Drug Design: Optimizing Access to the Target PDF

5. Drug design: optimizing access to the target In Chapter 4, we looked at drug design strategies aimed at optimizing the binding interactions of a drug with its target. However, the compound with the best binding interactions is not necessarily the best drug to use in medicine. The drug n...

5. Drug design: optimizing access to the target In Chapter 4, we looked at drug design strategies aimed at optimizing the binding interactions of a drug with its target. However, the compound with the best binding interactions is not necessarily the best drug to use in medicine. The drug needs to overcome many barriers if it is to reach its target in the body. In this chapter, we shall study design strategies which can be used to counter such barriers, and which involve modification of the drug itself. There are other methods of aiding a drug in reaching its target, which include linking the drug to polymers or antibodies, or encapsulating it within a polymeric carrier. In general, the aim is to design drugs that will be absorbed into the blood supply, will reach their target efficiently, be stable enough to survive the journey, and will be eliminated in a reasonable period of time. This all comes under the banner of a drug’s pharmacokinetics. 1 5.1. Optimizing hydrophilic or hydrophobic properties The relative hydrophilic/hydrophobic properties of a drug are crucial in influencing its solubility, absorption, distribution, metabolism, and excretion (ADME). Drugs which are too polar or too hydrophilic do not cross the cell membranes of the gut wall easily. One way round this is to inject them, but they cannot be used against intracellular targets as they will not cross cell membranes. They are also likely to have polar functional groups which will make them prone to plasma protein binding, metabolic phase II conjugation reactions, and rapid excretion. Very hydrophobic drugs fare no better. If they are administered orally, they are likely to be dissolved in fat globules in the gut and will be poorly absorbed. If they are injected, they are poorly soluble in blood and are likely to be taken up by fat tissue, resulting in low circulating levels. It has also been observed that toxic metabolites are more likely to be formed from hydrophobic drugs. 2  The hydrophobic character of a drug can be measured experimentally by testing the drug’s relative distribution in an n-octanol/water mixture. Hydrophobic molecules will prefer to dissolve in the n-octanol layer of this two phase system, whereas hydrophilic molecules will prefer the aqueous layer. The relative distribution is known as the partition coefficient (P) and is obtained from the following equation: Hydrophobic compounds have a high P value, whereas hydrophilic compounds have a low P value. In fact, log P values are normally used as a measure of hydrophobicity. 3  Many drugs can exist as an equilibrium between an ionized and an un-ionized form. However, log P measures only the relative distribution of the un-ionized species between water and octanol. The relative distribution of all species (both ionized and un-ionized) is given by log D. As a postscript, the hydrophilic/hydrophobic properties of a drug are not the only factors that influence drug absorption and oral bioavailability. Molecular flexibility also has an important role in oral bioavailability, and so the tactics of rigidification described in Chapter 4 can be useful in improving drug absorption. 4 5.1.1. Masking polar functional groups to decrease polarity Molecules can be made less polar by masking a polar functional group with an alkyl or acyl group.  For example, an alcohol or a phenol can be converted to an ether or ester, a carboxylic acid can be converted to an ester or amide, and primary and secondary amines can be converted to amides or to secondary and tertiary amines. Polarity is decreased not only by masking the polar group, but by the addition of an extra hydrophobic alkyl group – larger alkyl groups having a greater hydrophobic effect. One has to be careful in masking polar groups, though, as they may be important in binding the drug to its target. Masking such groups would decrease binding interactions and lower activity. If this is the case, it is often useful to mask the polar group temporarily such that the mask is removed once the drug is absorbed. 5 5.1.2. Adding or removing polar functional groups to vary polarity A polar functional group could be added to a drug to increase its polarity.  For example, the antifungal agent tioconazole is only used for skin infections because it is non-polar and poorly soluble in blood.  Introducing a polar hydroxyl group and more polar heterocyclic rings led to the orally active antifungal agent fluconazole, with improved solubility and enhanced activity against systemic infection (i.e. in the blood supply) (Figure 23). 6  Nitrogen-containing heterocycles (e.g. morpholine or pyridine) are often added to drugs in order to increase their polarity and water solubility. This is because the nitrogen is basic in character and it is possible to form water-soluble salts. If a polar group is added in order to increase water solubility, it is preferable to add it to the molecule in such a way that it is still exposed to surrounding water when the drug is bound to the target binding site. This means that energy does not have to be expended in desolvation. The polarity of an excessively polar drug can be lowered by removing polar functional groups. This strategy has been particularly successful with lead compounds derived from natural sources (e.g. alkaloids or endogenous peptides). It is important, though, not to remove functional groups which are important to the drug’s binding interactions with its target. 7  In some cases, a drug may have too many essential polar groups.  For example, the antibacterial agent shown in figure 24 has good in vitro activity but poor in vivo activity, because of the large number of polar groups.  Some of these groups can be removed or masked, but most of them are required for activity.  As a result, the drug cannot be used clinically. 8 5.1.3. Varying hydrophobic substituents to vary polarity Polarity can be varied by the addition, removal, or variation of suitable hydrophobic substituents.  For example, extra alkyl groups could be included within the carbon skeleton of the molecule to increase hydrophobicity if the synthetic route permits. Alternatively, alkyl groups already present might be replaced with larger groups. If the molecule is not sufficiently polar, then the opposite strategy can be used (i.e. replacing large alkyl groups with smaller alkyl groups or removing them entirely). Sometimes there is a benefit in increasing the size of one alkyl group and decreasing the size of another. This is called a methylene shuffle and has been found to modify the hydrophobicity of a compound. The addition of halogen substituents also increases hydrophobicity. Chloro or fluoro substituents are commonly used, and, less commonly, a bromo substituent. 9 5.1.4. Variation of N-alkyl substituents to vary pKa Drugs with a pKa outside the range 6 – 9 tend to be too strongly ionized and are poorly absorbed through cell membranes. The pKa can often be altered to bring it into the preferred range. For example, this can be done by varying any N-alkyl substituents that are present. However, it is sometimes difficult to predict how such variations will affect the pKa. Extra N-alkyl groups or larger N-alkyl groups have an increased electron-donating effect which should increase basicity, but increasing the size or number of alkyl groups increases the steric bulk around the nitrogen atom. This hinders water molecules from solvating the ionized form of the base and prevents stabilization of the ion. This, in turn, decreases the basicity of the amine. Therefore, there are two different effects acting against each other. Nevertheless, varying alkyl substituents is a useful tactic to try. 10  A variation of this tactic is to ‘wrap up’ a basic nitrogen within a ring.  For example, the benzamidine structure (I in figure 25) has anti-thrombotic activity, but the amidine group present is too basic for effective absorption.  Incorporating the group into an isoquinoline ring system (PRO 3112) reduced basicity and increased absorption. 11 5.1.5. Variation of aromatic substituents to vary pKa The pKa of an aromatic amine or carboxylic acid can be varied by adding electron- donating or electron-withdrawing substituents to the ring. The position of the substituent relative to the amine or carboxylic acid is important if the substituent interacts with the ring through resonance. 12 5.1.6. Bioisosteres for polar groups The use of bioisosteres has already been described in the design of compounds with improved target interactions. Bioisosteres have also been used as substitutes for important functional groups that are required for target interactions, but which pose pharmacokinetic problems. For example, a carboxylic acid is a highly polar group which can ionize and hinder absorption of any drug containing it. One way of getting around this problem is to mask it as an ester prodrug. Another strategy is to replace it with a bioisostere which has similar physicochemical properties, but which offers some advantage over the original carboxylic acid. Several bioisosteres have been used for carboxylic acids, but among the most popular is a 5-substituted tetrazole ring (Figure 26). 13  Like carboxylic acids, tetrazoles contain an acidic proton and are ionized at pH 7.4. They are also planar in structure. However, they have an advantage in that the tetrazole anion is 10 times more lipophilic than a carboxylate anion and drug absorption is enhanced as a result. They are also resistant to many of the metabolic reactions that occur on carboxylic acids. N-Acylsulphonamides have also been used as bioisosteres for carboxylic acids. Phenol groups are also commonly present in drugs but are susceptible to metabolic conjugation reactions. Various bioisosteres involving amides, sulphonamides, or heterocyclic rings have been used where an N-H group mimics the phenol O-H group. 14 5.2. Making drugs more resistant to chemical and enzymatic degradation There are various strategies that can be used to make drugs more resistant to hydrolysis and drug metabolism, and thus prolong their activity. 15 5.2.1. Steric shields Some functional groups are more susceptible to chemical and enzymatic degradation than others.  For example, esters and amides are particularly prone to hydrolysis.  A common strategy that is used to protect such groups is to add steric shields, designed to hinder the approach of a nucleophile or an enzyme to the susceptible group.  These usually involve the addition of a bulky alkyl group close to the functional group.  For example, the t-butyl group in the anti-rheumatic agent D 1927 serves as a steric shield and blocks hydrolysis of the terminal peptide bond (Figure 27). 16 5.2.2. Electronic effects of bioisosteres Another popular tactic used to protect a labile functional group is to stabilize the group electronically using a bioisostere. Isosteres and non-classical isosteres are frequently used as bioisosteres. For example, replacing the methyl group of an ethanoate ester with NH2 results in a urethane functional group which is more stable than the original ester. The NH2 group is the same valency and size as the methyl group and, therefore, has no steric effect, but it has totally different electronic properties as it can feed electrons into the carboxyl group and stabilize it from hydrolysis. 17  The cholinergic agonist carbachol is stabilized in this way, as is the cephalosporin cefoxitin. Alternatively, a labile ester group could be replaced by an amide group (NH replacing O).  Amides are more resistant to chemical hydrolysis, due, again, to the lone pair of the nitrogen feeding its electrons into the carbonyl group and making it less electrophilic. 18  It is important to realize that bioisosteres are often specific to a particular area of medicinal chemistry. Replacing an ester with a urethane or an amide may work in one category of drugs but not another. One must also appreciate that bioisoteres are different from isosteres. It is the retention of important biological activity that determines whether a group is a bioisostere, not the valency. Therefore, non-isosteric groups can be used as bioisosteres. One is not confined to the use of bioisosteres to increase stability. Groups or substituents having an inductive electronic effect have frequently been incorporated into molecules to increase the stability of a labile functional group. 19 5.2.3. Steric and electronic modifications Steric hindrance and electronic stabilization have often been used together to stabilize labile groups.  For example, procaine (Figure 28) is a good, but short-lasting, local anaesthetic because its ester group is quickly hydrolysed.  Changing the ester group to the less reactive amide group reduces susceptibility to chemical hydrolysis.  Furthermore, the presence of two ortho-methyl groups on the aromatic ring helps to shield the carbonyl group from attack by nucleophiles or enzymes.  This results in the longer-acting local anaesthetic lidocaine. 20 21 5.2.4. Metabolic blockers Some drugs are metabolized by the introduction of polar groups at particular positions in their skeleton.  For example, steroids can be oxidized at position 6 of the tetracyclic ring to introduce a polar hydroxyl group.  The introduction of this group allows the formation of polar conjugates which can be eliminated quickly from the system.  By introducing a methyl group at position 6, metabolism is blocked and the activity of the steroid is prolonged. 22  The oral contraceptive megestrol acetate is an agent which contains a 6-methyl blocking group. 23  On the same lines, a popular method of protecting aromatic rings from metabolism at the para-position is to introduce a fluoro substituent.  For example, CGP 52411 is an enzyme inhibitor which acts on the kinase-active site of the epidermal growth factor receptor.  It went forward for clinical trials as an anticancer agent and was found to undergo oxidative metabolism at the para-position of the aromatic rings.  Fluoro-substituents were successfully added in the analogue CGP 53353 to block this metabolism. 24 5.2.5. Removal or replacement of susceptible metabolic groups Certain chemical groups are particularly susceptible to metabolic enzymes.  For example, methyl groups on aromatic rings are often oxidized to carboxylic acids.  These acids can then be quickly eliminated from the body.  Other common metabolic reactions include aliphatic and aromatic C-hydroxylations, N- and S-oxidations, O- and S-dealkylations, and deaminations. 25  Susceptible groups can sometimes be removed or replaced by groups that are stable to oxidation in order to prolong the lifetime of the drug.  For example, the aromatic methyl substituent of the antidiabetic tolbutamide was replaced by a chloro substituent to give chlorpropamide, which is much longer lasting. An alternative strategy which is often tried is to replace the susceptible methyl group with CF3, CHF2, or CH2F. The fluorine atoms alter the oxidation potential of the methyl group and make it more resistant to oxidation. 26 5.2.6. Group shifts Removing or replacing a metabolically vulnerable group is feasible if the group concerned is not involved in important binding interactions with the binding site. If the group is important, then we have to use a different strategy. There are two solutions. We can either mask the vulnerable group on a temporary basis by using a prodrug or we can try shifting the vulnerable group within the molecular skeleton. 27  The latter tactic was used in the development of salbutamol.  Salbutamol was introduced in 1969 for the treatment of asthma and is an analogue of the neurotransmitter noradrenaline – a catechol structure containing ortho-phenolic groups. 28  One of the problems faced by catechol compounds is metabolic methylation of one of the phenolic groups. As both phenol groups are involved in hydrogen bonds to the receptor, methylation of one of the phenol groups disrupts the hydrogen bonding and makes the compound inactive. For example, the noradrenaline analogue (I in figure 30) has useful anti-asthmatic activity, but the effect is of short duration because the compound is rapidly metabolized to the inactive methyl ether (II in figure 30). 29  Removing the OH or replacing it with a methyl group prevents metabolism, but also prevents the important hydrogen bonding interactions with the binding site. So how can this problem be solved? The answer was to move the vulnerable hydroxyl group out from the ring by one carbon unit. This was enough to make the compound unrecognizable to the metabolic enzyme, but not to the receptor binding site. Fortunately, the receptor appears to be quite lenient over the position of this hydrogen bonding group and it is interesting to note that a hydroxyethyl group is also acceptable (Figure 31). Beyond that, activity is lost because the OH group is out of range or the substituent is too large to fit. These results demonstrate that it is better to consider a binding region within the receptor binding site as an available volume, rather than imagining it as being fixed at one spot. A drug can then be designed such that the relevant binding group is positioned in any part of that available volume. 30  Shifting an important binding group that is metabolically susceptible cannot be guaranteed to work in every situation. It may well make the molecule unrecognizable both to its target and to the metabolic enzyme. 31 5.2.7. Ring variation and ring substituents Certain ring systems may be susceptible to metabolism and so varying the ring might improve metabolic stability. This can be done by adding a nitrogen into the ring to lower the electron density of the ring system.  For example, the imidazole ring of the antifungal agent tioconazole mentioned previously is susceptible to metabolism, but replacement with a 1,2,4-triazole ring, as in fluconazole, results in improved stability. Electron-rich aromatic rings, such as phenyl groups, are particularly prone to oxidative metabolism, but can be stabilized by replacing them with nitrogen-containing heterocyclic rings, such as pyridine and pyrimidine. Alternatively, electron-withdrawing substituents could be added to the aromatic ring to lower the electron density. 32  Ring variation can also help to stabilize metabolically susceptible aromatic or heteroaromatic methyl substituents. Such substituents could be replaced with more stable substituents, but sometimes the methyl substituent has to be retained for good activity. In such cases, introducing a nitrogen atom into the aromatic/heteroaromatic ring can be beneficial, as lowering the electron density in the ring also helps to make the methyl substituent more resistant to metabolism.  For example, F13640 underwent phase II clinical trials as an analgesic (Figure 32).  The methyl substituent on the pyridine ring is susceptible to oxidation and is converted to a carboxylic acid, which is inactive.  The methyl group plays an important binding role and has to be present. 33  Therefore, the pyridine ring was changed to a pyrimidine ring resulting in a compound (F15599) that has increased metabolic stability without affecting binding affinity. 34 5.3. Making drugs less resistant to drug metabolism So far, we have looked at how the activity of drugs can be prolonged by inhibiting their metabolism. However, a drug that is extremely stable to metabolism and is very slowly excreted can pose just as many problems as one that is susceptible to metabolism. It is usually desirable to have a drug that does what it is meant to do, then stops doing it within a reasonable time. If not, the effects of the drug could last too long and cause toxicity and lingering side effects. Therefore, designing drugs with decreased chemical and metabolic stability can sometimes be useful. 35 5.3.1. Introducing metabolically susceptible groups Introducing groups that are susceptible to metabolism is a good way of shortening the lifetime of a drug.  For example, a methyl group was introduced to the anti-arthritic agent L 787257 to shorten its lifetime.  The methyl group of the resulting compound (L 791456) was metabolically oxidized to a polar alcohol, as well as to a carboxylic acid (Figure 33). 36 5.3.2. Self-destruct drugs A self-destruct drug is one which is chemically stable under one set of conditions, but becomes unstable and degrades spontaneously under another set of conditions. The advantage of a self-destruct drug is that inactivation does not depend on the activity of metabolic enzymes, which could vary from patient to patient. The best example of a self-destruct drug is the neuromuscular blocking agent atracurium, which is stable at acid pH but self-destructs when it meets the slightly alkaline conditions of the blood. This means that the drug has a short duration of action, allowing anaesthetists to control its blood levels during surgery by providing it as a continuous, intravenous drip. 37 38 5.4. Targeting drugs One of the major goals in drug design is to find ways of targeting drugs to the exact locations in the body where they are most needed. The principle of targeting drugs can be traced back to Paul Ehrlich, who developed antimicrobial drugs that were selectively toxic for microbial cells over human cells. Drugs can also be made more selective to distinguish between different targets within the body. Here, we discuss other tactics related to the targeting of drugs. 39 5.4.1. Targeting tumour cells: ‘search and destroy’ drugs A major goal in cancer chemotherapy is to target drugs efficiently against tumour cells rather than normal cells. One method of achieving this is to design drugs which can make use of specific molecular transport systems. The idea is to attach the active drug to an important ‘building block’ molecule that is needed in large amounts by the rapidly dividing tumour cells. This could be an amino acid or a nucleic acid base. Of course, normal cells require these building blocks as well, but tumour cells often grow more quickly than normal cells and require the building blocks more urgently. Therefore, the uptake is greater in tumour cells. 40  A more recent idea has been to attach the active drug (or a poison such as ricin) to monoclonal antibodies which can recognize antigens unique to the tumour cell. Once the antibody binds to the antigen, the drug or poison is released to kill the cell. The difficulties in this approach include the identification of suitable antigens and the production of antibodies in significant quantity. Nevertheless, the approach has great promise for the future. Another tactic which has been used to target anticancer drugs is to administer an enzyme- antibody conjugate where the enzyme serves to activate an anticancer prodrug, and the antibody directs the enzyme to the tumour. This is a strategy known as ADEPT (Antibody-directed enzyme prodrug therapy). 41 5.4.2. Targeting gastrointestinal infections If a drug is to be targeted against an infection of the gastrointestinal tract, it must be prevented from being absorbed into the blood supply. This can be done by using a fully ionized drug that is incapable of crossing cell membranes.  For example, highly ionized sulphonamides are used against gastrointestinal infections. 42 5.4.3. Targeting peripheral regions rather than the central nervous system It is often possible to target drugs such that they act peripherally and not in the central nervous system (CNS). By increasing the polarity of drugs, they are less likely to cross the blood-brain barrier, and this means they are less likely to have CNS side effects. Achieving selectivity for the CNS over the peripheral regions of the body is not straightforward. In order to achieve that, the drug would have to be designed to cross the blood-brain barrier efficiently, while being metabolized rapidly to inactive metabolites in the peripheral system. 43 5.4.4. Targeting with membrane tethers Several drug targets are associated with cell membranes and one way of targeting drugs to these targets is to attach membrane tethers to the drug such that the molecule is anchored in the membrane close to the target.  The antibacterial agent teicoplanin is one such example.  Another membrane-tethered drug has been designed to inhibit the enzyme b-secretase, with the ultimate aim of treating Alzheimer’s disease (AD).  This enzyme generates the proteins that are responsible for the toxic protein aggregates found in the brains of AD sufferers, and does so mainly in cellular organelles called endosomes.  A peptide transition-state inhibitor has been linked to a sterol such that it is taken into endosomes by endocytosis.  This sterol then acts as the membrane tether to lock the drug in position, such that it targets b-secretase in endosomes rather than b-secretase in other locations. 44  Potential agents for AD treatment are also being targeted to mitochondria where AD leads to the generation of radicals and oxidation reactions that are damaging to the cell. MitoQ is an agent undergoing clinical trials which contains an antioxidant prodrug linked to a hydrophobic triphenylphosphine moiety. The latter group aids the drug’s entry into mitochondria, then tethers it to the phospholipid bilayers of the mitochondria membrane. The quinone ring system is reduced rapidly to the active quinol form which can then act as an antioxidant to neutralize free radicals. 45 5.5. Reducing toxicity It is often found that a drug fails clinical trials because of toxic side effects. This may be due to toxic metabolites, in which case the drug should be made more resistant to metabolism as described earlier. It is also worth checking to see whether there are any functional groups present that are particularly prone to producing toxic metabolites.  For example, it is known that functional groups, such as aromatic nitro groups, aromatic amines, bromoarenes, hydrazines, hydroxylamines, or polyhalogenated groups, are often metabolized to toxic products. 46  Side effects might also be reduced or eliminated by varying apparently harmless substituents.  For example, the halogen substituents of the antifungal agent UK 47265 were varied in order to find a compound that was less toxic to the liver.  This led to the successful antifungal agent fluconazole (Figure 34). 47  Varying the position of substituents can sometimes reduce or eliminate side effects.  For example, the dopamine antagonist SB 269652 inhibits cytochrome P450 enzymes as a side effect.  Placing the cyano group at a different position prevented this inhibition (Figure 35). 48 5.6. Prodrugs Prodrugs are compounds which are inactive in themselves, but which are converted in the body to the active drug. They have been useful in tackling problems such as acid sensitivity, poor membrane permeability, drug toxicity, bad taste, and short duration of action. Usually, a metabolic enzyme is involved in converting the prodrug to the active drug, and so a good knowledge of drug metabolism and the enzymes involved allows the medicinal chemist to design a suitable prodrug which turns drug metabolism into an advantage rather than a problem. Prodrugs have been designed to be activated by a variety of metabolic enzymes. 49  Ester prodrugs which are hydrolysed by esterase enzymes are particularly common, but prodrugs have also been designed which are activated by N-demethylation, decarboxylation, and the hydrolysis of amides and phosphates. Not all prodrugs are activated by metabolic enzymes, however.  For example, photodynamic therapy involves the use of an external light source to activate prodrugs. When designing prodrugs, it is important to ensure that the prodrug is effectively converted to the active drug once it has been absorbed into the blood supply, but it is also important to ensure that any groups that are cleaved from the molecule are non-toxic. 50 5.6.1. Prodrugs to improve membrane permeability 5.6.1.1. Esters as prodrugs Prodrugs have proved very useful in temporarily masking an ‘awkward’ functional group which is important to target binding but which hinders the drug from crossing the cell membranes of the gut wall.  For example, a carboxylic acid functional group may have an important role to play in binding a drug to its binding site via ionic or hydrogen bonding.  However, the very fact that it is an ionisable group may prevent it from crossing a fatty cell membrane.  The answer is to protect the acid function as an ester.  The less polar ester can cross fatty cell membranes and, once it is in the blood stream, it is hydrolysed back to the free acid by esterases in the blood. 51  Examples of ester prodrugs used to aid membrane permeability include enalapril, which is the prodrug for the anti-hypertensive agent enalaprilate, and pivampicillin, talampicillin and bacampicillin, which are prodrugs for ampicillin. 52  Not all esters are hydrolysed equally efficiently and a range of esters may need to be tried to find the best one. It is possible to make esters more susceptible to hydrolysis by introducing electron- withdrawing groups to the alcohol moiety (e.g. OCH2CF3, OCH2CO2R, OCONR2, OAr). The inductive effect of these groups aids the hydrolytic mechanism by stabilizing the alkoxide leaving group. Care has to be taken, however, not to make the ester too reactive in case it becomes chemically unstable and is hydrolysed by the acid conditions of the stomach or the more alkaline conditions of the intestine before it reaches the blood supply. 53 5.6.1.2. N-methylated prodrugs N-demethylation is a common metabolic reaction in the liver, so polar amines can be N- methylated to reduce polarity and improve membrane permeability. Several hypnotics and anti-epileptics take advantage of this reaction, for example hexobarbitone (Figure 36). 54 5.6.1.3. Trojan horse approach for transport proteins Another way round the problem of membrane permeability is to design a prodrug which can take advantage of transport proteins in the cell membrane, such as the ones responsible for carrying amino acids into a cell. A well-known example of such a prodrug is levodopa.  Levodopa is a prodrug for the neurotransmitter dopamine and has been used in the treatment of Parkinson’s disease – a condition due primarily to a deficiency of that neurotransmitter in the brain.  Dopamine itself cannot be used as it is too polar to cross the blood-brain barrier. 55  Levodopa is even more polar and seems an unlikely prodrug, but it is also an amino acid, and so it is recognized by the transport proteins for amino acids which carry it across the cell membrane. Once in the brain, a decarboxylase enzyme removes the acid group and generates dopamine. 56 5.6.2. Prodrugs to prolong drug activity Sometimes prodrugs are designed to be converted slowly to the active drug, thus prolonging a drug’s activity.  For example, 6-mercaptopurine (Figure 37) suppresses the body’s immune response and is, therefore, useful in protecting donor grafts.  Unfortunately, the drug tends to be eliminated from the body too quickly.  The prodrug azathioprine has the advantage that it is slowly converted to 6- mercaptopurine by being attacked by glutathione, allowing a more sustained activity. 57  The rate of conversion can be altered, depending on the electron-withdrawing ability of the heterocyclic group. The greater the electron-withdrawing power, the faster the breakdown.  The NO2 group is therefore present to ensure an efficient conversion to 6-mercaptopurine, as it is strongly electron-withdrawing on the heterocyclic ring. 58  Another approach to maintaining a sustained level of drug over long periods is to deliberately associate a very lipophilic group to the drug. This means that most of the drug is stored in fat tissue from where it is steadily and slowly released into the blood stream. The antimalarial agent Cycloguanil pamoate is one such agent. The active drug is bound ionically to an anion containing a large lipophilic group and is only released into the blood supply following slow dissociation of the ion complex. Similarly, lipophilic esters of the antipsychotic drug fluphenazine are used to prolong its action.  The prodrug is given by intramuscular injection and slowly diffuses from fat tissue into the blood supply, where it is rapidly hydrolysed. 59 60 5.6.3. Prodrugs masking drug toxicity and side effects Prodrugs can be used to mask the side effects and toxicity of drugs.  For example, salicylic acid is a good painkiller, but causes gastric bleeding because of the free phenolic group. This is overcome by masking the phenol as an ester (aspirin). The ester is later hydrolysed to free the active drug. Prodrugs can be used to give a slow release of drugs that would be too toxic to give directly. Propiolaldehyde is useful in the aversion therapy of alcohol, but is not used itself because it is an irritant. The prodrug pargyline can be converted to propiolaldehyde by enzymes in the liver. 61  Cyclophosphamide (Figure 38) is a successful, non-toxic prodrug which can be safely taken orally. Once absorbed, it is metabolized in the liver to a toxic alkylating agent which is useful in the treatment of cancer. 62 5.6.4. Prodrugs to lower water solubility Some drugs have a revolting taste! One way to avoid this problem is to reduce their water solubility to prevent them dissolving on the tongue.  For example, the bitter taste of the antibiotic chloramphenicol can be avoided by using the palmitate ester.  This is more hydrophobic because of the masked alcohol and the long chain fatty group that is present.  It does not dissolve easily on the tongue and is quickly hydrolysed once swallowed. 63 5.6.5. Prodrugs to improve water solubility Prodrugs have been used to increase the water solubility of drugs. This is particularly useful for drugs which are given intravenously, as it means that higher concentrations and smaller volumes can be used.  For example, the succinate ester of chloramphenicol increases the latter’s water solubility because of the extra carboxylic acid that is present.  Once the ester is hydrolysed, chloramphenicol is released along with succinic acid, which is naturally present in the body. 64  Prodrugs designed to increase water solubility have proved useful in preventing the pain associated with some injections, which is caused by the poor solubility of the drug at the site of injection.  For example, the antibacterial agent clindamycin is painful when injected, but this is avoided by using a phosphate ester prodrug which has much better solubility because of the ionic phosphate group. 65 5.6.6. Prodrugs used in the targeting of drugs Methenamine is a stable, inactive compound when the pH is more than 5. At a more acidic pH, however, the compound degrades spontaneously to generate formaldehyde, which has antibacterial properties. This is useful in the treatment of urinary tract infections. The normal pH of blood is slightly alkaline (7.4) and so methenamine passes round the body unchanged. However, once it is excreted into the infected urinary tract, it encounters urine which is acidic as a result of certain bacterial infections. Consequently, methenamine degrades to generate formaldehyde just where it is needed. 66  Prodrugs of sulphonamides have also been used to target intestinal infections.  For example, succinyl sulphathiazole is a prodrug of sulphathiazole (Figure 39).  The succinyl moiety contains an acidic group which means that the prodrug is ionized in the intestine.  As a result, it is not absorbed into the bloodstream and is retained in the intestine.  Slow enzymatic hydrolysis of the succinyl group then releases the active sulphathiazole where it is needed. 67 5.6.7. Prodrugs to increase chemical stability The antibacterial agent ampicillin decomposes in concentrated aqueous solution as a result of intramolecular attack of the side chain amino group on the lactam ring. Hetacillin is a prodrug which locks up the offending nitrogen in a ring and prevents this reaction. Once the prodrug has been administered, hetacillin slowly decomposes to release ampicillin and acetone. 68 5.6.8. Prodrugs activated by external influence (sleeping agents) Conventional prodrugs are inactive compounds which are normally metabolized in the body to the active form. A variation of the prodrug approach is the concept of a ‘sleeping agent’. This is an inactive compound which is only converted to the active drug by some form of external influence. The best example of this approach is the use of photosensitizing agents (such as porphyrins or chlorins in cancer treatment) – a strategy known as photodynamic therapy. Given intravenously, these agents accumulate within cells and have some selectivity for tumour cells. By themselves, these agents have little effect, but if the cancer cells are irradiated with light, the porphyrins are converted to an excited state and react with molecular oxygen to produce highly toxic singlet oxygen. 69 5.7. Endogenous compounds as drugs Endogenous compounds are molecules which occur naturally in the body. Many of these could be extremely useful in medicine. For example, the body’s hormones are natural chemical messengers, so why not use them as medicines instead of synthetic drugs that are foreign to the body? In this section, we look at important molecules, such as neurotransmitters, hormones, peptides, proteins, and antibodies, to see how feasible it is to use them as drugs. 70 5.7.1. Neurotransmitters Many non-peptide neurotransmitters are simple molecules which can be prepared easily in the laboratory, so why are these not used commonly as drugs? For example, if there is a shortage of dopamine in the brain, why not administer more dopamine to make up the balance? Unfortunately, this is not possible for a number of reasons. Many neurotransmitters are not stable enough to survive the acid of the stomach and would have to be injected. Even if they were injected, there is little chance that they would survive to reach their target receptors. The body has efficient mechanisms which inactivate neurotransmitters as soon as they have passed on their message from nerve to target cell. Therefore, any neurotransmitter injected into the blood supply would be swiftly inactivated by enzymes, or taken up by cells via transport proteins. 71  Even if they were not inactivated or removed, they would be poor drugs indeed, leading to many undesirable side effects. For example, the shortage of neurotransmitter may only be at one small area in the brain; the situation may be normal elsewhere. If we gave the natural neurotransmitter, how would we stop it producing an overdose of transmitter at these other sites? Of course, this is a problem with all drugs, but it has been discovered that the receptors for a specific neurotransmitter are not all identical. There are different types and subtypes of a particular receptor, and their distribution around the body is not uniform. One subtype of receptor may be common in one tissue, whereas a different subtype is common in another tissue. The medicinal chemist can design synthetic drugs which can take advantage of that difference, ignoring receptor subtypes which the natural neurotransmitter would not. In this respect, the medicinal chemist has actually improved on nature. 72  We cannot even assume that the body’s own neurotransmitters are perfectly safe, and free from the horrors of tolerance and addiction associated with drugs such as heroin. It is quite possible to be addicted to one’s own neurotransmitters and hormones. For example, some people are addicted to exercise and are compelled to exercise long hours each day in order to feel good. The very process of exercise leads to the release of hormones and neurotransmitters which can produce a ‘high’, and this drives susceptible people to exercise more and more. If they stop exercising, they suffer withdrawal symptoms, such as deep depression. To conclude, many of the body’s own neurotransmitters are known and can be synthesized easily, but they cannot be used effectively as medicines. 73 5.7.2. Natural hormones, peptides, and proteins as drugs Unlike neurotransmitters, natural hormones have potential in drug therapy as they normally circulate around the body and behave like drugs. Indeed, adrenaline is commonly used in medicine to treat (among other things) severe allergic reactions. Most hormones are peptides and proteins, and some naturally occurring peptide and protein hormones are already used in medicine. These include insulin, calcitonin, erythropoietin, human growth factor, interferons, and colony stimulating factors. 74  The availability of many protein hormones owes a great deal to genetic engineering. It is extremely tedious and expensive to obtain substantial quantities of these proteins by other means. For example, isolating and purifying a hormone from blood samples is impractical because of the tiny quantities of hormone present. It is far more practical to use recombinant DNA techniques, whereby the human genes for the protein are cloned and then incorporated into the DNA of fast-growing bacterial, yeast, or mammalian cells. These cells then produce sufficient quantities of the protein. 75  Using these techniques, it is also possible to produce ‘cut down’ versions of important body proteins and polypeptides which can also be used therapeutically. For example, teriparatide is a polypeptide which has been approved for the treatment of osteoporosis, and was produced by recombinant DNA technology using a genetically modified strain of the bacterium Escherichia coli. It consists of 34 amino acids that represent the N-terminal end of human parathyroid hormone (consisting of 84 amino acids). Another recombinant protein that has been approved is etanercept, which is used for the treatment of rheumatoid arthritis. More than 80 polypeptide drugs have reached the market as a result of the biotechnology revolution, with more to come. 76  Recombinant enzymes have also been produced.  For example, glucarpidase is a carboxypeptidase enzyme which was recently approved in 2012.  It is administered to cancer patients with failed kidneys when they are taking the anticancer drug methotrexate.  The enzyme serves to metabolize methotrexate and prevent it from reaching toxic levels. Many endogenous peptides and proteins have proved ineffective though. This is because peptides and proteins suffer serious drawbacks, such as susceptibility to digestive and metabolic enzymes, poor absorption from the gut, and rapid clearance from the body. Furthermore, proteins are large molecules that could possibly induce an adverse immunological response. This involves the body producing antibodies against the proteins, resulting in serious side effects. 77  Solutions to some of these problems are appearing, though. It has been found that linking the polymer polyethylene glycol (PEG) to a protein can increase the latter’s solubility and stability, as well as decreasing the likelihood of an immune response. PEGylation, as it is called, also prevents the removal of small proteins from the blood supply by the kidneys or the reticuloendothelial system. The increased size of the PEGylated protein means that it is not filtered into the kidney nephrons and remains in the blood supply. The PEG molecules surrounding the protein can be viewed as a kind of hydrophilic, polymeric shield which both protects and disguises the protein. The PEG polymer has the added advantage that it shows little toxicity. The enzymes L-asparaginase and adenosine deaminase have been treated in this way to give protein-PEG conjugates called pegaspargase and pegademase, which have been used for the treatment of leukaemia and severe combined immunodeficiency (SCID) syndrome respectively. 78  SCID is an immunological defect associated with a lack of adenosine deaminase. The conjugates have longer plasma half-lives than the enzymes alone and are less likely to produce an immune response. Interferon has similarly been PEGylated to give a preparation called peginteferon a2b which is used for the treatment of hepatitis C. 79 5.7.3. Antibodies as drugs Biotechnology companies are producing an ever increasing number of antibodies and antibody-based drugs with the aid of genetic engineering and monoclonal antibody technology. Because antibodies can recognize the chemical signature of a particular cell or macromolecule, they have great potential in targeting cancer cells or viruses. Alternatively, they could be used to carry drugs or poisons to specific targets. Antibodies that recognize a particular antigen are generated by exposing a mouse to the antigen so that the mouse produces the desired antibodies (known as murine antibodies). 80  However, the antibodies themselves are not isolated. Antibodies are produced by cells called B lymphocytes, and it is a mixture of B lymphocytes that is isolated from the mouse. The next task is to find the B lymphocyte responsible for producing the desired antibody. This is done by fusing the mixture with immortal (cancerous) human B lymphocytes to produce cells called hybridomas. These are then separated and cultured. The culture that produces the desired antibody can then be identified by its ability to bind to the antigen, and is then used to produce antibody on a large scale. As all the cells in this culture are identical, the antibodies produced are also identical and are called monoclonal antibodies. 81  There was great excitement when this technology appeared in the 1980s which spawned an expectation that antibodies would be the magic bullet to tackle many diseases. Unfortunately, the early antibodies failed to reach the clinic, because they triggered an immune response in patients which resulted in antibodies being generated against the antibodies! In hindsight, this is not surprising: the antibodies were mouse-like in character and were identified as ‘foreign’ by the human immune system, resulting in the production of human anti-mouse antibodies (the HAMA response). In order to tackle this problem, chimeric antibodies have been produced which are part human (66%) and part mouse in origin, to make them less ‘foreign’. Genetic engineering has also been used to generate humanized antibodies which are 90% human in nature. 82  In another approach, genetic engineering has been used to insert the human genes responsible for antibodies into mice, such that the mice (transgenic mice) produce human antibodies rather than murine antibodies when they are exposed to the antigen. As a result of these efforts, 10 antibodies had reached the clinic in 2002 and were being used as immunosuppressants, antiviral agents, and anticancer agents. Many others are in the pipeline.  Omalizumab is an example of a recombinant humanized monoclonal antibody which targets immunoglobulin E (IgE) and was approved in 2003 for the treatment of allergic asthmatic disease.  It is known that exposure to allergens results in increased levels of IgE, which triggers the release of many of the chemicals responsible for the symptoms of asthma.  Omalizumab works by binding to IgE, thus preventing it from acting this way. 83  Another example is adalimumab, which was launched in 2003 and was the first fully humanized antibody to be approved. It is used for the treatment of rheumatoid arthritis and works by binding to an inflammatory molecule called a cytokine, specifically one called tumour necrosis factor (TNF-a). Molecules such as these are overproduced in arthritis, leading to chronic inflammation. By binding to the cytokine, the antibody prevents it interacting with the receptor. The antibody can also tag cells that are producing the chemical messenger, leading to the cell’s destruction by the body’s immune system. Other monoclonal antibodies undergoing clinical trials include reslizumab for the treatment of asthma. 84  Denosumab is a fully humanized monoclonal antibody that was approved in 2009 for the treatment of osteoporosis. Belimumab was approved in 2011 for the treatment of lupus – an autoimmune disease, while natalizumab was approved in 2004 for the treatment of multiple sclerosis. Work on the large-scale production of antibodies has also been continuing. They have traditionally been produced using hybridoma cells in bioreactors, but, more recently, companies have been looking at the possibility of using transgenic animals in order to collect antibodies in milk. Another possibility is to harvest transgenic plants which produce the antibody in their leaves or seeds. 85