Crystallisation of the β2AR-Gs complex

X-ray Crystallography is a fantastic technique, showing us a detailed insight into the structure of a protein. In conjunction with other methods, crystallography can allow us to form an accurate model of the protein in question. In this experiment, the B2AR was the target protein. X-ray crystallography was carried out on a B2AR-Gs complex in the hopes of understanding the structural basis behind its intricate signalling function. Crystallography is a well-established technique and there have been countless experiments carried out using it, on GPCR’s especially. However, this is not to say that one can simply follow a guide and form high quality diffraction crystals first time round. No. Every experiment has thousands of different parameters that can be changed and different chemicals that can be used, and every target protein prefers different conditions. The article in discussion mainly focuses on the problems encountered during crystallization of the B2AR-Gs complex and how these challenges were overcome in order to produce a high quality diffraction crystal end product.

 

Aim of the experiment

The aim of their experiment is to find the crystal structure of the active-state B2AR, bound to the Gs and compare it to previously obtained structure for inactive B2AR so that the 2 states can be compared and conformational changes can be detected in the B2AR in an attempt to understand how structure influences signalling. (Lower down I have interpreted the full methods(click here)used in the article into a slightly simplified protocol, but first I will discuss the main challenges they were faced with during crystallization and how they overcame them.)

Forming a stable B2AR-Gs complex in detergent solution was the first challenge encountered. 

Figure 2. Comparison of 2 detergent monomers. a. MNG-3, containing
a quaternary carbon structure (depicted by the cross) distinguishing
it from conventional detergents. This allows it to form a monomer comprising
of 2 hydrophilic head groups and 2 hydrophobic chains. This structure
is thought to give MNG-3 higher stabilizing abilities realtive toconventional
detergents, when purifying membrane proteins (ref1) b. DDM monomer

Why was this an issue? Well crystallization comes in two main steps; purification of the protein to obtain a pure sample for crystallization, and then crystallization itself. Purification and Crystallization require 2 very different environments, and it’s the transition between the two steps that can often lead to destabilization of the protein. In this case, the B2AR is a membrane protein and a relatively harsh detergent must be used to purify it from the membrane. However, after the protein is purified using this harsh detergent, it is no longer in its native membrane environment and must be exchanged to a less harsh detergent so that it does not immediately unfold. As you can see, it’s a detergent balancing act. The detergent must be harsh enough to purify the protein yet still provide a stable environment to retain the proteins integrity.                                      

Figure 3. Structure of a detergent micelle. Detergents can

be found as monomers, but when added in excess they

form micelles. 

  

How are detergents able to purify membrane proteins? Detergents achieve purification by disrupting the bonds formed between the target protein and the membrane, but care must be taken to not use a very harsh detergent, such as SDS, which would also disrupt intramolecular bonds within our protein, destroying its integrity. Detergents, like lipids, are amphiphiles; they comprise of a polar hydrophilic head-group and a non-polar hydrophobic hydrocarbon chain. When added in excess, detergents form micelles (figure 3), like lipids, an it is these micelles
that are used to disrupt membrane 
interactions with the target protein,
purifying it.

Figure 4. Structure of a detergent micelle with the protein,
that it purified from the membrane, trapped inside it.

After purification, the proteins are held in these detergent micelles, that mimic the lipid environment (figure 4), but often the environment in the micelle is too harsh for the protein, leading to destabilization and unfolding. Some detergents are better at stabilizing the target protein than others, for example in this experiment the DDM detergent used initially for purification was too harsh for B2AR so the purified complex was exchanged into the less harsh, more stabilising MNG-3 detergent (figure 2).

The next problem encountered was protein shielding by large detergent micelles

A further issue that is often encountered with detergents is a high level of shielding of the protein from other proteins by the micelle. Detergent micelles can be relatively large and therefore shield all the hydrophobic surface of the protein and the majority of the extracellular hydrophilic surface. This is ideal for purification as it shields the protein away from the membrane, but not so ideal for crystallisation where the protein must be relatively exposed in order to form crystal-crystal contacts. There are many ways that this problem can be overcome and 2 were tested in this experiment. Initially, they attempted using antibodies specific to the complex to stabilise the protein and extend its hydrophilic surface to aid crystal-crystal contacts. However, this did not work so they tried another method. The idea was to genetically engineer a T4-lysozyme, a highly crystallisable globular protein, to the B2AR in order to aid stabilisation in the detergent micelle and extend its polar surface. This second method was successful, discussed in more detail in the methods section. 

Methods for purification, stabilization and crystalization of the B2AR-Gs complex

B2AR

• They were able to stabilize the B2AR complex using T4-lysozyme which was amino-terminally fused to the B2AR. This was accomplished by genetically engineering an amino-terminally fused T4-L-B2AR construct and infecting Sf9 insect cells with the construct via a recombinant baculovirus. The culture was left to grow an express the B2AR complex, it was then solubilized in the detergent DDM.


• Following this, various purification methods were used to try and isolate the B2AR complex. These included flag affinity chromatography followed by alprenolol-Sepharose chromatography. During a proceeding chromatography step the alprenolol bound to the receptor was exchanged for a high affinity agonist (BI-167107). This agonist-bound receptor was then eluted with buffer and treated with lambda phosphatase. Throughout experimentation, the purified receptor was analysed by SDS-PAGE/coomassie brilliant blue staining.

Gs

• The alpha, beta and gamma subunits of the Gs-protein were expressed in insect cells via individual viruses for each of the 3 components. The cultures were left to incubate, harvested and then underwent multiple centrifugation and re-suspending steps, with the addition of various chemicals, until purified Gs was obtained.    

Nanobody (Nb35)

• Nb35 was expressed in the E. coli strain WK6, extracted and purified by nickel affinity chromatography.

Complex formation, stabilization and purification

• The purified Gs heterotrimer was mixed with a molar excess of the agonist-bound T4L-B2AR and buffer (containing DDM detergent) and left to incubate at room temperature. BI-167107 was chosen as the agonist in this experiment over others due to its long half-life which allowed the complex to be stable for longer.


• Apyrase was then added in order to hydrolyse any released GDP from the complex. This is essential as GDP can distrupt the high affinity interaction between B2AR and Gs, destabilizing our complex. The GMP formed due to GDP hydrolyses has a low affinity for Gs and therefore would not cause dissociation of the complex.


• So far, DDM was the detergent of choice, however, analysis showed that DDM did not stabilize the complex well enough and there was significant dissociation. After experimenting with different detergents, MNG-3 was identified as a substantially better stabilizing detergent, therefore, the complex was exchanged into MNG-3.


• After this process, the mixture contained B2AR-Gs complex and free Gs/B2AR. Proceeding purification steps were undertaken in order to separate out the complex from the free proteins until a final yield of purified B2AR-Gs complex was obtained.

Genetic engineering of the T4-lysozyme to the B2 Adrenergic Receptor

• In a bid to increase stability of the protein complex, antibodies specific to the receptor were created and added. Unfortunately, this method did not work so they took a different approach. Using genetic engineering, the flexible, unstructured amino terminal of the B2AR was replaced by a globular T4-lysozyme, with a linker, containing a protease recognition sequence, between them (figure 5). The amino terminus resides in the extracellular side of the receptor and therefore the T4-lysozyme resides there too. The T4-lysozyme, a highly crystallisable protein, stabilizes the protein by extending the receptors extracellular polar surface, allowing it to form more crystal-crystal interactions and stabilizing it on transfer from the detergent micelle to the LCP. 

Figure 5. Crystal structure of the active state agonist bound B2AR with T4-L N-terminally attached. Everything else in the complex has been hidden for clarity. From the crystal structure, it looks as if the T4-L and B2AR are 2 different entities, however, they are not. The linker connecting the two structures together is far too flexible to be seen by X-ray crystallography, however, the link can be seen using electron microscopy analysis. 

Stabilisation with nanobodies

• When tested, the B2AR crystals formed were of low quality. After analyses of the crystals, by electron microscopy imaging, they observed that the alpha-helical domain of the GalphaS subunit was flexible, and it was suggested that this was a possible cause for poor crystal quality. In an attempt to stabilize the alpha-helical subunit, RNA encoding specific nanobodies were obtained from llamas after immunization with the target B2AR-Gs-agonist complex. This RNA was used to prepare cDNA which allowed them to construct a Nanobody library from which they isolated Nb35 which proved to stabilize the alpha-helical Gas subunit best (figure 1 and 2 under the 'Structure of activated G2' tab).

Crystalization

• The agonist-bound T4L-B2AR-Gs complex was mixed with Nb35 and incubated, allowing agonist-bound T4L-B2AR-Gs-Nb35 complexes to form. As shown earlier, the complex was previously exchanged into MNG-3. MNG-3 has a high affinity for the receptor, allowing it to maintain structural integrity even when diluted below MNG-3’s CMC and providing the complex with great stability during it’s exchange into the LCP. MNG-3 was chosen over DDM, because experiments showed that the complex unfolded rapidly in DDM, unlike in MNG-3.


•  40nl drops of the LCP-protein mixture were delivered into 24-well glass sandwhich plates by an LCP dispensing robot and left to crystallise. The crystals reached their maximum size after 3-4 days, were collected and analysed. After refinement, diffraction data was obtained from these crystals at an overall resolution of 3.2A, and was delivered to a phaser which solved the crystal structure by molecular replacement. We created a short movie of the T4L-B2AR-Gs-Nb35 complex, using the crystal structure obtained in the experiment (video 1).

Video 1. A short movie showing a full rotation of the T4L-B2AR-Gs-Nb35 complex. 

Allowing you to get a better idea of the structure!

References used on this page:
Ref 1. Structure of MNG-3 was taken from http://www.nature.com/nmeth/journal/v7/n12/fig_tab/nmeth.1526_F1.html