Lead Compound Identification from Endogenous Sources

In the attempt to identify logical or rational methods for designing and discovering lead compounds, the notion of exploiting endogenous molecules quickly comes to mind.

Human disease arises from perturbations of normal biochemical processes. A logical therapeutic approach involves the administration of one or more of these naturally occurring endogenous biochemical molecules, or analogs thereof. In addition, certain human diseases seem to arise from a deficiency of a certain endogenous molecule. It is reasonable to assume that such diseases could be cured or at least helped by the admin-istration of the missing molecule.

Medicinal chemistry has many examples of the development of successful thera-peutics based on an exploration of endogenous compounds. The treatment of diabetes mellitus, for example, is based upon the administration of insulin, the hormone that is functionally deficient in this disease. The current treatment of Parkinson’s disease is based upon the observation that the symptoms of Parkinson’s disease arise from a deficiency of dopamine, an endogenous molecule within the human brain. Since dopamine cannot be given as a drug since it fails to cross the blood–brain barrier and enter the brain, its biosynthetic precursor, L-DOPA, has been successfully developed as an anti-Parkinson’s drug. Analogously, the symptoms of Alzheimer’s disease arise from a relative deficiency of acetylcholine within the brain. Current therapies for Alzheimer’s-type dementia are based upon the administration of cholinesterase

enzyme inhibitors that prolong the effective half-life of remaining acetylcholine molecules within the brain.

As discussed in chapter 1, the human body contains many different molecules and thus offers many opportunities for the discovery of lead compounds based on endogenous molecules. Nowhere is this opportunity more apparent than in the area of peptide neu-rotransmitters and peptide hormones (see chapters 4 and 5). Neuneu-rotransmitters and hor-mones are endogenous messengers, controlling diverse biochemical processes within the body. Not surprisingly, they have the capacity to be ideal starting points in the drug discovery process. However, there are a number of major problems that must be con-fronted when exploiting peptides or proteins as lead compounds for drug discovery.

Peptidomimetic chemistry is an attempt to address these problems.

3.2.3.1 Peptidomimetic Chemistry as a Source of Lead Compounds

Although they are potent endogenous bioactive molecules, peptides rarely make good drugs. There are several reasons for the failings of peptides as drugs:

1. Peptides are too big (molecular weight frequently over 1000 dalton).

2. Peptides are often too flexible (thus binding with too many receptors, leading to toxicity).

3. Peptides contain amide bonds that can be metabolized by hydrolysis.

4. Peptides cannot be given orally as drugs (they tend to be digested).

5. Peptides do not readily cross the blood–brain barrier to enter the brain.

Despite these obvious deficiencies, peptides have a number of properties that make them attractive as starting points in drug design:

1. Peptides contain numerous stereogenic (chiral) centers (an excellent starting point when designing stereoselective drugs).

2. Peptides contain many functional groups (e.g., carboxylate, ammonium, hydroxyl, thiol) that can readily constitute functional groups within a pharmacophore; since receptors are usually proteins, peptides are good starting points for designing a molecule to interact with a receptor, owing to the energetically favorable nature of peptide–peptide interactions.

3. Peptides are easily synthesized and many analogs can be readily produced.

4. Peptides can have their conformation and geometries easily optimized by energy minimization calculations using current computational methods (e.g., molecular mechanics); this makes subsequent modeling studies easy.

5. Peptides function as neurotransmitters and hormones and thus are good starting materials when designing bioactive molecules.

Since peptides are ideal starting molecules that cannot be turned into successful peptidic drugs, the specialty area of peptidomimetic chemistry has emerged. The goal of pep-tidomimetic chemistry is to design small, conformationally constrained, non-peptidic organic molecules that possess the biological properties of a peptide. Hopefully, this will retain the strength of the peptide as a putative drug while eliminating the problems. There are two approaches whereby peptidomimetic chemistry can achieve this design goal.

The first approach is shown in figure 3.1. This approach uses various techniques (e.g., alanine scanning) to identify the smallest peptide segment with biological activity within the overall peptide. This “minimal bioactive segment” may be cyclized or have its stereochemistry altered in order to attain restriction of conformational freedom and Figure 3.1 Peptidomimetic chemistry attempts to produce a non-peptidic drug to mimic a bioactive peptide. In Step A, the smallest bioactive fragment of the larger peptide is identified; in Step B, a process such as an alanine scan is used to identify which of the amino acids are impor-tant for bioactivity; in Step C, individual amino acids have their configuration changed from the naturally occurring L-configuration to the unnatural D-configuration (in an attempt to make the peptide less “naturally peptidic”); in Step D, individual amino acids are replaced with atypical unnatural amino acids and amino acid mimics; in Step E the peptide is cyclized to constrain it con-formationally; finally, in Step F, fragments of the cyclic peptide are replaced with bioisosteres in an attempt to make a non-peptidic organic molecule.

to render the molecule less like a naturally occurring peptide. Next, this segment is then rebuilt isosteric fragment by isosteric fragment, gradually replacing each portion of the molecule in a stepwise fashion. For example, the amide bond may be replaced by a bioisosterically equivalent amide bioisostere. In this fashion, an equivalent but non-peptidic organic molecule drug eventually emerges.

An alternative approach is a little less plodding and perhaps a little more elegant. The three-dimensional structure of the peptide is determined using either theoretical (molec-ular mechanics, molec(molec-ular orbital calculations) or experimental (X-ray crystallographic, NMR spectroscopic) methods. Next, an educated guess (hopefully based on some exper-imental data) is made to suggest which portion of the peptide is the pharmacophore. The geometries of the functional groups within the pharmacophore are then measured from the theoretical and experimental studies of the peptide’s geometry and conformation. For example, these data may show that the peptide pharmacophore contains a carboxylate group located 4.6 Å from a hydroxyl group, which in turn is 5.1 Å from a phenyl group.

Using these precise data, databases of known organic molecules are then computation-ally searched to identify an organic molecule with similar functional groups held in the same position in three-dimensional space. Hopefully, this will yield a non-peptidic but bioactive organic molecule drug.

3.2.3.2 Other Endogenous Drug Lead Platforms: Carbohydrates, Nucleic Acids Although peptides have been studied the most extensively, there are other endogenous molecules within the human body worthy of exploitation as drug discovery platforms.

These include nucleic acids, lipids, and carbohydrates, which are discussed in detail in chapter 8. These molecules share the same potential strengths and weaknesses as do peptides. Likewise, there is a need to develop small organic molecules as mimetics of these other endogenous molecules. Although not as clearly defined as peptidomimetic chemistry, ultimately, “nucleotidomimetic” or “carbohydromimetic” chemistries may eventually emerge as new design strategies for lead compound identification.

3.2.4 Lead Compound Identification from Exogenous