Examining The Covalent Modifier In Drug Discovery Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Pharmaceutical industry usually disregard or filter out screening hits containing potentially reactive functionality behalf of compounds that modulate target proteins through noncovalent interactions. Indeed, the potential indiscriminant covalent interactions of these reactive drug candidates with biological nucleophiles like glutathione (GSH) or with unwanted biological target (off-target proteins, DNA, etc.) could results to unfavourable toxicological outcomes. [1] These adverse responses can be observed early as well as delayed toxicity due to either immunological or histological factors (Table 1). [2] 

These may occur either acutely or as a delayed response In the case of immunological (allergic) response, either a drug-protein complex or a degradation product of such a complex can act as a stimulant for the immune system (Figure 1).

Figure 1. Postulated mechanism for drug hypersensitivity reactions.

This pathway can be achieved by either a reactive parent compound or a metabolite.

Even when compounds present no grave toxicological outcome in preclinical models, idiosyncratic reactions can be manifested in human clinical trials or arise once the entity is exposed to a larger patient pool. Unfortunately, predicting life-threatening idiosyncratic adverse events in humans at the preclinical stage has been evasive. As a result, there is often resistance toward the development of drugs containing reactive functionality, even when the reactivity is modest and confined to the biochemical target.

Alternatively, there are instances where controlled, targets specific covalent modification has proven to be useful to the development of drugs, pharmacological tools or biomarkers for biological assay. Indeed, an increasing numbers of drugs exert their pharmacological functions on enzymes, receptors or structural protein by covalent modification of target. [3] All of these compounds presents a molecular structure in which a chemically reactive fragment "warhead" is able to establish covalent interactions, reversible or not, with one or more nucleophilic residues exposed on protein's surface or into an inner protein cavity.

As a drug, the covalent binding of one of these compounds to the desired biological target usually, but not exclusively, provides prolonged and irreversible target inactivation at low micromolar or nanomolar range. Their target selectivity depend by their structure as well as by their warhead reactivity. While the former is important for drug-target recognition and for proper warhead positioning, the latter is fundamental to ensure an effective covalent binding to the target without react with off-target nuclephyles. The best pharmacological profile, with high selectivity and bioavailability, is achieved when the covalent modifier is typically poorly reactive with solution nucleophiles under physiological conditions but yet upon appropriate positioning will selectively react with a nucleophile within the target protein. [4] 

Compounds able to selectively bind the thiol portion of cysteine residues on proteins without react with other nucleophiles are attracting the interest of scientific community around the world. Indeed, if compared with other nucleophiles-reactive compounds (such as serine or tyrosine-reactive molecules), the cysteine oriented reactivity of these kind of compounds could furnish an additional source of selectivity that is beyond of fitting optimization provided by proper structural modulation. This is because of cysteine residues in a primary protein structure are usually few than the number of other structural amino acids. In this way, the positions of cysteines on folded protein structure can be used as specific binding point for cysteine trapping covalent modifiers able, with a proper structural optimization, to discriminate all other undesired cysteine exposing proteins.

This concept has been applied to the development of drugs such as cysteine-protease inhibitors (cathepsin and caspase inhibitors), tyrosine kinase inhibitors (EGFR inhibitors) and lipase inhibitors (MGL cysteine trappers inhibitors) but also to the development of pharmacological tools to localize, study and validate targets.

1.1.1 Covalent drug-target interaction: an orthogonal approach to design pharmacological tools or drugs.

The examples presented herein provide strong evidence that covalent modifiers can be safe and effective therapeutics. While in many instances the mechanism of inhibition was determined after efficacy was realized, one could adopt a covalent modifier approach from the beginning of a program.

One key success factor for this approach is the proper selection of the warhead moiety. Although there are examples of compounds containing very active functionality, such as aspirin (activated ester) and fosfomycin 1i (epoxide), a majority of the successful drugs contain functionality whose reactivity is attenuated to achieve targeted modulation. For example, the binding of rivastigmine 1d to acetylcholinesterase activates the carbamate toward cleavage by the active site serine of the catalytic triad. Another elegant example is finasteride 1u, which acts as a selective hydride acceptor from NADPH only when bound to 5R-reductase.65 In addition, the Cat K inhibitor odanacatib 1x highlights the reversible nucleophilic addition of an active site thiol to a nitrile. These examples illustrate how the location of the warhead within a structural motif can deliver both the desired therapeutic effect and safety profile.

Additionally, the prodrug approach is also valid but arguably more challenging. There are several drugs that utilize a masked warhead as the electrophilic component such as the H+/K+ ATPase inhibitors (exemplified by omeprazole 1n), where the reactive species is generated in the acidic environment of the stomach where the drug exercises its antisecretory effect. This target-localized formation of the reactive intermediate reduces systemic exposure and potential for off-target toxicities.66 The blockbuster drug clopidogrel 1o is converted to an active metabolite that is hypothesized to react preferentially with P2Y12 to prevent stoke.

Whether these successful drugs were discovered serendipitously or by design, we can use the insight provided by the available mechanistic and/or structural information to enable future de novo design of selective covalent modifiers. Paramount for success is the availability of detailed structural information on protein-ligand interaction, such as that derived from of X-ray crystallography, to facilitate the refinement of compound design and warhead placement. This approach is elegantly illustrated by the EGFR inhibitor 1t, where an appropriately placed Michael acceptor reacts readily with a nucleophilic amino acid side chain when facilitated by assistance from an internal basic amine moiety.

A systematic review of the known covalently modulated targets reveals several trends (Table 3, Charts 1 and 2). It is no surprise that the most prevalent covalently modified targets identified are enzymes (Chart 1). As a subset of the overall targets, the cysteine and serine residues are primarily modified, with few examples of other nucleophilic amino acid residues (Chart 2). Among the enzymes, proteases or hydrolases appear frequently. In addition, cofactor mediated enzymes are also represented. These data indicate that cofactor mediated enzymes or enzymes bearing an active site cysteine or serine represent attractive targets for covalent modification. The strategy to drug a target through employing covalent modifying approach could provide advantages under certain scenarios. There is typically a cost to improving the potency of lead structures that bind through noncovalent interactions. This endeavor must balance increases in molecule weight, lipophilicity, and hydrogen bonding functionality that can be detrimental to other important properties such as pharmacokinetics and ancillary pharmacology. In contrast, when a significant amount of binding energy is derived from the drug-protein covalent bond, there should be a reduction in the number noncovalent interactions needed to achieve desired potency. In the case of irreversible binders, drug concentrations in systemic circulation need only be available for a long enough period to achieve target coverage, potentially deemphasizing the need for a high, prolonged systemic drug load and therefore potentially mitigating off-target activity.67 Also, the half-life of the compound need not be long in order to achieve once a day or twice a day dosing. Certainly, reversible noncovalent inhibitors that display slow off-rates would also provide a similar benefit. While there will always be a healthy debate about pursuing molecules that bind covalently, this risk may be minimized by pursuing covalent modifiers that would be administered acutely or to patients with a life threatening disease.

Analysis of the pharmacodynamic needs of a particular therapy may lead one to consider irreversible covalent inhibition. For many diseases pharmacodynamic activity is correlated to the degree of target inhibition or occupancy. For therapies that require a high target occupancy for effective treatment, such as cancer or antibacterial therapeutics (where in the absence of high target coverage mutations may occur),68 irreversible covalent modulation could be the most effective means of treatment.

Conversely, there are therapeutic axes that would not benefit from complete covalent inhibition, wherein the complete shutdown of a primary pathway would lead to on-target toxicities. In these instances, irreversible covalent inhibition may not be appropriate. For example, in the case of warfarin, it is known that using the drug for an extended period of time (or at a high dose) can cause fatal bleeding. For this reason, warfarin is recommended for short-term use; when warfarin is used for long-term thrombosis therapy, patients are closely monitored.

That said, the industry is still searching for a safe and effective alternative to warfarin. Whether medicinal chemists pursue covalent or noncovalent modifiers, compounds should be selective for the desired target. This selectivity encompasses related pharmacological targets, as well as other endogenous nucleophilic moieties such as proteins, peptides (such as glutathione), and DNA. In any drug discovery program ancillary pharmacology studies are conducted to assess the potential liability for observing off-target toxicities in addition to in vitro safety studies. While selectivity criteria are identical for programs striving to develop either a covalent or noncovalent modifier, one might consider conducting studies to determine promiscuous binding earlier in a program utilizing a potentially reactive functional group.

It is interesting to consider how an organization might become better positioned to exploit covalent modification as a more general approach to drug discovery. For instance, one may consider building a focused screening set that would be populated with low molecular weight compounds that possess "low to moderately" reactive functionality. A lead identified from this collection could be optimized with information from crystallography and modeling studies. Medicinal chemists could further "fine-tune" reactivity, if needed, so covalent adduction is confined to the target protein. Of course opinions regarding an acceptable level of reactivity for a lead structure will always be defined differently throughout the industry. In addition, identification of functional groups beyond those mentioned in this review that selectively form covalent adducts could further enable this strategy.

There are a number of covalent modifiers in preclinical or early clinical investigation that will continue to offer insight to this drug discovery strategy, including ones that target the caspases,69 MMP13,70 thyroid hormone receptor,71 and FAAH.72

Certainly, the presence of small screening or fragment sets comprising compounds with low to moderately reactive functionality will be crucial for providing starting points. Alternatively, one could look to strategically position a warhead within a lead compound. Hopefully, the compilation of examples in this review will inspire drug discovery scientists to consider pursuing covalent modifiers in the future.

1.1.2 From reactants to drugs.

1.2 Covalent modifiers.

In most instances this has not been the strategy but rather discovered in hindsight.

Development of highly potent and selective covalent cysteine-protease inhibitors showed that many chemically different warheads are able to react with specific catalytic or non-catalytic cysteine residues in cysteine-protease active sites by covalent interaction. [5] 

The Cysteine-trapping warheads described in the literature can be classified on the basis of their "cysteine interaction" mechanism:

alkylating warheads. [6] 

Nucleophile substitution-based warheads.

Michael addition-based warheads.

Acylating warheads (as -lactons [7] and -lactams).

disulfide bond forming warheads. [8] 

covalent but reversible warheads: an example of this kind of warhead can be represented by N-Cyanomethylamides that are able to covalently interact with a cysteine group on the target, forming a reversible thioimidate complex.

The covalent inhibition of a biological target (i.e. a receptor or an enzyme) presents some advantages with respect to the reversible one. Indeed, an irreversible inhibitor doesn't need prolonged circulating blood levels to achieve a desired biological effect. Once the target is deactivated by covalent bond formation, the biological effect should persist even after the drug leaves the circulation. As a result, the duration of action of such a drug will be long-lasting because will be a function of the rate of enzyme turnover.

However, the intrinsic reactivity of warheads often gives rise to low druggability of these molecules because they are often liable of a heavy metabolic degradation (low bioavailability) or toxicity for lack of target-specificity.

In order to overcome these drawbacks, two strategies could be used to improve the drug-likeliness of a reactive compound:

Modulate the "warhead" reactivity to decrease metabolic degradation and/or to optimize its selectivity.

Link the "warhead" fragment to "driver groups" that assure an increase of specificity by optimized fit at the active site of the target, in which the reactive centres are held in close proximity and proper orientation for a covalent interaction to ensue.

Both strategies have been used to develop cysteine-binder molecules that are still under investigation as cysteine-protease inhibitors or irreversible ligands of other cysteine-exposing targets.

Among covalent cysteine-protease inhibitors, some compounds have reached preclinical and clinical phases demonstrating that reactive warheads can be inserted in drug-like compounds. Presently, cathepsine inhibitors as APC-3328 or CRA-013783 have reached the preclinical phase as promising anti-osteoporosis drugs, while calpain inhibitor A-705253 and caspase-1 inhibitor Vx-765 are still under evaluation respectively as neuroprotecive and anti-inflammatory agents (Figure 1).

Figure 1

It is possible to change the target selectivity of synthesized cysteine-reactive compounds.

1.2.1 Warhead reactivity and target selectivity.

1.2.2 Serine and Tyrosine reactive compounds.

1.2.3 Cysteine reactive compounds.

Pharmacologically important target for cysteine-trapping agents