点击显示 收起
【关键词】 Targeting
The G protein β5 subunit differs from other β subunits in having divergent sequence and subcellular localization patterns. Although β52 modulates effectors, β5 associates with R7 family regulators of G protein signaling (RGS) proteins when purified from tissues. To investigate β5 complex formation in vivo, we used multicolor bimolecular fluorescence complementation in human embryonic kidney 293 cells to compare the abilities of 7 subunits and RGS7 to compete for interaction with β5. Among the subunits, β5 interacted preferentially with 2, followed by 7, and efficacy of phospholipase C-β2 activation correlated with amount of β5 complex formation. β5 also slightly preferred 2 over RGS7. In the presence of coexpressed R7 family binding protein (R7BP), β5 interacted similarly with 2 and RGS7. Moreover, 2 interacted preferentially with β1 rather than β5. These results suggest that multiple coexpressed proteins influence β5 complex formation. Fluorescent β52 labeled discrete intracellular structures including the endoplasmic reticulum and Golgi apparatus, whereas β5RGS7 stained the cytoplasm diffusely. Coexpression of o targeted both β5 complexes to the plasma membrane, and q also targeted β52 to the plasma membrane. The constitutively activated o mutant, oR179C, produced greater targeting of β5RGS7 and less of β52 than did o. These results suggest that o may cycle between interactions with β52 or other β complexes when inactive, and β5RGS7 when active. Moreover, the ability of β52 to be targeted to the plasma membrane by subunits suggests that functional β52 complexes can form in intact cells and mediate signaling by G protein-coupled receptors.
In contrast to the other four members of the G protein β subunit family, which share 80% amino acid sequence identity, β5 shares only 50% amino acid identity with these other β subunits and exhibits less association with cell membranes (Watson et al., 1994; Jones et al., 2004). Also, unlike the other β subunits, β5 can associate with RGS proteins in the R7 family (RGS6, RGS7, RGS9, and RGS11), which interact with β5 via their G protein -like (GGL) domain (Jones et al., 2004). β5R7 complexes can activate the GTPase activity of o (Posner et al., 1999; Hooks et al., 2003) and accelerate both the activation and deactivation kinetics of GIRK channels (Kovoor et al., 2000; Drenan et al., 2006). Coexpression of β5 and RGS7 increases the expression levels of both proteins compared with when they are expressed individually (Witherow et al., 2000), and mice lacking β5 have reduced levels of R7 family RGS proteins (Chen et al., 2003), suggesting that these proteins are obligate dimers.
Whether β5 also interacts with G protein subunits in vivo is controversial. β52 can activate phospholipase C-β2 (Watson et al., 1994; Zhang et al., 1996; Lindorfer et al., 1998) and inhibit GIRK channels (Mirshahi et al., 2002; Lei et al., 2003) and N-type Ca+2 channels (Zhou et al., 2000). However, when purified from native tissues, β5 is associated with R7 family RGS proteins rather than subunits (Witherow et al., 2000). Complicating the issue, β52 dimers are unstable under nondenaturing buffer conditions (Jones and Garrison, 1999; Jones et al., 2004), which could explain why they have yet to be isolated.
Because G protein-coupled receptors and G protein subunits localize predominantly to the plasma membrane, complexes between β5 and either R7 family proteins or subunits would be expected to localize there as well to modulate signaling. Plasma membrane targeting of β5R7 complexes is promoted by association with both o (Takida et al., 2005) and R7BP (Drenan et al., 2006). Using BiFC, which involves the reconstitution of a fluorescent signal from nonfluorescent fragments of YFP or CFP when they are fused to interacting proteins (Kerppola, 2006), we previously visualized complexes between β5 and 1, 2, or 7 and found that they localized intracellularly rather than at the plasma membrane (Hynes et al., 2004b). This indicated that the β subunit could regulate targeting of β complexes, because these same subunits localized to the plasma membrane when associated with other β subunits. The β subunit, unlike the subunit, is not known to contain modifications that cause membrane targeting (Wedegaertner et al., 1995). However, one means by which β subunits could regulate targeting would be via association with subunits. Because o could target β5RGS7 to the plasma membrane (Takida et al., 2005), we hypothesized that coexpression of o and/or other subunits might lead to plasma membrane targeting of β5 complexes.
Here, using live cell-based assays, we address the issues of which proteins β5 forms complexes with, how complex formation and functionality are related, which subunits β5 complexes interact with, when in the GTPase cycle these interactions take place, and how the localization of β5 complexes is regulated. Using multicolor BiFC, we compare the abilities of seven subunits (1, 2, 5, 7, 10, 11, and 12) and RGS7 to compete for interaction with β5. Using a plasma membrane targeting assay, we compare the abilities of active and inactive o to target β52 and β5RGS7 to the plasma membrane and of o and q to target β52 to the plasma membrane. These studies demonstrate and quantify interactions that have not been detected using in vitro approaches and lead to a model for the roles of β5 complexes in regulating G protein signaling.
Production of Fluorescent Fusion Protein and Subunit Constructs. YFP-N-β1 was produced as described previously (Hynes et al., 2004b). Cer-N-β1 and Cer-N-β5 were produced in the same manner as YFP-N-β1 using the human β1 and β5 cDNAs and Cer(1–158)pcDNAI/Amp, which was produced as described previously (Mervine et al., 2006) using monomeric Cerulean (Rizzo et al., 2004) (obtained from David Piston, Vanderbilt University, Nashville, TN), which contains S72A, Y145A, H148D, and A206K substitutions in ECFP. Cer-N-RGS7 was produced in the same manner as Cer-N-β1 and Cer-N-β5 using human RGS7-S2 (Guthrie cDNA Resource Center, Sayre, PA). For CFP-N-RGS7t, the procedure was the same except that the sequence amino terminal to the GGL domain was deleted by amplifying RGS7 residues 202 to 479 and the polymerase chain reaction product was subcloned into CFP(1–158)pcDNAI/Amp. Cer-N- constructs and CFP-C-β1 were produced as described previously (Mervine et al., 2006). YFP-N-2 was produced in the same manner as the Cer-N- constructs, using YFP(1–158)pcDNAI/Amp (Hynes et al., 2004b). CFP-C-β5 and CFP-C-2 were produced in the same manner as CFP-C-β1, using the human β5 and 2 cDNAs, respectively. Cer-C-β5 was produced in the same manner as CFP-C-β5, using Cer(159–238)pcDNAI/Amp.
mCherry-Mem was produced as described for mRFP-Mem (Mervine et al., 2006) except that mCherry (Shaner et al., 2004) (obtained from Roger Tsien, University of California, San Diego, CA) was used as the polymerase chain reaction template. pEYFP-Golgi, encoding a fusion protein consisting of EYFP and the amino-terminal 81 residues of human beta1,4-galactosyltransferase, which targets to the trans-medial region of the Golgi apparatus, was obtained from Clontech (Mountain View, CA). pEYFP-ER, encoding a fusion protein consisting of EYFP with the ER targeting sequence of calreticulin at the amino terminal end and the ER retrieval sequence, KDEL, at the carboxyl terminal end, was obtained from Clontech. pGM130-EGFP, encoding a fusion protein consisting of EGFP and GM130, a cis-Golgi matrix protein, was obtained from Graham Warren (Yale University, New Haven, CT).
3FLAG-R7BP, consisting of the R7BP coding region subcloned into p3FLAG-CMV10 (Sigma-Aldrich, St. Louis, MO) was obtained from Kendall Blumer (Washington University, St. Louis, MO). The human phospholipase C-β2 cDNA in pRc/CMV (Invitrogen, Carlsbad, CA) was obtained from Ravi Iyengar (Mount Sinai School of Medicine, New York, NY).
The EE epitope (EYMPTE) was introduced into the rat o-1 cDNA by replacing Asp167 with Glu and Gln169 with Met and Arg179 in o-EE was replaced by Cys to produce oR179C-EE by oligonucleotide-directed in vitro mutagenesis using the Bio-Rad Muta-Gene kit. s-YFP was produced as described for s-CFP (Hynes et al., 2004a) except that EYFP (Clontech) containing a substitution of Met for Gln69 was substituted for ECFP. q-YFPpcDNAI/Amp was produced from q-GFP/pcDNAI/Amp (Hughes et al., 2001). EYFP (Clontech) containing a substitution of Met for Gln69 and including S-G-G-G-G-S linkers on each end was substituted for GFP containing the same linkers as a BamHI/SacI cassette. This substitution was performed after the other BamHI and SacI sites in q-GFP were removed by silent mutations using oligonucleotide-directed in vitro mutagenesis and q-GFP was subcloned as a NotI insert into a modified version of pGEM-HE (Hughes et al., 2001) containing no BamHI or SacI sites. The resultant q-YFP cDNA was then subcloned into pcDNAI/Amp as a NotI insert. To produce o-YFP, a BglII site in the 5' untranslated region of o-EE/pcDNAI/Amp was removed by digestion with T4 DNA polymerase and religation and then a unique BglII site was introduced in frame between Pro119 and Phe120 in the B/C loop of the helical domain, analogous to the YFP insertion site in q-YFP, using polymerase chain reactions that produced DNA fragments with overlapping ends that were combined subsequently in a fusion polymerase chain reaction. EYFP (Clontech, Moutain View, CA) containing a substitution of Met for Gln69 and including S-G-G-G-G-S linkers on each end was then subcloned into the BglII site as a BamHI cassette. All subunit constructs used in this study contain the EE epitope. Henceforth in the text the EE designation is omitted for simplicity. All of the above constructs were verified by DNA sequencing.
Imaging of Transfected Cells by Spinning Disc Confocal Microscopy. HEK-293 cells (American Type Culture Collection, Manassas, VA) were plated at a density of 2 x 105 cells per well on four-well chambered coverslips (Lab-Tek II; Nalge Nunc International, Rochester, NY). On the following day, the cells were transiently transfected using 0.25 µl of LipofectAMINE 2000 Reagent (Invitrogen). Plasmids were transfected as described in the legends to Figs. 1, 5, 7, 9, and 10. A membrane marker (YFP-Mem or mCherry-Mem) was included in all transfections.
Fig. 1. Complexes of β5 with 1, 2, 5, 7, 10, 11, and 12 exhibit distinct localization patterns. A–I, images of HEK-293 cells expressing the indicated Cer-C-β5Cer-N- complexes (A–G), s-YFP (H), or mCherry (I). HEK-293 cells were transfected with the following quantities of plasmids: Cer-C-β5, 0.075 µg; Cer-N- subunits, 0.075 µg; YFP-Mem, 0.0025 µg; mCherry, 0.0125 µg (A–G); s-YFP, 0.15 µg; mCherry-Mem, 0.0025 µg (H); and mCherry, 0.0125 µg (I). The top numbers in the images are the ratios of average pixel intensity in the nucleus to that in the cytoplasm and the bottom numbers are the normalized cytoplasmic standard deviations of pixel intensity for the cells shown. J, ratios of average pixel intensity in the nucleus compared with the cytoplasm. K, normalized cytoplasmic standard deviations of pixel intensity for the indicated constructs. Values represent the means ± S.E. The numbers of cells analyzed are as follows: β51, 39; β52, 62; β55, 45; β57, 54; β510, 44; β511, 41; β512, 50; s, 19; and mCherry, 335. Scale bar, 10 µm.
Fig. 5. Both 2 and RGS7 form complexes with β5 that can be imaged simultaneously in the same cells using BiFC. YFP (A) and CFP (B) images from the same cell expressing CFP-C-β5YFP-N-2 and CFP-C-β5Cer-N-RGS7. HEK-293 cells were transfected with the following quantities of plasmids: CFP-C-β5, 0.3 µg; YFP-N-2 and Cer-N-RGS7, 0.15 µg; and mCherry-Mem, 0.0025 µg. YFP (C) and CFP (D) images from the same cell expressing CFP-C-β5YFP-N-2 and CFP-C-β5CFP-N-RGS7-t, which contains a truncated form of RGS7 lacking the DEP domain. HEK-293 cells were transfected with the following quantities of plasmids: CFP-C-β5, 0.3 µg; YFP-N-2 and CFP-N-RGS7-t, 0.15 µg; and mCherry-Mem, 0.0025 µg. The top numbers in the images are the ratios of average pixel intensity in the nucleus to that in the cytoplasm and the bottom numbers are the normalized cytoplasmic standard deviations of pixel intensity for the cells shown. E, ratios of average pixel intensity in the nucleus compared with the cytoplasm; F, normalized cytoplasmic standard deviations of pixel intensity for CFP-C-β5YFP-N-2, CFP-C-β5Cer-N-RGS7, and CFP-C-β5CFP-N-RGS7-t. Values represent the mean ± S.E. n = 113 for CFP-C-β5YFP-N-2, 64 for CFP-C-β5Cer-N-RGS7, and 49 for CFP-C-β5CFP-N-RGS7-t. Scale bar, 10 µm.
Fig. 7. Competition between 2 and RGS7 for limiting amounts of β5. A and B, the intensity of CFP-C-β5YFP-N-2 was measured in the presence of Cer-N-2, Cer-N-RGS7, or empty vector. 1.6 x 106 HEK-293 cells were transfected with 0.6 µg each of plasmids expressing CFP-C-β5 and YFP-N-2, and the indicated amount of Cer-N-2 or Cer-N-RGS7. The total amount of plasmid in each transfection was maintained at 3.63 µg using pcDNAI/Amp. A, CFP-C-β5YFP-N-2 intensity expressed as a function of µg of cotransfected Cer-N-2 or Cer-N-RGS7 plasmid. Values represent the means ± S.E. from seven experiments performed in duplicate. B, CFP-C-β5YFP-N-2 intensity expressed as a function of the relative amounts of coexpressed Cer-N-2 or Cer-N-RGS7 plasmid. Expression levels were determined in 1.6 x 106 HEK-293 cells transfected with 0.6 µg each of plasmids expressing CFP-C-β5 and pcDNAI/Amp, and 0.03, 0.09, 0.27, 0.81, or 2.43 µg of Cer-N-2 or Cer-N-RGS7 plasmid. The total amount of plasmid in each transfection was maintained at 3.63 µg using pcDNAI/Amp. The expression level of Cer-N-2 was 3.08-fold greater than that of Cer-N-RGS7 under these conditions (S.E. = 0.54, n = 4). Consequently, the plasmid amounts used for Cer-N-2 expression in A were multiplied by this factor to normalize the expression levels of Cer-N-2 and Cer-N-RGS7. C–E, R7BP targets CFP-C-β5Cer-N-RGS7 to the plasma membrane. HEK-293 cells (2 x 105) were transfected with 0.075 µg of CFP-C-β5 plasmid, 0.3038 µg of Cer-N-RGS7 plasmid, 0.0025 µg of mCherry-Mem plasmid, and, where indicated, 0.01875 µg of R7BP plasmid. Images of HEK-293 cells expressing CFP-C-β5Cer-N-RGS7 in the absence (C) or presence (D) of R7BP. Scale bar, 10 µm. E, plasma membrane fraction of Cer-C-β5Cer-N-RGS7 in the absence or presence of R7BP. Values represent the means ± S.E. n = 34 for CFP-C-β5Cer-N-RGS7 in the absence of R7BP and 38 for CFP-C-β5Cer-N-RGS7 in the presence of R7BP. F and G, in the presence of R7BP, CFP-C-β5 exhibits similar preferences for Cer-N-2 and Cer-N-RGS7. F, CFP-C-β5YFP-N-2 intensity is expressed as a function of the relative amounts of coexpressed Cer-N-2 or Cer-N-RGS7. HEK-293 cells were transfected as in A except that 0.15 µg of R7BP plasmid was also transfected. G, CFP-C-β5YFP-N-2 intensity is expressed as a function of the relative amounts of coexpressed Cer-N-2 or Cer-N-RGS7 plasmid. Values represent the means ± S.E of four experiments performed in duplicate. Expression levels were determined in HEK-293 cells transfected as in B except that 0.15 µg of R7BP plasmid was also transfected. The expression level of Cer-N-2 was 4.20-fold greater than that of Cer-N-RGS7 under these conditions (S.E. = 0.59, n = 4). Consequently, the plasmid amounts used for Cer-N-2 expression in F were multiplied by this factor to normalize the expression levels of Cer-N-2 and Cer-N-RGS7. CC indicates CFP-C and YN indicates YFP-N.
Fig. 9. Plasma membrane targeting of β52 and β5RGS7 by wild-type and constitutively activated o. HEK-293 cells were transfected with 0.075 µg of Cer-C-β5 plasmid and 0.0025 µg of mCherry-Mem plasmid and either 0.075 µg of Cer-N-2 plasmid (A and B) or 0.15 µg of Cer-N-RGS7 plasmid (C and D) in the absence or presence of 0.2 µg of plasmid encoding o or oR179C, as indicated. The plasma membrane fractions of Cer-C-β5Cer-N-2 and Cer-C-β5Cer-N-RGS7 for the cells shown are indicated on the images. A, images of cells expressing Cer-C-β5Cer-N-2 with plasma membrane fractions similar to the mean values in E. B, images of cells expressing Cer-C-β5Cer-N-2 with plasma membrane fractions in the 85th to 95th percentile of all values. C, images of cells expressing Cer-C-β5Cer-N-RGS7 with plasma membrane fractions similar to the mean values in F. D, images of cells expressing Cer-C-β5Cer-N-RGS7 with plasma membrane fractions in the 85th to 95th percentile of all values. Scale bar, 10 µm. E and F, plasma membrane fractions of Cer-C-β5Cer-N-2 (E) and Cer-C-β5Cer-N-RGS7 (F) in the absence (light gray bars) or presence (dark gray bars) of o or oR179C. Values represent the means ± S.E. from 64–128 cells. G, comparison of the expression levels of o and oR179C. Values represent the means ± S.E from three experiments.
Fig. 10. o-YFP and q-YFP exhibit equivalent abilities to target Cer-C-β5Cer-N-2 to the plasma membrane. HEK-293 cells were transfected with 0.075 µg each of Cer-C-β5 and Cer-N-2 plasmids, and 0.0025 µg of mCherry-Mem plasmid, and either 0, 0.0125, 0.025, 0.05, or 0.075 µg of o-YFP plasmid or 0.025, 0.05, or 0.1 µg of q-YFP plasmid. A and B, image of cell coexpressing Cer-C-β5Cer-N-2 (A) and mCherry-Mem (B). C and D, image of cell coexpressing Cer-C-β5Cer-N-2 (C) and o-YFP (D). E and F, image of cell coexpressing Cer-C-β5Cer-N-2 (E) and q-YFP (F). The plasma membrane fractions of the fluorescent proteins are indicated on the images. G, plot of plasma membrane fraction of Cer-C-β5Cer-N-2 as a function of the ratio of the average plasma membrane intensity of o-YFP (open circles, solid line) or q-YFP (filled circles, dashed line) to the average cellular intensity of Cer-C-β5Cer-N-2. Measurements from individual cells were sorted into bins of 10 cells along the x-axis and averaged. Values represent the means ± S.E. 160 cells coexpressing Cer-C-β5Cer-N-2 and o-YFP and 170 cells coexpressing Cer-C-β5Cer-N-2 and q-YFP were analyzed.
Cells were imaged 2 days after transfection using a white-light, spinning-disc confocal microscope composed of an Olympus IX81 inverted microscope, UIS2 60x 1.42 numerical aperture objective, IX2-DSU spinning disc system, 100-W mercury arc lamp, Hamamatsu C9100-02 electron multiplier camera, Ludl filter wheels, shutters, and xy stage, under the control of IPLab software (BD Biosciences, San Jose, CA). Excitation and emission filters for CFP (438/24, 483/32), YFP (504/12, 542/27), Red (589/15, 632/22), and a triple dichroic (FF444/521/608) were obtained from Semrock (Rochester, NY). One hour before imaging, the culture medium was replaced with 20 mM HEPES-buffered minimal essential medium with Earle's salts without bicarbonate. Cells were imaged at 25°C. For each condition, cells from at least three independent transfections were imaged.
The criteria for selecting cells for imaging were visible expression of all transfected fluorescent constructs, a clear section of plasma membrane border with adjacent region of cytoplasm, and a defined nucleus. The background intensity was determined by averaging the intensity of a region of pixels outside the cell and was subtracted from each image. All image processing was performed using IPlab software.
Normalized Cytoplasmic Standard Deviation. The normalized cytoplasmic standard deviation is a measure of the variation in pixel intensities within the cytoplasmic area of the cell. Using a Cintiq pen-based display screen (Wacom, Vancouver, WA), a membrane border 6 pixels wide and centered on the plasma membrane was drawn around the cell using the image of the plasma membrane marker. A separate nuclear border was drawn just inside the nucleus excluding any intensity in the nuclear membrane. The standard deviation of the intensities of pixels inside the membrane border and outside the nuclear border was calculated and normalized by dividing by the average intensity of the cytoplasmic pixels to correct for differences in the intensities of fluorescent complexes.
Nuclear-to-Cytoplasmic Intensity Ratio. The nuclear-to-cytoplasmic intensity ratio is a measure of the distribution of the labeled protein between the nuclear and cytoplasmic compartments and was determined using the membrane and nuclear borders defined above. The nuclear intensity was calculated as the average intensity of pixels in the nucleus including the border. The cytoplasmic intensity was calculated as the average intensity of pixels inside the membrane border and outside the nuclear border. The ratio is the nuclear intensity divided by the cytoplasmic intensity.
Plasma Membrane Fraction. The plasma membrane fraction is a measurement of the distribution of a labeled protein between the plasma membrane and cytoplasm and the method of its determination has been described in detail previously (Mervine et al., 2006). In brief, the plasma membrane-to-cytoplasm intensity ratio of the protein of interest is compared with that of plasma membrane and cytoplasm markers. A value of 0 corresponds to a completely cytoplasmic distribution, and a value of 1 corresponds to a completely plasma membrane distribution.
Colocalization of β52 with ER and Golgi Markers. To visualize colocalization of β52 with the ER or Golgi apparatus, Cer-C-β5Cer-N-2 was coexpressed in HEK-293 cells with YFP-ER, YFP-trans-medial Golgi, or GFP-cis-Golgi markers as described in the legend to Fig. 2. 3D Z-stacks (16 slices, 0.6 µm/slice) of live cells were collected on a Leica TCS SP2 confocal microscope using 458 nm and 514 nm laser lines for excitation of CFP and YFP. 3D Z-sections were analyzed because the regions of a cell with clear ER or Golgi structures often occur at different levels of focus. The Z-sections displayed in Fig. 2 were selected to highlight the structure of the coexpressed marker. Two color laser TIRF images were collected on a Nikon TE200-E microscope equipped with Perfect Focus, TIRF-2 illuminator, and 440 nm and 514 nm laser lines for excitation of CFP and YFP. To insure image registration, a triple-pass dichroic (Z442/514/594; Chroma, Brattleboro, VT) was used with emission filters in a motorized filter wheel (Ludl, Hawthorne, NY). The TIRF micrometer was motorized to control the TIRF angle for each laser line. Data collection was automated using IPLab software.
Fig. 2. β52 exhibits colocalization with the ER and the Golgi complex. HEK-293 cells were transfected with the following quantities of plasmids: Cer-C-β5, 0.075 µg; Cer-N-2, 0.075 µg; either YFP-ER marker, 0.0025 µg (A and B) or YFP-trans-medial Golgi marker, 0.0025 µg (C); mCherry, 0.0125 µg. Cells were imaged using a Leica TCS SP2 confocal microscope (A and C) or a two-color laser TIRF microscope (B). Cer-C-β5Cer-N-2 complexes (first column of images) were simultaneously imaged with YFP-ER or YFP-Golgi markers in the same cell (second column of images). The third column is a merge of each set of images and demonstrates colocalization (yellow) of Cer-C-β5Cer-N-2 (red) with the YFP markers (green). Images in the fourth column were produced by subtracting a percentage of the YFP-ER or YFP-Golgi marker image from the corresponding Cer-C-β5Cer-N- image as described under Materials and Methods. These images demonstrate that a substantial amount of the intracellular Cer-C-β5Cer-N-2 signal can be accounted for by either the ER or the Golgi apparatus. Scale bar, 10 µm. Images are representative of 10 cells (A), 16 cells (B), and 8 cells (C).
To visualize colocalization with each marker, a percentage of the 3D image of the ER or Golgi marker was subtracted from the β52 image to generate a 3D subtracted image illustrating the remaining protein distribution that was not associated with the marker. The percentage subtracted was the amount that minimized the standard deviation of pixel intensities in the subtracted image, in the cytoplasm excluding the Golgi region for the ER marker, and in the cytoplasm region that included the Golgi region for the Golgi marker. The standard deviation minimum indicated that pixel intensity variations resulting from visible ER or Golgi structures had been optimally subtracted.
Measurement of Fluorescence in Cell Populations. HEK-293 cells (1.6 x 106 per 60-mm dish) were transfected with plasmids as described in the legends to Figs. 3, 4, 6, 7, and 8 using 6 µl of Lipofectamine 2000 Reagent (Invitrogen) according to the manufacturer's instructions. Two days later, cells were washed once in 4 ml of HBSS + CaCl2 media (20 mM HEPES, pH 7.2, 118 mM NaCl, 4.6 mM KCl, 10 mM D-glucose, and 1 mM CaCl2). Two milliliters of HBSS + EDTA media (20 mM HEPES, pH 7.2, 118 mM NaCl, 4.6 mM KCl, 10 mM D-glucose, and 0.5 mM EDTA) were then added to the dish, and the cells were scraped off with a rubber policeman and resuspended in a 1-cm square glass cuvette with a magnetic stir bar.
Fig. 3. β5 interacts preferentially with 2 rather than 1, 5, 7, 10, 11, or 12. A, fluorescence intensities of CFP-C-β5Cer-N- complexes when CFP-C-β5 is not limiting. HEK-293 cells were transfected with 2.4 µg of plasmid expressing CFP-C-β5 and the indicated µg of Cer-N- plasmids. The total amount of plasmid in each transfection was maintained at 2.7 µg by making up the difference with pcDNAI/Amp. Values represent the means ± S.E. from three experiments performed in duplicate. CFP-C-β5Cer-N- intensities per microgram of Cer-N- plasmid, determined from linear fits to the data were as follows (means ± S.E, x 10-4): β51, 3.93 ± 0.27; β52, 408 ± 44.8; β55, 21.8 ± 2.36; β57, 106 ± 4.34; β510, 6.72 ± 0.94; β511, 11.3 ± 1.40; and β512, 22.0 ± 0.93. B, expression levels of Cer-N- subunits when coexpressed with an excess of CFP-C-β5. HEK-293 cells were transfected as in A. Values represent the means ± S.E from three experiments. Cer-N- expression per microgram of plasmid, determined from linear fits to the data were as follows (means ± S.E, x 10-6): 1, 1.82 ± 0.29; 2, 16.87 ± 1.50; 5, 4.77 ± 1.00; 7, 10.51 ± 2.13; 10, 2.78 ± 0.05; 11, 8.48 ± 0.28; and 12, 5.54 ± 0.59. C, normalization of CFP-C-β5Cer-N- intensities to Cer-N- expression levels. The intensity per microgram of Cer-N- plasmid of each CFP-C-β5Cer-N- complex was divided by the corresponding Cer-N- expression per microgram of plasmid. D, competition between Cer-N- subunits and YFP-N-2 for limiting amounts of CFP-C-β5. The intensity of CFP-C-β5YFP-N-2 was measured in the presence of each Cer-N- subunit or empty vector. HEK-293 cells were transfected with 0.6 µg each of plasmids expressing CFP-C-β5 and YFP-N-2, and the indicated micrograms of each Cer-N- plasmid. The total amount of plasmid in each transfection was maintained at 3.63 µg using pcDNAI/Amp. Values represent the means ± S.E. from three experiments performed in duplicate. E, CFP-C-β5YFP-N-2 intensity is expressed as a function of the relative amounts of coexpressed Cer-N-. Expression levels were determined in HEK-293 cells transfected with 0.6 µg each of plasmids expressing CFP-C-β5 and pcDNAI/Amp, and 0.03, 0.09, 0.27, or 2.43 µg of each Cer-N- plasmid. The total amount of plasmid in each transfection was maintained at 3.63 µg using pcDNAI/Amp. The expression levels of the Cer-N- subunits varied linearly and the data were fit by linear regressions. The plasmid amounts used in D were multiplied by Cer-N- expression per microgram of plasmid to yield the normalized amount of each Cer-N- subunit. CC indicates CFP-C and YN indicates YFP-N.
Fig. 4. Comparisons of the abilities of different β5 and combinations to activate phospholipase C-β2 and to form complexes. A, activation of phospholipase C-β2 in cells expressing CFP-C-β5 and Cer-N- subunits. HEK-293 cells were transfected with 3 µg of phospholipase C-β2 plasmid and, where indicated, 2.4 µg of CFP-C-β5 plasmid and 0.3 µg of Cer-N- plasmids. The total amount of plasmid in each transfection was maintained at 5.7 µg by making up the difference with pcDNAI/Amp. Values represent the means ± S.E. from six experiments performed in triplicate. B, fluorescence intensities of CFP-C-β5Cer-N- complexes expressed under the same transfection conditions as in A. Values represent the means ± S.E. from 3 experiments performed in duplicate. C, normalization of phospholipase C-β2 activation by coexpressed CFP-C-β5 and Cer-N- subunits to the fluorescence intensities of the CFP-C-β5Cer-N- complexes. Inositol phosphate levels obtained in cells transfected with pcDNAI/Amp (background) were subtracted from the levels in cells expressing CFP-C-β5 and either Cer-N-2, Cer-N-5, Cer-N-7, or Cer-N-12. These background-subtracted activities were divided by the corresponding CFP-C-β5Cer-N- intensities.
Fig. 6. Quantification of fluorescence and expression levels when Cer-N-2 and Cer-N-RGS7 were coexpressed with an excess of either CFP-C-β5 or CFP-C-β1. A, intensities of CFP-C-β5Cer-N-2 and CFP-C-β5Cer-N-RGS7 when CFP-C-β5 is not limiting. HEK-293 cells were transfected with 2.4 µg of plasmid expressing CFP-C-β5 and the indicated quantity of Cer-N-2 or Cer-N-RGS7. The total amount of plasmid in each transfection was maintained at 3 µg using pcDNAI/Amp. Values represent the means ± S.E of three experiments performed in duplicate. From linear fits to the data, CFP-C-β5Cer-N-2 intensity per microgram of Cer-N-2 plasmid (x 10-5) was 34.25 ± 2.89, and CFP-C-β5Cer-N-RGS7 intensity per microgram of Cer-N-RGS7 plasmid (x 10-5) was 1.84 ± 0.14. B, expression levels of Cer-N-2 and Cer-N-RGS7 when coexpressed with an excess of CFP-C-β5. HEK-293 cells were transfected as in A. Values represent the means ± S.E from three experiments. From linear fits to the data, Cer-N-2 expression per microgram of plasmid (x 10-6) was 27.67 ± 2.55, and Cer-N-RGS7 intensity per microgram of plasmid (x 10-6) was 1.26 ± 0.24. C, Normalization of CFP-C-β5Cer-N-2 and CFP-C-β5Cer-N-RGS7 intensities to the expression levels of Cer-N-2 and Cer-N-RGS7. The intensities per microgram of Cer-N-protein plasmid of CFP-C-β5Cer-N-2 and CFP-C-β5Cer-N-RGS7 were divided by the expression per microgram of plasmid of Cer-N-2 and Cer-N-RGS7, respectively. D, CFP-C-β1 forms a fluorescent complex with Cer-N-2 but not Cer-N-RGS7. HEK-293 cells were transfected as in A, except that CFP-C-β1-expressing plasmid was substituted for CFP-C-β5-expressing plasmid. Values represent the means ± S.E of three experiments performed in duplicate. From linear fits to the data, CFP-C-β1Cer-N-2 intensity per microgram of Cer-N-2 plasmid (x 10-5) was 18.02 ± 1.52, and CFP-C-β1Cer-N-RGS7 intensity per microgram of Cer-N-RGS7 plasmid (x 10-5) was 0.0714 ± 0.0233. E, expression levels of Cer-N-2 and Cer-N-RGS7 when coexpressed with an excess of CFP-C-β1. HEK-293 cells were transfected as in D. Values represent the means ± S.E from four experiments. From linear fits to the data, Cer-N-2 expression per microgram of plasmid (x 10-6) was 22.22 ± 4.20, and Cer-N-RGS7 intensity per microgram of plasmid (x 10-6) was 0.52 ± 0.085. F, normalization of CFP-C-β1Cer-N-2 and CFP-C-β1Cer-N-RGS7 intensities to the expression levels of Cer-N-2 and Cer-N-RGS7. The intensities per microgram of Cer-N-protein plasmid of CFP-C-β1Cer-N-2 and CFP-C-β1Cer-N-RGS7 were divided by the expression per microgram of plasmid of Cer-N-2 and Cer-N-RGS7, respectively.
Fig. 8. β1 competes more effectively than β5 for limiting amounts of 2. A, intensities of Cer-N-β1CFP-C-2 and Cer-N-β5CFP-C-2 when CFP-C-2 is not limiting. HEK-293 cells were transfected with 2.4 µg of plasmid expressing CFP-C-2 and the indicated µg of plasmid expressing Cer-N-β1 or Cer-N-β5. The total amount of plasmid in each transfection was maintained at 2.43 µg using pcDNAI/Amp. Values represent the means ± S.E of four experiments performed in duplicate. Cer-N-β-CFP-C-2 intensities per microgram of Cer-N-β plasmid, determined from linear fits to the data, were as follows (means ± S.E, x 10-5): β12, 8.22 ± 1.53; β52, 8.44 ± 1.59. B, expression levels of Cer-N-β1 and Cer-N-β5 when coexpressed with excess CFP-C-2. HEK-293 cells were transfected as in A. Values represent the means ± S.E from three experiments. Cer-N-β expression per microgram of plasmid, determined from linear fits to the data, were as follows (means ± S.E, x 10-8): β1, 2.41 ± 0.47; β5, 2.48 ± 0.37. C, normalization of Cer-N-βCFP-C-2 intensities to Cer-N-β expression levels. The intensity per microgram of Cer-N-β plasmid of each Cer-N-βCFP-C-2 complex was divided by the corresponding Cer-N-β expression per microgram of plasmid. D, competition between Cer-N-β1 or Cer-N-β5 and YFP-N-β1 for limiting amounts of CFP-C-2. The intensity of YFP-N-β1CFP-C-2 was measured in the presence of Cer-N-β1, Cer-N-β5, or empty vector. HEK-293 cells were transfected with 0.3 µg of CFP-C-2-expressing plasmid, 0.6 µg of YFP-N-β1-expressing plasmid, and the indicated µg of either Cer-N-β1 or Cer-N-β5 plasmid. The total amount of plasmid in each transfection was maintained at 9 µg using pcDNAI/Amp. Values represent the means ± S.E. from five experiments performed in duplicate. Expression levels were determined in HEK-293 cells transfected with 0.3 µg of CFP-C-2-expressing plasmid, 0.6 µg of pcDNAI/Amp, and the same range of micrograms of Cer-N-β1 or Cer-N-β5 plasmid, maintaining the total amount of plasmid in each transfection at 9 µg using pcDNAI/Amp. CC, CFP-C; YN, YFP-N.
Data were collected on a PC1 photon-counting spectrofluorometer (ISS, Champaign, IL) configured with motorized filter wheels on both the excitation path between the excitation monochrometer and the sample, and on the emission path between the sample and the emission monochrometer as described previously (Mervine et al., 2006).
In multicolor BiFC experiments, the IC50 for inhibition of association of YFP-N-2 with CFP-C-β5 by Cer-N- subunits or Cer-N-RGS7 was defined as micrograms of Cer-N- subunit or Cer-N-RGS7 plasmid that produced a 50% decrease in the intensity of CFP-C-β5YFP-N-2. To determine IC50 values, the data were fit, using Kaleidograph (Abelbeck/Synergy Software, Reading, PA), to Y = (100)/(1 + (X/a)b), where X is micrograms of of transfected Cer-N- or Cer-N-RGS7 plasmid, Y is the percentage of maximal fluorescence produced by CFP-C-β5YFP-N-2, a is the half-maximal inhibitory concentration (IC50) of the Cer-N- subunit or Cer-N-RGS7, and b is the slope factor. The IC50 for inhibition of association of YFP-N-β1 with CFP-C-2 by Cer-N-β subunits was determined in the same manner.
Immunoblots. The expression levels of Cer-N-proteins were determined in HEK-293 cells (1.6 x 106 per 60-mm dish) that were transfected as described in the legends to Figs. 3, 6, 7, and 8 using 6 µl of Lipofectamine 2000 Reagent. Two days after transfection, total cell lysates (7.5 or 15 µg) were resolved on NuPAGE Bis-Tris 4 to 12% gels (Invitrogen) and transferred to nitrocellulose. The expression levels of the Cer-N- subunits were determined for Fig. 3 by probing with a polyclonal antibody to residues 3 to 17 of GFP (Anti-GFP, N-terminal; Sigma-Aldrich, St. Louis, MO), and the expression levels of Cer-N-2 and Cer-N-RGS7 were determined for Figs. 6 and 7 by probing with a polyclonal antibody to full-length GFP (Rockland Immunochemicals, Gilbertsville, PA). The antigen-antibody complexes were detected according to the ECL Western blotting protocol (GE Healthcare, Chalfont St. Giles, Buckinghamshire, UK). Chemiluminescence was imaged using a Lumi-Imager (Roche Applied Science, Indianapolis, IN). The expression levels of Cer-N-β1 and Cer-N-β5 were determined for Fig. 8 by probing with a polyclonal antibody to full-length GFP (Rockland Immunochemicals). The antigen-antibody complexes were detected using SuperSignal West Pico Chemiluminescent Substrate (Pierce Biotechnology, Rockford, IL). Chemiluminescence was imaged using a FluorChem SP Imaging System (Alpha Innotech, San Leandro, California). Bands in the images were quantified using IPLab software.
The expression levels of EE-tagged subunits were determined for Fig. 9 using membranes prepared as described previously (Medina et al., 1996) 2 days after transfection of HEK-293 cells (4.45 x 106/100-mm dish) with 4.45 µg of each subunit plasmid using 5.56 µl of LipofectAMINE 2000 Reagent. 50 µg of membrane proteins were resolved by SDS-polyacrylamide electrophoresis (10%), transferred to nitrocellulose, and probed with a monoclonal antibody to the EE epitope. The antigen-antibody complexes were detected according to the ECL Western blotting protocol and chemiluminescence was imaged using a Lumi-Imager. Bands in the images were quantified using IPLab software.
Assay for Inositol Phosphate Accumulation in Transiently Transfected Cells. HEK-293 cells (1.6 x 106 per 60-mm dish) were transfected with plasmids as described in the legend to Fig. 4 using 6 µl of Lipofectamine 2000 Reagent according to the manufacturer's instructions. Twenty-four hours after transfection, the cells were replated in 24-well plates and labeled with [3H]inositol (GE Healthcare). After an additional 24 h, inositol phosphate levels were determined in the presence of 5 mM LiCl as described previously (Medina et al., 1996).
Complexes of β5 with Different Subunits Exhibited Distinct Localization Patterns. We investigated the ability of β5 to form complexes with 1, 2, 5, 7, 10, 11, and 12 in HEK-293 cells. β5 (Wang et al., 1999a) and each of these subunits, with the exception of 1 (Wang et al., 1997), have been detected at the protein level in these cells. β5 complexes were imaged using BiFC, which involves the production of fluorescence by two nonfluorescent fragments of CFP or YFP when they are brought together by interactions between proteins fused to each fragment. In contrast to fluorescence resonance energy transfer (FRET), in which the intensity of the signal depends on the distance between and relative orientation of two fluorophores, BiFC is based on the formation of a fluorescent complex from nonfluorescent constituents and does not require that the interacting proteins position the fluorescent protein fragments in a specific orientation or within a fixed distance from each other (Kerppola, 2006). However, steric constraints can prevent the association of the fluorescent protein fragments within a complex, in which case inserting peptide linkers between the fragments and the interacting proteins may enable association of the fluorescent fragments (Kerppola, 2006). We applied the BiFC approach previously to image β1 complexes, which were demonstrated to be functional by their abilities to potentiate activation of adenylyl cyclase by s in COS-7 cells (Hynes et al., 2004b) and to internalize in response to stimulation of the β2-adrenergic receptor in HEK-293 cells (Hynes et al., 2004a). For live cell imaging, we fused a carboxyl-terminal fragment (residues 159–238) of cerulean, an engineered form of ECFP that is 2.5-fold brighter than ECFP (Rizzo et al., 2004), to β5 to produce Cer-C-β5, and an aminoterminal cerulean fragment (residues 1–158) to the subunits, producing Cer-N- subunits.
Each of the Cer-C-β5Cer-N- complexes produced a fluorescent signal (Fig. 1, A–G). However, Cer-C-β5 and Cer-N-2, which produced one of the brightest signals when coexpressed, were not fluorescent when expressed individually (data not shown). The localization patterns of the β5 complexes varied depending on the associated subunit (Fig. 1), in agreement with a previous study of a subset of these β5 complexes (Hynes et al., 2004b), and in contrast to the corresponding β1 complexes, which localized predominantly to the plasma membrane (Mervine et al., 2006). The β5 complexes exhibited very little plasma membrane signal and varied in their distribution between the cytoplasm and the nucleus (Fig. 1, A–G, and J). β51, β55, β510, and β511 exhibited relatively high ratios of nuclear to cytoplasmic signal (0.63–0.88) (Fig. 1A, C, E, F, and J), whereas β52, β57, and β512 exhibited lower ratios of nuclear-to-cytoplasmic signal (0.36–0.45) (Fig. 1B, D, G, and J). For comparison, s, which, when over-expressed without exogenous β, labels the cytoplasm diffusely (Mervine et al., 2006) (Fig. 1H), yielded a ratio of nuclear-to-cytoplasmic signal (0.34) similar to that of β52 and β57 (Fig. 1J), suggesting that this amount of nuclear signal represents background labeling by proteins that are excluded from the nucleus. In contrast, mCherry, which diffuses freely between the nucleus and cytoplasm (Fig. 1I), exhibited a nuclear-to-cytoplasmic signal ratio of 0.97 (Fig. 1J).
The β5 complexes also varied in the degree to which their cytoplasmic signals were diffuse or associated with discrete intracellular structures (Fig. 1, A–G). This aspect of the cytoplasmic β5 signals was quantified by determining normalized cytoplasmic standard deviations of pixel intensity as described under Materials and Methods. This measurement indicates the extent to which labeled proteins in the cytoplasm are distributed evenly as free soluble proteins (low standard deviation), as opposed to being localized on discrete vesicles, membranes, or other structures that would increase the range of pixel intensities significantly (high standard deviation). Lower standard deviations were associated with the diffuse labeling patterns of s-YFP and mCherry (Fig. 1, H, I, and K). This analysis showed that β51 exhibited by far the lowest standard deviation and was comparable with s-YFP and mCherry (Fig. 1K). In contrast, β52 and β57 had the greatest standard deviation. The other β5 complexes had values that were closer to those of β52 and β57 than to that of β51. The diffuse nature of the β51 signal may be due, in part, to the fact that 1 is farnesylated, rather than geranylgeranylated (Wedegaertner et al., 1995). However, because 11 is also farnesylated and β511 exhibited much more discrete staining than did β51, additional differences between 1 and the other subunits seem to be important in determining the nature of the signal. In summary, these results show that the partitioning of β5 complexes between the cytoplasm and the nucleus and the nature of their distribution in the cytoplasm (diffuse or discrete) are determined by the subunit component.
The discrete cytoplasmic labeling observed with some of the β5 complexes, notably β52, seemed to reside on a number of intracellular structures, primarily the ER, Golgi apparatus, and nuclear membrane. To define the localization of these complexes more precisely, 3D stacks of images of β52 coexpressed with markers for the ER or the Golgi apparatus were collected on a laser-scanning confocal microscope (Fig. 2, A and C). Colocalization of β52 with both the ER and trans-medial Golgi apparatus was observed in the merge images (Fig. 2, A and C). Colocalization with a cis-Golgi marker was similar to that seen with the trans-medial Golgi marker (data not shown). To visualize β52 distribution not associated with the coexpressed markers, the marker images were subtracted from the β52 images as described under Materials and Methods. After subtraction of the ER images from the β52 images, significant intensity remained in the perinuclear region, as well as some diffuse cytoplasmic and nuclear membrane intensity (Fig. 2A). The perinuclear intensity of β52 was due primarily to the Golgi apparatus, with clear colocalization in the merge image and very little signal visible above the surrounding intensity of the ER and diffuse cytoplasm staining in the subtracted image (Fig. 2C). Colocalization of β52 with the ER marker was also observed using a two-color laser TIRF microscope, where the densely packed folds of the ER membranes seen in the confocal cross-sections were visualized distinctly (Fig. 2B).
β5 Interacted Preferentially with 2 Compared with 1, 5, 7, 10, 11, and 12. The intensities of CFP-C-β5Cer-N- complexes were quantified in HEK-293 cell populations using a spectrofluorometer. In the presence of an excess of CFP-C-β5, the intensities of the CFP-C-β5Cer-N- complexes varied over a 100-fold range, with CFP-C-β5Cer-N-2 and CFP-C-β5Cer-N-1 being the most and least intense, respectively (Fig. 3A). This range was much greater than the 3-fold range seen previously when the intensities of the corresponding CFP-C-β1Cer-N- complexes were compared under the same conditions (Mervine et al., 2006). The expression levels of the Cer-N- subunits, when coexpressed with excess CFP-C-β5, were compared by immunoblotting total cell lysates with an antibody to the amino terminus of GFP (Fig. 3B). There was a larger range in expression levels than when the same Cer-N- subunits were coexpressed with an excess of CFP-C-β1 (Mervine et al., 2006), suggesting that β1 and β5 differ in their abilities to stabilize these subunits. However, the range in Cer-N- expression levels (Fig. 3B) was narrower than the range in intensities of the CFP-C-β5Cer-N- complexes (Fig. 3A). CFP-C-β5Cer-N-2 exhibited by far the highest ratio of CFP-C-β5Cer-N- intensity to Cer-N- expression level, followed by CFP-C-β5Cer-N-7 (Fig. 3C). The CFP-C-β5Cer-N- intensity-to-Cer-N- expression ratio of CFP-C-β5Cer-N-2 was 18.6-fold greater than that of CFP-C-β5Cer-N-11, which had the lowest ratio (Fig. 3C). In contrast, the intensity-to-Cer-N- expression ratios of the corresponding CFP-C-β1Cer-N- complexes varied by 2-fold or less (Mervine et al., 2006).
Given that cells coexpress multiple isoforms of β and subunits, the predominance of particular β complexes will be influenced both by the relative expression levels and the association preferences of the expressed β and subunits. Although CFP-C-β5Cer-N-2 was clearly the most intense complex when CFP-C-β5 was not limiting, we sought to determine whether this was also the preferred complex when different subunits were coexpressed with a limiting amount of β5 and whether differences between the less preferred subunits would be revealed under these conditions. Multicolor BiFC makes it possible to simultaneously image multiple complexes and quantify the abilities of different proteins to compete for a limiting amount of a common binding partner, because the amino terminal fragment of the fluorescent protein determines the spectral properties of the complex (Grinberg et al., 2004). We found previously that the intensities of CFP-C-β1Cer-N- complexes were similar when CFP-C-β1 was not limiting, but when the intensities of coexpressed CFP-C-β1Cer-N- (cyan) and CFP-C-β1YFP-N-2 (yellow) complexes were compared under conditions in which CFP-C-β1 was limiting, the Cer-N- subunits exhibited an approximately 4.5-fold range in their abilities to compete with YFP-N-2 for association with CFP-C-β1 (Mervine et al., 2006).
The abilities of the Cer-N- subunits to compete with YFP-N-2 for association with limiting amounts of CFP-C-β5 were compared by determining the amounts of each Cer-N- subunit that decreased the intensity of CFP-C-β5YFP-N-2 by 50%. Cer-N-2 competed 18-fold more effectively than Cer-N-10 (the least effective competitor) and 2.7-fold more effectively than the next best competitor, Cer-N-7 (Fig. 3D, Table 1). When the amounts of transfected Cer-N- plasmids were normalized to their relative expression levels in the presence of limiting amounts of CFP-C-β5, Cer-N-2 was still the most effective Cer-N- subunit, competing 23-fold more effectively than Cer-N-11, the least effective subunit (Fig. 3E, Table 1). When expression levels were corrected for, Cer-N-1 became almost as effective in competition as Cer-N-7. Cer-N-2 was 4-fold more effective than Cer-N-7 and 6.5-fold more effective than Cer-N-1(Fig. 3E, Table 1). This range in the abilities of the Cer-N- subunits to compete for limiting amounts of CFP-C-β5 was much greater than that of their abilities to compete for CFP-C-β1 (Mervine et al., 2006) and the relative efficacies of the Cer-N- subunits were different, as described under Discussion.
TABLE 1 Competition of CerN- subunits with YN-2 for dimerization with CC-β5 in live HEK-293 cells Values represent the mean ± S.E. from three experiments.
Efficacy of Phospholipase C-β2 Activation by β5 Combinations Was Correlated with the Amount of Complex Formation. Previous comparisons of the abilities of β5 complexes to modulate effectors demonstrated that cells expressing β52 exhibited greater phospholipase C-β2 activity than did cells expressing β51, β53, β54, β55, or β57 (Watson et al., 1994; Watson et al., 1996) and N-type Ca+2 channel inhibition was obtained in cells expressing β52 but not β51 or β53 (Zhou et al., 2000). These results indicated either that β52 was more effective than the other β5 complexes at modulating these effectors or that β52 complexes formed preferentially relative to the other β5 combinations. To distinguish between these two alternatives, we compared phospholipase C-β2 activity in cells expressing CFP-C-β5 and different Cer-N- subunits with the relative amounts of CFP-C-β5Cer-N- complex formation detected as BiFC. Four of the CFP-C-β5Cer-N- complexes (those containing Cer-N-2, Cer-N-5, Cer-N-7, or Cer-N-12) activated coexpressed phospholipase C-β2, whereas the other three complexes (those containing Cer-N-1, Cer-N-10, or Cer-N-11) produced no activation above that seen in cells transfected with empty vector (Fig. 4A). CFP-C-β5Cer-N-2 and CFP-C-β5Cer-N-7 exhibited the greatest activity. No activity was obtained when CFP-C-β5 was expressed without a Cer-N- subunit or when any of the Cer-N- subunits was expressed without CFP-C-β5 (Fig. 4A). The three CFP-C-β5Cer-N- complexes that did not stimulate phospholipase C-β2 exhibited only minimal fluorescence, indicating that lack of activity was due to ineffective complex formation (Fig. 4B). For the 4 CFP-C-β5Cer-N- complexes that activated phospholipase C-β2, the ratios of CFP-C-β5Cer-N--stimulated activity (Fig. 4A) to amount of complex formation (Fig. 4B) were similar (Fig. 4C), indicating that the different efficacies of the CFP-C-β5Cer-N- combinations were due primarily to different amounts of complex formation. This is the first time that β function in intact cells has been correlated directly with the amount of β complex formation.
β52 and β5RGS7 Complexes in the Same Cell Could Be Imaged Simultaneously using BiFC. The above studies demonstrated that β5 associates preferentially with 2 compared with the other subunits tested. To determine the association preference of β5 for subunits versus R7 family RGS proteins and to compare the localization patterns of these β5 complexes, we expressed fluorescent β52 and β5RGS7 complexes in the same cells. CFP-C-β5 was coexpressed with Cer-N-RGS7 and YFP-N-2 to produce CFP-C-β5Cer-N-RGS7 (cyan) and CFP-C-β5YFP-N-2 (yellow). Both complexes exhibited minimal localization to the nucleus (Fig. 5, A–C and E), compared with β51, β55, β510, and β511 (Fig. 1J). However, CFP-C-β5Cer-N-RGS7t, containing a truncated form of RGS7 in which the DEP domain was deleted, localized preferentially in the nucleus (Fig. 5, D and E), in agreement with a previous study of RGS6 splice variants that demonstrated that the DEP domain can function as a cytoplasmic retention signal (Chatterjee et al., 2003). In contrast to the distribution of CFP-C-β5YFP-N-2 on discrete structures in the cytoplasm (Fig. 5, A and C), the cytoplasmic signals of CFP-C-β5Cer-N-RGS7 and CFP-C-β5Cer-N-RGS7t were diffuse (Fig. 5, B and D). The different types of cytoplasmic signals were confirmed and quantified by the higher normalized standard deviation of cytoplasmic pixel intensity of CFP-C-β5YFP-N-2 compared with CFP-C-β5Cer-N-RGS7 and CFP-C-β5Cer-N-RGS7t (Fig. 5F). These results indicate that both subunits and R7 family RGS proteins can dictate the localization pattern of β5. Fluorescence was not obtained when CFP-C-β5, Cer-N-RGS7, or YFP-2 were expressed alone (data not shown), and minimal fluorescence was obtained when CFP-C-β1 and Cer-N-RGS7 were coexpressed (Fig. 6D), consistent with previous reports that β1 and R7 family RGS proteins do not interact (Snow et al., 1998; Posner et al., 1999).
β52 and β5RGS7 Complexes Formed with Equal Efficiency When Excess β5 Was Coexpressed with Either 2 or RGS7. In the presence of excess cotransfected CFP-C-β5 plasmid, linear relationships between the amounts of transfected Cer-N-2 and Cer-N-RGS7 plasmids and the intensities of CFP-C-β5Cer-N-2 and CFP-C-β5Cer-N-RGS7 complexes, respectively, were obtained (Fig. 6A). The intensity of CFP-C-β5Cer-N-2 was 19-fold greater than that of CFP-C-β5Cer-N-RGS7, based on the slopes of linear fits to the data. To determine whether this difference was due to a greater ability of CFP-C-β5 to form fluorescent complexes with Cer-N-2 compared with Cer-N-RGS7 or to differences in expression of the Cer-N fusion proteins, the expression levels of Cer-N-2 and Cer-N-RGS7, when coexpressed with excess CFP-C-β5, were determined using immunoblots. The slope of the linear fit to the Cer-N-2 data was 22-fold greater than that for Cer-N-RGS7 (Fig. 6B). The ratios of CFP-C-β5Cer-N-2 intensity to Cer-N-2 expression level and of CFP-C-β5Cer-N-RGS7 intensity to Cer-N-RGS7 expression level were used to normalize the β5-interacting abilities of 2 and RGS7 to their expression levels. As shown in Fig. 6C, these ratios were the same, indicating that 2 and RGS7 exhibit the same ability to interact with an excess of β5 when tested one at a time. In contrast, minimal fluorescence intensity was obtained with CFP-C-β1Cer-N-RGS7. When Cer-N-2 and Cer-N-RGS7 were coexpressed with an excess of CFP-C-β1, the intensity of CFP-C-β1Cer-N-2 was 254-fold greater than that of CFP-C-β1Cer-N-RGS7 (Fig. 6D). Under these conditions, the expression level of Cer-N-2 was 43-fold greater than that of Cer-N-RGS7 (Fig. 6E). The decreased expression of Cer-N-RGS7 relative to Cer-N-2 when coexpressed with CFP-C-β1 rather than CFP-C-β5 suggests that, despite the Cer-N tag, the stability of Cer-N-RGS7 is at least partially dependent on interaction with β5, consistent with losses of R7 family RGS proteins in β5 knockout mice (Chen et al., 2003). In addition, the ratio of CFP-C-β1Cer-N-2 intensity to Cer-N-2 expression level was 6-fold greater than that of CFP-C-β1Cer-N-RGS7 intensity to Cer-N-RGS7 expression level (Fig. 6F).
β5 Exhibited a Slight Preference for 2 over RGS7 That Was Eliminated in the Presence of R7BP. To determine whether preferential association of β5 with 2 or RGS7 would be revealed when a limiting amount of β5 was coexpressed with both 2 and RGS7 at the same time, we compared the abilities of Cer-N-RGS7 and Cer-N-2 to compete with YFP-N-2 for binding to CFP-C-β5. The amount of yellow fluorescence obtained from CFP-C-β5YFP-N-2 was measured when a range of amounts of either Cer-N-RGS7 or Cer-N-2 plasmid was coexpressed. The amount of Cer-N-RGS7 plasmid required to reduce the CFP-C-β5YFP-N-2 intensity by 50% was 8-fold higher than that of Cer-N-2 (Fig. 7A). When the amounts of Cer-N-RGS7 and Cer-N-2 plasmids used were normalized to their relative expression levels in the presence of limiting amounts of CFP-C-β5, three times as much Cer-N-RGS7 compared with Cer-N-2 was required to reduce the intensity of CFP-C-β5YFP-N-2 by 50% (Fig. 7B). Thus, β52 and β5RGS7 complexes can form simultaneously in intact cells, and β5 exhibits a slight preference for 2 over RGS7 when coexpressed with both potential binding partners.
β5RGS7 complexes can be targeted to the plasma membrane by R7BP, which, like β5 and RGS7, is highly expressed in the nervous system (Drenan et al., 2006). To determine whether R7BP can influence the formation of β52 and β5RGS7 complexes, we investigated whether coexpression of R7BP, at levels which targeted β5RGS7 to the plasma membrane (Fig. 7, C–E), affected competition between RGS7 and 2 for β5. We found that coexpression of R7BP decreased the preference of β5 for 2 over RGS7. In the presence of R7BP, the amount of Cer-N-RGS7 plasmid required to reduce the CFP-C-β5YFP-N-2 intensity by 50% was 5-fold higher than that of Cer-N-2 (Fig. 7F). When the amounts of Cer-N-RGS7 and Cer-N-2 plasmids used were normalized to their relative expression levels in the presence of R7BP and limiting amounts of CFP-C-β5, approximately the same amounts of Cer-N-RGS7 and Cer-N-2 reduced the intensity of CFP-C-β5YFP-N-2 by 50% (Fig. 7G).
2 Exhibited a Preference for β1 over β5 When the Three Proteins Were Coexpressed. The above studies compared the preferences of β5 for different interaction partners and demonstrated that 2 was preferred over the other six subunits tested and over RGS7 in the absence of R7BP. Because the prevalence of particular β5 complexes will reflect the interaction preferences of both β5 and its potential interaction partners, we investigated the interaction preferences of 2. The intensities of Cer-N-β1CFP-C-2 and Cer-N-β5CFP-C-2 complexes were compared as well as the abilities of Cer-N-β1 and Cer-N-β5 to compete with YFP-N-β1 for interaction with CFP-C-2.
In the presence of an excess of cotransfected CFP-C-2 plasmid, linear relationships between the amounts of transfected Cer-N-β1 and Cer-N-β5 plasmids and the intensities of Cer-N-β1CFP-C-2 and Cer-N-β5CFP-C-2 complexes, respectively, were obtained, and the intensities of the complexes were similar (Fig. 8A). The relationships between the expression levels of Cer-N-β1 and Cer-N-β5 and the amounts of transfected plasmid under these expression conditions were also linear and similar (Fig. 8B), resulting in similar ratios of Cer-N-βCFP-C-2 intensities to Cer-N-β subunit expression levels (Fig. 8C).
When the abilities of Cer-N-β1 and Cer-N-β5 to compete with YFP-N-β1 for interaction with CFP-C-2 were compared, the amount of Cer-N-β5 plasmid required to reduce the YFP-N-β1CFP-C-2 intensity by 50% was 4.2-fold higher than of that of Cer-N-β1 (Fig. 8D). The expression level of Cer-N-β1 was 0.98-fold of that of Cer-N-β5 in the presence of limiting amounts of CFP-C-2 (S.E. = 0.10, n = 3). This preferential interaction of 2 with β1 rather than β5 might be expected to work against the preference of β5 for 2 over RGS7 in cells coexpressing β1, β5, 2, and RGS7 by diverting some of the available 2 away from β5.
β52 and β5RGS7 Were Targeted Preferentially to the Plasma Membrane by Inactive and Activated o, Respectively. Although β12 and β22 localize to the plasma membrane, β52 accumulates on intracellular membranes, including the ER and the Golgi apparatus (Hynes et al., 2004b) (Figs. 1, 2, and 5). The ability of the β subunit to influence targeting of the subunit was surprising, because the β subunit is not known to have a membrane-targeting signal, such as the prenyl group on the subunit (Wedegaertner et al., 1995). Because o was reported recently to target β5RGS7 to the plasma membrane (Takida et al., 2005), we hypothesized that plasma membrane targeting of β52 also might require o, which is not expressed in HEK-293 cells (Wang et al., 1999b). Indeed, we found that coexpressed o did target both Cer-C-β5Cer-N-2 and Cer-C-β5Cer-N-RGS7 to the plasma membrane (Fig. 9, A–F). Because subunits in the inactive rather than the activated state have a higher affinity for β complexes, whereas RGS proteins interact preferentially with the activated form of subunits, we investigated how targeting of β52 and β5RGS7 were affected by an o mutant, oR179C, that is constitutively activated as a result of decreased GTPase activity. In agreement with the expected preferences of β52 and β5RGS7, oR179C was less effective than o at targeting β52 and more effective than o at targeting β5RGS7 (Fig. 9, A–F). The expression levels of o and oR179C, determined by immunoblotting membrane preparations using an antibody to the EE epitope included in both constructs, were similar (Fig. 9G). These results suggest that the inactive form of o interacts preferentially with β52, whereas activated o interacts preferentially with β5RGS7.
o and q Exhibited Similar Abilities to Target β52 to the Plasma Membrane. Our observation that o can target β52 to the plasma membranes of live cells is consistent with a previous in vitro study demonstrating that o can bind to β52 and prevent activation of phospholipase C-β2 (Yoshikawa et al., 2000) but contrasts with two other studies in reconstituted systems suggesting that β52 interacts with q but not other subunits (Fletcher et al., 1998; Lindorfer et al., 1998). To investigate the subunit specificity of β52 in live cells, we compared the abilities of o and q to target β52 to the plasma membrane using fluorescent versions of these subunits in which YFP was inserted at the homologous location in each. This made it possible to compare Cer-C-β5Cer-N-2 targ
【参考文献】
Ballon DR, Flanary PL, Gladue DP, Konopka JB, Dohlman HG, and Thorner J (2006) DEP-domain-mediated regulation of GPCR signaling responses. Cell 126: 1079-1093.
Benzing T, Kottgen M, Johnson M, Schermer B, Zentgraf H, Walz G, and Kim E (2002) Interaction of 14-3-3 protein with regulator of G protein signaling 7 is dynamically regulated by tumor necrosis factor-. J Biol Chem 277: 32954-32962.[Abstract/Free Full Text]
Burchett SA (2003) In through the out door: nuclear localization of the regulators of G protein signaling. J Neurochem 87: 551-559.
Chatterjee TK, Liu Z, and Fisher RA (2003) Human RGS6 gene structure, complex alternative splicing, and role of N terminus and G protein -subunit-like (GGL) domain in subcellular localization of RGS6 splice variants. J Biol Chem 278: 30261-30271.[Abstract/Free Full Text]
Chen CK, Eversole-Cire P, Zhang H, Mancino V, Chen YJ, He W, Wensel TG, and Simon MI (2003) Instability of GGL domain-containing RGS proteins in mice lacking the G protein β-subunit Gβ5. Proc Natl Acad Sci U S A 100: 6604-6609.[Abstract/Free Full Text]
Drenan RM, Doupnik CA, Jayaraman M, Buchwalter AL, Kaltenbronn KM, Huettner JE, Linder ME, and Blumer KJ (2006) R7BP augments the function of RGS7-Gβ5 complexes by a plasma membrane-targeting mechanism. J Biol Chem 281: 28222-28231.[Abstract/Free Full Text]
Fletcher JE, Lindorfer MA, DeFilippo JM, Yasuda H, Guilmard M, and Garrison JC (1998) The G protein β5 subunit interacts selectively with the Gq subunit. J Biol Chem 273: 636-644.[Abstract/Free Full Text]
Grinberg AV, Hu CD, and Kerppola TK (2004) Visualization of Myc/Max/Mad family dimers and the competition for dimerization in living cells. Mol Cell Biol 24: 4294-4308.[Abstract/Free Full Text]
Hooks SB, Waldo GL, Corbitt J, Bodor ET, Krumins AM, and Harden TK (2003) RGS6, RGS7, RGS9, and RGS11 stimulate GTPase activity of Gi family G-proteins with differential selectivity and maximal activity. J Biol Chem 278: 10087-10093.[Abstract/Free Full Text]
Hughes TE, Zhang H, Logothetis DE, and Berlot CH (2001) Visualization of a functional Gq-green fluorescent protein fusion in living cells. Association with the plasma membrane is disrupted by mutational activation and by elimination of palmitoylation sites, but not by activation mediated by receptors or . J Biol Chem 276: 4227-4235.[Abstract/Free Full Text]
Hynes TR, Mervine SM, Yost EA, Sabo JL, and Berlot CH (2004a) Live cell imaging of Gs and the β2-adrenergic receptor demonstrates that both s and β17 internalize upon stimulation and exhibit similar trafficking patterns that differ from that of the β2-adrenergic receptor. J Biol Chem 279: 44101-44112.[Abstract/Free Full Text]
Hynes TR, Tang L, Mervine SM, Sabo JL, Yost EA, Devreotes PN, and Berlot CH (2004b) Visualization of G protein β dimers using bimolecular fluorescence complementation demonstrates roles for both β and in subcellular targeting. J Biol Chem 279: 30279-30286.[Abstract/Free Full Text]
Jones MB and Garrison JC (1999) Instability of the G-protein β5 subunit in detergent. Anal Biochem 268: 126-133.
Jones MB, Siderovski DP, and Hooks SB (2004) The G β dimer as a novel source of selectivity in G-protein signaling: GGL-ing at convention. Mol Interv 4: 200-214.[Abstract/Free Full Text]
Kerppola TK (2006) Visualization of molecular interactions by fluorescence complementation. Nat Rev Mol Cell Biol 7: 449-456.
Kovoor A, Chen CK, He W, Wensel TG, Simon MI, and Lester HA (2000) Coexpression of Gβ5 enhances the function of two G subunit-like domain-containing regulators of G protein signaling proteins. J Biol Chem 275: 3397-3402.[Abstract/Free Full Text]
Kovoor A, Seyffarth P, Ebert J, Barghshoon S, Chen CK, Schwarz S, Axelrod JD, Cheyette BN, Simon MI, Lester HA, et al. (2005) D2 dopamine receptors colocalize regulator of G-protein signaling 9–2 (RGS9–2) via the RGS9 DEP domain, and RGS9 knock-out mice develop dyskinesias associated with dopamine pathways. J Neurosci 25: 2157-2165.[Abstract/Free Full Text]
Lei Q, Jones MB, Talley EM, Garrison JC, and Bayliss DA (2003) Molecular mechanisms mediating inhibition of G protein-coupled inwardly-rectifying K+ channels. Mol Cells 15: 1-9.
Lindorfer MA, Myung CS, Savino Y, Yasuda H, Khazan R, and Garrison JC (1998) Differential activity of the G protein β52 subunit at receptors and effectors. J Biol Chem 273: 34429-34436.[Abstract/Free Full Text]
Medina R, Grishina G, Meloni EG, Muth TR, and Berlot CH (1996) Localization of the effector-specifying regions of Gi2 and Gq. J Biol Chem 271: 24720-24727.[Abstract/Free Full Text]
Mervine SM, Yost EA, Sabo JL, Hynes TR, and Berlot CH (2006) Analysis of G Protein β dimer formation in live cells using multicolor bimolecular fluorescence complementation demonstrates preferences of β1 for particular gamma subunits. Mol Pharmacol 70: 194-205.[Abstract/Free Full Text]
Michaelson D, Ahearn I, Bergo M, Young S, and Philips M (2002) Membrane trafficking of heterotrimeric G proteins via the endoplasmic reticulum and Golgi. Mol Biol Cell 13: 3294-3302.[Abstract/Free Full Text]
Mirshahi T, Robillard L, Zhang H, Hebert TE, and Logothetis DE (2002) Gβ residues that do not interact with G underlie agonist-independent activity of K+ channels. J Biol Chem 277: 7348-7355.[Abstract/Free Full Text]
Posner BA, Gilman AG, and Harris BA (1999) Regulators of G protein signaling 6 and 7. Purification of complexes with Gβ5 and assessment of their effects on G protein-mediated signaling pathways. J Biol Chem 274: 31087-31093.[Abstract/Free Full Text]
Rizzo MA, Springer GH, Granada B, and Piston DW (2004) An improved cyan fluorescent protein variant useful for FRET. Nat Biotechnol 22: 445-449.
Shaner NC, Campbell RE, Steinbach PA, Giepmans BN, Palmer AE, and Tsien RY (2004) Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein. Nat Biotechnol 22: 1567-1572.
Snow BE, Krumins AM, Brothers GM, Lee SF, Wall MA, Chung S, Mangion J, Arya S, Gilman AG, and Siderovski DP (1998) A G protein subunit-like domain shared between RGS11 and other RGS proteins specifies binding to Gβ5 subunits. Proc Natl Acad Sci U S A 95: 13307-13312.[Abstract/Free Full Text]
S?derberg O, Gullberg M, Jarvius M, Ridderstrale K, Leuchowius KJ, Jarvius J, Wester K, Hydbring P, Bahram F, Larsson LG, et al. (2006) Direct observation of individual endogenous protein complexes in situ by proximity ligation. Nat Methods 3: 995-1000.
Takida S, Fischer CC, and Wedegaertner PB (2005) Palmitoylation and plasma membrane targeting of RGS7 are promoted by o. Mol Pharmacol 67: 132-139.[Abstract/Free Full Text]
Takida S and Wedegaertner PB (2003) Heterotrimer formation, together with isoprenylation, is required for plasma membrane targeting of Gβ. J Biol Chem 278: 17284-17290.[Abstract/Free Full Text]
Wang Q, Mullah B, Hansen C, Asundi J, and Robishaw JD (1997) Ribozyme-mediated suppression of the G protein 7 subunit suggests a role in hormone regulation of adenylylcyclase activity. J Biol Chem 272: 26040-26048.[Abstract/Free Full Text]
Wang Q, Mullah BK, and Robishaw JD (1999a) Ribozyme approach identifies a functional association between the G protein β17 subunits in the β-adrenergic receptor signaling pathway. J Biol Chem 274: 17365-17371.[Abstract/Free Full Text]
Wang Y, Windh RT, Chen CA and Manning DR (1999b) N-Myristoylation and β play roles beyond anchorage in the palmitoylation of the G protein o subunit. J Biol Chem 274: 37435-37442.[Abstract/Free Full Text]
Watson AJ, Aragay AM, Slepak VZ, and Simon MI (1996) A novel form of the G protein β subunit Gβ5 is specifically expressed in the vertebrate retina. J Biol Chem 271: 28154-28160.[Abstract/Free Full Text]
Watson AJ, Katz A, and Simon MI (1994) A fifth member of the mammalian G-protein β-subunit family. Expression in brain and activation of the β2 isotype of phospholipase C. J Biol Chem 269: 22150-22156.[Abstract/Free Full Text]
Wedegaertner PB, Wilson PT, and Bourne HR (1995) Lipid modifications of trimeric G proteins. J Biol Chem 270: 503-506.[Free Full Text]
Witherow DS, Wang Q, Levay K, Cabrera JL, Chen J, Willars GB, and Slepak VZ (2000) Complexes of the G protein subunit Gβ5 with the regulators of G protein signaling RGS7 and RGS9. Characterization in native tissues and in transfected cells. J Biol Chem 275: 24872-24880.[Abstract/Free Full Text]
Yoshikawa DM, Hatwar M, and Smrcka AV (2000) G protein β5 subunit interactions with alpha subunits and effectors. Biochemistry 39: 11340-11347.
Zhang S, Coso OA, Lee C, Gutkind JS, and Simonds WF (1996) Selective activation of effector pathways by brain-specific G protein β5. J Biol Chem 271: 33575-33579.[Abstract/Free Full Text]
Zhou JY, Siderovski DP, and Miller RJ (2000) Selective regulation of N-type Ca channels by different combinations of G-protein β subunits and RGS proteins. J Neurosci 20: 7143-7148.[Abstract/Free Full Text]
作者单位：Weis Center for Research, Geisinger Clinic, Danville, Pennsylvania