We are searching data for your request:
Upon completion, a link will appear to access the found materials.
I have a protein that acts as a monomer, but appears to have a relatively high self-affinity in vivo (measured using transient transfection of a plasmid coding for the protein), which is annoying because I'm trying to engineer conditional self-association like with FRB-FKBP w/ rapamycin. Have there been any papers where self-clumping of a protein has been engineered away?
My current approach is to use threading to identify a protein structure and a docking program to identify sites of self-interaction and mutate those sites away.
G-Protein-coupled receptor oligomerization and its potential for drug discovery
The human genome encodes more than 1,000 G-protein-coupled receptors (GPCRs), making these protein the largest class of drug targets, and it has been estimated that 50% of all modern drugs modulate GPCR activity. Despite the recent realization that these receptors form homo-oligomeric and hetero-oligomeric complexes, virtually all therapeutics directed towards GPCRs have been designed using assays that presume these receptors are monomeric, and this important aspect of GPCR biology remains largely unincorporated into schemes to search for new therapeutics.
Although it is not known how large GPCR oligomers are or whether receptors can exist in multiple oligomeric states, the assembly of correctly formed oligomers seems to be a requirement for proper cellular transport to the cell surface. Oligomerization of a receptor with mutant or variant forms of itself can result in attenuation of receptor expression, and might, therefore, represent a potential disease mechanism or a means of cellular self-regulatory modulation of receptor function.
Traditionally, signal amplification is generally thought to occur only at the level of the G protein or the effector, and not at the receptor level. However, both homo- and hetero-oligomerization might provide a means for signal amplification through the activation of many receptors by the action of a single ligand molecule.
GPCR hetero-oligomerization can result in the formation of receptor complexes that have ligand-binding and signalling properties distinct from their constituent receptors. It represents a novel aspect of GPCR biology that has exciting potential to generate new drug targets. Heteromeric interactions might also mask the individual properties of one of the constituent receptors.
Altered levels of GPCR hetero-oligomerization could represent the molecular basis of some clinical disorders.
Little is known about the dynamics and regulation of GPCR oligomer formation that is, whether ligands promote association or dissociation of oligomers, or whether they bind to pre-formed oligomers and change the oligomeric receptor conformation. However, resonance-energy-transfer assays in live cells have indicated that agonist-induced oligomerization might occur.
There is evidence that the oligomeric nature of GPCRs can be exploited to improve drugs by developing dimeric ligands. Several dimeric ligands have been shown to have increased affinity and altered potency compared with their constituent ligands, potentially because dimeric ligands might more readily induce or stabilize the dimeric conformation of the receptors, which in some manner increases the efficacy of signal transduction.
Contemporary drug discovery for GPCRs has largely been based on using a single GPCR of interest expressed in a recombinant cell line. For future lead-compound identification, the current understanding of GPCR oligomerization mandates that hetero-oligomeric receptors must be considered as novel targets in the screening of compounds as drug candidates. Drugs that can enhance or disrupt GPCR oligomer formation as a means to regulate oligomerization-dependent functions will also have to be explored.
Given what is known about GPCR oligomerization so far, it is plausible that the development of 'new' therapies could be as simple as creating novel regimens with 'old' drugs.
The consideration of GPCR quaternary structure has been slow to permeate into the thinking of the drug discovery mainstream despite the potential to exploit it for improved therapies.
The formation of transporter oligomers has been studied by a range of methods, including fluorescence resonance energy transfer (FRET) , cross-linking , pull-downs  and co-immunoprecipitation , as well as biophysical characterization using, for example, size-exclusion chromatography–multiangle light scattering (SEC–MALS) . While such approaches have effectively shown the formation of transporter oligomers, more recent investigations using dominant mutants and high-resolution structural studies have started revealing insights into the precise roles of oligomerization for transporter trafficking, function and regulation of function. This review uses many particularly well-studied transporters and transporter families to provide a brief overview of current understanding of the roles of transporter oligomerization. There are additional examples where high-resolution structures have revealed that the interface between transporter protomers in an oligomeric arrangement forms the substrate-binding site and translocation channel. Examples include the small multidrug transporter, EmrE  and ABC transporters [7,8]. In these cases, oligomerization is responsible for generating the correct architecture for both substrate binding and transport. These transporters are not covered in this review. Additionally, we have not included transporters where the sole purpose of oligomerization seems to be for stability.
Results and Discussion
Dimerization breaks long-time noise correlations in autogenous circuit
To evaluate the dynamic effects of protein-protein binding in positive-autoregulation gene circuits, we construct several alternative models of positive autogenous circuits. Each model emphasizes a different combination of possible feedback mechanisms, and the network topologies considered can be grouped into the two classes of monomer-only (MO) and dimer-allowed (DA) circuits, according to the availability of a protein-dimer state (color coding in Fig. 1). We further group the DA circuits into three variations, DA1 through DA3, depending on which form of the protein is the functional transcription factor (TF) and where the dimerization occurs. For DA1, we only allow the dimer to bind with the DNA-operator sequence (dimeric transcription factor, DTF), while for DA2 dimerization occurs through sequential binding of monomers on the DNA. In DA3, the protein-DNA binding kinetics is the same as in the MO circuit, hence monomeric transcription factor (MTF), with the addition of a cytosolic protein dimer state. While we will only present results for DA1 in this paper, there is no significant difference for DA2 and DA3 [Additional file 1].
Schematic of model autoregulation gene circuit. The DNA binding status is indicated by Dxy, where x corresponds to the operator region (empty = 0, monomer = 1, dimer = 2), and y to the promoter region (empty = 0, RNA polymerase bound = 1). C represents the open complex of DNA-RNAp holoenzyme with the promoter sequence just cleared of RNAp and is subject to transcription elongation. Finally, M, P1 and P2 correspond to mRNA, protein monomer, and dimer, respectively. The network topologies can be grouped into two classes, monomer-only (MO) or dimer-allowed (DA) circuits. We have studied DA1 (red lines), which only allows the dimer to bind with the DNA-operator sequence, DA2 (green) with sequential binding of monomers on the DNA, and DA3 (blue), which shares protein-DNA binding kinetics with MO while allowing dimerization in the cytosol. Note that for topology DA2, we have chosen K31 = K30 (see text for details) We have assumed cells to be in the exponential growth phase and the number of RNAp (R) constant.
Note that the feedback loop is not explicit in Fig. 1 but implicitly included through the dependence of RNAp-promoter binding equilibrium on the binding status of the TF-operator pair. The sign (positive or negative) and strength of the feedback control is determined by the relative magnitude of the dissociation constants between RNAp and DNA which is either free or TF-bound. For instance, topology DA1 has positive feedback control if K30 = k30/q30 > K32 = k32/q32, and K30 corresponds to the level of constitutive transcription (transcription initiation in the absence of bound transcription factor). For each topology, we study the dependence of noise characteristics on the kinetic rates by varying the dimer lifetime, binding affinity, and the individual association/dissociation rates (see Table and Fig. 1). While we only discuss positive feedback control of the autogenous circuit in this paper, we have obtained corresponding results for negative feedback control [Additional file 1].
Fig. 2 shows a sample of ten representative time courses for the protein abundance. The effect of stochastic fluctuations is marked in the MO circuit. However, in all the DA circuits where the protein may form a cytosolic dimer we observe a significantly reduced level of noise in the monomer abundance. The suppression of fluctuations persists throughout the range of kinetic parameters that (so far) is known to be physiologically relevant (see Table 1).
Ten independent time courses of the abundance of protein monomers in the (positive) autoregulatory circuit. The availability of a cytosolic dimer state (red, using circuit topology DA1) significantly reduces the copy-number fluctuations of the monomer compared to the monomer-only (MO) circuit (blue). All corresponding MO and DA1 parameters have the same values. In the ensuing simulations initial conditions are chosen to be the steady state solution of the corresponding deterministic rate equation so that the transient behavior should be minimized.
Calculating the steady-state distribution for the monomer and dimer abundances (Fig. 3) we observe a clear trend that the monomer Fano factor (variance-to-mean ratio) is reduced as the binding equilibrium is shifted towards the dimer. This trend is conserved for all the investigated DA topologies (see Supplementary Information). As long as dimerization is allowed in the cytosol, the fast-binding equilibrium absorbs long-time fluctuations stemming from bursty synthesis or decay of the monomer. When a random fluctuation brings about a sudden change in the monomer copy number, dimerization provides a buffering pool that absorbs the sudden change. Otherwise, random bursts in the monomer abundance will propagate to the transcriptional activity of the promoter, leading to erratic control of protein expression. It should be emphasized that this has nothing to do with the sign of regulation and is in agreement with the observations of Ref.  for negative autoregulation. Surprisingly, the magnitude of noise reduction in the positive autoregulatory circuit is nearly the same as that for negative autoregulation which is typically considered a highly stable construct [Additional file 1].
Stationary state distribution of monomer (black) and dimer (orange) protein abundance in the positive autogenous circuits. The left (right) column corresponds to a ratio of the dimer and monomer decay rates of γ2/γ1 = 1/10 (γ2/γ1 = 1/2). The molecular copy numbers are collected at a fixed time interval (5·10 3 sec) after the steady state has been reached. Here K1 ≡ q1/k1 is the dissociation constant of the protein dimer. As the binding equilibrium is shifted towards the dimer state (decreasing K1), the noise level is monotonically reduced (see Table 2). Note that the prolonged protein lifetime due to the complex formation (left column) affects the noise level.
A heuristic explanation can be found from Jacobian analysis of a deterministic dynamical system, which is justified for small perturbations around a steady state. When a random fluctuation shifts the monomer copy number away from its steady-state value, the decay toward the steady state can be described by the system Jacobian. The disparity in the magnitude of the (negative) eigenvalues of the Jacobian matrix for the MO versus the DA circuits signifies that the perturbed state is buffered by fast settlement of the monomer-dimer equilibrium. This buffering occurs before random fluctuation can accumulate, possibly with catastrophic physiological effects, explaining the coarse long-time patterns observed in the MO model in contrast with the DA circuits (Fig. 2).
Frequency-selective whitening of Brownian noise
The dimerization process itself generates stochastic fluctuations on a short time scale. However, since this time scale is essentially separated from that of monomer synthesis and decay (orders of magnitude faster), dimerization effectively mitigates monomer-level fluctuations. The frequency content of the fluctuations is best studied by an analysis of the power spectral density (PSD), which is defined as the Fourier transform of the autocorrelation function , originally introduced for signal processing. Fig. 4 shows the noise power spectra of DA1, and the distinction between the MO circuit and the DA topology is immediately evident. In particular, we note the following two features. (i) A power-law decay with increasing frequency and (ii) a horizontal plateau for the DA circuits. The power-law feature is explained by the "random walk" nature of protein synthesis and decay: The power-law exponent is approximately 2, which is reminiscent of Brownian motion (a Wiener process) in the limit of large molecular copy numbers. Compared to other commonly observed signals, such as white (uncorrelated) noise or 1/f noise, protein synthesis/decay has a longer correlation time. If the autocorrelation function of a time course is characterized by a single exponential decay, as is the case for Brownian noise, the PSD is given by a Lorentzian profile, and thus, well approximated by an inverse-square law in the low-frequency regime. We do not observe a saturation value for the MO circuit, and it is likely not in the frequency window of physiological interest. This may especially be the case for circuits where the correlation times are long.
Power spectral density (PSD) of fluctuations in protein abundance. The PSD of the MO circuit clearly displays a power-law behavior. All other model systems with an available cytosolic protein dimer state (DA1 shown here) develop a plateau in the mid-frequency region regardless of the model details (see Supplementary Information). As the dimer binding affinity increases, the noise level is further reduced. We have included the MO result in the dimer panel (right) for reference. Datasets with solid (empty) symbols correspond to γ2/γ1 = 1/10 (γ2/γ1 = 1/2).
The noise reduction is in the physiologically relevant low-frequency regime, and in Fig. 4 we have indicated the typical values for a cell cycle and mRNA lifetime. Although stochastic fluctuations impose a fundamental limit in cellular information processing, multiple noise sources may affect cellular physiology non-additively. For a living cell, fluctuations are especially relevant when their correlation time is comparable to, or longer than, the cell cycle. At the same time, short-time scale fluctuations (relative to the cell cycle) are more easily attenuated or do not propagate . Additionally, the observed flat region in the PSD of the DA circuits implies that as far as mid-range frequency fluctuations are concerned, we can safely approximate them as a white noise. This insight may shed light on the reliability of approximation schemes for effective stochastic dynamics in protein-only models.
Increased lifetime of dimer plays an important role
The virtue of the cytosolic dimer state is also directly related to the extended lifetime of proteins when in a complex. Except for the degradation tagging for active proteolysis, a much slower turnover of protein oligomers is the norm. This is partly explained by the common observation that monomers have largely unfolded structures, which are prone to be target of proteolysis . It has also been pointed out that the prolonged lifetime of the oligomeric form is a critical factor for enhancing the feasible parameter ranges of gene circuits . As seen from Fig. 3 (also Table 2), the fold change of the noise reduction, while still significant, is not as strong for the (hypothetical) case of dimer lifetime being the same as that of the monomer (γ2/γ1 = 1/2). However, the low-frequency power spectra still exhibit almost an order-of-magnitude smaller noise power than in the MO circuit with the same rate parameters (Fig. 4). Hence, the noise reduction capability holds good as long as the dimer lifetime is kept sufficiently long compared with the monomer-dimer transition.
Effects of homo-dimerization in genetic toggle switch
The exceptionally stable lysogeny of the phage λ, for which the spontaneous loss rate is ≲ 10 -7 per cell per generation [31, 32], has motivated the synthesis of a genetic toggle switch . Toggle switch is constructed from a pair of genes, which we will denote as gene A and B, that transcriptionally repress each other's expression. This mutual negative regulation can be considered an effective positive feedback loop and provides the basis for the multiple steady states. The existence of multistability, in turn, may be exploited as a device for epigenetic memory or for decision making .
As the general attributes of positive feedback with cooperativity suggest, a genetic toggle switch responds to external cues in an ultrasensitive way: When the strength of a signal approaches a threshold value, the gene expression state can be flipped by a small change in the signal. For example, the concentration of protein A (B) may rapidly switch from high to low and vice versa. However, previous studies of a synthetic toggle switch have shown that the noise-induced state switching is a rare event [15, 34, 35]. In the ensuing analysis, we aim to delineate the origin of this exceptional stability.
In a simple model, the monomer-only (MO) toggle, regulatory proteins only exist in monomeric form. Although an external signal is not explicitly included, random fluctuations in the abundance of the circuit's molecular components will occasionally flip the toggle-state for the two protein species. Drawing on the results from our analysis of positive autoregulatory gene circuits, we hypothesize that dimerization in the regulatory proteins of the toggle switch will serve to stabilize its performance against noise. We allow the protein products of each gene to form a homodimer, being either AA or BB, which is similar to the cI-cro system in phage λ . The dissociation constant for the dimers is defined as K1 = q1/k1, where k1 is the rate of two monomers forming a complex, and q1 the rate of the complex breaking up into its two constituents.
We evaluate the effect of the fast protein binding-unbinding dynamics on the toggle switch performance by using either (i) the monomers or (ii) the homodimers as the functional form of the repressor. Fig. 5 shows, for selected values of the dissociation constant K1, representative time series of the protein monomer (left) and dimer (right) abundances for the case of (a) monomeric or (b) dimeric transcription factors, respectively. A careful analysis of the phase space (in presence of noise) for our chosen set of parameters confirms that the studied toggle-switch systems are in the bistable region .
Sample time series of monomer and dimer copy numbers in genetic toggle switch. (a) MTF circuit, where monomer is the functional form of the repressor. (b) DTF circuit, where dimer is the functional form of the repressor. The left (right) column shows the number of the two monomer molecules A and B (dimers AA and BB), and the initial state is always with species A (red) in high abundance. Note that the switching frequency depends on the binding affinity of protein dimer.
When monomer is the functional form of the repressor molecule (Fig. 5(a)) and K1 is large (limit of low dimer affinity), the protein populations are dominated by monomers. Hence, the circuit effectively behaves as an MO toggle. As K1 decreases, we see that the level of random switching is suppressed: Analogous to the autogenous circuit, the dimer pool stabilizes the protein monomer population. However, the noise suppression is not monotonic with increasing dimer binding affinity. Indeed, for very large binding affinities (small K1), the number of random switching events is increased since the monomer is only available in low copy numbers. Consequently in this limit, it becomes more likely that a small fluctuation in the monomer abundance can cause a dramatic change in the overall gene expression profile. The noise-stabilizing effect of dimerization is also reflected in the corresponding PSDs [Additional file 1]. For instance, we observe a marked suppression of low-frequency fluctuations in the monomer abundance with increasing K1.
In Fig. 5(b) we show corresponding sample time series for the case of a dimeric repressor, all other properties being the same as in (a). While the overall trends are similar, we do note the following difference. Contrary to the monomeric repressor case, there are very few toggle events in the strong binding limit: Since the signaling molecules (dimers) of the dominant gene (the "on"-gene) tend to exist in large copy numbers, a significant fluctuation is needed to flip the state of the toggle switch. In the case of monomeric repression, the signaling molecule exists in low abundance in this limit. Thus, the dominant protein species in the dimeric-repressor system is able to maintain much better control over the state of the toggle switch.
In Fig. 6, we show the distribution (N A- N B), the difference in molecule abundance for the two protein species in the case of monomeric (left) and dimeric (right) transcription factor. The asymmetry with respect to the zero axis is caused by our choice of initial conditions (protein species A in high concentration and species B in low concentration), as well as the finite length of the time series. For monomeric transcription, the presence of dimers with moderate binding affinity sharpens the monomer abundance distribution while accentuating its bimodal character. This is in agreement with the qualitative observation from Fig. 5 on switching stability. For dimeric transcription, we clearly observe that the symmetry of the system is broken for small values of K1, indicating that the state of the toggle switch is extremely stable, and hence, likely determined by the choice of the initial conditions.
Distribution of monomer abundance differences between protein species A and B. The asymmetry with respect to the zero axis is due to the choice of initial state (species A high) and the finite time span of simulations.
To systematically quantify our observations on the interplay between dimer-binding affinity and the functional stability of the toggle switch, we generated long time series (≈3·10 7 sec) to measure the average spontaneous switching rate. In Fig. 7, we show the average toggle frequency relative to that of the MO toggle for the binding affinities K1/nM = <2, 20, 100, 1000>, and the average MO switching rate is 7.5 × 10 -6 /hour. As expected, we find that intermediate values of K1 are able to stabilize the toggle switch. Fig. 7 also highlights the increased stability of the toggle switch for a dimeric versus monomeric transcription factor, the dimeric switching rates always being lower and approaching zero for strong dimer binding.
Random switching rates of genetic toggle switches. Ordinate is the ratio of the random switching rates of various toggle switches to that of the monomer-only (MO) circuit, 7.5 × 10 -6 /hour. MTF, monomeric transcription factor DTF, dimeric transcription factor Het-MTF, monomeric transcription factor with deactivated heterodimer state.
Heterodimerization in genetic toggle switch
We have also considered the case of heterodimerization in the toggle switch, since the noise- and functional stabilization of the switch may be directly affected by the composition and source of the dimers. Note that, the gene-regulation activity is conferred by the two monomer proteins A and B and not the heterodimer AB. However, we find that the presence of (inactive) heterodimers gives rise to very similar noise-stabilizing effects as that of homodimers (Fig. 7). In fact, the existence of heterodimer state allows the dominant protein species to effectively suppress the (active) monomers of the minority species. Thus the heterodimer circuit shows dramatically enhanced functional stability as compared to the case of homodimeric repressors, not sharing the discussed vulnerability of MO circuit to intrinsic noise. Although, to our knowledge, this is a purely hypothetical toggle-switch design, it provides a general strategy for noise control in synthetic gene circuits, along with previously proposed approach of overlapping upstream regulatory domains .
Oligomerization States of FL_RAGE
In order to maintain FL_RAGE in solution, we applied a detergent environment capable of mimicking the native lipid environment in which the TM region is normally embedded. Therefore, FL_RAGE was purified and characterized in the presence of the non-ionic detergent n-Dodecyl β-D-maltoside (DDM). The SEC-MALS analysis of FL_RAGE in the presence of DDM revealed a well-defined peak (Fig. 3a). In the case of the protein–detergent complex, however, the interpretation of the corresponding masses is not straightforward. The amount of detergent bound to the protein may differ depending on both the protein structure and the nature of the detergent. Therefore, the mass ratio between the protein and detergent in the micelle can vary dramatically. A dedicated approach is required in the data analysis step, which considers the presence of the detergent 25 . The “protein conjugate” module of ASTRA software (Wyatt Technologies) enables the calculation of the molecular weight of the whole complex as well as its protein component and the detergent micelle mass associated with the protein.
SEC-MALS and AUC analysis of full-length RAGE. (a) SEC-MALS of FL_RAGE. Chromatograms show the detector readings of the LS and UV detectors in red and black, respectively. Scales for LS and UV detectors are shown on the left-hand axis scale on right-hand axis represents the molecular mass. The green, blue, and cyan lines across the peaks indicate calculated molecular masses of the protein/detergent complex (green), the protein (blue), and the attached detergent micelle (cyan). (b) Sedimentation velocity experiment of DDM-solubilized FL_RAGE (absorption at 280 nm in red and interference in green). The black trace represents the interference data of the buffer including DDM micelles. (c) Example of the analysis of the dataset of FL_RAGE (panel b) using type f/fo GUSSI Membrane Protein Calculation module. Red, green, and blue lines represent the high, optimal, and low limits of f/fo for the protein/micelle fraction, respectively (see text). The pictures show the analysis of the 5.8S peak and indicate the presence of a dimer. The 4.4S and 7.4S peaks represent monomers and trimers, respectively (see Fig. S5 in Supplementary information).
The SEC-MALS chromatogram of FL_RAGE (Fig. 3a) showed a main protein peak eluting at
13 ml. The molar mass of the complex, estimated by ASTRA, was heterogeneous across the peak, spanning the range of masses of the protein–detergent micelle complex (
120–140 kDa). The calculated molar mass of the protein was also heterogeneous (
47–75 kDa), indicating a substantial non-monomeric fraction in the samples of FL_RAGE (the expected molecular mass of FL_RAGE is 41.8 kDa).
The sedimentation velocity of FL_RAGE (Fig. 3b) was investigated by monitoring the absorbance and interference data. The absorbance at 280 nm displayed a broad signal between 3.5S and 8.0S, with the maximum around 4.4S and a trace amount of higher molecular weight species. At the same time, the interference, characterized by a better spatial resolution as compared to the absorption, showed three well-defined peaks (4.4S, 5.8S, and 7.4S) located in the same range of 3.5–8.0S. The additional peak at 3.2S clearly corresponded to the interference boundary observed for the buffer. This peak was not detected when the buffer alone was investigated with absorbance (data not shown) indicating the presence of free DDM. Similar signal was also observed by other authors 26 . Interpreting these findings in terms of the oligomerization status of FL_RAGE required a more detailed analysis. For this purpose, we used the f/fo analysis procedure described in the GUSSI Membrane Protein Module package by Brautigam 27 . Following this, the results indicated that the 4.4S, 5.8S, and 7.4S peaks corresponded to monomer, dimer, and trimer, respectively, as only in this case the f/fo values were within the allowed range of 1.2–1.7. The assumption of another degree of oligomerization led to f/fo values deviated from the permissible range: >1.9 and <1 for higher order and smaller oligomers, respectively, which is not possible. Figure 2c illustrates this type of analysis based on peak number 2 (5.8S). These results of the AUC experiment revealed the presence of a mixture of monomers, dimers, trimers and even a small fraction of higher order oligomers, which is in good agreement with the SEC-MALS result. Obviously, a rapid exchange between monomeric and dimeric species resulted in a broad peak at 280 nm.
For native MS analysis, FL_RAGE was solubilized in DDM or in Triton X-100 at a concentration of two times the critical micelle concentration (CMC). Despite the stability of FL_RAGE in the presence of DDM during protein purification, the buffer exchange to an MS-friendly ammonium acetate (AmAc) solution, which also contained 2x CMC of DDM, led to protein precipitation, even at a quite high concentration of AmAc (1M). The precipitation of FL_RAGE was not observed when Triton X-100 (2x CMC) was used instead of DDM. Following this, the native MS spectrum acquired at 600 mM AmAc revealed monomeric, dimeric, trimeric, and even tetrameric signals (Fig. 4a). Both monomeric and dimeric species are close to equilibrium and accounted for the vast majority of the signal. Interestingly, the purification of FL_RAGE using SEC instead of ion exchange chromatography enabled a better preservation of higher-order oligomers, e.g. trimers and tetramers (Fig. S3), with however some impurities (unidentified contaminants) and a more challenging membrane protein liberation in the gas phase (poor resolution). The addition of 0.1% formic acid and 50% acetonitrile to FL_RAGE led to the disappearance of signals stemming from oligomeric species (Fig. 4b), indicating a non-covalent character of the oligomerization. As native MS of membrane proteins depends on a collision-activated liberation of the protein from the detergent micelle cluster, we required a minimum of (15V) 30V collision energy along with an activation of 200V in the sampling cone in order to observe (monomeric) oligomeric species of FL-RAGE in the spectrum.
Mass spectra of FL_RAGE under native (a) and (b) denaturing conditions. Peaks are marked by colored dots and the corresponding charge state for monomers and multimers of FL_RAGE accordingly.
GPCR oligomerization is now well recognized, with the potential for significant functional and regulatory impact (Cvejic and Devi 1997 Pascal and Milligan 2005 Rocheville et al. 2000a). The most extensive analysis of GPCR oligomerization has been performed on Family A receptors, and this receptor subfamily is also the best characterized with regard to structure (Milligan 2007). Family B GPCRs also form homo-oligomeric (Harikumar et al. 2006b Lisenbee and Miller 2006 Seck et al. 2003) and hetero-oligomeric complexes (Harikumar et al. 2006b). The structure of receptors in this family has been predicted to be distinct from those in Family A (Dong et al. 2007 Salom et al. 2006), with different signature sequences and even predicted differences in the helical bundle.
Family B GPCRs have a long amino-terminal extracellular tail region that includes six conserved cysteine residues that are involved in three conserved disulfide bonds (Lisenbee et al. 2005). This region is important for natural peptide ligand binding (Dong and Miller 2002). This region, however, is not important for secretin receptor oligomerization (Lisenbee and Miller 2006). The intracellular carboxyl-terminal tail of the secretin receptor can also be truncated without effect on its oligomerization (Lisenbee and Miller 2006). In a recent study, the major determinant for secretin receptor homo-oligomerization was the lipid-exposed face of the fourth transmembrane segment (Harikumar et al. 2007).
In the current work, we have systematically explored the ability of a broad variety of Family B GPCRs to form hetero-oligomeric complexes with the secretin receptor. We already knew that selected Family A GPCRs did not form such complexes (Harikumar et al. 2006a) and that the most closely related Family B GPCRs, the VPAC1 and VPAC2 receptors, were able to form oligomers (Harikumar et al. 2006b). In this studies, we extended this survey across a wide range of Family B peptide hormone receptors. Of note, each of these receptors except for the calcitonin receptor was able to form hetero-oligomers with the secretin receptor indicating that the potential for hetero-oligomerization is likely to be extensive across Family B receptors. In evaluating the fourth transmembrane segment of these receptors, it is interesting that the human calcitonin receptor includes two residues within this segment that are not present in any other receptors in this series. These are Arg at equivalent position to Tyr 233 of the secretin receptor and Thr at equivalent position to Ala 250 of the secretin receptor these residues are predicted to be at either end of the transmembrane segment, facing the lipid bilayer. One (or both) of these residues could contribute to the distinct hetero-oligomerization behavior exhibited by the calcitonin receptor in the current series of experiments. However, it is not clear if the same sequence determinant responsible for homo-oligomerization is also the primary sequence involved in hetero-oligomerization. Future work will address the physiological significance of heterologous interaction across Family B receptors.
Get full journal access for 1 year
All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.
Get time limited or full article access on ReadCube.
All prices are NET prices.
Site-specific labelling of purified monomeric R and reconstitution into lipid vesicles
The goal of our studies was to monitor self-association of β2AR following reconstitution in lipid vesicles and to obtain information about the relative orientation of protomers in oligomers. FRET using small-molecular-weight fluorescent probes is an ideal tool for these studies because it provides relative distance information yet requires relatively small amounts of protein (Mansoor et al, 2006). To achieve site-specific labelling of the β2AR, we generated modified receptors having single-reactive cysteines that can be chemically modified with sulphydryl-reactive fluorophores. Mutants were made on a minimal cysteine background in which the five chemically reactive cysteines, out of 13, were mutated (see section Materials and methods). These mutations have no effect on ligand binding or G protein coupling. The remaining three cysteines that are not palmitoylated or are part of disulphide bonds are not reactive due to their location in the hydrophobic core ( Figure 1A ). We initially constructed 18 single-cysteine mutants in the cytoplasmic domains of the receptor. Three were chosen based on their functional properties, chemical reactivity and their distribution ( Figure 1B ): 㥅-T66C (intracellular loop-1, ICL1), 㥅-A265C (transmembrane domain-6, TM6) and 㥅-R333C (helix-8, H8) (Cherezov et al, 2007 Rasmussen et al, 2007 Rosenbaum et al, 2007). This spatial distribution of the labelling sites was designed to provide information about the orientation of protomers relative to each other.
β2AR single-cysteine constructs and FRET donorptor pair. (A) Three single-reactive cysteines constructs were generated on a minimal cysteine background (㥅-β2AR). The labelling sites were placed in the first ICL, 㥅-β2AR-T66C, at the cytoplasmic end of the sixth transmembrane segment, 㥅-β2AR-A265C, and helix eight, 㥅-β2AR-R333C. (B) Intracellular 3D view of the distribution of regions chosen for single-cysteine mutants, α-carbons are depicted. (C) FRET donor (λex=549 nm λem=570 nm) and acceptor pair (λex=650 nm λem=670 nm).
Modified receptors were expressed in Sf9 cells using recombinant baculovirus and purified using sequential antibody and alprenolol affinity chromatography. We have shown previously that this purification protocol produces monomeric, detergent-solubilized β2AR (Whorton et al, 2007). Purified, detergent-solubilized β2AR was labelled with Cy3 or Cy5 maleimide. These fluorophores were chosen for FRET studies because they possess an R0 value (Förster critical distance where 50% of energy transfer occurs) in the range of 37 to 56 Å depending on the experimental system (Mansoor et al, 2006 Massey et al, 2006). This is ideal for studying receptor–receptor interactions since a monomer has approximate dimensions of 30 Å × 40 Å × 70 Å. β2ARs were labelled with stoichiometric amounts of either Cy3 or Cy5, and the efficiency of labelling was determined by absorption spectroscopy (Supplementary Table I). Labelled β2AR was reconstituted into a mixture of 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) and cholesterol hemisuccinate (CHS) lipids. Three samples were generated for each experiment: (1) reconstituted Cy3-labelled β2AR (2) reconstituted Cy5-labelled β2AR and (3) Cy3-labelled β2AR mixed with Cy5-labelled β2AR and then reconstituted. The final lipid-to-receptor molar ratio (mol-to-mol) was 1000:1 unless otherwise indicated. The same samples were prepared for controls, but were maintained in 0.1% DM and not reconstituted into vesicles.
Orientation of R in lipid vesicles
Knowing the orientation of β2AR in our lipid vesicles is essential for interpreting FRET measurements. Random orientation would generate potential non-physiological (antiparallel) oligomers. While random orientation might be expected, previous studies have shown that rhodopsin orients predominantly in one direction following reconstitution (Niu et al, 2002). We used several complementary strategies to determine the orientation of β2AR in phospholipid vesicles ( Figure 2A ). Factor Xa is a protease that selectively cleaves the β2AR within the third ICL (ICL3). Receptors oriented inside-out (ICL3 outside of lipid vesicle) will be susceptible to Factor Xa, whereas those oriented outside-out will not ( Figure 2A ). Approximately 90% of reconstituted β2AR was resistant to Factor Xa, whereas all of the receptor was cleaved in the presence of 0.1% DM ( Figure 2B ), a concentration of detergent, which permeabilizes the vesicles. These results are consistent with a predominantly outside-out orientation. PNGase F is an enzyme that cleaves asparagine-linked oligosaccharides on the extracellular N-terminus ( Figure 2A ). Treatment of reconstituted receptor with PNGase F led to a mobility shift that was indistinguishable from that observed in the presence of 0.1% DM, consistent with predominantly outside-out orientation ( Figure 2C ). The orientation was further confirmed using NHS-PEO4-biotin to chemically modify the N-terminus ( Figure 2A ). This polar compound would not be expected to cross the lipid bilayer. Chemical modification of the N-terminal FLAG epitope results in loss of reactivity to the M1 antibody treatment of vesicles following reconstitution resulted in the loss of M1 reactivity for more than 90% of reconstituted β2AR ( Figure 2D and E ). Finally, we labelled A265C on the cytoplasmic side of the β2AR with monobromobimane (mBBr) and examined the ability of tryptophan in solution to quench bimane fluorescence. We observed no quenching of reconstituted, bimane-labelled β2AR with 1 mM tryptophan. However, solubilization of vesicles using 0.2% DM resulted in significant quenching (Supplementary Figure 1). Taken together, these studies show that the β2AR is predominantly oriented with the extracellular domains on the outside of the vesicle.
β2ARs are predominantly oriented outside-out in lipid bilayers. (A) Strategies for determining orientation of β2AR in lipid bilayers. (B) Purified receptors were reconstituted as described under Materials and methods and then subjected to treatment with Factor Xa and resolved by 10% SDS–PAGE and transferred onto nitrocellulose. The presence of β2AR was determined by probing with an M1 antibody conjugated with Alexa-680. (C) Samples subjected to PNGase F were prepared and imaged as in panel A. (D, E) Reconstituted samples were treated with the hydrophilic, amine-reactive, alkylating reagent NHS-PEO4-biotin that disrupts binding of the M1 monoclonal antibody to the FLAG epitope. Samples were assessed for reactivity to M1 antibody (D) and an antibody that recognizes the C-terminal six-histidine tag (E). All data are representative of three independent experiments.
Distribution of R in lipid vesicles
In studying oligomerization, it is important to avoid forcing protein together by inhomogeneous reconstitution, that is, trapping the majority of the receptor molecules in a minor population of lipid vesicles. For instance, it has been shown previously that 90% of rhodopsin molecules were incorporated into only 10% of vesicles (Mansoor et al, 2006). We used isopycnic density centrifugation to assess the distribution of β2ARs in lipid vesicles as previously described for rhodopsin (Mansoor et al, 2006). Cy5-labelled β2ARs were reconstituted at a lipid-to-receptor ratio of 1000:1 in lipids containing NBD–phosphocholine (at a final of 0.4% of total lipid content). This allowed us to analyze samples subjected to a discontinuous sucrose density gradient by following Cy5 fluorescence (for the presence of β2AR) and NBD fluorescence (for the presence of lipid vesicles). Our results show nearly perfect correlation between Cy5 fluorescence and NBD fluorescence at every fraction analyzed, suggesting that β2AR molecules are uniformly distributed in these vesicles ( Figure 3A ). Similar results were obtained with β2AR reconstituted at a 10 000:1 lipid-to-receptor ratio (Supplementary Figure 2).
β2ARs are homogenously distributed in lipid vesicles. (A) To determine the distribution of β2ARs in lipid vesicles, sucrose density gradients of samples containing 0.4% NBD–phosphocholine and Cy5–β2ARs reconstituted at a lipid-to-receptor ratio of 1000:1 were performed as described in the Supplementary data. Detection of lipid fractions was performed by following NBD fluorescence (λex=460 nm) and receptor fractions by following Cy5 fluorescence (λex=649 nm). (B) Reconstituted β2ARs were imaged using a negative staining protocol as described in the Supplementary data to determine the size distribution of vesicles and the number of receptors per vesicle. Scale bar length represents 200 nm. Data are representative of three independent experiments.
To assess the density of β2ARs in the lipid vesicles, we used electron microscopy to determine the average diameter of our β2AR-containing lipid vesicle preparations. Using a negative staining protocol, we determined that the average diameter of our vesicle preparations at a lipid-to-receptor ratio of 1000:1 was 83 nm넒 nm ( Figure 3B ). Using the calculations detailed in the Supplementary Materials and methods section, we concluded that there are 50 β2ARs per lipid vesicle, with the majority oriented in an outside-out manner.
Functional characterization of R in lipid vesicles
We performed saturation binding on purified, reconstituted receptor to determine the affinity of all three single-cysteine mutants for the antagonist [ 3 H]-dihydroalprenolol (DHA). We observed no significant difference between the three modified β2ARs and wild-type β2ARs ( Table I and Supplementary Figure 3). Competition binding studies with [ 3 H]-DHA were used to determine the Ki values for the agonist isoproterenol (Iso) and the inverse agonist ICI 118,551 (ICI). As shown in Table I and Figure 4 , the values for the single-cysteine mutants are comparable to those obtained for wild-type β2AR, suggesting that introduction of the single-reactive cysteines and reconstitution of purified β2AR into lipid vesicles does not alter the pharmacology of the receptor.
Single-reactive cysteine mutants are fully functional. The affinity of the agonist isoproterenol (A) and the inverse agonist ICI 118,551 (B) was measured for all three single-cysteine mutants (㥅-T66C, 㥅-A265C and 㥅-R333C) and wild-type receptor by competitive binding of [ 3 H]-DHA. Results are expressed as percent of radio-ligand bound in the absence of competitor. (C) Functionality of the three single-cysteine mutants, unlabelled or labelled with Cy5, and wild-type receptor was determined by GTPγS binding as described in the Supplementary data. [ 35 S]-GTPγS-specific binding induced by 10 μ M isoproterenol (agonist response) or by 10 μ M ICI 118,551 (inverse agonist response) is shown as fold over basal. All functional data represent the mean±s.e.m. of three independent experiments performed in triplicate.
Agonist, antagonist and inverse agonist binding properties for the single-reactive cysteine receptors a
|β2AR||Ki [s.e. interval] (n M )||Kd±s.e.m.|
|Mutant||(−)-Isoproterenol||ICI 118,551||[ 3 H-DHA]|
|Wild type||355 ||1.17 [0.77𠄱.77]||1.3ଐ.16|
|㥅-T66C||388 ||2.09 [1.86𠄲.35]||1.8ଐ.21|
|㥅-A265C||298 ||1.84 [1.45𠄲.33]||2.5ଐ.23|
|㥅-R333C||214 ||1.90 [1.59𠄲.27]||2.4ଐ.24|
|a Saturation and competition binding were performed as described under Materials and methods. Data represent the mean±s.e.m. of at least three independent experiments.|
Functionality for G protein coupling of the three single-cysteine mutants was addressed by [ 35 S]-GTPγS binding. This assay involves reconstituting purified β2AR with purified Tet-Gαs as previously described (Swaminath et al, 2005 Granier et al, 2007). Agonist binding to all three β2AR single-cysteine mutants led to significant stimulation of G protein coupling that was similar to wild-type receptor ( Figure 4C ). Treatment of samples with the inverse agonist ICI led to decreases in basal activity similar to that observed for wild-type receptor ( Figure 4C ). Modification of the single cysteines with Cy5–maleimide fluorophore had no significant effect on G protein coupling ( Figure 4C PϠ.05).
FRET analysis of fluorophore-labelled Rs in lipid bilayers
We first determined FRET between receptors labelled at the same position. 㥅-A265C labelled with Cy3 was reconstituted with an equivalent amount of 㥅-A265C labelled with Cy5 in order to monitor TM6/TM6 interactions. This was repeated for 㥅-T66C and 㥅-R333C, as reporters for ICL1/ICL1 and H8/H8 interactions, respectively. Figure 5A shows an example of a typical experiment performed on 㥅-T66C. FRET between Cy3- and Cy5-labelled receptors (30.3.2%) is only observed after receptor reconstitution into a lipid bilayer, but not when receptors remain solubilized in detergent ( Figure 5A and Table II ). Similar observations were made for Cy3- and Cy5-labelled 㥅-A265C (16.7.3%) and for Cy3- and Cy5-labelled 㥅-R333C (26.9ଐ.8% Table II ).
Intermolecular FRET between Cy3- and Cy5-labelled β2AR is independent of other cellular proteins and is specific. (A) Purified, detergent-solubilized receptor protein was labelled with Cy3 or Cy5 maleimide and unreacted fluorophore was quenched with cysteine and separated from protein by gel filtration as described under Materials and methods. Cy3- and Cy5-labelled protein samples were mixed at a 1:1 molar ratio and reconstituted into phospholipids bilayers or maintained in detergent. Subtraction of the proper controls and normalization of the raw traces is described in the Supplementary data. Labelled β2ARs were reconstituted at a 10-fold higher lipid-to-receptor ratio (10 000:1) and FRET efficiency was measured for ICL1/ICL1 (B), TM6/TM6 (C) and H8/H8 (D) interactions. Data are representative of at least three independent experiments (A) or represent the mean±s.e.m. of at least three independent experiments (B𠄽).
FRET efficiencies in the absence of ligand and upon binding of agonist, neutral antagonist or inverse agonist a
|β2AR region||No ligand±s.e.m.||+Isoproterenol±s.e.m.||P-value||ʺlprenolol±s.e.m.||P-value||+ICI 118,551±s.e.m.||P-value|
|TM-6/H-8||30.98ଐ.7||35.27.5||0.01 *||33.63ଐ.2||0.05 *||25.33.0||0.005 **|
|a FRET efficiencies were calculated as described under Materials and methods. Data represent the mean±s.e.m. of at least three independent experiments. P-values refer to statistical comparisons between no ligand and three different ligands: isoproterenol, alprenolol and ICI 118,551.|
|* Pπ.05 ** Pπ.005.|
To provide additional information about the relative orientation of β2AR protomers, we investigated FRET between different labelling sites. For example, 㥅-T66C labelled with Cy3 was reconstituted with an equivalent amount of 㥅-A265C labelled with Cy5 in order to examine ICL1/TM6 interactions. The same approach was followed for the other possible combinations, ICL1/H8 and TM6/H8 ( Table II ). The observation of different FRET efficiencies for different labelling pairs suggests a specific arrangement of receptors in the lipid bilayers rather than nonspecific aggregation. To further rule out the possibility that the FRET observed in these studies is simply due to crowding of labelled receptors at the lipid bilayer, a 10-fold higher molar concentration of lipids (a final lipid-to-receptor ratio of 10 000:1) was used in order to reduce the number of receptors per unit area of lipid bilayer. FRET efficiencies observed at a lipid-to-receptor ratio of 10 000:1 were not significantly different from those obtained at a ratio of 1000:1 ( Figure 5B𠄽 PϠ.05).
FRET saturation of fluorophore-labelled R oligomers
To further investigate the specificity of the observed oligomerization, as well as the stoichiometry of the oligomers, we performed FRET saturation experiments where the ratio of acceptor fluorophore (Cy5-labelled β2AR) to donor fluorophore (Cy3-labelled β2AR) is increased, while maintaining the overall receptor concentration and lipid-to-receptor ratio constant. If the energy transfer is due to specific receptor–receptor interactions, FRET efficiency will saturate as the Cy5/Cy3 ratio is increased. In contrast, random collisions should yield a quasi-linear relationship (Mercier et al, 2002 James et al, 2006 Harikumar et al, 2008). We observe FRET saturation for all three β2AR labelling sites ( Figure 6A𠄼 ), demonstrating the specific nature of the interactions.
Specificity of β2AR oligomerization as assessed by FRET saturation. FRET saturation involved varying the ratio of Cy5- to Cy3-labelled β2ARs over a range of 1:1 to 10:1 (Cy5:Cy3), while the overall β2AR concentration was kept constant. Saturable FRET is observed for ICL1/ICL1 (A), TM6/TM6 (B) and H8/H8 (C). FRET measurements were performed and calculated as described in the Supplementary data. Data represent the mean±s.e.m. of at least three independent experiments. (D) FRET saturation data from all three constructs (A𠄼 above) was normalized to maximal FRET efficiency and then averaged and plotted together with theoretical curves (dashed lines) for dimer, trimer, tetramer and higher-order oligomer that were generated using equation (1) in the Supplementary data.
In addition, FRET saturation can provide insight into the number of protomers per oligomer. Our FRET saturation results were compared with a well described mathematical model (Veatch and Stryer, 1977 Mercier et al, 2002 James et al, 2006 Harikumar et al, 2008 Harding et al, 2009) that has been used to predict the maximal energy transfer expected in energy transfer saturation experiments (FRET or BRET) for dimers, trimers, tetramers, etc. It follows that saturation will occur at a lower acceptor/donor ratio for higher-order oligomers than for simple dimers. We normalized our FRET saturation data for all three constructs and compared them with models for dimers, trimers, tetramers and higher-order oligomers (eight protomers), and found that our data are superimposed on the theoretical curve for tetramers ( Figure 6D ).
The effect of ligand efficacy on R oligomerization
We examined the effects of three classes of GPCR drugs: a full agonist (isoproterenol), a neutral antagonist (alprenolol) and an inverse agonist (ICI) on FRET efficiency between different labelling sites. Upon treatment with saturating amounts (10 μM) of the full agonist isoproterenol, a small, but significant, increase in FRET was observed between TM6 and H8 ( Figure 7A and Table II Pπ.05). At saturating concentrations (500 nM), alprenolol produced a similar result between TM6 and H8 ( Figure 7A and Table II Pπ.05). It is not possible to say if these small changes are due to subtle changes in the relative arrangement of protomers, or small conformational changes in the receptor.
β2AR oligomers are regulated by inverse agonists. (A) Treatment of FRET samples with saturating amounts of the inverse agonist ICI 118,551, agonist isoproterenol and neutral antagonist alprenolol. (B) FRET saturation in the presence of ligands. Isoproterenol and alprenolol led to no observable difference from the unliganded FRET saturation curve, whereas ICI 118,551 yielded to a curve that is more consistent with higher-order oligomers. (C) Cross-linking of reconstituted Cy5-labelled β2AR samples in the presence or absence of isoproterenol or ICI 118,551 was carried out as described in the Supplementary data. (D) FRET saturation in the presence of the inverse agonists carvedilol (red) and carazolol (green). All data are reported as mean±s.e.m. (A, B, D) or are representative of at least three independent experiments (C). * (Pπ.05) and ** (Pπ.005).
In contrast to the small changes observed with the agonist and neutral antagonist, much larger changes were observed following exposure to the inverse agonist ICI ( Figure 7A and Table II ). Inverse agonists include many compounds that were originally classified as antagonists, ligands that occupy the orthosteric binding site, but do not alter receptor function. Instead, inverse agonists inhibit basal agonist-independent activity exhibited by many GPCRs, including the β2AR (Galandrin and Bouvier, 2006). Interestingly, at a saturating concentration (500 nM) of ICI, significant changes in FRET efficiency were observed for four of the six labelling pairs ( Figure 7A and Table II ). The ICI-induced changes in FRET reach a maximum at 10 min (Supplementary Figure 4).
The changes in FRET observed with ICI could reflect changes in the orientation of protomers or the number of protomers in the oligomeric complex. However, these changes could also be due to ligand-induced changes in the conformation of receptors that influence the mobility of the fluorophore or the polarity of its environment. We, therefore, examined the effect of isoproterenol and ICI on the intensity of the fluorescence and on the anisotropy of the fluorophores in labelled receptors. For both Cy3- and Cy5-labelled β2AR reconstituted individually, treatment with either isoproterenol or ICI did not induce significant changes in the intensity of the fluorescence (data not shown) or anisotropy of the fluorophores, suggesting that the change in FRET efficiencies observed upon ICI treatment are not a result of conformational changes in protomers (Supplementary Figure 5).
The ICI-induced changes in FRET efficiency may be attributed to several additional factors: reorientation of protomers in the oligomers, a tighter packing of the protomers, an increase in the temporal stability of the oligomers (assuming there is an equilibrium between monomers and oligomers), and an increase in the stoichiometry of the oligomeric state (e.g., going from dimers or tetramers to higher-order oligomers). To distinguish between these possibilities, we performed FRET saturation in the presence and absence of ICI, alprenolol or isoproterenol. Samples were incubated with ligands for 30 min at room temperature and measurements were taken. Results show that in the presence of ICI the saturation curve is more similar to a model for higher-order oligomers, while alprenolol and isoproterenol appear to have no effect on the apparent oligomeric state of the receptor ( Figure 7B ). Higher-order oligomerization was also observed for the inverse agonists carazolol and carvedilol ( Figure 7D ).
Cross-linking was used to further address the state of multimeric assembly of the β2AR. We used Bis(NHS)PEO5, a homobifunctional cross-linker with a spacer length of 21.7 Å that covalently modifies ɛ-amines of lysine residues and α-amine groups at the N-termini, effectively trapping receptors that come within interacting distances. Although concerns have been raised about the potential for cross-linkers to trap transiently interacting proteins (Brett and Findlay, 1979 Downer, 1985 Medina et al, 2004), it is evident that pre-incubation of samples with ICI leads to more extensive cross-linking and trapping of higher-order oligomers of reconstituted β2AR when compared with the unliganded, the agonist and the antagonist treated samples ( Figure 7C and Supplementary Figure 6). Taken together, these results suggest that the β2AR forms higher-order oligomers in the presence of the inverse agonists ICI, carazolol and carvedilol.
To investigate the effect of ICI on the stability of interactions between protomers, we monitored FRET following addition of 0.2% DDM, a concentration of detergent that solubilizes the vesicles. We found that the decline in FRET following the addition of detergent was slower and less complete in samples pre-incubated with ICI compared with unliganded samples (Supplementary Figure 7, Pπ.05), providing evidence that ICI also stabilizes interactions between protomers.
The effect of Gs on R oligomerization
To investigate the effect of G protein coupling on oligomerization, we performed FRET saturation experiments by reconstituting β2AR with a three-fold molar excess of purified Gs heterotrimer ( Figure 8 ). This concentration of G protein was chosen to ensure that sufficient G protein would be incorporated into vesicles while having a minimal effect on the reconstitution. The inclusion of Gs did not alter the orientation of the β2AR as determined by the susceptibility to PNGase F ( Figure 8A ). We observed a statistically significant (Pπ.008) decrease in FRET saturation in the presence of Gs as compared with reconstitutions in the absence of Gs ( Figure 8B ) for Cy5/Cy3 of 0.25, 0.5 and 1. To determine whether the effect of Gs on FRET saturation was due to nonspecific effects of reconstituting with another membrane-associated protein, we performed FRET saturation of β2AR in the presence of Gs and GTPγS, which uncouples β2AR and Gs. As shown in Figure 8C , the presence of GTPγS increases, in a statistically significant manner (Pπ.04), FRET saturation to the values observed for β2AR alone. To estimate the fraction of β2AR that couples to Gs under these reconstitution conditions, we labelled C265 at the cytoplasmic end of TM6 with mBBr, an environmentally sensitive fluorescent probe. We previously showed that maximal coupling of Gs to β2AR reconstituted into HDL particles results in a decrease in the fluorescence intensity and an 18-nm shift in the maximal emission wavelength (λMAX) of mBBr–β2AR (mBBr–β2AR Yao et al, 2009). As shown in Figure 8D , under reconstitution conditions used for FRET saturation experiments, Gs induced a decrease in intensity and a 4-nm shift in λMAX of mBBr–β2AR relative to the same reconstitution in the presence of GTPγS. Based on the shift of λMAX we estimate that approximately 20% of the reconstituted β2AR is coupled to Gs.
Effect of the G protein Gs on FRET saturation of Cy5- and Cy3-labelled β2AR. FRET saturation was performed by varying the ratio of Cy5- to Cy3-labelled β2AR-R333C over a range of 1:4 to 4:1 (Cy5:Cy3), while the overall β2AR concentration was kept constant. Purified Gs heterotrimer was added at a molar ratio of 3 Gs:1 β2AR before reconstitution. (A) The inclusion of Gs in the reconstitution did not alter the orientation of β2AR in vesicles as determined by the susceptibility of reconstituted β2AR to PNGase F (see Figure 3C ). FRET saturation was significantly lower in the presence of Gs compared with β2AR alone (B) or β2AR and Gs with 10 μ M GTPγS (C). (D) β2AR was labelled on C265 at the cytoplasmic end of TM6 with mBBr–β2AR and reconstituted with Gs under the same conditions that were used for FRET saturation experiments. Gs induced a decrease in intensity and a 4-nm shift in λMAX of mBBr–β2AR relative to the same reconstitution in the presence of GTPγS. A two-way ANOVA was used to compare FRET values for β2AR, β2AR+Gs and β2AR+Gs+GTPγS at the different Cy5:Cy3 ratios. A posteriori statistical analysis showed significant decrease in FRET between β2AR and β2AR+Gs (Pπ.008), and a significant increase in FRET between β2AR+Gs and β2AR+Gs+GTPγS (Pπ.04) for all Cy5:Cy3 ratios except 2 and 4. No statistical differences are found between β2AR and β2AR+Gs+GTPγS.
The PI3K–Akt is a key regulator in mediating cell survival ( Franke et al, 2003 ). We have now delineated a novel pathway linking PI3K–Akt with a BRCT domain protein TopBP1 for the control of E2F1 and Miz1 activities. Through the Akt–TopBP1 pathway, both E2F1-induced apoptosis and Miz1-induced cell cycle arrest can be controlled to ensure progression of normal proliferation. Moreover, a new mechanism of regulation of TopBP1 function through Akt-dependent oligomerization has also been elucidated.
Akt–TopBP1 pathway in the control of E2F1-mediated apoptosis
We provide compelling evidence that Akt phosphorylates TopBP1 at a single residue Ser 1159 . This phosphorylation event is necessary for TopBP1 to repress the transcriptional and proapoptotic activities of E2F1 under physiological conditions. We also demonstrate that regulation of E2F1 by TopBP1 is independent of all pocket proteins. Taken together, we now propose a pathway involving Akt–TopBP1 in parallel to the classical Cdk–Rb pathway (Figure 7F) for the control of E2F1-dependent apoptosis. In response to growth factor stimulation, activation of Cdk leads to pRb phosphorylation and release of free E2F to induce S-phase entry. Activation of PI3K–Akt is important to hold in check the proapoptotic activity of E2F1 through regulation of TopBP1 during G1/S transition. To ensure proper control of apoptosis, there exist several feedback loops. TopBP1 in fact is an E2F target and its level peaks at G1/S and S phases ( Liu et al, 2004 Yoshida and Inoue, 2004 ). E2F also activates Gab2 and upregulates Akt activity ( Chaussepied and Ginsberg, 2004 ). During S phase, E2F1 activity is inhibited by cyclin A/Cdk2 ( Krek et al, 1994 Xu et al, 1994 ). Together, these mechanisms provide a multi-layered regulation to control E2F1-mediated apoptosis during cellular proliferation.
The Akt–TopBP1 pathway may control E2F1 during DNA damage as well. The interaction between TopBP1 and E2F1 is induced upon phosphorylation of E2F1 at Ser 31 by ATM ( Liu et al, 2003 ). TopBP1 can be phosphorylated by ATM ( Yamane et al, 2002 ) and the level of TopBP1 protein is induced by the radiomimetic agent neocarzinostain (NCS) ( Liu et al, 2003 ). Ionizing radiation activates Akt ( Contessa et al, 2002 Viniegra et al, 2005 ), which was confirmed in our hands by NCS treatment (data not shown). ATM mediates the phosphorylation of Akt at Ser 473 and contributes to activation of Akt in response to ionizing radiation ( Viniegra et al, 2005 ). Activation of Akt during adriamycin treatment inhibits the expression of E2F1 target genes through TopBP1 action (Figure 3B). Hence, in response to ionizing radiation, ATM kinase may target the Akt–TopBP1 pathway to control E2F1-mediated apoptosis and allow DNA repair to complete.
Akt–TopBP1 pathway in the control of HPV 16 E2 and Miz1
The interaction between TopBP1 and HPV16 E2 or Miz1 requires the Akt phosphorylation site as well. These results suggest that Akt–TopBP1 also regulates other biological processes. HPV16 E2 is essential for the life cycle of HPVs. It regulates transcription of the viral genome. It also interacts with viral replication factor E1 and recruits E1 to replication origins for viral replication. TopBP1 interacts with HPV16 E2 and enhances its ability to activate transcription and replication ( Boner et al, 2002 ). During HPV infection, HPV E7 targets members of the pocket protein family such as pRb to release E2Fs. It also activates Akt through an interaction with PP2A, preventing dephosphorylation of activated Akt ( Pim et al, 2005 ). Thus, the activation of Akt by E7 could regulate TopBP1 and enhance E2-mediated activation of viral transcription and replication.
Miz1 is a Myc-associated zinc-finger transcription factor. Miz1 is involved in cell cycle arrest through its activity to transactivate p15 INK4b and p21 Cip1 ( Wanzel et al, 2003 ). Myc interacts with Miz1 and represses transcriptional activation by Miz1, at least in part through competition with p300 for binding to Miz1 ( Staller et al, 2001 ). The interaction with Miz1 plays a crucial role in mediating transcriptional repression by Myc. TopBP1 interacts with Miz1 through its BRCT7 and BRCT8 domains, and represses Miz1 transcriptional activity ( Herold et al, 2002 ). Our data further suggest that phosphorylation by Akt is required for TopBP1 to bind and repress Miz1. In addition to the control of Miz1 by Akt–TopBP1, Akt also phosphorylates Miz1 directly this phosphorylation leads to binding of 14-3-3η to the Miz1 DNA binding domain and inhibits Miz1 function ( Wanzel et al, 2005 ). Therefore, Akt regulates Miz1 to control cell cycle arrest directly through phosphorylation of Miz1 as well as indirectly through phosphorylation of TopBP1 (Figure 7F). The 14-3-3η-mediated and TopBP1-mediated inhibition of Miz1 are likely two alternative processes, as there is no detectable interaction between TopBP1 and 14-3-3 to suggest a trimeric Miz1/TopBP1/14-3-3 complex formation (Supplementary Figure 6). The inhibition of p21 expression during adriamycin treatment by ca-Akt is abolished by TopBP1 siRNA (Figure 3B), suggesting that the regulation of Miz1 by Akt, at least in the control of p21 during DNA damage, is mainly mediated by TopBP1. A role of Miz1 for G1 arrest has been demonstrated in the context of DNA damage ( Seoane et al, 2002 Wanzel et al, 2005 ) and TGFβ signaling ( Seoane et al, 2001 ). The regulation of Miz1 activity by Akt–TopBP1 suggests that Miz1 may be under constant repression by Akt–TopBP1 during normal proliferation, similar to the control for E2F1-mediated apoptosis. The hypothesis is supported by the fact that Akt–TopBP1 is required to control the expression of p21, one of the Miz1 target genes, during normal growth (Figures 3A and 7E). This may explain the lack of measurable effect on gene expression in unstressed cells by depletion of Miz1 ( Wanzel et al, 2005 ).
Akt phosphorylation regulates oligomerization of TopBP1 protein
Akt generally regulates its substrates by altering their enzymatic activity or increasing the affinity of the substrates for 14-3-3 proteins, thereby retaining phosphorylated substrates in the cytosol. The nuclear localization of TopBP1 is not regulated by Akt, nor can we detect the interaction between 14-3-3 and TopBP1 in growing HEK293 cells (Supplementary Figure 6). We present evidence that phosphorylation of TopBP1 by Akt leads to direct self-association of TopBP1 BRCT domains. To our knowledge, this represents the first example that phosphorylation by Akt induces protein oligomerization as a regulatory mechanism. Considering that BACH1 phosphopeptide binds to the BRCT repeats of BRCA1 through specific interactions with a surface cleft at the junction of the two BRCT repeats ( Shiozaki et al, 2004 ), the TopBP1 BRCT6 domain alone may not be sufficient to foster stable interaction between TopBP1 and E2F1. The oligomerization of TopBP1 could bring two BRCT6 to proximity to stabilize its interaction with E2F1. It is not clear at this moment whether TopBP1 forms as dimers, trimers, tetramers or higher-order multimers within the cells. It would be very informative to analyze the multimeric structure of the TopBP1 complex. The structural basis for the role of phosphorylated Ser 1159 in TopBP1 self-association is also a very interesting question. The phosphopeptide-binding activity of BRCT78 toward a pSer 1159 -peptide suggests a model that phosphorylation of Ser 1159 creates a binding site for the BRCT78 of other TopBP1 molecules and induces oligomerization.
Although TopBP1 is involved in several aspects of DNA metabolism, protein oligomerization appears to be specifically required for its interaction with transcription factors. Given the conservation of Ser 1159 and its surrounding sequence in metazoan TopBP1, the Akt-dependent oligomerization is likely conserved as well. The carboxyl oligomerization domain of TopBP1 is not found in yeast homologs. Thus, it appears that both oligomerization and the function of transcriptional regulation were acquired in metazoan TopBP1 during evolution. It is tempting to speculate that Akt might selectively regulate the transcriptional function of TopBP1. Although the Akt phosphorylation site is not required for its binding to Rad9 or topoisomerase IIβ, it would be very interesting to test whether oligomerization of TopBP1 is involved in checkpoint activation or DNA replication. It is worth noting that the interaction between TopBP1 and Rad9 depends on the fourth and fifth BRCT motifs, and does not require the carboxyl oligomerization domain. TopBP1 has also been shown to be an ATR activator ( Kumagai et al, 2006 ). Although the ATR-activating domain (residues 978–1286 of human TopBP1) contains Ser 1159 residue, it only contains a portion of the seventh BRCT motif. Thus, it is unlikely that oligomerization is involved in this process. Whether Akt phosphorylation is involved in ATR activation warrants further investigation. The identification of Akt phosphorylation and TopBP1 oligomerization will greatly facilitate future studies in TopBP1 functions.
The role of TopBP1 self-association in E2F1 control may have therapeutic implications. Most cancer cells harbor excessive E2F1 activities due to de-regulation of the Rb pathway. Downregulation of TopBP1 induces E2F1-dependent apoptosis ( Liu et al, 2004 ) therefore, cells containing higher E2F1 levels would be more susceptible to TopBP1 inhibitors. The oligomerization domain of TopBP1 could be considered as a therapeutic target to de-repress E2F1 activity and potentiate E2F1-mediated apoptosis during chemotherapy.
Terrillon S and Bouvier M (2004) Roles of G-protein-coupled receptor dimerization. EMBO Rep 5: 30–34
Milligan G (2004) G protein-coupled receptor dimerization: function and ligand pharmacology. Mol Pharmacol 66: 1–7
Park PS et al. (2004) Oligomerization of G protein-coupled receptors: past, present, and future. Biochemistry 43: 15643–15656
Rocheville M et al. (2000) Subtypes of the somatostatin receptor assemble as functional homo- and heterodimers. J Biol Chem 275: 7862–7869
Pfeiffer M et al. (2001) Homo- and heterodimerization of somatostatin receptor subtypes. Inactivation of SST(3) receptor function by heterodimerization with SST(2A). J Biol Chem 276: 14027–14036
Grant M et al. (2004) Agonist-dependent dissociation of human somatostatin receptor 2 dimers: a role in receptor trafficking. J Biol Chem 279: 36179–36183
Monnot C et al. (1996) Polar residues in the transmembrane domains of the Type 1 angiotensin II receptor are required for binding and coupling. J Biol Chem 271: 1507–1513
Abdalla S et al. (2004) Factor XIIIA transglutaminase crosslinks AT(1) receptor dimers of monocytes at the onset of atherosclerosis. Cell 119: 343–354
Miura S et al. (2005) Constitutively active homo-oligomeric angiotensin II type 2 receptor induces cell signaling independent of receptor conformation and ligand stimulation. J Biol Chem 280: 18237–18244
Terrillon S et al. (2003) Oxytocin and vasopressin V1a and V2 receptors form constitutive homo- and heterodimers during biosynthesis. Mol Endocrinol 17: 677–691
Zhu X and Wess J (1998) Truncated V2 vasopressin receptors as negative regulators of wild-type V2 receptor function. Biochemistry 37: 15773–15784
Kroeger KM et al. (2001) Constitutive and agonist-dependent homo-oligomerization of the thyrotropin-releasing hormone receptor. Detection in living cells using bioluminescence resonance energy transfer. J Biol Chem 276: 12736–12743
Song GJ et al. (2005) Regulated dimerization of the thyrotropin-releasing hormone receptor affects receptor trafficking but not signaling. Mol Endocrinol 19: 2859–2870
Latif R et al. (2001) Oligomerization of the human thyrotropin receptor: fluorescent protein-tagged hTSHR reveals post-translational complexes. J Biol Chem 276: 45217–45224
Latif R et al. (2002) Ligand-dependent inhibition of oligomerization at the human thyrotropin receptor. J Biol Chem 277: 45059–45067
Urizar E et al. (2005) Glycoprotein hormone receptors: link between receptor homodimerization and negative cooperativity. EMBO J 24: 1954–1964
Calebiro D et al. (2005) Intracellular entrapment of wild-type TSH receptor by oligomerization with mutants linked to dominant TSH resistance. Hum Mol Genet 14: 2991–3002
Roess DA et al. (2000) Luteinizing hormone receptors are self-associated in the plasma membrane. Endocrinology 141: 4518–4523
Horvat RD et al. (2001) Luteinizing hormone receptors are self-associated in slowly diffusing complexes during receptor desensitization. Mol Endocrinol 15: 534–542
Osuga Y et al. (1997) Co-expression of defective luteinizing hormone receptor fragments partially reconstitutes ligand-induced signal generation. J Biol Chem 272: 25006–25012
Ji I et al. (2002) Cis- and trans-activation of hormone receptors: the LH receptor. Mol Endocrinol 16: 1299–1308
Tao YX et al. (2004) Constitutive and agonist-dependent self-association of the cell surface human lutropin receptor. J Biol Chem 279: 5904–5914
Ji I et al. (2004) Trans-activation of mutant follicle-stimulating hormone receptors selectively generates only one of two hormone signals. Mol Endocrinol 18: 968–978
Fan QR et al. (2005) Structure of human follicle stimulating hormone in complex with its receptor. Nature 433: 269–277
Cornea A et al. (2001) Gonadotropin-releasing hormone receptor microaggregation. Rate monitored by fluorescence resonance energy transfer. J Biol Chem 276: 2153–2158
Horvat RD et al. (2001) Binding of agonist but not antagonist leads to fluorescence resonance energy transfer between intrinsically fluorescent gonadotropin-releasing hormone receptors. Mol Endocrinol 15: 695–703
Brothers SP et al. (2004) Human 'loss-of-function' GnRH receptor mutants retain wild type receptors in the endoplasmic reticulum: molecular basis of the dominant-negative effect. Mol Endocrinol 18: 1787–1797
Berglund MM et al. (2003) Neuropeptide Y Y4 receptor homodimers dissociate upon agonist stimulation. J Pharmacol Exp Ther 307: 1120–1126
Dinger MC et al. (2003) Homodimerization of neuropeptide Y receptors investigated by fluorescence resonance energy transfer in living cells. J Biol Chem 278: 10562–10571
Mandrika I et al. (2005) Melanocortin receptors form constitutive homo- and heterodimers. Biochem Biophys Res Commun 326: 349–354
Biebermann H et al. (2003) Autosomal-dominant mode of inheritance of a melanocortin-4 receptor mutation in a patient with severe early-onset obesity is due to a dominant-negative effect caused by receptor dimerization. Diabetes 52: 2984–2988
Bai M et al. (1998) Dimerization of the extracellular calcium-sensing receptor (CaR) on the cell surface of CaR-transfected HEK293 cells. J Biol Chem 273: 23605–23610
Bai M et al. (1999) Intermolecular interactions between dimeric calcium-sensing receptor monomers are important for its normal function. Proc Natl Acad Sci USA 96: 2834–2839
Jensen AA et al. (2002) Probing intermolecular protein-protein interactions in the calcium-sensing receptor homodimer using bioluminescence resonance energy transfer (BRET). Eur J Biochem 269: 5076–5087
Pin JP et al. (2005) Allosteric functioning of dimeric class C G-protein-coupled receptors. FEBS J 272: 2947–2955
Patel RC et al. (2002) Photobleaching fluorescence resonance energy transfer reveals ligand-induced oligomer formation of human somatostatin receptor subtypes. Methods 27: 340–348
Rocheville M et al. (2000) Receptors for dopamine and somatostatin: formation of hetero-oligomers with enhanced functional activity. Science 288: 154–157
Terrillon S et al. (2004) Heterodimerization of V1a and V2 vasopressin receptors determines the interaction with β-arrestin and their trafficking patterns. Proc Natl Acad Sci USA 101: 1548–1553
Hanyaloglu AC et al. (2002) Homo- and hetero-oligomerization of thyrotropin-releasing hormone (TRH) receptor subtypes. Differential regulation of beta-arrestins 1 and 2. J Biol Chem 277: 50422–50430
Saveanu A et al. (2002) Demonstration of enhanced potency of a chimeric somatostatin-dopamine molecule, BIM-23A387, in suppressing growth hormone and prolactin secretion from human pituitary somatotroph adenoma cells. J Clin Endocrinol Metab 87: 5545–5552
Milligan G and Bouvier M (2005) Methods to monitor the quaternary structure of G protein-coupled receptors. FEBS J 272: 2914–2925
Eidne KA et al. (2002) Applications of novel resonance energy transfer techniques to study dynamic hormone receptor interactions in living cells. Trends Endocrinol Metab 13: 415–421
Förster T (1948) Intermolecular energy migration and fluorescence. Ann Phys (Leipzig) 2: 55–75
Suhling K et al. (2005) Time-resolved fluorescence microscopy. Photochem Photobiol Sci 4: 13–22
Maurel D et al. (2004) Cell surface detection of membrane protein interaction with homogenous time-resolved fluorescence resonance energy transfer technology. Anal Biochem 329: 253–262
Xu Y et al. (1999) A bioluminescence resonance energy transfer (BRET) system: application to interacting circadian clock proteins. Proc Natl Acad Sci USA 96: 151–156
Ban T et al. (1992) Specific antibody to the thyrotropin receptor identifies multiple receptor forms in membranes of cells transfected with wild-type receptor complementary deoxyribonucleic acid: characterization of their relevance to receptor synthesis, processing, structure, and function. Endocrinology 131: 815–829
Grossman RF et al. (1995) Immunoprecipitation isolates multiple TSH receptor forms from human thyroid tissue. Thyroid 5: 101–105
Graves PN et al. (1996) Multimeric complex formation by the thyrotropin receptor in solubilized thyroid membranes. Endocrinology 137: 3915–3920
Alberti L et al. (2002) Germline mutations of TSH receptor gene as cause of nonautoimmune subclinical hypothyroidism. J Clin Endocrinol Metab 87: 2549–2555
Refetoff S (2003) Resistance to thyrotropin. J Endocrinol Invest 26: 770–779
Persani L et al. (2004) Different forms of resistance to thyrotropin (TSH) action. In Syndromes of Hormone Resistance in the Hypothalamic-Pituitary-Thyroid Axis, 177–192 (Ed. Beck-Peccoz P) Norwell, MA: Kluwer Academic