DOI: 10.31038/JPPR.2019233

Abstract

Chagas disease, a parasitic, endemic infection affects about 10–15 million of people all over Latin America. In addition, Chagas disease has an increased global impact in nonendemic areas due to immigration patterns. More than 10, 000 deaths are caused by this disease each year, and nearly 70 million people are susceptible to infection. Although Chagas disease was identified more than 100 years ago, current therapeutic options for this condition are limited mainly to two ineffective drugs, benznidazole and nifurtimox. These two drugs are not optimal as currently applied because they show substantial toxicity, require long courses of administration and have inconsistent efficacy. In light of these problems, safer and more effective therapeutic options for patients with Chagas disease are clearly needed. Many candidate drugs are currently being tested. Among them, cruzapain inhibitors are considered promising agents that have been shown to have potential for more effective treatment of the chronic form of Chagas disease. In this work we review briefly the most significant advance in the design of new molecules able to kill Trypanosoma cruzi via an inhibition of the enzyme cruzain (also referred to as cruzipain, the full-length native enzyme).

Keywords

Chagas disease; Anti-chagasic drugs; Cruzain; Irreversible cruzain inhibitors; Reversible cruzain inhibitors; Vinyl sulfones; Vinyl nitroalkenes; Amidonitriles; Lipinski’ parameters and in silico study

Introduction

Among 17 Neglected Tropical Diseases (NTDs), Chagas Disease (CD, American trypanosomiasis) has been considered one of the most important tropical protozoan afflictions [1–3]. To treat parasitic infections like CD, there are four general tools, e.g. diagnostics, vector control agents, vaccines and drugs. Nevertheless, these disease-specific technologies are still deficient and have narrow use because of their inadequate efficacies and toxicities, or because they do not avoid reinfection. This also applies to Chagas disease, an important health problem in Central and South America [4, 5]. Since 1908 it is well-known that the CD is caused by infection with the haemoflagellate protozoan Trypanosoma cruzi, which is transmitted in rural areas to humans and other mammals by reduviid bugs such as Rhodnius prolixus, Triatoma dimidiate and Triatoma infestans, so-called kissing bugs, the most important vectors. The latter insects were usually reported in sub-Amazonian endemic regions (southern South America), while R. prolixus and T. dimidiata habit northern South America and Central America. Despite to the morbidity and mortality of CD, until now two old drugs are used to treat Chagas disease, namely the nitro derivatives based on imidazole and furan rings, – benznidazole (1) and nifurtimox (2), discovered in 1965 and in 1971, respectively. These drugs are known under the names Rochagan^® and Lampit^®, respectively (Figure 1).

Benznidazole and nifurtimox require long-term administration and often show efficacy problems, including insufficient activity in established, chronic forms of the disease (very low cure rates in the chronic stage). Having a bad reputation both in terms of safety and efficacy, these old medications must be replaced by new safer and more efficacious drugs. Some developed molecules are being investigated in clinical trials. Among them, it should be mentioned new 2-substituted 5-nitroimidazole, fexinidazole (3) and repurposed antifungal triazole, posaconazole (4) (Figure 1).

Figure 1. Chemical structures of anti-chagasic drugs and drug candidates based on azole scaffold.

Fexinidazole has been identified as a new promising drug candidate for treatment of African trypanosomiasis [6], but it was demonstrated that it could be used in the chemotherapy for Chagas disease [7, 8]. Successful treatment with posaconazole of a patient with chronic Chagas disease and systemic lupus erythematosus was reported in 2010 [9].

Since this time, posaconazole is the drug with more advanced development in both experimental model and clinical trial [10, 11]. However, despite the promising results obtained in the animal model, the clinical trials of this drug candidate did not meet expectations. Testing a combination of posaconazole and beznidazole called chagasazol, it was concluded that this triazole is not effective as mono-pharmaceuticals in the treatment of patients in chronic asymptomatic phase with respect to beznidazole[12].

While the antiprotozoal toxicity of nitroimidazoles (1, 3) and nitrofurans (2) are thiol-dependent on reduction by nitroreductases of the nitro group, which generates cytotoxic species that cause damage to DNA, lipids and proteins [13], triazoles like posaconazole (4) are inhibitors of the sterol biosynthesis pathway, in particular C14-α-demethylase inhibitors [7, 14].

Thus, the approach to antiparasitic chemotherapy is usually based on two broad classes of targets, – trypanothione reductasa (TR) and sterol 14α-demethylase (CYP 51), which are specific to the parasite. Other alternative potential targets of responsible parasite for CD, T. cruzi are trans-sialidase (TS), tubulin, kinetoplastid DNA and cruzain (also referred to as cruzipain, the full-length native enzyme) [15, 16].

Cruzain and Identification of its Inhibitors

Cruzain (EC 3.4.22.51), the major papain-like cysteine peptidase of T. cruzi, is a principal virulence factor and a chemotherapeutic target with remarkable pre-clinical validation evidence [17–19].

The studies of the basic biochemistry of T. cruzi have allowed the identification of novel targets for CD chemotherapy. Killing T. cruzi through an inhibition of the enzyme cruzain is a different approach in therapies for Chagas disease. Being a member of the papain C1 family of cysteine protease, this enzyme hydrolyzes chromogenic peptides at the carboxyl arginine or lysine residue and requires at least one more amino acid between the terminal arginine or lysine and the amino-blocking group.

Thus, cruzain enzyme is essential for the viability and virulence of T. cruzi [20]. From currently available information in the Protein Data Bank (PDB) [21] it is known that cruzain consists of two domains: one with α-helices form and the other of antiparallel β-sheets. The active site, located at the interface between the two domains, encloses the catalytic triad formed by Cys25, His159 and Asn175. Cruzain has four relevant subsites for ligand binding (S1, S2, S3 and S1’) [22, 23].

During early development of cysteine protease inhibitors for CD therapy different class of many small peptidyl molecules were designed, synthetized and tested; among them, there were diazomethane inhibitors, fluoromethyl ketones, oxygen-containing heterocycles and vinyl sulfones. All these molecules bind a peptide or pseudopeptide backbone of cysteine protease (i.e., cruzain) to form a covalent adduct that inactivates the enzyme. Unfortunately, irreversible inhibitors may also inhibit host protease enzymes and possibly cause unacceptable side effects. Such irreversible inhibitors, especially peptidyl vinyl sulfones have become interesting leads in clinical trials [24].

The most outstanding vinyl sulfone, which was efficacious in pre-clinical models of T. cruzi infection, including immunocompetant and immunodeficient mice and dogs, is the irreversible cruzain inhibitor K-777(5) [25, 26]. Vinyl sulfone K-777 (5, known also as K11777), a potent cruzain inhibitor (IC₅₀ = 5 nM) that inhibits the enzyme by attaching irreversibly to the catalytic Cys25 thiol group: first, conjugate addition of the thiolate of the cysteine at the active site to the double bond occurs that leads to the formation of carbanion, and then the resulting carbanion is subsequently protonated driving the process thermodynamically to the more stable enzyme−inhibitor complex (Figure 2).

Figure 2. Structure of K-777, three-dimensional structural representation of cruzain crystallized with this inhibitor and simplified inhibition scheme.²¹

These results indicated that there were no laboratory abnormalities, histologic abnormalities, or auto-immune phenomena. However, in 2005 it was announced that due to severe problems related to elevated hepatotoxicity of K-777 could lead to the abandonment of this project [27]. Since then, a number of more rigorous follow-up safety studies were done. Thus, while K777 continues to undergo preclinical evaluation, which may lead towards a possible human clinical trial [28], more selective, reversible inhibitors would be more suitable, the ideal drugs.

In general, three are two approaches to designing reversible inhibitors: (a) modifying K777 structure and (b) modifying structure of active inhibitors of human cathepsins that were designed for other diseases related to these enzymes. It should be evoked that cruzain is closely related to the human cysteine protease family of cathepsins, especially cathepsins L, B and K, which is the primary enzyme involved in bone formation process.

Both routes have active development. Indeed, modifying K777 structure (approach a), recently diverse reversible inhibitors, which contain electrophilic groups that serve as a ‘warhead’ that binds to cruzipain through a covalent bond to the active site cysteine, have been designed. Among them were first epoxysulfones and then halogenated vinyl sulfones [29, 30] and dipeptidyl nitro alkenes [31]. The latter molecules like a nitro alkene (6) (Figure 3) resulted in a new type of highly potent covalent reversible inhibitors of cysteine proteases. The results showed that this dipeptidyl nitro alkene (6) should bind in a comparable manner as the corresponding vinyl sulfones. However, chemical nature of the arising nitroalkane carbanion is quite different to the one of agent K-777, since the acidity of the corresponding acid, namely, the nitro alkane, is higher, and thus, the basicity of the carbanion is lower [30]. Moreover, the authors commented that the docking experiments indicate that a reaction, i.e. a nucleophilic attack at the double bond, should be possible.

Figure 3. Potent inhibitor (6) of cruzain with K_i values of 0.44 nM.

However, it should be noted that this molecular inhibitor exhibited a high affinity for human cathepsins, cathepsin B and cathepsin L showing K_i values of 8 and 11 nM, respectively. Thus, this cruzain inhibitor lucks a good selectivity index (SI) of human cathepsins/cruzain ratio.

An excellent example of the approach (b) is the development of new cathepsin K inhibitors based on amidoacetonitrile moiety [23, 32–34] that led to the identification of a peptidomimetic α-amido nitrile MK-0822 called odanacatib (7) (Figure 4), which was selected in the clinical development for osteoporosis and bone metastasis in 2008 [35].

Figure 4. Structure of odanacatib and simplified scheme of its specific interaction with cruzain that occurs in the binding pocket with Cys-SH.

Unfortunately, in 2016 odanacatib was dropped because of adverse events [36]. However, specific interaction of this molecule and related aminonitrile compounds with cysteine proteases consists of the nucleophilic attack of the thiolate moiety (Cys-SH) at the triple C ≡ N bond and thus the reversible formation of thioimidate adducts suggested that reversible cruzain inhibitors could be successful anti-T. cruzi agents [37].

Given the structural similarities between cruzain and the cathepsins, a drug discovery effort was mounted to identify an odanacatib analogs locking for the best selectivity (human cathepsins/cruzain). Division of the scaffold of this drug into three portions that it was identified as P1, P2, and P3 aided much to find new suitable cruzain inhibitors. Realizing focused screening of diverse cysteine protease inhibitors collection, first it could be identified amidonitrile (8) with basic moieties in P3 portion of model drug odanacatib (7) [38].

However, the basic nature of (8) is likely associated with lysosomotropism in which the accumulation of the compound in lysosomes could lead to unexpected activity against human cathepsins. Diverse related studies confirmed the potential of cruzain inhibitors as anti-T. cruzi therapies and allowed to identify two orally active, reversible cruzain inhibitors Cz007 (9) and Cz008 (10) (Figure 5).

Figure 5. Drug odanacatib as a model for developing new reversible cruzain inhibitors.

These structural non peptidic amidonitrile analogs of odanacatib have been found to be potent cruzain inhibitors displaying in vivo activity in a murine model of acute T. cruzi infection [39, 40]. Recently, several large libraries of α-amidoacetonitriles were also prepared and tested their inhibitory activity against the T. cruzi cysteine protease cruzain and anti-trypanosomal effects. A K_i value of 16 nM and an EC₅₀ value of 28 μM were observed for the most potent of these inhibitors [41].

In silico computational studies through Molinspiration software

Being small nitrogen-containing molecules, the above mentioned compounds (1–10) are capable to exert non-covalent interactions (hydrophobic and electrostatic interactions as well as hydrogen bonds, metal co-ordination, van der Waals forces, etc.) with some active sites in organisms that ends with pharmacological activity manifestation. This pharmacological activity is also influenced by certain simple physicochemical parameters (descriptors).

Usually, four most important parameters of small molecules are used to predict their oral bioavailability; these descriptors are molecular weight (MW), lipophilicity (Log P), H-bond donors (HBD) and H-bond acceptors (HBA). All these are involved in the Lipinski “rule of five” (RO5), which states that good absorption or permeation of a compound is more likely when there are less than 5 H-bond donors, 10 H-bond acceptors, the MW is smaller than 500 Da and the calculated Log P (cLog P) is smaller than 5 [42]. These simple filters can help in selecting drug-like compounds. However, RO5 compliance is not a guarantee that a compound will be drug-like. Additionally, other descriptors such as number of rotatable bonds (ROTB), topological polar surface area (TPSA) and molecular volume (MV) are considered very useful parameters in activity–structure relationship (SAR) studies. All these properties could be easily calculated by diverse software available online.

Considering above backgrounds and paying attention to the importance of knowing these physicochemical properties, we used Molinspiration software [43] for actual anti-chagastic drugs 1–4 and candidate drug, promising cruzain inhibitors 5–10 (Table 1).

Table 1. Molecular descriptors calculated for drugs and drug candidates (1–10) according to the Lipinski’s Rule using the Molinspiration Cheminformatics software.

^a Molecular Weight (g/mol); ^b Logarithm of the partition coefficient between n-octanol and water; ^c Logarithm of the solubility measured in mol/L; ^d Number of hydrogen-bond acceptors; ^e Number of hydrogen-bond donors; ^f Number of rotatable bonds; ^g Polar Surface Area (Ų); ^h Molecular volume.

Analyzing the results obtained, it can be noted that two classic anti-chagastic drugs, benznidazole (1) and nifurtimox (2), and drug candidate fexinidazole (3) are very small, highly soluble molecules with poor capacity of permeability through membranes (weak lipophilicity); its cLogP are 0.78, 0.71 and 2.58, respectively. In addition, they are molecules with low flexibility properties (ROTB = 3–5). In contrast, drug candidates (4–10) are more lipophilic molecules (6.15 <cLogP > 3.81) with elevated MW (383.45–762.86) and MV (359.60–678.93).

Among them, antifungal drug, posaconazole (4) that was proposed in CD treatment, resulted to be the biggest molecule with the highest lipophilicity parameter (cLogP = 6.15) and flexibility properties (ROTB = 13). Additionally, posaconazole has a better ability of intermolecular hydrogen bonding (HBA = 12), which is partially responsible for the interactions with secondary and tertiary structures of proteins and nucleic acids.

The physicochemical descriptors of cruzain inhibitor amidonitriles (8, 9) derived from drug odanacatib (7), are very close to the one of posaconazole that inhibits the ergosterol production by binding and inhibiting the lanosterol-14α-demethylase. Noteworthy, these cruzain inhibitors and posaconazole are structurally different, very lipophilic molecules with more than two RO5 violations.

Own odanacatib molecule, possessing MW more than 500 g/mol, has one RO5 violation. Moreover, its more elaborated amidonitrile derivative (10) does not show any RO5 violation like dipeptidyl nitro alkene (6), a K777 analog (Table 1). However, it should be commented that there are plenty of examples available for such prediction violation amongst the existing drugs [44].

Thus, both approaches to designing reversible inhibitors (modifying from drug odanacatib and agent K777) result in promising cruzain inhibitors as anti-T. cruzi therapies. However, it is believed that α-amidoacetonitriles derived from odanacatib have more outlooks in anti-chagastic drug research [45].

Interesting results were obtained also analyzing bioactivity score for these compounds. According to the predicted bioactivity score, drug and drug candidates (1–4) do not make to switch any tested biological targets, while especially designed cruzain inhibitor agents (5–10) may alter selectively proteases (Table 2) that confirm generally its biochemical behavior. Unfortunately, simple in silico evaluation of these drug candidates does not allow anticipating their SI properties.

Table 2. Molinspiration bioactivity score for compounds (1–10)*

* Marked in bold parameters indicated a pronounced bioactivity score

Additionally, it can be noted that tested inhibitors (5–8) could modify other important biological targets: all four molecules would affect G protein–coupled receptor (GPCR) (score = 0.19–0.34), drug odanacatib (7) would work also as an enzyme inhibitor (score = 0.27). These protein covalent binding properties possibly cause unacceptable side effects.

Conclusion

Chagas disease represents an emerging worldwide challenge and there is an urgent, unmet need for safe and effective medication and thus, the last few years have seen a revolutionary improvement of the in vitro and in vivo screening methodologies to search for new anti-Chagas therapies.

The studies on reversible cruzain α-amidoacetonitrile-based inhibitors represent a valuable step in the search for new drugs for the treatment of a neglected disease that continues to affect the lives of millions of people.

The RO5 methodology appears to be as useful today in defining therapeutically relevant pharmacokinetic drugability and could help in this search. This simple and brief in silico analysis not only could help to design better new molecules with suitable pharmacological properties, but also comprehend its possible mechanism of action. It is evident that one of the important rational approaches for the fast advance of novel trypanocidal chemotherapy would be one focused on drugs ready to enter to clinical trials or on the clinical evaluation of drug association with existing trypanocidal agents to get extra effectiveness and fewer secondary effects [46].

References

1. Rassi A Jr, Rassi A, Marin-Neto JA (2010) Chagas disease. Lancet 375: 1388–1402. [crossref]
2. Bern C (2015) Chagas’ Disease. N Engl J Med 373: 456–466. [crossref]
3. Cavalli A, Bolognesi ML (2009) Neglected tropical diseases: multi-target-directed ligands in the search for novel lead candidates against Trypanosoma and Leishmania. J Med Chem 52: 7339–7359.
4. Chatelain E (2015) Chagas disease drug discovery: toward a new era. J Biomol Screen 20: 22–35. [crossref]
5. Bonney KM (2014) Chagas disease in the 21st century: a public health success or an emerging threat? Parasite 21: 11. [crossref]
6. Torreele E, Trunz BB, Tweats D, Kaiser M, Brun R, et al. (2010) Fexinidazole: a new oral nitroimidazole drug candidate entering clinical development for the treatment of sleeping sickness. PLoS Negl Trop Dis 4: 923.
7. Caldas S, Caldas IS, Cecílio AB, Diniz LF, Talvani A, et al. (2014) Therapeutic responses to different anti-Trypanosoma cruzi drugs in experimental infection by benznidazole-resistant parasite stock. Parasitology 141: 1628–1637.
8. Bahia MT, de Andrade IM, Martins TA, do Nascimento ÁF, Diniz Lde F, et al. (2012) Fexinidazole: a potential new drug candidate for Chagas disease. PLoS Negl Trop Dis 6: 1870. [crossref]
9. Pinazo MJ, Espinosa G, Gállego M, López-Chejade PL, Urbina JA, et al. (2010) Successful treatment with posaconazole of a patient with chronic Chagas disease and systemic lupus erythematosus. Am J Trop Med Hyg 82: 583–587.
10. Molina I, Salvador F, Sánchez-Montalvá A (2015) The use of posaconazole against Chagas disease. Curr Opin Infect Dis 28: 397–407.
11. Molina I, Gómez i Prat J, Salvador F, Treviño B, Sulleiro E, et al. (2014) Randomized trial of posaconazole and benznidazole for chronic Chagas’ disease. N Engl J Med 370: 1899–1908. [crossref]
12. Alonso-Padilla J, Rodríguez A (2014) High throughput screening for anti-Trypanosoma cruzi drug discovery. PLoS Negl Trop Dis 8: 3259. [crossref]
13. Hall BS (2011) Wilkinson SR. Activation of benznidazole by trypanosomal type I nitroreductases results in glyoxal formation. Antimicrob Agents Chemother 56: 115–123.
14. Diniz LF, Urbina JA, de Andrade IM, Mazzetti AL, Martins TAF, et al. (2013) Benznidazole and posaconazole in experimental Chagas disease: positive interaction in concomitant and sequential treatments. PLoS Negl Trop Dis 7: 2367.
15. Sueth-Santiago V, Decote-Ricardo D, Morrot A, Freire-de-Lima CG, Lima MEF (2017) Challenges in the chemotherapy of Chagas disease: Looking for possibilities related to the differences and similarities between the parasite and host. World J Biol Chem 8: 57–80.
16. Lepesheva GI (2013) Design or screening of drugs for the treatment of Chagas disease: what shows the most promise? Expert Opin Drug Discov 8: 1479–1489. [crossref]
17. Du X, Guo C, Hansell E, Doyle PS, Caffrey CR, et al. (2002) Synthesis and structure-activity relationship study of potent trypanocidal thio semicarbazone inhibitors of the trypanosomal cysteine protease cruzain. J Med Chem 45: 2695–2707.
18. Cazzulo J, Stoka V, Turk V (2001) The major cysteine proteinase of Trypanosoma cruzi: a valid target for chemotherapy of Chagas disease. Curr Pharm Design 7: 1143–1156.
19. Branquinha MH, Oliveira SSC, Sangenito LS, Sodre CL, Kneipp LF, det al. (2015) Cruzipain: an update on its potential as chemotherapy target against the human pathogen Trypanosoma cruzi. Curr Med Chem 22: 2225–2235.
20. Duschak VG, Couto AS (2009) Cruzipain, the major cysteine protease of Trypanosoma cruzi: a sulfated glycoprotein antigen as relevant candidate for vaccine development and drug target. A review. Curr Med Chem 16: 3174–3202.
21. Protein Data Bank, http://www.rcsb.org/pdb.
22. Gillmor SA, Craik C, Fletterick RJ (1997) Structural determinants of specificity in the cysteine protease cruzain. Protein Sci 6: 1603–1611.
23. Martinez-Mayorga K, Byler KG, Ramirez-Hernandez AI, Terrazas-Alvares DE (2015) Cruzain inhibitors: efforts made, current leads and a structural outlook of new hits. Drug Discov Today 20: 890–898.
24. McKerrow JH (1999) Development of cysteine protease inhibitors as chemotherapy for parasitic diseases: insights on safety, target validation, and mechanism of action. Int J Parasitol 29: 833–837.
25. Doyle PS, Zhou YM, Engel JC, McKerrow JH (2007) A cysteine protease inhibitor cures Chagas’ disease in an immunodeficient-mouse model of infection. Antimicrob Agents Chemother 51: 3932–3939.
26. Barr SC, Warner KL, Kornreic BG, Piscitelli J, Wolfe A, et al. (2005) A cysteine protease inhibitor protects dogs from cardiac damage during infection by Trypanosoma cruzi. Antimicrob Agents Chemother 49: 5160–5161.
27. Urbina JA (2010) Specific chemotherapy of Chagas disease: relevance, current limitations and new approaches. Acta Tropica 115: 55–68.
28. Sajid M, Robertson SA, Brinen LS, McKerrow JH (2011) Cruzain: the path from target validation to the clinic. In Cysteine Proteases of Pathogenic Organisms. Robinson MW, and John P. Dalton JP, Eds. Adv Exp Med Biol. Boston: Springer 712: 100–115.
29. Doherty W, James J, Evans P, Martin L, Adler N, et al. (2014) Preparation, anti-trypanosomal activity and localisation of a series of dipeptide-based vinyl sulfones. Org Biomol Chem 12: 7561–7571.
30. Schirmeister T, Kesselring J, Jung S, Schneider T, Weickert A, et al. (2016) Quantum chemical-based protocol for the rational design of covalent inhibitors. J Am Chem Soc 138: 8332–8335.
31. Latorre A, Schirmeister T, Kesselring J, Jung S, Johe´ P, et al. (2016) Dipeptidyl nitroalkenes as potent reversible inhibitors of cysteine proteases rhodesain and cruzain. ACS Med Chem Lett 7: 1073–1076.
32. Ferreira LG, Andricopulo AD (2017) Targeting cysteine proteases in trypanosomatid disease drug discovery. Pharm Therap180: 49–61.
33. Frizler M, Lohr F, Furtmann N, Kla¨s J, Gu¨tschow M (2010) Structural optimization of azadipeptide nitriles strongly increases association rates and allows the development of selective cathepsin inhibitors. J Med Chem 54: 396–400.
34. Frizler M, Stirnberg M, Sisay MT, Gutschow M (2010) Development of nitrile-based peptidic inhibitors of cysteine cathepsins. Curr Top Med Chem 10: 294–322.
35. Gauthier JY, Chauret N, Cromlish W, Desmarais S, Duong LT, et al. (2008) The discovery of odanacatib (MK-0822), a selective inhibitor of cathepsin K. Bioorg Med Chem Lett 18: 923–928. [crossref]
36. Mullard A (2016) Merck &Co. drops osteoporosis drug odanacatib. Nat Rev Drug Discov 15: 669. [crossref]
37. Nicoll-Griffith DA (2012) Use of cysteine-reactive small molecules in drug discovery for trypanosomal disease. Expert Opin Drug Discov 7: 353–366.
38. Black WC, Bayly CI, Davis DE, Desmarais S, Falgueyret JP, et al. (2005) Trifluoroethylamines as amide isosteres in inhibitors of cathepsin K. Bioorg Med Chem Lett 15: 4741–4744. [crossref]
39. Beaulieu C, Isabel E, Fortier A, Masse F, Mellon C, et al. (2010) Identification of potent and reversible cruzipain inhibitors for the treatment of Chagas disease. Bioorg Med Chem Lett 20: 7444–7449.
40. Ndao M, Beaulieu C, Black WC, Isabel E, Vasquez-Camargo F, et al. (2014) Reversible Cysteine Protease Inhibitors Show Promise for a Chagas Disease Cure. Antimicrob Agents Chemother 58: 1167–1178.
41. Avelar LA, Camilo CD, de Albuquerque S, Fernandes WB, Gonçalez C, et al. (2015) Molecular Design, Synthesis and Trypanocidal Activity of Dipeptidyl Nitriles as Cruzain Inhibitors. PLoS Negl Trop Dis 9: 0003916.
42. Lipinski CA (2004) Lead- and drug-like compounds: the rule-of-five revolution. Drug Discov Today Technol 1: 337–341. [crossref]
43. Molinspiration Cheminformatics Software, https://www.molinspiration.com .
44. Benet LZ, Hosey CM, Ursu O, Oprea TI (2016) BDDCS, the Rule of 5 and drugability. Adv Drug Deliv Rev 101: 89–98. [crossref]
45. Ang KK, Ratnam J, Gut J, Legac J, Hansell E, et al. (2011) Mining a cathepsin inhibitor library for new antiparasitic drug leads. PLoS Negl Trop Dis 5: e1023. [crossref]
46. Duschak, VG (2016) Targets and patented drugs for chemotherapy of Chagas disease in the last 15 years-period. Recent Patentes Anti-infect Drug Discov 11: 74–173.

DOI: 10.31038/JPPR.2019232

Abstract

We present a way to understand what messages may resonate with ordinary people regarding the negative effects of drugs. We identify six Questions or categories of statements, mix and match them, and present the combinations (vignettes) to ordinary respondents, over the web. The responses to the vignettes are deconstructed into the contribution of each individual element. We find clear differences by gender, age, and education, but NO patterns which lead to general knowledge. We divide the respondents by the pattern of their individual responses to the elements, which division reveals two equal size mind-sets. Mind-set 1 responds to scare tactics, facts. There is little in the way of convincing these respondents. Mind-set 2 responds to what it will to them, personally, at an emotional level. These respondents can be reached and convinced by the proper messaging.

Introduction

The notion of ‘addiction’ is not a new one. The notion of a drug problem is certainly not a recent development. Historically, the drugs involved in the world of ‘drug problems’ have been related to the products created out of poppy seeds, products that promised relief from pain, but were found to be addictive. The literature, stories, poetry, plays, is replete with depictions of people in opium dens (ref.) Famous characters in the literature of the time were featured as addicts to one or another opium-derived product, such as the legendary detective created by Arthur Conan Doyle, namely Sherlock Holmes [1, 2]. Wars have been fought over drugs, such as the Opium Wars in China.

For a long time, the topic of addiction has been one applying to alcohol, tobacco, and drugs [3] As a species, homo sapiens appears to have a predilection to become addicted to something or other. People, in fact, can be described as being addicted to gambling, addicted to a sport, a hobby, even addicted to work [4]. We hear of addiction to foods, to sweets, with the word ‘crave’ used to describe the desire for food in the same way that the drug addict ‘craves’ a ‘fix.’ People can be even be said to be addicted to each other, especially at the early stages of infatuation. This paper focuses on addiction to drugs, specifically on what one communicates to avoid addiction.

The scholarly literature is replete with studies documenting the nature of addiction. These studies usually concern the nature of addiction of people at different stages of life, in different circumstances. There is a sense that the pattern of addiction may co-vary with who the person is, and the situation in which the finds himself or herself [5–7]. There is a notion, however, that the mind of the individual may be very important in the nature of the addiction to which a person is subject [8]. The point of view about the link between personality and addiction, and the possibly ‘addiction-leaning personality’ can be found in the research reports going back three quarters of a century, to 1944 [9], and undoubtedly earlier with the works of psychoanalysts such as Freud and Jung [10].

On almost a daily basis we hear stories about the opioid crisis, a drug problem which has taken the place of previous big stories about crack and heroin, and before that morphine and opium. As a species, we humans can become addicted quite easily to many of these drugs, which cause an initial rush or feeling of wellness, but which can slowly become the master of the person, forcing that person to do anything to get the daily ‘fix, ’ (Kolodny et. al., 2015.) Addiction has become more than a problem. It has become an entire causa-belli, war cry, as activists blame food companies for making food ‘addicting’ [12].

In Western Europe drugs have been a problem in the real life, as products like heroin and morphine, so touted for health and freedom from agonizing pain have been found to be remarkably addictive. The many studies on these drugs could themselves occupy several books, just for a literature review. The social issues of addiction, the change in personality, and the creation of entire subcultures of addiction have been documented again and again. One need only go to sources such as Google^® Scholar to see the dramatic, ever increasing number of papers devoted to drug dependency.

In recent years the focus on opium-based drugs has taken a back seat to the emerging problem of opiods, available by prescription (e.g., oxycontin), given for pain relief, and powerfully addictive. What started out as a new way to get rid of pain has, in turn, morphed into a social problem, involving over-prescriptions, unnecessary prescriptions to satisfy addiction, and a decline in the moral character of some individuals in the medical professional and cooperating pharmacies.

What is missing in all of these studies is a profound understanding of how to prevent drug addiction. There is now widespread appreciation of the addiction problem, especially its prevalence, and the group mantras of messages which are supposed to communicate the will to resist drugs and addictions, such as ‘Say NO to Drugs.’ These are important slogans, analogous to patriotic war slogans, memorable, rallying cries, designed to inspire. Yet there is a gaping lack of knowledge. What inspires an individual person? What specific messages can one use to communicate about avoiding drugs?

We deal here with applying the method of experimental design of messages to identify specific messages which might work. This study is not designed to provide the ultimate data, but rather designed to explore whether the methods subsumed in the science of Mind Genomics can be used to convey messages about drugs with the same efficacy as sending messages about food. The latter, messages about food and drink, are designed to persuade consumers to buy the product. The former, message about drugs and addiction, are designed to inform about the perils of addiction, with the hope of reducing drug abuse.

As part of a new effort to understand the mind of the next generation, the so-called ‘millennials, ’ we embarked on a series of projects in which millennial respondents of college age were to select a topic of interest to predefined groups of four researchers. It would be the task of this ad hoc ‘research group’ to generate the different messages to be tested within the structure of a Mind Genomics experiment. That is, we changed the typical structure of research, wherein professionals design the study, and test with the appropriate sample of respondents.

The rationale for letting the millennials become the researchers comes from the desire to understand how they respond about the topic of drugs. By instructing the millennials to come up with the test stimuli, and then test those stimuli, we discover both what is important to the millennial from the messages they choose, and how these messages convince respondents in the general population.

Method

The foundations of Mind Genomics have been treated in a variety of papers and books, to which the reader is referred [13–15] Briefly, Mind Genomics begins with a topic, asks a series of questions which ‘tell a story, ’ provides a set of answers to each question (messages), combines these messages into combinations, presents the combinations to respondents, obtains ratings, deconstructs the ratings into the contribution of the individual messages, and finally creates both new-to-the-world mind-sets, as well as a PVI, a personal viewpoint identifier, to assign new people to one of the mind-sets previously discovered through the Mind Genomics experiment.

We begin with the set of questions and answers, i.e., Questions or categories of ideas, and then the specific elements pertaining to each Question. The Questions and the elements appear in Table 1. We have phrased the Question as a question, which is how the development of the test elements occurs. The research group works with a topic, comes up with the requisite number of questions or Questions (here six), the order of the questions or Questions which thus ‘tell a story, ’ and then the relevant elements or answers to each question.

Table 1. The test stimuli.

When we look at the six questions in Table 1, we see that the millennials focus on the damage done by drugs. This insight, simply by itself, is valuable. It means that the younger people focus on drugs and damage, on negative effects. One might say that their questions are obvious, but not necessarily? Would, in fact, a PR company given the same task come up with the same Questions as the millennials? We don’t know. We do know, however, that this type of exercise forces the millennials to think about drug abuse in their own way, from their own perspective.

Creating vignettes by experimental design

Mind Genomics uncovers the contribution of the different elements by incorporating these elements into combinations, test vignettes. The vignettes comprise either three or four elements, a maximum of one element from a Question, but often NO element from a Question. The composition of the vignettes is dictated by an underlying experimental design, ensuring that the 36 elements are statistically independent across the set of 48 vignettes created for each respondent. Each element appears five times across the 48 vignettes, and thus is missing from 43 of the 48 vignettes. Finally, each respondent evaluates a unique set of 48 vignettes, created according to the same experimental design, but with the combinations different [13] This strategy ensures that the test vignettes cover a wide number of possible combinations. Furthermore, the strategy means that one need not know which combinations to create ahead of time. The more conventional approach would create a fixed set of 48 vignettes, present this fixed set of vignettes in different orders to respondents, average the ratings, and by so doing get a stable data set. Mind Genomics dispenses with that single set of vignettes.

The interview

The actual experiment begins with an invitation to panelists who are members of an on-line panel. These individuals have agreed to participate, in return for rewards dispensed by the company which provides the panel. Working with a panel company makes it easy to do these experiments, in return for a modest payment per respondent for the participation.

When the respondent agrees to participate, the respondent hits the embedded link in the email invitation, from which the interaction begins. Figure 1 shows the orientation page for the study. The orientation page tells the respondent a little about the study but provides NO information about what is expected. That lack of information is important. Anything in the orientation page providing greater detail is likely to set up the expectations of a ‘correct answer’ and thus bias the data in ways that are unknown.

Figure 1. The respondent introduction to the study on abusing drugs.

Each respondent evaluated 48 different vignettes, rating each vignette on two rating scale attributes. This paper will present results only from the first scale (How likely are you to say NO to drugs based on this information.) Figure 2 shows an example of a vignette. The experimental design dictated that 36 of the 48 vignettes would comprise four elements, one from each of four Questions, and that the remaining 12 of the 48 vignettes would comprise three elements, one from each of three Questions.

Figure 2. Example of a four-element vignette created according to the experimental design, for a specific respondent.

Transformation of the ratings from a 9-point to a binary scale and regression modeling

One of the ongoing ‘discoveries’ emerging from Mind Genomics is that users of the data report that they have difficulty understanding exactly what a scale point means. They understand the anchors of the scale (not likely and very likely, respectively), but they do not know how to interpret the values on the scale. Rather than label each point, a task which itself is fraught with biases of interpretation, we transform the rating scale into a binary scale, with ratings 1–6 transformed to the number 0, and ratings 7–9 transformed to 100. This transformation has been consistently used since 1985, and appears to work well, except for some countries wherein the respondents often ‘up-rate’ the vignettes, and which require a slightly more stringent transformation (1–7 transformed to 0, 8–9 transformed to 100.)

The transformation accomplishes its purpose, which is to make the user’s interpretation of the results simple. The numbers now deal with either NO (not interested in saying NO to drug) or yes (interested in saying NO to drugs). In order to ensure that the OLS (ordinary least-squares) regression will not crash, we add a small random number (<10^–5) to each transformed number. The random number does not materially affect the results, but does ensure that the OLS regression will work for a respondent who limits his or her ratings either to the low end of the scale (1–6, all transformed to 0), or the high end of the scale *7–9, all transformed to 100).

OLS regression is the preferred way to analyze the data. OLS regression deconstructs the binary rating into the contribution of both the additive constant (basic interest in avoiding drugs, in the absence of elements, viz., an estimated baseline), and the contribution of the 36 individual elements. Even when the respondent avers that she or he simply cannot state how much of the rating is due to each element, the OLS regression reveals that contribution.

How the individual regression models fit the original 9-point ratings

During the past two decades, numerous complaints have emerged among communities of professionals engaged in polling consumers. The complaints emerge out of the reality that consumers are increasing being asked to fill out questionnaires, virtually for every action they take in the commercial environment. When a consumer purchases a product on line or in a store, when a consumer uses a bank service or a medical service, when a consumer travels and stays at a hotel, one can be fairly certain that in an increasing proportion of those transactions there will be a follow-up request for the consumer to rate the transaction and even to provide deeper responses, e.g., through an automated interview.

The consequence is that the respondents are increasingly hesitant to participate in the studies, and the quality of the data from those who do participate have dropped. In previous studies with Mind Genomics, we have presented the distribution of goodness of fit statistics across individuals to assess the quality of data. The statistic used, the multiple R-squared, shows the proportion of variability in the ratings accounted for by the multiple linear regression.

Figure 3 shows the distribution of the R-squared values across the 50 respondents. What is remarkable is the very high individual models for goodness of fit. We attribute the quality of data to the importance of the topic. For other topics of far less social relevance, such as a credit card, the good of fit statistics are far poorer. It should be noted that whether we average good fitting models (such as those for this study) or average ‘noisy’ models (such as those for a credit card), the measure of central tendency, will still reveal what is important, and what is not important. Our focus here is simply to show how respondents pay attention to this important topic.

Figure 3. Distribution of the goodness-of-fit of the 50 individual models for Say NO to Drugs. The model relates the presence/absence of the elements to the 9-point rating (before the binary transformation).

How the individual elements drive the interest in Saying NO to Drugs

Our first substantive analysis deals with the degree to which the 36 different messages drive the response to say ‘I’ll say not to drugs.’ (see Table 2.) Recall that the dependent variable, the 9-point scale, was transformed to a binary scale, after which the OLS regression was run on the 50 individual data sets, one from each respondent, with each data set comprising 48 cases or observations.

Table 2. Performance of all of the elements as drivers of ‘saying NO to drugs’.

Table 3. The strong performing elements, by gender, as drivers of ‘saying NO to drugs’.

When we average the individual models from the 50 respondents, we see the following:

1. Additive constant = 48 = conditional probability that the respondent will say ‘NO to drugs” in the absence of any elements in a vignette. For the total the additive constant is 48, meaning that in the absence of elements, we already have a moderate predisposition to say that they will say NO to drugs. The experimental design ensures that each vignette comprises 3–4 elements, so the additive constant is an estimated parameter. The additional motivation will come from the elements themselves.
2. The sorted elements. Previous studies suggest that coefficients about 8 or higher correspond to elements which have a meaningful effect when used. We use the term ‘meaningful’ rather than statistically significant because a coefficient can have the value of 2 or 3 and be statistically significant owing to the large number of ratings, and thus the increased likelihood that even a small coefficient will be statistically significant.
3. The two elements which score strongly combine both the positive and the negative. It may well be that the key to increasing donation is a combination of the so-called ‘carrot and stick, ’ and not just the ‘stick’ alone.

   Everyone has hopes and dreams for the future, but for addicts those hopes and dreams only focus down to where the next score is coming from

   Enjoy every day of your life, without having to rely on drugs to make you feel normal

Gender differences

We now focus only on those elements which perform strongly (coefficient > 6.51) for either males or females. Our data suggests dramatic differences, and reveals a new, unexpected pattern emerging from the additive constant and the performance of the elements.

1. The study incorporated more females than males. That, however, is simply a observation, and does not affect the differences between the genders:
2. The additive constant is far higher for the females than for the males (54 vs 29.) In practical terms, this means that the interesting in ‘staying NO to drugs’ is basically higher among females than it is among males. In the absence of elements, the conditional probability of a female saying “NO” is almost twice as high as the conditional probability for a male.
3. Males respond very strong to some of messages, however, with coefficients of 15 or higher, a value that has been operationally accepted as a very strong driver of the dependent variable, in our case the likelihood of saying NO.
4. The strong messages for male are primarily emotional and personal, bringing in the respondent, but also talking about the negatives of drugs. The only exception is the element talking about ethnic and religious conflicts, which do not bring in the respondent to the element, but simply state a fact.

   Enjoy every day of your life, without having to rely on drugs to make you feel normal

   Drugs can limit the friends you keep

   Everyone has hopes and dreams for the future, but for addicts those hopes and dreams only focus down to where the next score is coming from

   Some of the most bitter ethnic and religious conflicts worldwide are fueled by drug income

   Drug use can make you ill and an overdose can kill

   Babies of drug addicts are far more likely to be underweight and to suffer from birth complications

5. There is only one strong message for females, the one which talks about friendship.

   Friends are there to help when you’re down … Drugs drag you down further

Age differences

Do we see the same strong differences by age that we did by gender? Table 4 shows us that the three age groups exhibit virtually the SAME additive constant, 47–49. The quite large differences emerge in the performance of the elements.

Table 4. The strong performing elements, by age, as drivers of ‘saying NO to drugs’.

1. The types of messages which drive the different ages to say NO suggest patterns, but there may be some elements which perform well but depart from that pattern. Nonetheless, Table 4 suggests different types of information to which the three age-groups respond:
2. Age 18–29: They respond to concrete information, messages which paint a clear ‘word-picture.’
3. Age 30–52: They respond to messages which talk about the loss of ‘feeling good.’
4. Age 53+: They respond to messages which sound very much like a parent or older adult might convey to a younger person.

Where the person lives

The three groups differ in the additive constant (especially city versus rural and suburban), as well aa differing dramatically in the messages which drive each group to say ‘NO to drugs’ (Table 5.)

Table 5. The strong performing elements, by where one lives, as drivers of ‘saying NO to drugs’.

1. The additive constant, the basic proclivity to say ‘NO to drugs’ is higher for those respondents who live in the city, lower for those in a rural area.
2. Rural – talk about drugs and bad endings.
3. Suburban – talk about the financial problems with drugs, and the difficulties with family and social life.
4. City – very little reaches them except other than it eventuates in a less pleasant social life.

The respondent’s education

We see substantial differences by education, as shown by in
Table 6. The differences do not tell as clear a story as did the differences among the previous complementary subgroups.

Table 6. The strong performing elements, by one’s education, as drivers of ‘saying NO to drugs’.

The key patterns are the following:

1. Those who finished their formal education in high school show the highest additive constant, 64. They will say NO to almost everything about drugs. The specific elements do not make much of a difference, and certainly do not drive a particularly strong response.
2. Those who have had some college, but did not finish, show a low additive constant (30), but they respond strongly to some elements, shown below. The elements do not suggest a specific pattern, but they seem to be the type of advice a parent might give to a child

   Drugs can limit the friends you keep

   Enjoy every day of your life, without having to rely on drugs to make you feel normal

   Addiction does more than harm your body…it can do lasting damage to your pocketbook and future earnings

   Every minute spent using drugs is a minute lost with a loved one

3. Those who have finished college and vocational school respond to drugs as destroying a person’s opportunities.

   Everyone has hopes and dreams for the future, but for addicts those hopes and dreams only focus down to where the next score is coming from

(At least) two mind-sets for giving to ‘Say NO to drugs’

One of the key tenets of Mind Genomics is that within any sphere of experience where people make judgments, there are differences among people which are systematic, and t, for the particular topic, may be likened to color primaries, namely red, yellow, and blue, respectively. It is not that people are different, which of course they are, but rather that there are groups of ideas which ‘travel together. A person can be characterized as more likely to respond in one pattern or another. The pattern most resembling the way a person responds is the person’s mind-set.

The above-mentioned description of the mind-sets for a sphere of experience means that there are no single mind-sets across all behavior, but rather each individual sphere of experience can be described as having its own particular experience-relevant set of primaries. Furthermore, these primaries emerge from the actual topic-specific study. That is, to discover the primaries or mind genomes for ‘Say NO to drugs’ we have to do the study, and let these genomes emerge.

The statistics to uncover these genomes or primaries is quite simple:

1. For each respondent, create the model relating the presence/absence of the 36 elements to the binary response, which we did before.
2. Array the models as a series of rows, considering only the coefficients, and not the additive constant. Its will be the pattern of coefficients which interests us, not the magnitudes of the individual coefficients.
3. Define a distance all pairs of respondents, based upon the pattern of their 36 coefficients. The distance we choose is the quantity (1-R), where R is the Pearson correlation computed using the two sets of coefficients, one for each respondent. The Pearson correlation goes from a low of -1 (perfect inverse correlation, and thus maximal distance of 2) to no relation (distance of 1), to a perfect linear relation (distance of 0.)
4. Array the respondents in two groups so that the variability between the averages of the 36 coefficients in each group is high, and the variability with the groups is low. Today’s statistical packages do these analyses quickly.
5. Compute the average coefficient for each element for each of the two segments or clusters. If there is a clear story, not necessarily perfect, then stop at two clusters or segments, interpret the meaning of the clusters.
6. If there is no clear story, and if there is a sense that the clusters are jamming together different types of messages, then move to three segments, and look for the story.
7. The objective of the clustering is to create a minimal set of reasonably different groups, with ‘different’ defined as groups about which one can write a different story. By the tine one reaches four segments, most meaningful stories emerge.
8. The results from our 50 respondents suggest two equal sized groups, although equality of size is not necessary for meaningful segmentation.
9. The two segments represent two mind-sets. The two mind-sets show radically different responses to the elements, as the segmentation is set up to produce. The interesting part of the mind-sets is what they are, and how does one recognize a person as having a specific mind-set. The mind-sets are not simply distributed by gender, age, where the respondent lives, and the level of education that the respondent has attained.
10. Mind-set 1 comprises individuals who respond to ‘raw facts.’ They will be hard to convince. Their additive constant is 49, but there are only two strong elements

    Drug use can make you ill and an overdose can kill

    You have NO idea how many people died getting that drug into America

11. Mind-set 2 comprises individuals who respond strongly to word pictures about what drugs do, and what drugs will do to them. The key is that they respond to the more elaborate messages.

Identifying the messages and what to say

The emergence of at least two mind-sets suggests that it will be possible to adjust the message to be more effective for at least one of the two mind-sets, the Personalizer (Mind-Set 2.) These respondents react to descriptions of what drugs do, and how drugs affect their lives. These respondents react strongly to the messages that have been used to combat drug dependency. Sadly, however, they comprise only half the population, and generally are difficult to discover if all one knows is age, gender, and where one lives. Just because a person has a certain profile in terms of geo-demographics or even in terms of interests, attitudes, and behaviors does not mean that one will belong to one mind-set or the other.

What is needed is a way to penetrate the mind of a person, early on, uncover the mind-set of a person, and make sure that the person, as an individual, is targeted with the messages appropriate for the person’s mind-set. We see from this small study that only half of the respondents react to the standard messages used in public-service campaigns. The other half of the respondents simply do not react. Further work is needed to discover what messages are appropriate for the non-responsive mind-set.

One initial way to make great progress in finding the messaging to help ‘Say NO to drugs’ is to assign a person to one of the two mind-sets. If the person is in the responsive Mind-Set, one should simply give that individual the ‘message’ appropriate for the mind-set. If the person is in the other mind-set, i.e., the non-responsive group, the messages won’t work, but this respondent could be invited to participate to future studies of the type reported here, such studies to be done with members of the non-responsive mind-set until the proper, motivating messages are discovered.

How then does one find these people who belong to the non-responsive mind-set? One way creates a Personal Viewpoint Analyzer (PVI), a set of questions, the pattern of answers to which assigns a person to one of the two mind-sets. The method uses the pattern of coefficients in Table 7, selecting the messages which best differentiate between the two mind-sets. Figure 4 shows the PVI created specifically for this study, a device which can begin to separate new people into one of the two mind-sets. A working version of the PVI as of this writing (Spring, 2019) is located at this website: http://162.243.165.37:3838/TT28/.

Table 7. The strong performing elements, by two emergent mind-sets, as drivers of ‘saying NO to drugs’.

Figure 4. The Personal Viewpoint Identifier (PVI) to assign new individuals to one of the two mind-sets. The pattern of responses to the five questions identifies a person as belonging to one of the two mind-sets.

Table 8. Number of mentions in the open-end question regarding reasons for avoiding drugs.

As in any work in progress, this PVI is simply a first approximation of what could be accomplished. The ideal approach would be to run 2–5 or more of these small-scale studies, changing the elements which don’t work, until one finds the elements which best put a group of 50–100 respondents into two or more clearly different, hopefully totally different mind-sets, strongly responsive to different messages. In our preliminary study we find only one of these groups. If circumstance dictate stopping here with the discovery of only one group, and messaging proceeds, that messaging should comprise elements which appeal to the other group (Mind-Set 1), which we label as ‘non-responsive’ and ‘only the facts.’

Open End Analysis

At the end of the session, and just before the study was over, the respondents were asked to write a short paragraph about the reasons that they find most compelling for avoiding drugs. The open-ends are not part of the study, per se, but rather enable the researcher to get additional information about drugs from individuals who have just been exposed to different sets of messages about what’s bad about drug.

The pattern of open-end responses suggests respondents say that they are most likely to avoid drugs because of the harmful effects (personal health) to themselves as well as to their families and friends. Cost is the next most frequently mentioned reasons. Other rationales for avoiding drugs include legal ramifications, past addiction, and a harmful example to children.

Discussion and conclusions

As of this writing (Spring, 2019), it is becoming increasingly clear that the so-called ‘drug problem’ is far more serious and insidious than ever before, primarily due to the increasing legality of some prescription drugs which cause addiction, e.g., opioids.

This paper presents an inexpensive, rapid, scalable way to attack the drug issue by understanding the mind of the individual, not treated like an addict, but rather treated like a consumer. That is, the objective is to understand how to communicate to ordinary people, before they become addicted, sending them messages. If we are to believe the preliminary numbers from a base of 50 respondents, then there are at two mind-sets. The first mind-set (‘The Raw Facts’) reacts to very little. More work is needed to find the deterring messages. The other mind-set reacts (‘Personalizer’) reacts strongly to messages about what the drug will do to them.

As we look at the increasing magnitude of the drug problem, could it be that the messaging is inadequate, and geared effectively to Mind-Set 2, the ‘Personalizer?’ Furthermore, could it be that we have not yet found the appropriate messages, if any, for Mind-Set 1, the ‘Raw Facts.’ Certainly, only one message or element among the 36 tested by this small sample suggests the need for more studies, studies that are remarkably inexpensive, easy to iterate, and scalable. Fortunately, the PVI, the personal viewpoint identifier, allows us to identify the individuals who belong to each mind-set, and thus enable us to classify each respondent ahead of time.

It would be ironic if the tools of today’s marketing, specifically the notion of mind-types and consumer segmentation, could be extended to the understanding and amelioration of addiction. The irony, of course, is that the same science used to drive consumers to purchase and to consume, themselves often called addiction in the most severe cases, would now be turned into the very tool used to message people to avoid sources of addiction. Certainly, the affordability of the Mind Genomics science deserves more application to the social and personal tragedies of addiction.

References

1. Dalby JT (1991) Sherlock Holmes’s cocaine habit. Irish Journal of Psychological Medicine 8: 73–74.
2. Peschel RE, Peschel E (1989) What physicians have in common with Sherlock Holmes: discussion paper. J R Soc Med 82: 33–36. [crossref] 
3. Carver H, Elliott L, Kennedy C. and Hanley J (2017) Parent–child connectedness and communication in relation to alcohol, tobacco and drug use in adolescence: An integrative review of the literature. Drugs: education, prevention and policy 24: 119–133.
4. Vitaro F, Brendgen M, Ladouceur R & Tremblay RE (2001) Gambling, delinquency, and drug use during adolescence: Mutual influences and common risk factors. Journal of gambling studies 17: 171–190.
5. Hanneman GJ (1973) Communicating drug-abuse information among college students. Public Opinion Quarterly 37: 171–191.
6. Kandel DB & Logan JA (1984) Patterns of drug use from adolescence to young adulthood: I. Periods of risk for initiation, continued use, and discontinuation. American journal of public health 74: 660–666.
7. Yamaguchi K, Kandel DB (1984) Patterns of drug use from adolescence to young adulthood: II. Sequences of progression. Am J Public Health 74: 668–672. [crossref] 
8. Kido T & Swan M (2013) March. Exploring the Mind with the Aid of Personal Genome-Citizen Science Genetics to Promote Positive Well-Being. In AAAI Spring Symposium: Data Driven Wellness.
9. Felix RH (1944) An appraisal of the personality types of the addict. American Journal of Psychiatry 100, 462–467.
10. Crowley RM (1939) Psychoanalytic literature on drug addiction and alcoholism. Psychoanalytic Review 26: 39–54.
11. Yamaguchi, K. & Kandel, D.B., 1984. Patterns of drug use from adolescence to young adulthood: III. Predictors of progression. American Journal of Public Health, 74, 673–681.
12. Lindgren E, Gray K, Miller G, Tyler R, Wiers CE, et al (2018) Food addiction: A common neurobiological mechanism with drug abuse. Froniers of Bioscience (Landmark Ed) 23: 811–836.
13. Gofman A & Moskowitz HR eds., (2012) Rule Developing Experimentation: A Systematic Approach to Understand & Engineer the Consumer Mind. Bentham Science Publishers.
14. Moskowitz HR (2012) ‘Mind genomics’: The experimental, inductive science of the ordinary, and its application to aspects of food and feeding. Physiology & behavior 107: 606–613.
15. Moskowitz HR & Gofman A, Inovation Inc, (2011) System and method for performing conjoint analysis. U.S. Patent 7,941,335.
16. Moskowitz HR, Gofman A, Beckley J & Ashman H (2006) Founding a new science: Mind genomics. Journal of sensory studies 21: 266–307.