Evidence That The Membrane Estradiolbinding Sites In The Cns Are Physiologically Relevant Proteins

Progress in this phase of our research has been mainly achieved by the use of an affinity chromatography column constructed with the conjugate E-6-BSA attached to an agarose matrix as reported earlier (11,40). Using this procedure, which leads to at least 1000-fold purification in a single passage of a detergent-solubilized P2 fraction from the rat brain or cerebellum, we were able to identify two proteins in the retained fraction: 1. corresponding to OSCP, a 23 kDa subunit of the ATP synthase complex of mitochondrial origin, and 2. a protein of 18 kDa, also a subunit of this same multimeric enzyme. The OSCP in the eluted fraction retained the capacity to bind E-6-125 I-BSA with nanomolar affinity and it appears to be the only subunit of the ATP synthase to be recognized by the

Fig. 17. (A) SDS-PAGE of recombinant bovine OSCP (rbOSCP, lanes 1 and 2, 1 ^g and 5 ^g, respectively) and E-6-BSA affinity column-retained proteins from digitonin-solubilized B-P2 (lanes 3 and 4, 2 ^g and 10 ^g, respectively). (B) RbOSCP (5 ¡xg, lanes 5 and 7) and purified B-P2 (10 ^g, lanes 6 and 8) were separated in SDS-PAGE and transferred to a nitrocellulose membrane. The binding proteins were detected by incubating the membrane with E-6-125I-BSA (1 x 106 cpm/mL, 1.6 nM) in the absence (lanes 5 and 6) and presences (lanes 7 and 8) of 1 ^M unlabeled E-6-BSA. Curiously, in lane 6 not only the OSCP band was specifically identified, the ligand also labeled a 130 kDa protein. This may represent contamination from plasma membrane. (C) Ligand blotting of rbOSCP (1 ^g in each lane) using E-6-125I-BSA (1 x 106 cpm/mL, 1.6 nM) in the absence of competitors (lane 9), and in the presence of 0.5 \x.M E-6-BSA (lane 10), 0.5 ^M 17a-E-6-BSA (11), and 12 ^M BSA (lane 12). The broad range Mr markers used in (A) are from high to low molecular weight: myosin, P-galactosidase, phosphorylase B, BSA, ovalbumin, bovine carbonic anhydrase, soybean trypsin inhibitor, lysozyme, and aprotinin. In (B) and (C) and in Figs. 18 and 19, similar but prestained broad range markers were used.

Fig. 17. (A) SDS-PAGE of recombinant bovine OSCP (rbOSCP, lanes 1 and 2, 1 ^g and 5 ^g, respectively) and E-6-BSA affinity column-retained proteins from digitonin-solubilized B-P2 (lanes 3 and 4, 2 ^g and 10 ^g, respectively). (B) RbOSCP (5 ¡xg, lanes 5 and 7) and purified B-P2 (10 ^g, lanes 6 and 8) were separated in SDS-PAGE and transferred to a nitrocellulose membrane. The binding proteins were detected by incubating the membrane with E-6-125I-BSA (1 x 106 cpm/mL, 1.6 nM) in the absence (lanes 5 and 6) and presences (lanes 7 and 8) of 1 ^M unlabeled E-6-BSA. Curiously, in lane 6 not only the OSCP band was specifically identified, the ligand also labeled a 130 kDa protein. This may represent contamination from plasma membrane. (C) Ligand blotting of rbOSCP (1 ^g in each lane) using E-6-125I-BSA (1 x 106 cpm/mL, 1.6 nM) in the absence of competitors (lane 9), and in the presence of 0.5 \x.M E-6-BSA (lane 10), 0.5 ^M 17a-E-6-BSA (11), and 12 ^M BSA (lane 12). The broad range Mr markers used in (A) are from high to low molecular weight: myosin, P-galactosidase, phosphorylase B, BSA, ovalbumin, bovine carbonic anhydrase, soybean trypsin inhibitor, lysozyme, and aprotinin. In (B) and (C) and in Figs. 18 and 19, similar but prestained broad range markers were used.

ligand (1,11,12). A further confirmation that the 23 kDa protein is OSCP and that this subunit has high affinity for E comes from recent experiments using a recombinant OSCP of bovine origin (rbOSCP) (kindly provided to us by Dr. Y. Hatefi et al., Scripps Research Institute, La Jolla, CA). This protein in sodium dodecyl sulfate-polyacrimide gel electrophoresis (SDS-PAGE) migrates as a single band in the same position as the 23 kDa protein of the purified brain-P2 (B-P2) fraction as shown in Fig. 17A. Importantly, the E-6-125 I-BSA binds specifically to both the rbOSCP as well as the affinity purified 23 kDa protein from B-P2, because both proteins are displaced by 1 ^M unlabeled ligand (Fig. 17B). It is noteworthy that in the purified fraction of the B-P2, a strong band in the 130 kDa range is also labeled by the ligand, probably an estradiol protein binder from the cell membranes contaminating the B-P2 fraction. The binding is quite selective and stereospecific because BSA does not compete and the 17a-isomer of E-6-BSA is a poor competitor (Fig. 17C).

The presence of E binding sites in mitochondria has been implicated in early experiments that showed that about 10-20% of total estradiol binding sites in immature female rats were detected in a mitochondrial fraction (41,42). Relevant to this finding is the fact that micromolar concentrations of DES, a synthetic estradiol with similar structure as E, also interacted with F0F1ATP-synthase/ATPase in the rat liver to modulate proton transport (43,44). The OSCP subunit is usually considered an essential contributor to the stalk that links the F0 sector to F1 sector, as well as a transducer that transforms the proton gradient across F0 to ATP synthesis on F1. It is conceivable that OSCP is responsible for

Fig. 18. Dose-dependent E-6- I-BSA binding to CB-P2 in nanomolar. CB-P2 proteins were separated by SDS-PAGE and electroblotted to a nitrocellulose membrane using Towbin buffer containing 0.005% SDS. The different lanes of the blotted nitrocellulose membrane were incubated for 45 min at 4°C with 0.30-10.5 nM (3.0-74 x 105 cpm/mL) E-6-125I-BSA. The blots were exposed to Kodak SB film for 2 h and then developed for 1 min. The prestained protein standards are shown on the left of the figure.

Fig. 18. Dose-dependent E-6- I-BSA binding to CB-P2 in nanomolar. CB-P2 proteins were separated by SDS-PAGE and electroblotted to a nitrocellulose membrane using Towbin buffer containing 0.005% SDS. The different lanes of the blotted nitrocellulose membrane were incubated for 45 min at 4°C with 0.30-10.5 nM (3.0-74 x 105 cpm/mL) E-6-125I-BSA. The blots were exposed to Kodak SB film for 2 h and then developed for 1 min. The prestained protein standards are shown on the left of the figure.

these early observations on E-mitochondria interactions. OSCP could be also involved in some of the rapid or nongenomic effects of E previously mentioned (1,45) by regulating ATP synthesis and concentration, since ATP not only serves as the master energy donor (46-48), but is also an important neurotransmitter that can be released by neurons to regulate varieties of brain functions (49). Previous studies have established that E did decrease the ATP concentration of uterus in ovariectomized or immature female rats (50), though this decrease was believed to be mainly owing to E-induced, nuclear estradiol receptor-mediated RNA synthesis (50). However, the involvement of E via mitochondria in decreasing or increasing ATP levels cannot be excluded. Furthermore, this mechanism could participate in the nuclear estradiol receptor-mediated genomic action because several recent studies have emphasized the importance of ATP in the molecular events associated with interaction among steroid hormone, heat shock proteins, and nuclear receptors (51,52).

It remains to be shown that rbOSCP binds [3H]-estradiol. Though rbOSCP binds [3H]-estradiol (about 4% of total), the binding conditions need to be optimized since this protein behaves differently than the classical estradiol receptor from the uterine cytosolic preparation using size exclusion chromatography or filtration assays (unpublished data).

Our results and the aforementioned data suggest a new mechanism by which E can affect the function of diverse cells through binding to F0F1 ATP synthase/ATPase, a key enzyme of the cell energy machinery, as discussed earlier (11). However, this mechanism cannot explain the direct effects of E at the plasma membrane demonstrated previously (for review, see refs. 1 and 13).

So, in our research for the putative mER we ended up isolating an estradiol protein binder that belongs to the multimeric ATP synthase enzyme of mitochondrial origin. This unexpected result was most likely owing to the high level of damaged mitochondrial organelles present in the brain or cerebellum P2 fractions as shown above in the E.M. images of these preparations. In addition, a Western blotting assay of CB-P2 fractions revealed that in addition to the 23 kDa protein, other proteins of different sizes were also identified by this procedure as depicted in Fig. 18. These unidentified proteins may correspond to the mER or mERs. Therefore, recently we have focused our studies to

Fig. 19. The SDS-PAGE analysis of affinity column-purified mP and P . Digitonin-solubilized B-mP2 or B-P3 fractions were applied to the E-6-BSA affinity column and the retained proteins by the affinity column were separated by SDS-PAGE. Lane 1, broad range protein markers (BioRad). Lane 2, retained proteins from solubilized B-mP (5 ^g). Lane 3, retained proteins from solubilized B-P3 (5 ^g). See text for details.

Fig. 19. The SDS-PAGE analysis of affinity column-purified mP and P . Digitonin-solubilized B-mP2 or B-P3 fractions were applied to the E-6-BSA affinity column and the retained proteins by the affinity column were separated by SDS-PAGE. Lane 1, broad range protein markers (BioRad). Lane 2, retained proteins from solubilized B-mP (5 ^g). Lane 3, retained proteins from solubilized B-P3 (5 ^g). See text for details.

isolate these estradiol binding proteins from the plasmalemma-microsomal enriched P3 fraction. To this end, a similar E-6-BSA affinity chromatography column was loaded with a detergent-solubilized P3 fraction and the retained proteins were isolated by SDS-PAGE under reducing conditions. Figure 19 compares the size and mobility of these affinity purified protein in mP2 and P3 fractions from the female rat brain. Note that in the mP2 fraction we have a similar pattern of retained proteins in the columns as shown previously, with the 23 kDa protein as a major band. In contrast, the affinity purified P3 fraction contained two different major proteins, one corresponding to a MW of about 48 kDa and the other smaller, an 18 kDa protein. These proteins may correspond to similar estradiol binding proteins detected by Western blot in Fig. 18 from a CB-P2 crude fraction. Current efforts are aiming to microsequencing these protein.

Was this article helpful?

0 0

Post a comment