Pharmacological plasticity—How do you hit a moving target?

Abstract Paul Ehrlich's concept of the magic bullet, by which a single drug induces pharmacological effects by interacting with a single receptor has been a strong driving force in pharmacology for a century. It is continually thwarted, though, by the fact that the treated organism is highly dynamic and the target molecule(s) is (are) never static. In this article, we address some of the factors that modify and cause the mobility and plasticity of drug targets and their interactions with ligands and discuss how these can lead to unexpected (lack of) effects of drugs. These factors include genetic, epigenetic, and phenotypic variability, cellular plasticity, chronobiological rhythms, time, age and disease resolution, sex, drug metabolism, and distribution. We emphasize four existing approaches that can be taken, either singly or in combination, to try to minimize effects of pharmacological plasticity. These are firstly, to enhance specificity using target conditions close to those in diseases, secondly, by simultaneously or thirdly, sequentially aiming at multiple targets, and fourthly, in synchronization with concurrent dietary, psychological, training, and biorhythm‐synchronizing procedures to optimize drug therapy.

hit a target with the accuracy and speed of a chemical bullet, the therapeutic target has an irritating habit of moving around.
Scientific research is an essentially rational occupation and phenomena are frequently reduced to the simplest system that can be studied. Look in any leading biomedical journal and the at- What do we mean by a moving target? The molecule aimed at with a drug may move location, from organ to organ or be transported within the body so that delivery of the drug is crucially important. The molecule may be present at a higher expression rate in some individuals or as a result of disease. In others it may be absent as a consequence of genetic differences such as single nucleotide polymorphisms so that the tested drug has no effect. As a corollary, the target itself may not move, but the treated organism can modify the drug, for instance by degrading the drug, meaning the "bullet" loses momentum and barely reaches the target. In addition, transport proteins (eg P-glycoprotein) are expressed which carry the drug back outside the targeted cells. With some drugs, the effects observed may be unrelated to the target at which it is aimed, because unbeknown to the medical "hunter", the "shot" has been sliced and a totally unexpected target has been hit, leading to so-called "off-target effects". All these off-target changes have a marked effect on the precision of the pharmaceutical bullet, since the reproducibility of the treatment is considerably reduced. 3,4 Alternatively, the target molecule can transform, not just in an oncological sense. It may occur in different isoforms or conformations, be changed or differentially expressed in response to cell stimuli or pathophysiological processes and can be modified over time in culture, with the time of day or as a result of aging.
These changes in the nature of the target modify the accuracy of the drug treatment, 5 so that what appears to be a true assessment Positively, off-target effects have contributed in part to the growing interest in repurposing of known drugs for novel therapeutic indications. 6 But the outworking of a moving target becomes immediately obvious when our pharmacological bullet overshoots the target and hits something else and we end up with undesired toxicity. 7 In the last few years, another aspect of this sort of collateral damage has been revealed by the fact that so many studies-many published in highly reputable journals-are poorly reproducible, 8 Drawing to some extent on personal experience, particularly in the field of inflammation and (auto)immunity, we review here some of the reasons for the mobility of drug targets and their interactions with ligands ( Figure 3) that result in pharmacological plasticity. We do not take into consideration the inherent chemical dynamics of target molecules themselves nor do we address neuronal plasticity.
The CNS is inherently highly plastic. New neuronal connections and pathways are generated or lost in response to emotional experiences, environmental stress or learning, disease, trauma, or conditioning and this subject has been extensively addressed. 12,13 We do suggest, however, ways in which the drug target can be made to stand still or at least become easier to hit. Many of the approaches are already known. We emphasize the importance of taking most or all of the varied conditions into account, in a holistic manner, when developing drugs or explaining their effects in an attempt to tackle the "biological traffic melee" which most chronic diseases present.

| G ENE TI C AND PHENOT YPIC VARIE T Y
One of the most important ways in which target molecules can alter their shape or form or even disappear is by changes in their gene expression.

| Mutation
Gene mutation is one of the most widely appreciated sources of vari- to combat the problem. 23,24 The situation highlights the importance of taking a multi-target approach, but also increasing specificity and developing more effective and rapid diagnostic procedures to identify the potential etiological micro-organism(s).
G-protein-coupled receptors (GPCRs) in cell membranes represent a major family of molecular targets for a variety of different drugs. Recently, it has been shown, on the basis of an analysis of UK National Health Service drug prescription and sales data from 68 496 individuals, that an average of 3% of these individuals carry at least one allele with a mutation at the active site of a GPCR drug target. 25 For some GPCRs, the incidence of missense changes is much higher.
Thus, over 86% of individuals carry at least one missense mutated allele at the active site of the cannabinoid receptor 2 (CNR2) and 69% in the glucagon like peptide 1 (GLP1) receptor, respectively. These two GPCRs are targets of the common antiemetic, nabilone, and several antidiabetic drugs such as exenatide, respectively. For these and other classes of drugs, the incidence of mutations suggests that there is an extensive propensity in the population for genetically induced, differential or even absent responses to drugs acting on GPCRs.  26,27 Here again, these drugs affect a wide spectrum of cellular growth processes. Adverse effects of cytotoxic drugs, particularly on the gastrointestinal system are pronounced, but more sinister long-term genetic effects can arise. For instance, treatment of mice with the chemotherapeutic agents, cyclophosphamide, mitomycin C, or procarbazine caused apparent genome-wide instability with significant elevation of simple tandem repeat mutation frequencies in the sperm and bone marrow of their offspring. 28 Thus, intergenerational effects of anticancer therapy may potentially place the children of treated parents at risk, suggesting that a warning of potential effects on offspring should be given to patients of childbearing age to be treated with chemotherapy. As with infectious diseases, the collateral damage caused by nonspecific cytostatic therapy could well be a health threat to future generations and highly specific multi-targeting could well be a better option. In fact, the most recent informationbased research suggests that a variety of different tumor types exhibit a recurrent regulatory structure consisting of functional master regulator proteins which are dysregulated in a posttranslational manner. 29 Multiple targeting of this network of master regulator proteins, potentially starting also with screens in vitro, may be a way ahead for cancer therapy.

| Heterogeneity and dynamic variability of biological systems
For a color blind marksman, the use of red or green in delineating a target would clearly present him with difficulties in aiming. The genetic variation in a target molecule presents a far more complex challenge to the pharmacologist. Under many experimental conditions, the use of hybridoma cell lines or inbred laboratory animal species goes some way to ensure that the genetic background of the cells or organisms is similar, wherever the experiment is performed.
The expectation is that by ensuring that the genetic background remains reasonably constant, the cellular and physiological responses under standardized conditions will also remain relatively constant.
Unfortunately, this is not always the case.  36 Moreover, the polarization of macrophages to a particular phenotype is under tight metabolic regulation, so changes in glucose, pH and other nutrients can change the genetic expression profile of the cells. 37 This appears to be related to a "rewiring" of the citric acid (Krebs) cycle by inflammatory stimuli, simple products of the cycle such as citrate, succinate, and fumarate exerting remarkable changes in macrophage and dendritic cell function. 38 We have also shown recently

| Epigenetic changes
Gene regulation by epigenetic mechanisms, such as histone acetylation or methylation or DNA methylation, has been widely reported in many cells. Such mechanisms have been reported to play important roles in the etiology of cancer and have led to the introduction of the histone deacetylase (HDAC) inhibitors, vorinostat, romidepsin, belinostat, and panobinostat, as inhibitors of tumor proliferation in several types of cancer. 42 However, these drugs are pan-inhibitors of different HDAC types, each with different specificities and functions, so they have pronounced clinical side-effects. Moreover, in addition to histones, other proteins that are subject to acetylation might also be affected by HDAC activity. Interestingly, it appears that concomitant inhibition of cancer-associated inflammation with an inhibitor of the transcription factor, nuclear factor κB (NFκB), may act synergistically with HDAC inhibitors and reduce side-effects-a further indication of the potential benefit of combined drug therapy of disease. 43 Often the epigenetic modifications seen in healthy and transformed cells are studied experimentally only for a few hours or days.
In this case, their effects should be viewed rather as cell signaling responses than as inheritable changes. Assessment of long-term changes is a better way of assessing epigenetic inheritability. This can be done quite effectively in zebrafish or insects, because the generational turnover is much more rapid than in mammals. 44,45 Many drugs, particularly the anticancer agents discussed above, modify posttranslational changes. But other types of drugs, including analgesics, are able to modify epigenetic markers such as histone acetylation and DNA methylation. 46,47 Thus, the post-translational silencing or activation of genes may be continued to subsequent generations. In this respect, a warning note is raised by the finding discussed earlier that treatment of male mice with standard chemotherapeutic drugs results in transgenerational instability in their offspring. 28 Variability in natural populations is often driven by environmental or infection induced posttranslational genetic modifications and it is highly probable that drug treatment may need to be included among the modifying factors. 48

| AUTOIMMUNIT Y
Variability is an essential characteristic of the immune response,  mice. 51 The lesson drawn is that the genetic and phenotypic variety in experimental models can be taken into account partially using a range of genetically sensitive strains and a selection of translationally relevant assessment methods to generate an efficacy profile that provides a more precise picture of drug efficacy.

| Models of autoimmunity
As with viral susceptibility to immune attack, epitope drift or genetic spreading also occurs during the development of autoimmunity and can complicate interpretation of drug effects. Consequently, the immune system begins to recognize antigenic molecules beyond those of the specific, external chemical determinants to which a specific immune response was initially induced. 52 Not only does this lead to autoimmune tissue damage but also to the recruitment of a variety of different humoral, cytotoxic, and immunoregulatory mechanisms. At a later stage of many chronic autoimmune diseases it is, therefore, difficult to determine the basic underlying defect.
Genetic and biomarker studies may highlight common factors in the disease, such as HLA-DR4 variants or anti-citrullinated protein antibodies in rheumatoid arthritis, which help to stratify patients for therapy, 53 once again emphasizing the need for a range of disease assessments in evaluating effects of drug therapy. But epitope drift can also complicate the use of experimental models of autoimmunity. One of us (MJP) well remembers performing a series of studies years ago on the tissue injuring effects of lymph node lymphocytes from Lewis and Wistar rats with adjuvant-induced arthritis, 54,55 during which the reactivity of the lymphocytes was gradually lost with time, presumably due to epitope drift, and the project had to be stopped. These observations are supported by results of a study by other authors on adjuvant arthritis induced in Sprague Dawley rats from two different vendors. 56 The rats from the two sources varied in their susceptibility to arthritis, as well as in immune (various proinflammatory cytokines) and endocrine (plasma ACTH and corticosterone) responses. The authors suggested that different types of genetic drift in the two colonies were probably responsible, a possibility raised in a previous publication. 57 It would be well worthwhile to carry out a comparative study between different labs of the range of antigenic reactivity of lymphocytes from animals in which the same initial stimulus is used to induce the autoimmune response. One would guess that this reactivity would differ considerably, indicating that the researchers are not necessarily assessing the same response. supplements to strengthen the immune response. There is also increasing evidence that probiotics such as Lactobacillus can not only modify the constitution of the gut microbiome but also promote innate immunity. In this way, they also indirectly enhance adaptive immunity. [63][64][65] Even in laboratory animals, though, dietary changes can occur unexpectedly. In one study in which one of us (MJP) was involved, eicosanoid metabolites, including PGE2, were determined over a period of several years in platelets and inflammatory exudates from different experiments on laboratory rats fed an essential fatty acid deficient diet. With time, the quantities of eicosanoids generated enigmatically increased. 66 The source was eventually found to be an unannounced change made by the diet manufacturer in the source of the oil used to ensure essential fatty acid deficiency.

| Diet, microbiota, and immune response
In the meantime, it has been well established that changing the type of dietary fatty acid can have marked effects on the structural type of eicosanoid generated. For example, increasing dietary intake of omega-3 fatty acids, which occur in high quantities in fish, reduces the amounts and conversion of the n-6 fatty acid arachidonic acid to metabolites such as PGE2 and leads to synthesis of metabolites that either have different or limited biological activities, as well as exerting inhibitory effects pe se on inflammatory mediator production. 59,67 Moreover, omega-3 fatty acids modify the composition of the gut microbiome, promoting bacteria with a less proinflammatory profile. 68 In terms of (auto)immunity, gut colonization with commen- during pregnancy in the rat was reported to increase survival from embryotoxicity induced by concomitantly administered sodium salicylate, though the mechanism was unclear. 70 Effects on reactive oxygen species seem likely. In nude mice with experimentally induced pulmonary tumors placed on a high fat (12% wt/wt linoleic acid) diet, co-administration of the NSAID indomethacin was found to significantly reduce tumor growth. 71 However, in a longitudinal clinical study in 906 patients with colorectal cancer, a positive interaction between low-fat diet and aspirin administration on cancer incidence could not be observed. 72 Thus, while changes in dietary constituents can markedly influence immune responses, much remains to be done to assess whether such variation in diet modifies drug efficacy. Perhaps such steps could be viewed as an attempt to modify the fuel used by the biological traffic to a more "environmentally friendly" form.

| CHRONOB IOLOGY
The axis, so that sensitivity of tissues to therapeutic corticosteroids also changes. Since receptor sensitivity is low in the morning and high in the evening, it has been proposed that this may be a basis for adjusting the timing of corticosteroid therapy to reduce side effects. 78 Similar circadian rhythms have been observed in the efficacy and pharmacokinetics of NSAIDs, probably because of the changes in circulating leukocytes. 79 Moreover, blood pressure is also under control of circadian rhythms and it has been proposed that antihypertensive therapies should be given at bedtime to achieve optimal effects. 80 Pain also underlies a circadian rhythm and most importantly, melatonin itself is thought to exert analgesic activity. 81

| TIME , AG E , AND THE RE SOLUTI ON OF DISE A SE
Not only do underlying biorhythms alter the properties of drug targets, but physiological and pathological processes also change with time. Biomolecules, in particular proteins are very dynamic and interactive, alter their location and undergo modification. 87 Their movement generally slows down with time and they become less flexible or dysfunctional, as with tau proteins and prions in neurodegenerative disorders and lipoproteins in atherosclerosis, predominantly due to increasing oxidative stress. 88,89 As time progresses, tachyphylaxis, the loss of drug efficacy, can occur probably because of protein receptor desensitization or because of cellular senescence, so that even if the target is hit, an effect is no longer observed. 90,91 Using our traffic metaphor, some of the traffic starts moving slowly or breaks down, causing increased congestion. Thus, the most obvious effect of time is seen in the effect of aging, a major issue in societies with an aging population.
Elderly people are far more susceptible to disease, partially because of the slowing and deterioration of normal physiology (especially of renal function), but also because of dysfunctional defense and repair mechanisms. There is a gradual decrease in the barrier function of the mucosa in the elderly with development of innate immune senescence, which affects all the cells of the innate immune system. 92 Polypharmacy in these patients is very common and the heterogeneity of studies on drug use in this group of patients hinders the development of approaches toward risk reduction. 93 to whom two to four drugs were prescribed. 95 Often, additional drugs are prescribed to combat the side-effects of other previously prescribed drugs. This situation is akin to the example described above of taking a shotgun to hit a small target, resulting in unnecessary collateral damage. For the good of many elderly patients, a regular reassessment of the number and doses of drugs administered is highly advisable. The authors are personally aware of several elderly patients whose condition improved considerably when drug treatments were markedly reduced! Less is sometimes more. In fact, a recent report describes the improvement in glycemic control in diabetic patients following reduction of the complexity of their medication regimen. 96 Nevertheless, the authors are well aware that in elderly patients with multiple morbidities, "appropriate polypharmacy" may be needed to treat these patients adequately. 97 The problem is that there are only a couple of guidelines dealing with the situation of comorbidity. In most cases, in patients suffering concurrently from several diseases, it is inappropriate to simply combine the drugs that are recommended by the respective guidelines which cover only the individual diseases.
Most marketed drugs have been introduced because of a specific mechanism of action, often involving a single major target.
During the development of the drug, the pathophysiological roles of the relatively novel target mechanism may be inadequately un- In this case, only short-term use of the NSAIDs is advisable to reduce initial inflammatory symptoms and avoid their potential inhibition of actively generated, endogenous proresolving lipid mediators. 100 Drugs which do facilitate resolution of inflammation are the macrolide antibiotics such as azithromycin which promotes the generation of the proresolving macrophage M2 phenotype.
Macrolides are known to be particularly effective in the treatment of respiratory infections such as community acquired pneumonia and acute exacerbations of chronic obstructive pulmonary disease, as well as the rare inflammatory lung disorder, diffuse panbronchiolitis. The macrolide antibiotics accumulate in leukocytes, causing an initial stimulation of neutrophil function and thus, antibacterial activity. Subsequently they promote the generation of the M2 macrophage phenotype and the resolution of inflammation. 41 As a consequence, their antibacterial use, requiring only a few days administration, has been extended to more long-term use in a limited number of inflammatory conditions. Macrolides without antibacterial activity are also being developed for their immunomodulatory properties. 101,102 Moreover, the capacity of azithromycin to promote the M2 macrophage phenotype has also been shown to be of potential therapeutic benefit in the treatment of cerebral ischemic injury. 103 Recognition of the potential of a mechanism as a target for the drug treatment of a particular disease does not, therefore, mean it can be administered at any time to any patient with the disease. The timing and stage of the disease to be treated is also crucial for successful therapy.
The duration of drug action is a further factor in assessing the interaction with its target. We have avoided more than passing consideration in this article of intramolecular dynamics of drug target molecules, but it is worth mentioning one aspect of drugligand binding that can affect duration of drug action. The drugtarget residence time model has gained increasing acceptance over the last decade in describing the interaction of a drug with its molecular target. 104 While the affinity of a drug for its receptor has dominated theoretical considerations in the past, the new theory assesses drug-receptor interactions in terms of the time for which the drug engages its receptor. The slower the drug dissociates from the target receptor, the longer its duration of action.
Potentially, this would mean that retention of a drug at its site of action should prolong its effect. With azithromycin, this could indeed be the case, as it accumulates rapidly in leukocytes but it is only slowly released from the cells, allowing it time to exert proresolving, immunomodulatory effects. 105 This pattern with azithromycin is also the same in fetal tracheal epithelial cell lines. 106 It would definitely be of great interest-in particular for chronic disease therapy -to compare drugs with moderate intrinsic activity at receptors but with prolonged duration of action with the efficacy of potent, short-acting drugs with high intrinsic activity.
There could be some surprises!

| S E X
In the past, preclinical studies on drug candidates were performed exclusively on male animals, to avoid the "inconvenient" alterations in sensitivity resulting from endogenous hormonal changes. But it has now long been clear that female responses to drugs often vary from those in males. These differences start with sex differences in disease susceptibility, possibly in part because of sex chromosome linkage, but also due to hormonal and metabolic distinctions, physical constitution and gender-specific lifestyles. This is obvious in autoimmune diseases which are often more prevalent in females than in males. Systemic lupus erythematosus and Sjögren's disease occur mainly in females, while systemic sclerosis is fourfold, rheumatoid arthritis two-to threefold and multiple sclerosis twofold more frequent in females than males. 107,108 Females also typically develop higher antibody responses and experience more adverse reactions to vaccination than males. 109 Similar findings have recently been made in collagen type II arthritis in DA rats.
In females, the ratios of CD4 and IL-17-producing T cells to Treg cells were raised and the production of Ig2a immunoglobulins increased in females in comparison to males. 110 In contrast, sepsis is less pronounced in women 111 and physiological resolution of acute vascular inflammation in humans appears to be more effective in women than in men. 112 This suggests that the breakdown of such effective resolution mechanisms may account, at least partially, for the more severe sequellae of chronic inflammatory disorders in women.
Interestingly, it appears that effects of sex hormones on the neuroimmune system may also account for the well-known higher incidence of chronic pain and increased sensitivity to pain in women. 113 Schizophrenia, though, is more frequent in young males (2:1) than in females, whereas this ratio is reversed in adulthood, probably because of protective effects of estradiol. 114 A sex difference has also been repeatedly reported in the incidence of and drug efficacy in depression in men and women, whereas sex differences have also been found in animal models of depression.
Consequently, it is more relevant for subsequent clinical development, especially in terms of autoimmune and neurological diseases, to use female animals or at least animals of both sexes, in early compound testing. 115 In fact, the US National Institutes of Health have recommended that the sex of animals and of cells be balanced for preclinical research studies. 116 The US Federal Drug Administration encourages pharmaceutical and medical device companies to provide data from clinical trials derived from both men and women. 117 This, however, means that the number of animals or human subjects to be included in trials is increased for statistical reasons, a fact that needs to be taken into account by local regulatory authorities.
Pharmacological actions of drugs are also subject to sex differences and some of these, particularly in relation to cardiovascular effects, have recently been reviewed. 118 While this variability is partially genetically based, there are indications, for instance with regard to ß-receptor sensitivity, that a sex difference in receptor sensitivity exists. The distinction between males and females is especially pronounced in pharmacokinetic processes, since the expression of many drug metabolizing enzymes and the occurrence of adverse drug effects is subject to modification by female sex hormones. 118 The authors of the review emphasized that it is crucial to design clinical trials to be able to distinguish between responses based on sex and to clearly assess the confounding effects of sex hormones.

| DRUG ME TABOLIS M AND DIS TRIBUTION
The most well-established source of plasticity in responses to drugs and a major reason for drugs to miss their targets is their kinetic fate within the body. For instance, sex hormones affect not only the pH and motility of the gastric intestinal tract, but also the expression of oxidative metabolizing enzymes and membrane transporter proteins. 118  Nevertheless, in the rat 2,4,6-trinitrobenzenesulfonic acid (TNBS)induced colitis model, sulfasalazine itself is able to restore the TNBS-induced gut dysbiosis, as reflected by increasing counts of short-chain fatty acid (SCFA)-producing and lactic acid-producing bacteria, as well as decreasing counts of proteobacteria, thereby ameliorating colitis. 123 Enhancement of SCFA production by gut bacteria is also thought to be one of the mechanisms by which the anti-diabetic type 2 drug, metformin, modifies glucose metabolism and cardiovascular responses. 124 Thus, changes in the gut microbiome, caused by diet or lifestyle differences, can potentially alter the metabolism and thereby the efficacy of drugs, even in similar strains of rodents, since as we discussed previously, animals from different suppliers can differ in responses to xenobiotics. 56 Indeed, an extensive sequence analysis of the gut microbiome in three different mouse strains from two suppliers, revealed considerable differences in bacterial composition between the same strains from different suppliers. 125 Polymorphisms in drug transporter proteins are a more subtle source of concern, since these differences affect not only metabolic behavior of drugs but also their specific distribution to organs, tissues, and cells. On the one hand, efflux transport of drugs out of tumors and bacteria is a major reason for drug resistance, 126,127 necessitating continual development of new approaches to cancer and infections, preferably using the multitarget approach to minimize the potential for resistance. 126,127 On the other hand, while differences in drug transporters are known to affect hepatic transport and kidney excretion of drugs, it is also worthwhile noting that these transporter proteins can also affect drug transport into other tissues and organs. For instance, organic anion-transporting polypeptide (OATPs) regulate intestinal absorption of macrolide antibiotics. 128 OATP 1A2 is most highly expressed in the brain and appears to regulate drug transport across the blood-brain barrier. 129

| CON CLUS I ON S AND PER S PEC TIVE S
In view of the immense variability and plasticity of the mammalian organism in response to pharmacological agents, it is not entirely surprising that drugs do not always provide the benefit we expect.
In fact, it has to be admitted that even when a new drug is marketed we know relatively little about its efficacy in a large population and subsequently, many drugs are prescribed too often for unsuitable conditions and patients are too willing to take medicine for conditions which may not require such treatment. Aside from this, nonadherence is also a big problem.
Apart from highly specific drugs such as monoclonal antibodies, it may be worth reconsidering the rational approach to drug therapy, namely of using a single specific drug to hit a single target, like Ehrlich's magic bullet. Rather, we should consider the multiple factors which make the target difficult to hit and try using a progressive, Alternatively, an immune reaction or neurological deficiency could potentially be regulated by light or melatonin to make the final target more accessible to treatment. The growth of a dysregulated cancer or other cell could be slowed so that more time is available to target a more specific mechanism. In this way, the serial use of two to three pharmacological agents, preferably with long drug-target residence times, would probably have greater efficacy than a single potent agent that only proves to be effective in a limited number of patients because of the inherent instability of the target mechanism. This approach is already taken to good effect in the treatment of hypertension in which a diuretic, for instance, is combined with a sartan and/ or a calcium antagonist to consistently decrease blood pressure. 137 We emphasize four already existing approaches (Figure 4) that can be taken either singly or in combination, to try to minimize effects of pharmacological plasticity:  145 The challenge is to identify, possibly by reverse pharmacology through extrapolation backwards from the clinical condition, a phenotype or set of symptoms to use as potential functional targets for initial screening. Subsequently, when an active hit or lead compound is identified, target deconvolution has to be carried out to investigate potential mechanism(s) of action. Recently, proposals have been made for establishing a "chain of translatability," starting with the identification of a disease-associated molecular characteristic or signature such as a disease-associated gene expression profile. 146 This is then succeeded by cellular models which aim to reconstruct a cellular phenotype as similar as possible to that of the disease condition and subsequently, use of relevant animal models of disease.
Alternatively, a wide range of assays and readouts can be used and a systems biology approach taken to assess the data for patterns. This approach is also taken in our laboratories, using different test models with a range of functional, molecular, biochemical, and imaging methods, at various stages of compound testing, to reflect the clinical conditions. 50,147,148 Where the target consists of a complex of interacting mechanisms, as in pain, a computational functional genomics-based approach can help to improve accuracy, clarifying both the specific targets to aim at for defined pain types and also offering insight into the target mechanisms for pharmacological modification either by individual drugs or a combination of therapeutic agents. 149 Sequential-slowing down the target, before moving in to the hit/kill, by aiming sequentially at one or more targets upstream and then one or more targets downstream, using single drugs with multiple actions or a combination of drugs with varied actions. This is illustrated by the long-established combination in a single dosage form of the antibiotics sulfamethoxazole, an inhibitor of dihydropteroate synthetase, and trimethoprim, an inhibitor of dihydrofolate reductase. Together they act synergistically to inhibit the bacterial synthesis of tetrahydrofolic acid. A further example of this approach is the sequential combination of drugs acting at different stages of cell signal pathways. Several growth factors act through the phosphoinositol-3-kinase/protein kinase B pathway which is crucial for cell proliferation and angiogenesis. A downstream effector of this pathway is the regulatory protein, mammalian target of rapamycin (mTOR). Addition of the anticancer drug, docetaxol to adenocarcinoma cells in vitro for 24 hours, followed by the mTOR inhibitor, temsirolimus was highly synergistic in suppressing phosphorylation of mTOR as well as in suppressing proliferation in lung cancer cell lines. 150 The authors proposed that this sequential combination may be effective in overcoming resistance of tumors to mTOR inhibitors.
Using a similar rationale, a phase 2 study was carried out in which the tubulin polymerization inhibitor, BNC105P was administered together with the mTOR inhibitor, everolimus, to patients with metastatic renal cell carcinoma who were refractory to tyrosine kinase inhibitors. 151 While the primary endpoint was not reached, analysis of biomarkers suggested that further studies are warranted.
In chronic diseases, inhibition of epigenetic changes with the resulting slowing of the chromatin aging process may soon be possible, making patients more amenable to other types of therapy.
Synchronized-using concurrent dietary, psychological, training, and biorhythm-synchronizing procedures to optimize drug therapy, but taking into account (and standardizing) possible changes in response with time and age. A wide variety of different pharmacological, physical, nutritional, and other procedures are used in the hospital setting to treat the ill patient. However, a therapeutic program synchronized for the various factors we have discussed is rare. Considerable potential appears to exist in modifying the gut microbiota to improve drug efficacy, as discussed previously in relation to drug metabolism. Thus, by modulating the gut bacterial population there is increasing evidence that the efficacy of current anticancer chemotherapeutics can be enhanced and their toxicity reduced, 152 perhaps by simple measures such as regular use of probiotics. Obviously, in the treatment of all human diseases, drug therapy is usually just one of the measures taken to improve the health of the patient. But the optimal conditions for implementing supportive measures are not often assessed, certainly not in preclinical investigations. For instance, the efficacy of the appetite suppressant, diethylpropion, used clinically to cause short-term weight loss, was greater when given to rats at night when their activity was greatest and when the animals were also placed on high fat dietary restriction. 153 In many chronic diseases, muscle loss occurs and general metabolic deficiencies arise. In chronic obstructive pulmonary disease, for instance, there is increasing evidence that combined nutritional and exercise interventions can be effectively used in combination with drug therapy. 154 In contrast, in the treatment of cancer, caloric restriction, as with ketogenic diets, has been found to both enhance drug efficacy and reduce toxicity. 155 Measurements taken during such synchronized treatment studies should be made at set times and seasons, allowing for the fact that the duration of action of some drugs is long. Such synchronization of drug therapy with other factors will undoubtedly become possible in the near future, as the combined use of the 4Ds, drugs, diagnostics, devices and (big) data, becomes increasingly possible. 156 In any case, even with established drugs, it is crucial to be vigilant and to be aware of new findings that can impact the way in which drugs are prescribed and used.

ACK N OWLED G M ENTS
The work of the authors is supported by the LOEWE Research