Transcript Bioinformatik - Chair of Computational Biology

How many drug targets are there?
In 2002, after the sequencing of the human genome, others arrived at ~8,000
targets of pharmacological interest, of which nearly 5,000 could be potentially
hit by traditional drug substances, nearly 2,400 by antibodies and ~800 by protein
pharmaceuticals2. And on the basis of ligand-binding studies, 399 molecular
targets were identified belonging to 130 protein families, and ~3,000 targets for
small-molecule drugs were predicted to exist by extrapolations from the number
of currently identified such targets in the human genome.
25. Lecture WS 2006/07
Bioinformatics III
1
Drug Target: Enzymes
25. Lecture WS 2006/07
Bioinformatics III
2
Drug Target: Enzymes II
25. Lecture WS 2006/07
Bioinformatics III
3
Drug Target: Enzymes III
25. Lecture WS 2006/07
Bioinformatics III
4
Drug Target: Enzymes III
25. Lecture WS 2006/07
Bioinformatics III
5
Drug Target: Receptors I
25. Lecture WS 2006/07
Bioinformatics III
6
Drug Target: Receptors II
25. Lecture WS 2006/07
Bioinformatics III
7
Drug Target: Receptors III
25. Lecture WS 2006/07
Bioinformatics III
8
Drug Target: Receptors III
25. Lecture WS 2006/07
Bioinformatics III
9
Drug Target: Ion channels
25. Lecture WS 2006/07
Bioinformatics III
10
Drug Target: Transport proteins
25. Lecture WS 2006/07
Bioinformatics III
11
Drug Target: DNA/RNA and the ribosome
25. Lecture WS 2006/07
Bioinformatics III
12
Drug Target: Targets of monoclonal antibodies
25. Lecture WS 2006/07
Bioinformatics III
13
Drug Target: Various physicochemical mechanisms
25. Lecture WS 2006/07
Bioinformatics III
14
Outlook
A large part of this paper is concerned with the nature of drug targets and the need to
consider the dynamics of the drug–targets (plural intended) interactions, as these
considerations were used to define what we would eventually count.
Many successful drugs have emerged from the simplistic ‘one drug, one target, one disease’
approach that continues to dominate pharmaceutical thinking, and we have generally used
this approach when counting targets here. However, there is an increasing readiness to
challenge this paradigm. We have discussed its constraints and limitations in light of the
emerging network view of targets. The recent progress made in our understanding of
biochemical pathways and their interaction with drugs is impressive.
However, it may be that ‘the more you know, the harder it gets’. It is not the final number of
targets we counted that is the most important aspect of this Perspective; rather, we stress
how considerations about what to count can help us gauge the scope and limitations of our
understanding of molecular reaction partners of active pharmaceutical ingredients.
Targets are highly sophisticated, delicate regulatory pathways and feedback loops but, at
present, we are still mainly designing drugs that can single out and, as we tellingly say, ‘hit’
certain biochemical units — the simple definable, identifiable targets as described here.
This is not as much as we might have hoped for, but in keeping with the saying of one of
earliest medical practitioners, Hippocrates:
“Life is short, and art long; the crisis fleeting; experience perilous, and decision difficult.”
Humility remains important in medical pharmaceutical sciences and practice.
25. Lecture WS 2006/07
Bioinformatics III
15
Specific example: protein kinases
Phosphorylation of serine, threonine, and tyrosine residues is a primary mechanism for
regulating protein function in eukaryotic cells. Protein kinases, the enzymes that catalyze
these reactions, regulate essentially all cellular processes and have thus emerged as
therapeutic targets for many human diseases.
Small-molecule inhibitors of the Abelson tyrosine kinase (Abl) and the epidermal growth factor
receptor (EGFR) have been developed into clinically useful anticancer drugs. Selective
inhibitors can also increase our understanding of the cellular and organismal roles of protein
kinases. However, nearly all kinase inhibitors target the adenosine triphosphate (ATP) binding
site, which is well conserved even among distantly related kinase domains. For this reason,
rational design of inhibitors that selectively target even a subset of the 491 related human
kinase domains continues to be a daunting challenge.
Cohen et al. Science 308, 1318 (2005)
25. Lecture WS 2006/07
Bioinformatics III
16
Specific example: protein kinases
Structural and mutagenesis studies have revealed key determinants of kinase inhibitor
selectivity, including a widely exploited filter in the ATP binding site known as the „gatekeeper“.
A compact gatekeeper (such as threonine) allows bulky aromatic substituents, such as those
found in the Src family kinase inhibitors, PP1 and PP2, to enter a deep hydrophobic pocket. In
contrast, larger gatekeepers (methionine, leucine, isoleucine, or phenylalanine) restrict access
to this pocket. A small gatekeeper provides only partial discrimination between kinase active
sites, however, as ca. 20% of human kinases have a threonine at this position. Gleevec, a
drug used to treat chronic myelogenous leukemia, exploits a threonine gatekeeper in the Abl
kinase domain, yet it also potently inhibits the distantly related tyrosine kinase, c-KIT, as well
as the platelet-derived growth factor receptor (PDGFR).
25. Lecture WS 2006/07
Bioinformatics III
17
Selection of gatekeeper residue
25. Lecture WS 2006/07
Bioinformatics III
18
Outlook
In this study, we have rationally designed halomethylketone-substituted
inhibitors whose molecular recognition by protein kinases requires the
simultaneous presence of two selectivity filters: a cysteine following the
glycine-rich loop and a threonine in the gatekeeper position.
We estimate that ca. 20% of human kinases have a solvent-exposed cysteine in
the ATP pocket. Because of the structural conservation of the pocket, it should
be possible to predict the orientation of these cysteines.
In addition, there are many reversible kinase inhibitors whose binding modes
have been characterized by x-ray crystallography.
The integration of both types of information should allow the design of scaffolds
that exploit selectivity filters other than the gatekeeper, as well as the
appropriate sites for attaching electrophilic substituents.
25. Lecture WS 2006/07
Bioinformatics III
19
Small molecule-kinase interaction map
Figure 1. Competition binding assay for
measuring the interaction between unlinked,
unmodified ('free') small molecules and kinases.
(a) Schematic overview of the assay. The phagetagged kinase is shown in blue, 'free' test compound
in green and immobilized 'bait' ligand in red. (b)
Binding assay for p38 MAP kinase. The immobilized
ligand was biotinylated SB202190. The final
concentration of test compounds during the binding
reaction was 10 M. (c) Determination of quantitative
binding constants. Binding of tagged p38 to
immobilized SB202190 was measured as a function
of unlinked test compound concentration. Tagged
p38 kinase was quantified by real-time quantitative
PCR and the results normalized.
Fabian et al. Nature Biotech 23, 329 (2005)
25. Lecture WS 2006/07
Bioinformatics III
20
Small molecule-kinase interaction map
25. Lecture WS 2006/07
Bioinformatics III
21
Small molecule-kinase interaction map
Each kinase represented in the assay panel is
marked with a red circle. Gene symbols for
kinases in the panel are shown in Figure 5.
TK, nonreceptor tyrosine kinases;
RTK, receptor tyrosine kinases;
TKL, tyrosine kinase-like kinases;
CK, casein kinase family;
PKA, protein kinase A family;
CAMK, calcium/calmodulin dependent kinases;
CDK, cyclin dependent kinases;
MAPK, mitogen-activated protein kinases;
CLK, CDK-like kinases.
25. Lecture WS 2006/07
Bioinformatics III
22
Small molecule-kinase interaction map
Figure 3. Specificity profiles of
clinical kinase inhibitors.
Circle size is proportional to binding
affinity (on a log10 scale).
Binding constants were measured at
least in duplicate for each interaction
identified in the primary screen.
25. Lecture WS 2006/07
Bioinformatics III
23
Distribution of binding constants
For each compound the
pKd (-log Kd) was plotted for all
targets identified.
Primary targets, as shown in
Table 1, are in blue, and offtargets in red.
Staurosporine does not have a
particular primary target or
targets, and the primary targets
for BAY-43-9006 (RAF1) and LY333531 (PKC ) were not part of
the assay panel.
25. Lecture WS 2006/07
Bioinformatics III
24
Hierarchical cluster analysis of specificity profiles
Lighter colors correspond to
tighter interactions.
(a) Twenty kinase inhibitors
profiled against a panel of 113
different kinases.
25. Lecture WS 2006/07
Bioinformatics III
25
Small molecule-kinase interaction map
We have described a systematic small molecule−kinase interaction map for clinical kinase
inhibitors. Integration of the information provided here with results from cell-based or animal
studies, and ultimately with clinical observations, should enable a more complete understanding
of the biological consequences of inhibiting particular combinations of kinases.
Binding profiles for larger numbers of chemically diverse compounds, combined with the
phenotypes elicited by these compounds in biological systems, will help identify kinases whose
inhibition leads to adverse effects, kinases that are 'safe' to inhibit and combinations of kinases
whose inhibition can have a synergistic beneficial effect in particular disease states.
This knowledge should enable the development of inhibitors with 'appropriate' specificity that
target multiple kinases involved in the disease process while avoiding kinases implicated in side
effects. The ability to rapidly screen compounds against multiple kinases in parallel and the
incorporation of specificity profiling during initial lead discovery and optimization should greatly
facilitate and accelerate the drug development process.
25. Lecture WS 2006/07
Bioinformatics III
26
Small molecule-kinase interaction map
The kinase binding profiles also provide valuable information to guide structural studies.
In many cases kinases that tightly bind the same compound have no obvious sequence similarity
(for example, p38 and ABL(T315I) binding to BIRB-796).
In other cases, compounds can discriminate between kinases closely related by sequence, such
as imatinib binding to LCK but not SRC.
ABL and the imatinib-resistant ABL mutants are of particular structural interest because some
compounds bind with good affinity to all forms (e.g., ZD-6474), whereas BIRB-796 has a strong
preference for a particular mutant.
Key insights should result from an analysis of selected co-crystal structures of kinase-compound
combinations identified through profiling studies, and the large, uniform data set presented here
should serve as a valuable training set for computation-based inhibitor design.
Finally, the use of phage-tagged proteins in quantitative biochemical assays circumvents the
need for conventional protein production and purification, and should help reduce one of the
major bottlenecks in modern proteomics and drug discovery research.
25. Lecture WS 2006/07
Bioinformatics III
27
Multidrug treatments are increasingly important in medicine and for probing
biological systems. But little is known about the system properties of a full drug
interaction network.
Epistasis among mutations provides a basis for analysis of gene function.
Similarly, interactions among multiple drugs provide a means to understand their
mechanism of action.
Aim: derive a pairwise drug interaction network.
Yeh et al. Nature Genetics 38, 489 (2006)
25. Lecture WS 2006/07
Bioinformatics III
28
Different ways of drug interaction
Clustering of individual drugs
into functional classes solely on
the basis of properties of their
mutual interaction network.
Schematic illustration of
additive, synergistic and
antagonistic interactions
between drugs X and Y by
measurements of bacterial
growth under the following
conditions:
no drugs, drug X only, drug Y
only, and both drugs X and Y.
25. Lecture WS 2006/07
Additive: no interaction
Synergistic: larger-than-additive effect
Antagonistic: smaller-than-additive effect
Bioinformatics III
29
Classification of drug interactions
g , gX, gXY : growth of wild-type, with drug X,
and with drugs X and Y
WXY 
W
g XY
g
  ~XY
 WX WY
WXY  WX WY
~
WXY  min WX , WY  for WXY  WX WY and 0 otherwise
For WXY  min WX , WY 
~ 
WXY  min WX , WY 
1
1  min WX , WY 
This scale maps synthetic lethal interactions to  = -1,
additive interactions are mapped to  = 0,
antagonistic buffering to  = 1,
and antagonistic suppression to  > 1.
25. Lecture WS 2006/07
Bioinformatics III
30
The Prism algorithm
25. Lecture WS 2006/07
Bioinformatics III
31
Classification
(b–d) A network (b) of synergistic interactions (red lines) and antagonistic interactions (green lines)
between drugs (black circles) can be clustered into functional classes that interact with each other
monochromatically (that is, with purely synergistic or purely antagonistic interactions between any two
classes; c). This classification generates a system-level perspective of the drug network (d). (e,f) Two
independent observations indicate whether a new drug (Z) will be clustered into a particular drug class
(a, dashed oval): mixed synergistic and antagonistic intraclass interactions of Z with a (e, thin dotted
green and red lines) and nonconflicting interclass interactions of Z (e, dotted thin lines) and a (e, dotted
thick lines) with all other classes. Both intra and interclass indications are depicted in e, and the drug is
clustered (black arrow) with an existing class. If drug Z has no such intra- or interclass association with
any existing drug class, the drug will be clustered in a new class (f).
25. Lecture WS 2006/07
Bioinformatics III
32
Tested drugs
25. Lecture WS 2006/07
Bioinformatics III
33
Experimental classification of drug interaction
Figure 2 Experimental classification of drug interactions into four types using bioluminescence
measurements of bacterial growth in the presence of sublethal concentrations of antibiotics.
(a) The pairs of antibiotics illustrate synergistic interactions.
The number of bacteria (proportional to
bioluminescence counts per second (c.p.s.)
is shown from two replicates, for control with
no drugs (f, solid black lines), each single
drug (X, Y; blue and magenta lines) and the
double-drug combination (X + Y, dashed
black lines).
Insets: normalized growth rates (W) with
error bars for f, X, Y and X+Y, from left to
right, respectively. Note the contrast
between the interactions of piperacillin with
the 50S ribosomal subunit drug
erythromycin (a, ERY-PIP, synergistic) and
the 30S ribosomal subunit drug tetracycline
(c, TET-PIP, antagonistic).
25. Lecture WS 2006/07
Bioinformatics III
34
Different modes of interaction
The pairs of antibiotics illustrate synergistic (a), additive (b), antagonistic
buffering (c) and antagonistic suppression (d) interactions
25. Lecture WS 2006/07
Bioinformatics III
35
Systematic measurements of pairwise interactions between antibiotics
(a) Growth measurements and classification of interaction for all pairwise combinations of drugs X and Y.
Within each panel, the bars represent measured growth rates for, from left to right: no drugs (f), drug X only,
drug Y only and the combination of the two drugs X and Y (see inset).
Error bars represent variability in replicate measurements.
The background color of each graph designates the
form of epistasis according to the scale in b:
synergistic (red: emax < -0.5; pink: -0.5 < emax < 0.25), antagonistic buffering (green: 0.5 < emin <
1.15; light green: 0.25 < emin < 0.5), antagonistic
suppression (blue: emin > 1.15) or additive (white: 0.25 < emax < 0.5 and -0.5 < emin < 0.25). Cases that
do not fall into any of these categories are labeled
inconclusive (gray background).
25. Lecture WS 2006/07
Bioinformatics III
36
Classification into interaction classes
Unsupervised classification of the
antibiotic network into monochromatically
interacting classes of drugs with similar
mechanisms of action.
(a) The unclustered network of drug-drug
interactions with synergistic (red),
antagonistic buffering (green) and
antagonistic suppression (blue) links.
25. Lecture WS 2006/07
Bioinformatics III
37
Monochromatically interacting functional classes
(b) Prism algorithm classification of drugs
into monochromatically interacting
functional classes.
This unsupervised clustering shows good
agreement with known functional
mechanism of the drugs (single letter
inside each node; see Table 1).
Bleomycin (BLM), which is believed to
affect DNA synthesis, although its
mechanism is not well understood,
cannot be clustered monochromatically
with any other class. The multifunctional
drug nitrofurantoin (NIT) shows nonmonochromatic interactions.
25. Lecture WS 2006/07
Bioinformatics III
38
System-level interactions between the drug classes
(c) Larger ellipses show higher-level
classification of DNA gyrase inhibitors (D)
with inhibitors of biosynthesis of DNA
precursors (F) and classification of the
two subclasses of drugs involved in the
inhibition of protein synthesis via the 50S
ribosomal subunit (R).
25. Lecture WS 2006/07
Bioinformatics III
39
Outlook
We provide a complete and systematic analysis of a drug-drug interaction network.
Systems analysis of the interaction network demonstrates that drugs can be classified
according to their action mechanism based on their interactions with other functional drug
classes.
The ability to classify drug function based solely on phenotypic measurements and
without the tools of biochemistry or microscopy can provide a simple and powerful method
for screening new drugs with multiple or novel mechanisms of action. Our systems
approach is general in nature and could be applied to other biological systems.
It would be particularly useful if the approach could be generalized to in vivo studies and
to a wider range of phenotypes despite added complexity of host-drug interaction.
Furthermore, applying network approaches to drug interactions may help suggest new
drug combinations and highlight the importance of gene-environment interactions,
including, in particular, the resistance and persistence of bacteria to antibiotics and of
cancer cells to antitumor drugs.
25. Lecture WS 2006/07
Bioinformatics III
40