Fundamental structural biology questions for the future:

1. What are all the conformational states of macromolecules and their relative energies, how do these depend on their environment and ligands, and how can we use these to model their properties in detail?

2. What are the transient interactions between macromolecules inside and outside cells and how do these influence their functions?

3. What are the networks of signaling in a cell?

4. What is the importance of supramolecular structures, assemblies, and interactions in the functioning of cells?

5. What are the structures of all macromolecules and of all their interactions with each other and with other molecules?

Fundamental structural biology questions for the future:

1. What are all the conformational states of macromolecules and their relative energies, how do these depend on their environment and ligands, and how can we use these to model their properties in detail?

2. What are the transient interactions between macromolecules inside and outside cells and how do these influence their functions?

3. What are the networks of signaling in a cell?

4. What is the importance of supramolecular structures, assemblies, and interactions in the functioning of cells?

5. What are the structures of all macromolecules and of all their interactions with each other and with other molecules?

Great Answer! What do you think about the possible roles of complexes in coordinating steps and pathways and avoiding destructive interference? Is this likely to be a big or small deal in cells? This pertains in part to your point 4 but I am asking if pathway coordination and regulation might emerge either partly or primarily as an intrinsic property of component dynamic supramomolecular structures.

I think it will be fascinating to find out how much regulation and coordination happens because of physical proximity and shielding of intermediate products. As in your original question we will need new methods for detection of proximity and new tools for modeling the environments in cells to see how important this is! At a basic level an experimental question might be..."How much correlation in location is there between members of each possible pair (or group) of macromolecules?"

You both, I think, correctly focus on the issue of the structure/function relationship in vivo. I think this is the great question for biology as a whole, not just for structural biology: What is it really like in there?

While I think that is indeed the major scientific challenge for structural biology, I would also point out a cultural one:

How do we keep structural biology interesting? As the supply of structures increases dramatically, the value of any one structure decreases, and so does the value of those who only do structures. The commoditization of structural biology requires that those who care about structure and function and wish to remain relevant must become adept at much more than structure determination.

I agree, cells are complex and open systems and the new results from structural biology should come integrating a vast repertoire of techniques: molecular biology, biophysics, imaging from low to high resolution and theorethical models. Only in this way the emerging principles that rule the organization and function of a cell or of multiple interacting cells can be characterised possibly also in different spatial and temporary terms. Therefore not only structures, but a morpho-dynamic and quantitative description is already giving and will give more relevance to structural biology results. Naturally it is difficult to realize this for many labs, collaboration has to play a big role to go deeper in cell organization at different levels.

There is a basic tension between the desire for high-resolution descriptions of protein conformations and the desire for explicit descriptions of mechanism or function. All biological functions involve changes in either protein conformation, its surrounding milieu, or both; this means that the coherent reconciliation of dynamics and 'resolution' (atomic-level spatial precision) is a major conceptual challenge to structural biology.

I'm excited about a) better recipes for integrating NMR observations at multiple timescales, b) better sampling schemes for simulation approaches, and c) single-particle approaches, which *force* the experimenter to grapple with the single-conformation description of protein function.

There are still a lot of challenges on the protein side of structural biology, especially for membrane proteins. Some enzymes such as some lipases and enzymes involved in lipid synthesis are functional only in the environment of microsome or liposome with proper lipid composition. NMR and simulation will provide another window for protein dynamics. I especially interest to Ryan’s third point, forcing or locking the single-conformation with function. The rapid development of monoclonal antibody engineering and related protein engineering technologies may provide some exciting tools to structural biology.

The structures of the "so-called" intrinsically unstructured proteins that (may) adopt a secondary structure transiently to interact with its partner(s) will be challenging and interesting.

To add a small aspect not yet included in the above discussion:

Initially the hope was that high-resolution (X-Ray, NMR) structures of target proteins will allow "easy" structure based drug design (SBDD) of e.g. inhibitors

Up to now there are only rare examples where this really worked, telling us that a picture with a single rigid structure is not appropriate. This is why high-throughput screening still is so important in pharmaceutical research.

It was mentioned above that NMR provides in-depth, information on flexbility/plasticity of target proteins. However this flexibility information has not been used convincingly for SBDD.

For me the "efficient" transfer of structural/flexibility information of target proteins to drugs still would be very important!

It is clear that all this is related to function.

I agree with you. We have developed an anti-gout drug, febuxostat, that is a structure-based drug (see J.Biol.Chem., 278.1848-1855,2003) compare to allopurinol that is so called a kind of suicide inhibitor (see our paper in Analogue-based Drug Discovery III, Wiley 2003), but it was developed by screening various compound by try and error way in classic way. However, it was found that febuxostat does not inhibit practically the bacterial enzyme that has very similar three dimensional structure and docking study could not tell us why it does not inhibit efficiently. We performed MD simulation that showed difference in mobility of enzyme cavity ( Kikuchi et al., Sci. Res. 2, 331, 2012). We may be able to design the compound using MD simulation, but it is still practically try and error method.

These are some great answers that certainly highlight many of the most pressing challenges currently facing structural biologists. I think it is becoming increasingly clear that, as we learn more about the organization of proteins in their native cellular environment, that very few proteins act alone. To truly understand how structure relates to function in vivo we will have to begin paying more attention to protein complexes (both stable and transient) and the inherent structural changes and modifications that control their assembly and disassembly. This will no doubt require us pushing forward the technological frontiers of established methods but will also require the development of new tools and the cross-platform integration of these tools. As we begin to learn more about large assemblies of biomolecules and macromolecular machines, the marriage of traditional structural biology techniques with other approaches such as single molecule methods and super-resolution microscopy will enable us to answer questions at that resolution regime between the atomic scale and the micro-scale that we know so little about.

The ideas on the marriage of traditional structural biology with single molecule and other new methods and of keeping structural biology relevant go together in suggesting paths going forward. This is hopefully something that will find increasing support.