone. Therefore, although this manuscript does not yet establish if estren is a primary or indirect effector of sex steroid receptor activity, we can be certain that it confirms that its androgenic actions contribute greatly to its function in vivo.
Michael Centrella, Thomas L. McCarthy, and Richard B. Hochberg
Yale University School of Medicine, New Haven, CT, USA1. Krishnan V, et al. (2004) Mol Pharm doi: 10.1124/mol.104.005272.
2. Centrella M, et al. (2004) Mol Endocrinol18:1120-1130.
3. Kousteni S, et al. (2002) Science298:843-846.
4. Kousteni S, et al. (2003) J Clin Invest111:1651-1664.
5. Penning TM (2003) Hum Reprod Update 9:193-205.
6. Napoli JL (2001) Mol Cell Endocrinol171:103-109.
