Structural studies of GRK5

A conversation with JBC Paper of the Week authors

Ribbon representation of GRK5·AMP-PNP crystal structure. Full-length GRK5(1–590) was crystallized, and residues 15–543 are clearly resolved (the first and last residues are labeled).

The Journal of Biological Chemistry featured two Papers of the Week in August about structural studies of G-protein-coupled receptor kinase 5, or GRK5. The studies (Benovic et al and Tesmer et al) were authored by two groups that conducted their work separately and then later discovered that their structures of GRK5 were in agreement.

The JBC’s podcast host, Diedre Ribbens, interviewed the corresponding authors of the papers, Jeff Benovic of Thomas Jefferson University and John Tesmer at the University of Michigan, to hear more about their work on the GRK5 protein and how these two structures have affected the GRK field. Here, we’ve reprinted some of their conversation. You can listen to the full podcast or read the full transcript here.

Benovic and Tesmer both got their start studying GRKs with Robert Lefkowitz, a Howard Hughes Medical Institute investigator at Duke University Medical Center whose group pioneered the GRK field. (Lefkowitz won the Nobel Prize in chemistry in 2012.)

DIEDRE RIBBENS: G-protein-coupled receptor kinases, or GRKs, are a family of protein kinases that have a role in the desensitization of G-protein-coupled receptors. In particular, GRK5, one of the most widely expressed proteins in the GRK family, has been implicated in several diseases, such as cardiovascular disease, cancer, diabetes and Alzheimer’s. Although structures have been solved for other GRKs, such as GRK2 and GRK6, no structure has been solved for the clinically relevant GRK5. The study by Benovic’s group was able to resolve the structure of human GRK5 at a resolution of 1.8 Å in complex with either the ATP analog AMP-PNP or the nucleoside sangivamycin. The Tesmer group study also solved the GRK5 structure to a 2.4 Å resolution but complexed with an inhibitor called CCG215022. Importantly, the two groups realized that their structures captured GRK5 in a strikingly similar conformation at its C-terminus that was unique to this enzyme and not found in other GRKs. Ultimately, these structures will enable future studies probing the function of GRK5 and possibly lead to the design of selective inhibitors.

In my interview with Benovic and Tesmer, they shared how they became interested in studying GRKs, how their work on GRK5 addressed previously unknown questions in their field and where each of their research groups would like to go next.

BENOVIC: Well, for me, it actually goes back to my graduate work in the mid-’80s (when) I worked with Bob Lefkowitz, and we were trying to identify the kinase that phosphorylates the β2adrenergic receptor in an agonist-dependent manner. And then this … ultimately led to the cloning of cDNAs for what are now called GRK5 and GRK6. We published that in 1993, and then we really spent, not full time, but a good portion of our time over the last 20 years trying to understand how this kinase functioned.

RIBBENS: Tesmer also crossed paths with the Lefkowitz lab during his postdoc. As a collaborator, he worked on GRK2, which eventually led him to start his own research group focused on GRKs.

BENOVIC: (GRK5 is) one of the few kinases that John’s group hasn’t crystallized, but it’s interesting because it’s been implicated in a number of human diseases, including various cardiovascular diseases, prostate cancer, diabetes and a number of neurological disorders. So I think it’s an interesting enzyme and interesting potential therapeutic target.

TESMER: When I started my lab in the late ’90s, my initial project was to structurally characterize GRK2 … and we were interested in it in principle because it interacted with a lot of G-protein subunits, and that was my specialty as a postdoc. And, since then, I actually collaborated with Jeff to look at other members of the GRK family, and one that we worked together on was GRK6, which we published back in 2006. And GRK5 is a member of that subfamily to which GRK6 also belongs.

More recently, we decided to look at these kinases more in a translational sense because we had a good feel for what they looked like structurally; and of course GRK2 was one of them because of its involvement in heart failure; and then we also turned back to GRK5 because, as Jeff pointed out, it’s involved in a host of diseases including cardiac hypertrophy. And so it’s been actually something that we’ve been working on for a number of years and looking for the structure of, which we finally got to be able to publish with Jeff this year.

RIBBENS: Tesmer and Benovic crossed paths often, leading to the occasional collaborative study.

TESMER: I think the people in our field, at least I feel, have a habit of getting together when they need to and going their own ways when they feel like they’re able to … Jeff, I think we first started working together in 2003 or so …

BENOVIC: Right. Yeah. We had certainly had our own efforts actually in trying to crystallize GRK2, and they didn’t really go anywhere, and then that was a major focus of John’s group. And a former trainee of mine, Rachel Sterne-Marr – I kind of linked her up with John to help with some characterization of GRK2. So that was kind of our initial interaction on the crystallization side of it, and then we worked somewhat on GRK6, but it was really us just providing reagents to John to do the work – to facilitate the work. You know, so we’ve collaborated over the years. I think we’re trying to do some similar things, and a lot of that will be collaborative, and some will kind of be trying to do some other things and publishing on our own too, I’m sure.

TESMER: Mm hmm. It’s all good.

A, comparison with the GRK6 sangivamycin complex (PDB entry 3NYN) (13). CCG215022 (spheres with yellow carbons) binds in the active site of the GRK5 kinase domain. As in most other GRK structures, the AST region is disordered (last visible residues denoted by asterisks). The C-terminal region of GRK5 (royal blue) has a dramatically different conformation than observed for GRK6 sangivamycin (brown), despite the fact they are closely related enzymes (see C). Key residues in the C terminus are labeled to emphasize how they contribute to packing in each structure. Side chains shown with beige carbons are from the RH domain of GRK6 (same identity and numbering as in GRK5). GRK6-Pro-547 is analogous to GRK5-Pro-546, which was mutated in this study. B, close-up view of the interactions between the C-terminal region (royal blue) and the RH domain (green). Hydrogen bonds/salt bridges are shown as dashed lines.

RIBBENS: Benovic admitted that his group had never undertaken a solo effort to solve a GRK crystal structure before, but GRK5 seemed to present the perfect opportunity to make an attempt at doing so.

BENOVIC: Really, the field, at least the GRK crystallography field, which has given tremendous insight — in terms of how these proteins … function, how they fold — was almost completely driven by John’s work. He crystallized GRK2, GRK2 in various complexes, GRK1 and GRK6 and has multiple papers and has gained a lot of insight.

For us, the challenge was that we really had never tried to crystallize a GRK alone … But we thought really one of our goals was to better understand how GRKs interact with GPCRs and how this results in activation of the GRK … GRK5 was a very well-behaved protein, so we thought it might be a good model to really pursue that effort. And then when … a senior postdoc joined my lab a few years ago — Konstantin Komolov, who had a lot of experience with GRK1 — we decided that he would focus on trying to crystallize GRK5. And this paper for us is the culmination of … his initial efforts in this area.

And it turns out that he actually got crystals almost immediately once he purified enough protein … Then, once we had that, we really just used John’s GRK6 structure to do molecular replacement to solve the structure of GRK5.

RIBBENS: Tesmer’s group was going at the problem from a slightly different angle, as they wanted to solve the structure of GRK5 so his group could focus on the design of specific inhibitors of the enzyme.

TESMER: We decided to tackle the problem of developing selected inhibitors for GRKs, and there’s a need for this in a couple different camps, one of which is that a lot of electro-pharmacologists would like to know which of these particular kinases are responsible for certain phenotypes in cells, and there’s no chemical probes out there that are very good, in my opinion, that are selected for individual members of this family.

And, of course, there are unresolved issues … GRK5 is very closely related to GRK6, and there’s a bit of a controversy on what’s going on with its C-terminal structure, and its C-terminus is very important for its membrane targeting in cells. In prior structures of GRK6, it appeared to be in a conformation that wouldn’t permit it to interact with membranes, and so that was an unanswered question as to what was really going on in the C-terminus of the subfamily of GRKs.

RIBBENS: I asked the two researchers at what point they became aware of their simultaneous efforts to obtain a GRK5 crystal structure and how they came to publish in the same issue of the JBC.

BENOVIC: We certainly discussed it initially at an (American Society for Biochemistry and Molecular Biology) meeting a few years ago. We had had the structure at that point, and I’m not sure how far along John had been on his structure, whether they had their complex or not at that point … And then we’ve been in a few meetings together and have talked in more detail about it.

TESMER: Yeah, that’s exactly right. And I have to thank Jeff. I think his structure was done earlier than mine, and I think he actually delayed his publication so that we could resolve our structure fully and write the paper and publish them together, which I’m very thankful for.

RIBBENS: Working with the human GRK5, Benovic’s group seemed to solve the GRK5 without too many roadblocks, but that was not the case for Tesmer’s group working with the bovine form of the enzyme.

TESMER: The truth of the matter is we’ve been working on GRK5 for quite some time, and we had actually given up on it. And, of course, kudos to Jeff for getting it done with the human enzyme. We’re working with bovine. And as a consequence of our drug-design efforts, a postdoc in my lab, Kristoff Homan, noticed that one of the inhibitors that we had rationally designed as a GRK2 inhibitor … increased the thermostability of GRK5 enormously …

And that worked almost instantly after — I don’t know — four years of trying to crystallize it otherwise, and so that led us to the structure. And what the structure enabled us to do is look at how this drug interacts with GRK5. It verified our rational design strategies. We’re very pleased with that.

RIBBENS: Ultimately, publishing back-to-back papers allows the work of both groups to garner equal visibility from the field and draws attention to the important similarities between their two structures.

TESMER: The neat thing was that we both resolved the same C-terminal structure, and I believe I speak for both of us in that we both believe that this is the proper confirmation of the C-terminus when this enzyme would be interacting with a membrane. And that’s really the power of the two papers, I think. So often in crystal structures, the flexible parts or the movable parts and frankly the interesting parts end up trapped in crystal contacts or in weird conformations, and it’s sometimes hard to figure out if it’s functional or not. And when you get two independent structures, completely different crystal forms, it’s really a powerful confirmation that you’re looking at the right thing.

Listen to more JBC podcasts.

Diedre Ribbens Diedre Ribbensis is a science writer, educator and communicator based in Minneapolis. She earned her Ph.D. at Johns Hopkins School of Medicine.