06_mic_appnote_itc_0609.qxd

Isothermal TitrationCalorimetry and Drug Design Introduction
In structure-based drug design (SBDD), full structural knowl-
edge of the protein target molecule provides information on how
a potential drug interacts with the target. Computer modeling
algorithms and molecular simulations are used to predict quanti-
tative structure-activity relationships (QSAR) and these data are
used to optimize the design of new drugs. However, molecular
modeling is limited by a lack of experimental thermodynamic
data to verify these models.
Ultrasensitive Isothermal Titration Calorimetry (ITC) providescomplete thermodynamic characterization of the binding of adrug to its therapeutic target, including enthalpy (ΔH) and FIGURE 1. Thermodynamic signatures for three different drugs binding tosame target. Scheme A: good hydrogen bonding and a conformational entropy (TΔS), free energy (ΔG) and binding affinity (KB). These change. Scheme B: binding dominated by hydrophobic interactions.
thermodynamics parameters are related to each other by the fol- Scheme C: Favorable hydrogen bonding and hydrophobic interactions.
susceptibility to mutations that can cause drug resistance.
Scheme C has favorable ΔH and ΔS, the most favorable thermo- For spontaneous reactions, ΔG is negative, and ΔG is directly Effective drugs are expected to have high binding affinity and related to the binding affinity. The tighter the binding, the more selectivity for target. Drugs for infectious diseases also should be negative the ΔG. Enthalpy and entropy both contribute to ΔG. adaptive, and bind to a broad spectrum of targets, such as protein Enthalpic contributions: ΔH reflects the strength of the drug-
mutants and families of targets. ITC is utilized for the following: target interaction relative to those with solvent, primarily due to • Optimization of binding affinity: ITC determines thermody-
hydrogen bond formation, and van der Waals interactions. WhenΔ namic signatures of lead compounds binding to target. Choose H is negative, binding is enthalpically favored. Favorable lead compound(s) with favorable enthalpy and further optimize enthalpy requires correct placement of hydrogen bond acceptor them by the addition of hydrophobic regions to the core drug.
and donor groups at the binding interface. Drugs with high binding affinities have lower dosage require- Entropic contributions: A positive TΔS results in entropically
ments and consequently fewer side effects.
favored binding. Favorable entropy changes are primarily due to • Binding adaptability: Design drugs which can bind to mutant
hydrophobic interactions, due to an increase in solvent entropy forms of target proteins, thereby reducing drug resistance. Also from burial of hydrophobic groups and release of water upon design drugs which bind to a family of related targets.
binding, as well as minimal loss of conformational degrees offreedom. • Improvement of binding selectivity: Design drugs which bind
selectively to a specific target, bind less selectively to serum Figure 1 shows different thermodynamic signatures that might be proteins or other non-specific targets.
observed for three different drugs binding to the same target.
Each drug has identical ΔG and binding affinity for the target.
Scheme A has favorable ΔH, characteristic of hydrogen bondformation, and an unfavorable TΔS. Drugs showing this bindingprofile typically have large degree of flexibility, as well as highpolarity, which could cause problems with membrane permeabil-ity in vivo. Scheme B has a favorable TΔS, indicating that bind-ing is driven by hydrophobic interactions, and an unfavorableΔH. Drugs showing this binding profile are very hydrophobicand are poorly soluble, and there are also conformationalrestraints leading to lack of adaptability and consequently a high Ultrasensitive Calorimetry for the Life SciencesTM Optimization of Binding Affinity
In typical lead optimization, compounds are initially screenedfor binding affinity to target, with no information about thermo- One goal in drug design is to make drugs which bind to their tar- dynamics. Compounds with high affinity continue through the get with the highest binding affinity. Higher affinity results in discovery pipeline, but may not lead to a successful drug.
lower dosage requirement, greater specificity, better drug effica- Freire’s research indicates that a better strategy is identifying cy, reduced side effects, and less drug resistance. Research by lead compounds with highly favorable ΔH, and optimizing these Ernesto Freire’s group at Johns Hopkins University has shown drugs by addition of hydrophobic groups. This could result in a that thermodynamics from ITC data can be used to characterize highly-specific drug with tight binding affinity. Preliminary work HIV-1 protease inhibitors, and binding affinity is optimized by has been done on the design of SARS drugs using thermody- overcoming enthalpy-entropy compensation.1-5 namic signatures.6 The first generation inhibitors of SARS 3CLproprotease have enthalpically favorable binding, and these drugs Figure 2 shows the thermodynamic signatures of binding of sev- are being optimized to increase binding affinity.
eral inhibitors to wild-type HIV-1 protease. Binding of indinavir,nelfinavir, and saquinavir are driven by entropy, due to This ITC-based QSAR has also been used to develop new drugs hydrophobic interactions between the drug and the target.
which bind to DNA gyrase, by researchers at AstraZeneca.7 They Binding of these inhibitors have unfavorable enthalpy. Ritonavir, observed that triazine inhibitors of DNA gyrase tend to have amprenivir, lopinavir, and two experimental inhibitors KNI-272 similar binding energetics, and they looked for triazine deriva- and KNI-764 have favorable binding enthalpy, as well as a favor- tives which had a markedly different thermodynamic signature, able entropic contribution. Binding interactions with these especially in ΔH, and studied these different derivatives for their inhibitors have both favorable hydrogen bond formation as well structural differences and binding modes. This kind of screening as hydrophobic interactions. These inhibitors are second and can be useful in the discovery of new drugs.
third generation inhibitors. Loprinavir has the highest bindingaffinity of these inhibitors (KB of 1.2 x 1011 M-1), measured by Binding Adaptability
Drug resistance is a serious side effect associated with antiviral therapies, due to the appearance of viral strains withmutant forms of target protein. Mutants with reduced bindingaffinity for inhibitor typically maintain affinity for substrate.
Freire’s group has studied the thermodynamics of inhibitor bind-ing to mutant strains of HIV-1 protease.3-5 One multi-drug resis-tant mutant with six amino acid mutations of HIV-1 protease(MDR-HM) had binding affinity for inhibitor 2 to 3 orders ofmagnitude lower than wild-type.5 Study of these mutationsdemonstrate individual effects of each individual mutation, aswell as cooperative interactions between distal mutations. The determination of enthalpic and entropic components of thedrug resistant HIV-1 protease mutants relative to wild type, FIGURE 2. Thermodynamic signatures of binding of inhibitors to wild-type along with structural studies, was used to develop guidelines in HIV-1 protease. Indinavir, nelfinavir, saquinavir, ritonavir,amprenavir, and the design of inhibitors more adaptable to drug target mutations.5 lopinavir are in clinical use. KNI-272 and KNI-764 are experimental protease inhibitors. Data from References 4 and 5. Refer to Reference 5 for design ofITC experiments.
• Maximize ΔH contribution to ΔG for enthalpically favorable Hydrogen bond formation and favorable van der Waals interac-tions between drug and target results in favorable enthalpy.
• Design drug-target interactions with highest binding affinity Hydrogen bond formation is a result of optimal placement of and specific hydrogen bonding between core of the drug and hydrogen bond donor and acceptor groups on the drug and tar- highly-conserved region of the target’s binding site.
get, and is highly directional and specific. Hydrogen bond • Introduce non-constrained functional groups on inhibitor facing donors and receptors may be in close proximity in a crystal mutation-prone regions of binding site of target. Since this will structure but may have little effect on binding affinity, due to make ΔS of binding less favorable by loss of conformational degrees of freedom of inhibitor, inhibitor ΔH must be highly A common strategy in drug design is the addition of non-polar favorable to maintain high affinity.
groups to increase hydrophobic interactions between target andcompound. However, hydrophobic interactions are non-specificcompared to hydrogen bonding. Freire’s research demonstratesthat drugs with favorable enthalphy AND entropy have highestaffinity, due to specific hydrogen bond formation and van derWaals interactions as well as hydrophobic interactions with thetarget.
Thermodynamics and ITC have also been used to design adap- References
tive inhibitors. This is achieved by designing an inhibitor with atight and specific binding interaction with the conserved region 1. Velazquez-Campoy, A., Todd, M.J., Freire, E. (2000) HIV-1 protease of the binding site of the primary target, and introduces flexible inhibitors: Enthalpic versus entropic optimization of the binding
affinity. Biochemistry 39, 2201-2207.
asymmetric functional groups to be able to bind the other relatedtargets. This approach has been used to design anti-malarial 2. Velazquez-Campoy, A., Kiso, Y., Freire, E. (2001) The binding ener- drugs against proteases of Plasmodium falciparum, plasmepsin I, getics of first- and second-generation HIV-1 protease inhibitors: II, IV, and histo-asparyl protease (HAP).8,9 A tight-binding Implications for drug design. Arch. Biochem. Biophys. 390, 169-175.
inhibitor to plasmepsin II also inhibited secondary targets, due to 3. Velazquez-Campoy, A., Vega, S., Freire, E. (2002) Amplification of the adaptive group on the inhibitor.
the effects of drug resistance mutations by background polymor-
phisms in HIV-1 protease from African subtypes. Biochemistry 41,
8613-8619.
Binding Selectivity
4. Ohtaka, H., Velazquez-Campoy, A., Xie, D., Freire, E. (2002) Overcoming drug resistance in HIV-1 chemotherapy: The binding Binding selectivity is dependent on the affinity of drug to both thermodynamics of amprenavir and TMC-126 to wild-type and desired and unwanted targets, as well as the KB ratio of the two drug-resistant mutants of the HIV-1 protease. Protein Sci. 11,
binding events. Drugs can bind to non-specific targets, such as serum proteins, or to other target proteins with different bindingaffinities. When a drug binds serum proteins, the drug’s bioavail- 5. Ohtaka, H., Schön, A., Freire, E. (2003) Multidrug resistance to HIV-1 protease inhibition requires cooperative coupling between ablity and efficacy are reduced, requiring a higher dosage. ITC distal mutations. Biochemistry 42, 13659-13666.
has been used to study the binding of HIV-1 protease inhibitorsto human serum proteins.10 HIV-1 protease inhibitors had a sig- 6. Bacha, U.M., Barrila, J.A., Velazquez-Campoy, A., Leavitt, S., nificant binding affinity for α1-acid glycoprotein (AAG) and a Freire, E. (2004) Development of potent inhibitors of the SARS relatively low binding affinity for human serum albumin (HSA).
associated coronavirus protease 3CLpro. Biophys J. 86, 97a. (Poster at
Using AAG and HSA concentration simulating in vivo condi- tions, up to 10 times higher concentration was needed to inhibit 7. Ward, W.H.J., Holdgate, G.A. (2001) Isothermal titration calorime- HIV-1 protease relative to the amount needed to inhibit in try in drug discovery. Prog. Med. Chem. 38, 309-376.
absence of serum proteins. When drug-resistant HIV-1 protease 8. Nezami, A., Luque, I., Kimura, T., Kiso, Y., Freire, E. (2002) mutants were used, up to 2000-fold more inhibitor was required Identification and characterization of allophenylnorstatime-based to inhibit protease, and this is beyond the solubility limit of the inhibitors of plasmepsin II, and anti-malarial target. Biochemistry 41, 2273-2280.
ITC and structural studies can be used to develop new drugs to 9. Nezami, A., Kimura, T., Hidaka, K., Kiso, A., Liu, J., Kiso, Y., maximize selectivity and specific binding to intended target by Goldberg, D.E., Freire, E. (2003) High-affinity inhibition of a family optimization of enthalpy and entropy.
of Plasmodium falciparum protease by a designed adaptive inhibitor.
Biochemistry 42, 8459-8464.
10. Schön, A., Ingaramo, M.d.M., Freire, E. (2003) The binding of HIV- 1 protease inhibitors to human serum proteins. Biophys. Chem. 105,
221-230.
ITC studies on HIV-1 protease inhibitors, anti-malarial drugs,SARS drugs, and DNA gyrase inhibitors have generated thermo-dynamic signatures of drug-target interactions. These data, usedwith structural analysis, provide valuable information in thedesign of new drugs which bind with higher affinity and selectiv-ity, as well as adaptability. The introduction of ITC instrumenta-tion with increased throughput allows use of microcalorimetryearlier in drug discovery pipelines.
Acknowledgement: MicroCal, LLC thanks Dr. Ernesto Freire forreviewing this application note. Ultrasensitive Calorimetry for the Life SciencesTM MICROCAL, LLC • 22 Industrial Drive East • Northampton, MA 01060 • USA Toll-Free in North America: 800.633.3115 • Tel. 413.586.7720 • Fax 413.586.0149 web page: www.microcalorimetry.com • e-mail: [email protected] 2 Warren Yard • Warren Farm Office Village • Wolverton Mill • Milton Keynes MK12 5NW • United Kingdom Tel. +44.1908.576330 • Fax +44.1908.576339 • e-mail: [email protected]

Source: http://kitto.cm.utexas.edu/courses/ch395g/fall2009/MOL190/ITC.pdf