About this Author
DBL%20Hendrix%20small.png College chemistry, 1983

Derek Lowe The 2002 Model

Dbl%20new%20portrait%20B%26W.png After 10 years of blogging. . .

Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases. To contact Derek email him directly: Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
Not Voodoo

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
A New Merck, Reviewed
Liberal Arts Chemistry
Electron Pusher
All Things Metathesis
C&E News Blogs
Chemiotics II
Chemical Space
Noel O'Blog
In Vivo Blog
Terra Sigilatta
BBSRC/Douglas Kell
Realizations in Biostatistics
ChemSpider Blog
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Eye on FDA
Chemical Forums
Symyx Blog
Sceptical Chymist
Lamentations on Chemistry
Computational Organic Chemistry
Mining Drugs
Henry Rzepa

Science Blogs and News:
Bad Science
The Loom
Uncertain Principles
Fierce Biotech
Blogs for Industry
Omics! Omics!
Young Female Scientist
Notional Slurry
Nobel Intent
SciTech Daily
Science Blog
Gene Expression (I)
Gene Expression (II)
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net

Medical Blogs
DB's Medical Rants
Science-Based Medicine
Respectful Insolence
Diabetes Mine

Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem

Politics / Current Events
Virginia Postrel
Belmont Club
Mickey Kaus

Belles Lettres
Uncouth Reflections
Arts and Letters Daily
In the Pipeline: Don't miss Derek Lowe's excellent commentary on drug discovery and the pharma industry in general at In the Pipeline

In the Pipeline

« Do Scientists Read Business Books? (Plus: Add a Circle to Dante's Inferno!) | Main | A Med-Chem Book Recommendation »

February 11, 2009

Kinases: Hot or Not?

Email This Entry

Posted by Derek

For the last ten or fifteen years, untold amounts of time and money have been spent developing drugs to inhibit kinase enzymes. Just go take a look at KinasePro’s archives; that’ll give you the idea. Huge programs have been run at all the major drug companies, and any number of smaller ones have been founded just on the strength of one kinase inhibitor or another.

The enthusiasm isn’t hard to understand. For those of you outside the med-chem / biochem worlds, kinase enzymes are there to stick phosphate groups into other molecules, which is a very widely used signaling pathway. A phosphate completely changes the character of the part of a molecule where it’s attached, changing what other partners it will recognize and bind to. This takes place generally on to some sort of free OH group. That doesn’t narrow things down much, though, since there a lot of incredibly important small molecules with OH groups that get phosphorylated. Adding to the fun, several amino acids (serine, threonine, and tyrosine) have OH groups on them, and the means that nearly every decent-sized protein has plenty. The patterns of their phosphate groups turn their activities on and off, determine where they go and what they’ll recognize. It’s a major, major switching mechanism for protein activity – you can’t overstate its importance. Here's the classic family tree of the protein kinases, just to give you the idea. (And in case you’re wondering, there is indeed a whole different class of enzymes, the phosphatases, that take the things back off again - whole different bag of snakes, those guys).

There are hundreds and hundreds of kinase enzymes, and I think it’s safe to say that they’re involved in just about every important biochemical process you can think of. The downside of working on them is that, well, they’re involved in just about every important biochemical process you can think of. (Try this on for size, or this, to get the idea). How do you get them to do what you want?

Well, we’re still not sure about that. I go back far enough to remember when kinases were considered nearly impossible to work with as drug targets, because no one could figure out how you’d get selectivity. But once we figured out how to make molecules that recognized the “hinge” region common to most of these enzymes, the game was on. You can make blunderbuss molecules that inhibit dozens of enzymes at the same time, or (in some cases) you can narrow down on a mere handful, or on just one.

But how far do you want to go? That’s where we’re “over-asked”, as the German expression translates. The downstream effects of many of these enzymes are absolutely bewildering in normal cells, and the differences in disease states are even more of a tangle. It’s no surprise at all that most kinase inhibitors have shown up first in oncology, because that’s where you can get away with the most severe side effects. There are plenty of tempting opportunities in inflammation, diabetes, cardiovascular disease, and other areas, but those have been slower to come along.

The experience with the cancer-targeting drugs has been mixed. You have your Gleevec (imatinib) – pretty selective, works pretty well on a very limited group of patients. And you have your hand grenades, like Sutent (sunitinib) or Nexavar (sorafenib), which hit a lot of kinases and work (to some degree) on a lot of different things. But none of them are magic bullets, for sure. So do you want selectivity or not? The only answer we can offer is (still) “that depends”.

These days, there’s a distinct “kinase hangover” in the industry. It’s not as hot a field as it was. “Not again” is the usual feeling on seeing yet another patent or publication on yet another structure that inhibits XYZ kinase. It’s not as hot an area as it was a few years ago – the belief is that many of the best targets have either wiped out in the clinic, are being tried there now, or haven’t yielded reasonable chemical matter to even get there.

My guess is that we’re waiting, whether we know it or not, for our understanding of the biology to catch up. We have all these compounds, with all these different fingerprints, and we’ve generated this huge pile of mixed data that we can’t quite make sense of. That adds to the frustrated “been there” feeling. The cure for it is to have a better idea of what we’re doing and why, but that’s coming on much more slowly. And because that’s slow, the kinase field may never regain its hot status. But who knows, it may make it all the way to useful and valuable, bypassing “hot” completely.

Comments (19) + TrackBacks (0) | Category: Cancer | Drug Development | Drug Industry History


1. Wavefunction on February 11, 2009 9:31 AM writes...

I think you will find this interesting:
"Is there a case for selectively promiscuous anticancer drugs?"
Fernandez et al.
Drug Disc Today, Vol. 14, No. 1/2, January 2009

Permalink to Comment

2. Wavefunction on February 11, 2009 9:43 AM writes...

Another nice new kinase paper from Pfizer where they screened kinase inhibitors for finding antibacterials is here:

A class of selective antibacterials derived from a protein kinase inhibitor pharmacophore
PNAS 2009 106:1737-1742; published online before print January 22, 2009, doi:10.1073/pnas.0811275106

In the last kinase conference I attended, the conclusions at a round-table event on "Selective vs Multi-targeted kinase inhibitors", were basically what you said; "It depends". A lot still needs to be done by going back to the bench and doing target validation. Sometimes exclusively selective inhibitors seems to work well and sometimes multi-targeted ones. The biggest challenge of course is to target only a very specific predetermined subset of kinases. Again, the complexity of kinase signaling makes it difficult to determine which of the subsets would be most fruitfully engaged. However, I definitely see a wealth of opportunities in the area, both for target validation and for non-cancer kinase therapeutics which are still in the shadows.

Permalink to Comment

3. HelicalZz on February 11, 2009 10:15 AM writes...

Still hot IMO. I am encouraged that we will see more approvals in the once dismal area of oncology due to the focus on this area over the past decade or so. Of course, that also means the market will be more crowded once they get there (so those market forecasts will likely need revising down).

But, it isn't all about inhibition. Don't ignore kinase activation, such as in diabetes and other metabolic indications. Still very Wild West with potentially more 'confounding biology' issues than inhibition.


Permalink to Comment

4. Jose on February 11, 2009 11:03 AM writes...

One factor that hasn't been mentioned is the uber crowded IP space. Most kinase inhibitors look pretty damn similar, and with the obscene number of patents out there... well, the lawyers can fix anything I suppose.

Permalink to Comment

5. Cellbio on February 11, 2009 12:57 PM writes...

Yes, the central question of sepcificity- How far do you want to go? And indeed, the related question- Well, what does kinase x, y or z do anyway? Without answers to the biology, the operational response varies in therapeutic area and size of company. In a well funded environment, it is more likely that the selectivity campaign marches forward. In small companies, more likely that a small handful of counterscreens will be done, and a story spun about how the approach utilizes new and improved methodology and is structure-based so we only need to make a few compounds, and the really good one is coming next. In academia, more likely to see a nice compound with modest selectivity used to prove the value of a target. Shokat has a new paper in Plos on mTOR. Nice work, but not that selective a compound (hits PI3K, no surprise, and PKCs) used in the uM range.

So do we wait until the biology catches up? How about driving the biology with some new fangled systems biology, chemical genetics, or if your older, some good old fashioned pharmacology. Nothing like empirical data, IMO. With the advances in automation and analyte detection, one can screen whole collections of kinase inhibitors in multiple cell-based assays to provide a measure of biological selectivity. I have done this, and can say the results tell more about selective biology than do kinase counter screens. After all, the kinome isn't the only place to worry about selectivity. Not sure this effort will be resourced in today's environment, but I hope to see more of it in use prior to animal pharmacology and toxicology, and potentially as a vehicle for discovery in the first place.

Permalink to Comment

6. kinase guy on February 11, 2009 1:38 PM writes...

After working on several kinase projects as a med chemist at big pharma, it's become clear that everytime we've made exquisitely selective compounds, we've learned new/different biology about the targets and usually walked away from targets because they are not as advertized (either lack of efficacy or tox. reasons).

While in cancer it might be okay to take a multi-targeted agent forward not being sure if your assigned target is driving efficacy (although makes finding response/patient stratification biomarkers a real mess and complicates clinical development), this is likely a road to ruin in more healthy patients.

The answer to the "how selective do we need to be" question really depends on the stage of the program IMO. I'd argue that it's far more important to get high selectivity early to prove the target is worthwhile but to keep relaxed standards to permit optimization of other properties to make your drug. That way, you're as sure as you can be preclinically about the efficacy of your primary target without getting hung up about theoretical safety issues of off-targets.

Measuring and defining selectivity is another contentious issue that always seems to generate excessive churn and lacks standardization across the field.

Permalink to Comment

7. Anonymous on February 11, 2009 2:12 PM writes...

I think the traditional "hinge binder" scaffold model is getting quite worn out (IP, etc) even though it is surprising the level of selectivity you can sometimes achieve, breaking the field's 1980's-1990's preconceptions. As mentioned in previous comments, there is always a case by case basis how much and what type of selectivity you want in order to optimize the efficacy vs. toxicity profile.

I would argue the more interesting future in kinase research at this point is to go after non-ATP competitive inhibitors. Clearer IP space, presumably different clinical mutation profiles, and their different overall selectivity profiles would be useful to test biological, efficacy and toxicity hypotheses both preclinically and clinically.

A number of companies are advancing allosteric inhibitors. There is also at least one start-up (Kinex Pharmaceuticals) in clinical trials with a kinase inhibitor designed to be peptide substrate competitive.

Permalink to Comment

8. anonymous on February 11, 2009 2:18 PM writes...

Re #6:

I've had the same experience with developing relatively selective kinase inhibitors at big pharma and then having the target not live up to expectations.

I think part of this stems from the lack of good selective tool compounds available to the academic community, who are in many cases doing the basic research. It happens with alarming regularity that someone writes a paper claiming that kinase X has such-and-such a function, and they establish that "fact" by inhibiting said kinase with staurosporine (or even dasatinib). I don't know what's worse: that academics appear oblivious to or are just plain unconcerned with the selectivity of the compounds they're working with, or that the peer review and editorial process allows so much garbage to cross and become part of the literature.

It seems to me like a good case for a little more corporate-academic partnership. I'm sure academic labs would love to work with better or more selective compounds... if only pharma would let those guys have access to the compounds.

Permalink to Comment

9. Petros on February 11, 2009 3:57 PM writes...

I'd say they were still pretty hot. A report on kinase inhibitor pipelines has sold like hotcakes (at $3000 per copy list price).

The above points are good. specificity is clearly a major issue with chronic conditions, hence the limited progress with p38 inhibitors in RA after 20 years work. However, for cancer it may actually be undesirable becuase of the redundancy in signaling pathways. GSK's lapatanib is probably the most selective kinase inhibitor on the market but hasn't made the commercial break through after 2 years with 2008 sales of $189 million cf Gleevec's $3.7 billion.

And the IP space around some scaffolds must now be a minefield, both for those working in the area and those drafting new patent claims.

Permalink to Comment

10. milkshake on February 11, 2009 5:31 PM writes...

I have seen great discrepancies between purified protein kinase assays and the cell-based kinase assays - more than for other kinds of targets. I think the purified protein is often an artificial construct, and is typically lot easier to inhibit it than the real thing. Also, polyaromatic heterocyclic bricks preferred by kinase active sites can have hard time getting into cell.

It is important that once you get a cell potency, you go by by the cell data, and also develop disease model in the animals early in the kinase project.

In my opinion it is comparably easier to find a potent class of inhibitor for a kinases target (than other proteins) but there are always some difficulties later at the pre-clinical stage.

Permalink to Comment

11. Marilyn Mann on February 11, 2009 7:59 PM writes...

Statins inhibit rho kinase. James K. Liao at Brigham and Women's has done a lot of work on this.

Evidence for Statin Pleiotropy in Humans
Differential Effects of Statins and Ezetimibe on Rho-Associated Coiled-Coil Containing Protein Kinase Activity, Endothelial Function, and Inflammation
Circulation. 2009;119:131-138

He is also working on a lab test for rho kinase, as a more specific marker of vascular inflammation than hsCRP. Unlike Paul Ridker, he does not have a patent on the test.

Permalink to Comment

12. Marilyn Mann on February 11, 2009 8:00 PM writes...

Statins inhibit rho kinase. James K. Liao at Brigham and Women's has done a lot of work on this.

Evidence for Statin Pleiotropy in Humans
Differential Effects of Statins and Ezetimibe on Rho-Associated Coiled-Coil Containing Protein Kinase Activity, Endothelial Function, and Inflammation
Circulation. 2009;119:131-138

He is also working on a lab test for rho kinase, as a more specific marker of vascular inflammation than hsCRP. Unlike Paul Ridker, he does not have a patent on the test.

Permalink to Comment

13. Marilyn Mann on February 11, 2009 8:00 PM writes...

Statins inhibit rho kinase. James K. Liao at Brigham and Women's has done a lot of work on this.

Evidence for Statin Pleiotropy in Humans
Differential Effects of Statins and Ezetimibe on Rho-Associated Coiled-Coil Containing Protein Kinase Activity, Endothelial Function, and Inflammation
Circulation. 2009;119:131-138

He is also working on a lab test for rho kinase, as a more specific marker of vascular inflammation than hsCRP. Unlike Paul Ridker, he does not have a patent on the test.

Permalink to Comment

14. Marilyn Mann on February 11, 2009 8:02 PM writes...

Oops, please delete the duplicates. Sorry about that.

Permalink to Comment

15. Anonymous BMS Researcher on February 11, 2009 9:50 PM writes...

The classic kinase tree came out of Sugen, who got swallowed up by a succession of mergers, ending with a Really Big Company then most of their people got Pfired. Their database and the website landed at the Manning lab in the Salk Institute where it resides today.

Permalink to Comment

16. Anonymous on February 12, 2009 9:06 AM writes...

Re #10 - milkshake

Large disconnects between biochemical and cellular assays for kinases are extremely common. A large portion of this can be attributed to the often vast differences in ATP concentrations in a cell vs. in a biochemical assay. It is very common to run biochemical assays at the ATP Km for a given kinase, which is usually far, far lower than the intracellular concentration of ATP. For an ATP competitive inhibitor, this can cause drastic (but expected) shifts in IC50s.

The important thing is to look for a *correlation* between biochemical and cellular assays. You should be able to plot IC50s of one vs. the other and see something resembling a trend. A 10- or 100-fold shift in IC50 from biochemical to cell isn't troubling, as long as that shift is *consistent*. If that plot looks like a shotgun blast, then I completely agree that you have to start thinking about the physicochemical properties of your brick-like molecules. :)

Permalink to Comment

17. RTW on February 12, 2009 9:58 AM writes...

I think we have a way to go yet. I agree with Derek that we need to wait for the biology to catch up. I was involved in a number of Kinase projects over the years in CV and Oncology, all had promiss, but on top of the selectivity issues the DMPK characteristics of a lot of the really potent selective compounds where difficult to overcome.

Additionally I am not suprised at the asertion that Statins inhibit roh Kinase. I have a feeling that just about every small molecule drug on the market effects signaling in one way or another. Just might not be a tremendously potent inhibitor is all.

Permalink to Comment

18. pete on February 12, 2009 1:38 PM writes...

re: #10/milkshake and #16/anon

In addition to the [ATP] issue, my own experience seconds what milkshake mentioned: the over-touting of PK/PD data based on truncated kinase domains in vitro - as opposed to that based on full length protein in cells.

There are some key examples where compound IC50s toward highly related kinases can differ dramatically depending upon use of kinase domains vs full-length proteins.

Presumably there's wider recognition of this in 2009 - but I'm certain there are still powerpoint slides shown at investor meetings that promulgate the mickey-mouse snapshot of kinase inhibition.

Oh and FWIW, I vote "still hot" for kinase inhibitors - despite all the jarring bumps in the road.

Permalink to Comment

19. topo on February 13, 2009 10:36 AM writes...

Topoisomerase inhibitors worked better than kinase inhibitors in anti-cancer patients. Is there any success with MMP inhibitors apart from Judah Folkaman's seminal work.

Permalink to Comment


Remember Me?


Email this entry to:

Your email address:

Message (optional):

The Last Post
The GSK Layoffs Continue, By Proxy
The Move is Nigh
Another Alzheimer's IPO
Cutbacks at C&E News
Sanofi Pays to Get Back Into Oncology
An Irresponsible Statement About Curing Cancer
Oliver Sacks on Turning Back to Chemistry