Treating achondroplasia: the role of Tyrosine Kinase Inhibitors

Introduction

Achondroplasia is caused by a mutation in the gene that encodes (contains the chemical information to create) the protein called fibroblast growth factor receptor 3 (FGFR3). FGFR3 is a protein that is located across the cell membrane of the chondrocytes (Figure 1). It has three parts or domains: the part outside the cell that works like an antenna, receiving signals from FGFs; the transmembrane domain, which "anchors" the protein across the cell wall; and the intracellular domain, which is responsible for spreading the signal received outside the cell to other proteins inside the cell. 

Figure 1. FGFR3 structure

Proteins are large and complex molecules constituted by smaller parts called amino acids. These amino acids are like pieces in a Lego model: each one has distinct features which, combined with others, confer special shapes and properties to the protein (Figure 2). The right combination of amino acids is therefore essential for the normal function of the protein. If you are curious about amino acids, just visit the Glossary page of the blog for some basic information or go to Wikipedia)

Figure 2. Combining Lego parts

The right combination of amino acids in a protein is crucial for its proper function. In achondroplasia a single substitution (the mutation) of one of these amino acids gives more power (activity) to FGFR3 than it was supposed to have. The normal function of FGFR3 is to reduce the speed of bone growth at the level of some tiny structures in the end of the long bones called growth plates. Because of the mutation, FGFR3 is working too much and causes a significant impairment in the ability of the chondrocytes living in the growth plates to multiply and mature, leading to the clinical features of achondroplasia.

Why is that so?

We refer to proteins that participate in chemical reactions in the body as enzymes. Enzymes are very active and FGFR3 is not different. To control the enzymes' activities the cells have "cleaning" (janitor) systems that allow those enzymes, when turned on (or activated), to work only for a limited period of time. In achondroplasia, the mutation makes FGFR3 more stable and "resistant" to these janitors, so it keeps working longer.

Well, what does it have with the title of this article? 

Calm down, we are getting there.

An enzyme works by transferring electric charges to another protein (a client that could be another enzyme), causing changes in the client. The client will take the electric charge to another destination, for instance, another protein, a given spot inside the cell nucleus, etc. It will depend on their function(s). Think in a domino chain, a piece pushing the next one, etc. This is the way cells work. To help you understand enzymatic chemical chain reactions or signaling cascades, I invite you to watch the fine animation below provided by Cold Spring Harbor Laboratory that I published in a previous article of the blog (Animation 1). 

Animation 1. This animation shows how the body starts the healing process after a skin bruise. The relevant part starts at 2:50 min, showing how a ligand (think on a FGF) binds to a receptor enzyme on the surface of a cell (think on FGFR3) and starts the signaling or chemical cascade.

DNA Learning Center by Cold Spring Harbor Laboratory

Enzymes are capable to transfer those electric charges because some of the amino acids in their structures have the right chemical properties to do so. In the case of FGFR3, there are some spots in its part that is located inside the cell occupied by the amino acid tyrosine (see Figure 1). Tyrosine can bind to some reactive phosphorus carriers (ATP) fluctuating nearby inside the cell but will only do so if FGFR3 is activated (turned on). When FGFR3 is turned on outside the cell (did you see the animation 1?), through several complex shape changes, it exposes the tyrosines in its intracellular part, which in turn attract the phosphorus nearby, starting a series of chemical reactions. You must take in account that I am simplifying a lot the "activation" phenomenon here. 

The regions where the tyrosines are located in FGFR3 are called  "ATP pockets". These boxes are well used by many other enzymes with many different functions in the cells, so these structures are not exclusive of FGFR3. We also call these regions tyrosine kinase domains (TKD) (Figure 1). 

Ok, finally we see a connection with the title!

For decades, scientists know about these ATP pockets. They found that in cancer many of the enzymes like FGFR3 are used by the malignant cells to let them grow more and faster, so they started to search for approaches that would be capable to block them. In theory, without the chemical reactions provided by enzymes like FGFR3, many cancer cells would die and the disease could potentially be controlled.

In fact, scientists have already developed a lot of strategies to achieve this goal. One of them is exactly through creating molecules that can bind the TKD, preventing the enzyme to start chemical reactions. These molecules are called tyrosine kinase inhibitors (TKI). We have already reviewed many of those which have activity against FGFR3 here in the blog, but in Table 1 you will see a more recent list of them.

 Table 1. Some recent TKIs with activity against FGFRs.

To better understand how these small molecules work I invite you to watch this rich animation showing the mechanism of action of lapatinib (Animation 2), a TKI developed to block EGFRs, another class of enzymes that work in a similar way FGFRs do. You will see that lapatinib targets the ATP pockets in the intracellular part of EGFR.

Animation 2. Lapatinib mechanism of action (audio in Italian).
 

Using TKIs in achondroplasia?

Since there are so many molecules capable to block the activity of FGFR3, why don't we see more research addressing their use in achondroplasia? These drugs can be taken as pills once a day. Wouldn't it be perfect? The problem is that, as mentioned above, the ATP pockets (the tyrosine kinase domains) are very similar in a large number of families of enzymes that work like FGFRs. Did you notice that I didn't write FGFR3 in the Table 1 title? This is because neither of those compounds listed there are specific for FGFR3. They can also block the activity of other FGFRs and other cell enzymes. To illustrate this feature of the TKIs, take a look into the enzyme families' map below (Figure 3), showing the groups some TKIs are capable of blocking.

Figure 3. Kinome map showing enzymes affected by some old TKIs. Every colored point means that the given TKI will have some action on that enzyme. Note the way the map is designed, like a tree seen from above.

Note the large number of enzymes affected by these TKIs. Picture from: Stjepanovic N & Capdevila J. Biologics: Targets and Therapy 2014. Open access. Reproduced here for educational purposes only.

Affecting several different enzymes and their functions, at the same time, may pose a huge problem. Several undesired effects may arise when these drugs are used to treat cancer. However, some of the adverse effects are tolerable or acceptable because of the nature of the disease and also because, although affecting several enzymes at the same time, TKIs tend to be less toxic than the older chemotherapy arsenal that is still being used today to fight cancer.

In fact, many TKIs like these are already being used in the clinic to treat several forms of cancer, but only in adults. With the promiscuous activity they have we can predict TKIs could be dangerous to use in the growing bodies of children. These enzymes participate in fundamental development processes and blocking them in a child could cause many bad consequences.

Researchers like Dr Moosa Mohammadi, from University of New York, and Dr Kalina Hristova, from Johns Hopkins University, have been working hard to better understand the FGFR group of enzymes and the ATP pockets. They create computerized 3D models of the enzymes, trying to map where the amino acids are located so they can predict how these enzymes work and help new molecules to be developed to fit in the right places, but this is not easy to do. You can see the work by Dr Mohammadi doing a simple search in Pubmed.

Current anti-FGFRs cannot be used to treat achondroplasia

Perhaps because of the predicted risks it is unlikely that the current TKIs could be used in achondroplasia. There is evidence in lab tests that they work (1) but, because of their lack of specificity, TKIs are unlikely to reach the clinic for indications such as achondroplasia.

To bring more data about TKIs' utility for achondroplasia, the group leaded by Dr. Pavel Krejci, an enthusiastic researcher in the FGFR3 area, has just published a compelling study where they explored exactly the use of some of these TKIs known to be more specific to the FGFR family in chondrocytes and bone cultures (2). Some of the TKIs tested in their work are listed in Table 1 above. After performing a long series of tests, both in cell cultures, bone culture and animal models they concluded that the current TKIs in development would be risky to use in achondroplasia exactly because they block other enzymes important for the development of the growing body.

Exclusivity is difficult to achieve

To understand why these molecules created to block the ATP pockets affect several enzymes at the same time it is good to go back to the enzymes' map above and look their distribution. You will note that there are some family stems that arise from a main one, like in a tree. This means that although there are hundreds of different enzymes today, they all came from few ancestors, so they carry a lot of similarity, which is called by scientists as homology. The four FGFRs are very similar in structure, with more than 50% of their structure shared among the four members (3).

As pointed by several investigators, including Dr. Krejci, a successful molecule of the TKI group for the treatment of achondroplasia will have to be very selective for FGFR3, with no effect in the sister proteins and in enzymes of the other enzyme groups. This will take some time more to achieve, and will be based, at least in part, in perfecting the ability to design computer models of enzymes able to simulate their natural state and dynamics in the living cell. 

This just means that we need to be open to the several options available to tackle overactive enzymes like FGFR3 in achondroplasia. ATP pockets have been the most researched targets because they are easy to work with. However, there are other sites in the protein structure that can be approached. I invite you to review previous articles in the blog. You will find that there are many ideas around, waiting for a developer not thinking only in the low hanging fruit.

References

1. Jonquoy A et al. A novel tyrosine kinase inhibitor restores chondrocyte differentiation and promotes bone growth in a gain-of-function Fgfr3 mouse model. Hum Mol Genet 2012;21(4):841-51. Free access.

2. Gudernova I et al. Multikinase activity of fibroblast growth factor receptor (FGFR) inhibitors SU5402, PD173074, AZD1480, AZD4547 and BGJ398 compromises the use of small chemicals targeting FGFR catalytic activity for therapy of short stature syndromes. Hum Mol Genet 2015 Oct 22. pii: ddv441. [Epub ahead of print].

3. Ho HK et al. Current strategies for inhibiting FGFR activities in clinical applications: opportunities, challenges and toxicological considerations. Drug Discov Today 2014 Jan;19(1):51-62.