Thermochimica Acta 417 (2004) 283–294 Calorimetry and structure–activity relationships for a series of antimicrobial hydrazides M.L.C. Montanari a , A.D. Andricopulo b , C.A. Montanari c,∗ a Departamento de Qu´ ımica, Universidade Federal de São Carlos, Rodovia, Washington Luiz, Km 235, P.O. Box 676, 13565-905, Brazil b Centro de Biotecnologia Molecular Estrutural, CEPID/FAPESP, Instituto de F´ ısica de São Carlos, Universidade de São Paulo-USP, Av. Trabalhador São-carlense, 400, P.O. Box 369, 13560-970 São Carlos, SP, Brazil c Núcleo de Estudos em Qu´ ımica Medicinal—NEQUIM, Departamento de Qu´ ımica, Universidade Federal de Minas Gerais, Campus da Pampulha, 31270-901 Belo Horizonte, MG, Brazil Received 14 June 2003; received in revised form 17 July 2003; accepted 18 July 2003 Available online 27 February 2004 Abstract This paper presents some recent developments on the use of quantitative structure–activity relationships (QSAR) based on biological calorimetry. The calorimetric biological potency can be measured for structurally related compounds whose activity would not be easily determined with less accurate and precise methods. A series of antimicrobial hydrazides was assayed against two different cultured cell systems, Escherichia coli and Saccharomyces cerevisiae. The direct demonstration of a similar mode of action for the two biological systems was achieved with the use of calorimetry. The measured values were described in terms of 3D molecular interaction fields (MIF) by means of a recently developed GRID independent method (GRIND). The aim of this approach is to allow the analysis of a large number of quantitative descriptors by using chemometric tools such as partial least squares (PLS). The correlation between chemical structures and changes in bioactivity is described without the need for 3D molecular alignment according to a suitable conformational bioactive template. The proposed model for these molecular interaction fields has revealed the importance of the stereo-electronic properties on the cells metabolism. Throughout this paper, we describe the usefulness of the same cell systems in disclosing partitioning behaviour of study hydrazide antimicrobials employing the diffusion technique of Taylor–Aris. Since this variable may be of utility in pharmacokinetic studies, we have modelled and predicted it based on computed MIF and multivariate statistics by a procedure called GRID/VolSurf. This result was achieved with a small number of VolSurf descriptors encoding a balanced range of hydrophilic–lipophilic properties. © 2004 Elsevier B.V. All rights reserved. Keywords: Biological calorimetry; Taylor–Aris partitioning; QSAR; QSPR 1. Introduction It has been written that “science moves forward accord- ing to what it can measure”, and at present, there appear to be numerous promising advances among several analyti- cal techniques that can be useful to describe drug–receptor interactions [1]. Calorimetric techniques are very useful in the field of medicinal chemistry for studying these interac- tions on very small quantities of biological molecules [2]. The robustness and sensitivity of thermal analysis methods [3] with automation, and the availability of reliable soft- ware tools are especially useful for the behavioural study of ∗ Corresponding author. Tel.: +55-31-3499-5728; fax: +55-31-3499-5700. E-mail address: montana@dedalus.lcc.ufmg.br (C.A. Montanari). bioactive substances, excipients [4,5], and delivery systems [6]. Calorimetry is suitable for the investigation of the effect of drugs on microorganisms and animal cellular systems, and thus has been used as a method for the determination of bioactivity [7,8]. A number of previous studies have been performed as- sociated with the applicability to derive structure–activity relationships (SAR), which can in turn help medicinal chemists gain insight about the key interactions between drug and its receptor, with the aim of producing new, more powerful antimicrobials. On the other hand, studies sel- dom showed the use of calorimetry in deriving quantitative structure–activity relationships (QSAR), a field where it is possible not only to set information on SAR, but also insight into modes of action can be envisaged [9]. We have already shown that QSAR based on biological calorimetry for a set 0040-6031/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.tca.2003.07.024