Professional Documents
Culture Documents
The low permeability of the mycobacterial cell wall is thought to contribute to the intrinsic drug resistance
of mycobacteria. In this study, the permeability of the Mycobacterium tuberculosis cell wall is studied by
computer simulation. Thirteen known drugs with diverse chemical structures were modeled as solutes
undergoing transport across a model for the M. tuberculosis cell wall. The properties of the solute-membrane
complexes were investigated by means of molecular dynamics simulation, especially the diffusion coefficients
of the solute molecules inside the cell wall. The molecular shape of the solute was found to be an important
factor for permeation through the M. tuberculosis cell wall. Predominant lateral diffusion within, as opposed
to transverse diffusion across, the membrane/cell wall system was observed for some solutes. The extent of
lateral diffusion relative to transverse diffusion of a solute within a biological cell membrane may be an
important finding with respect to absorption distribution, metabolism, elimination, and toxicity properties
of drug candidates. Molecular similarity measures among the solutes were computed, and the results suggest
that compounds having high molecular similarity will display similar transport behavior in a common
membrane/cell wall environment. In addition, the diffusion coefficients of the solute molecules across the
M. tuberculosis cell wall model were compared to those across the monolayers of dipalmitoylphosphati-
dylethanolamine and dimyristoylphosphatidylcholine, are two common phospholipids in bacterial and animal
membranes. The differences among these three groups of diffusion coefficients were observed and analyzed.
Figure 1. Structures of mycolic acids found in the M. tuberculosis cell wall. The usual value for k is 23. The sum of n, m, and h is approximately
50. The shorter chain is the R branch, and the longer one is the mero chain. Both R and β carbons have R chirality. (See text for details.)
Reprinted with permission from Hong and Hopfinger (2004). Copyright American Chemical Society.
and keto forms are 76-82, 83-90, and 84-89, respec- from entering the cell, which attributes, at least in part, to
tively.20 the intrinsic drug resistance of mycobacteria.7
AG, the other main component of the mycobacterial cell In the study reported in the preceding paper, conforma-
wall, is a polysaccharide consisting of arabinose (Ara) and tional search and molecular dynamics simulation (MDS)
galactose (Gal).21-23 Within AG, all arabinose and galactose were employed to investigate the conformational profile of
residues are in the furanose form, and mycolic acids are AG and subsequently the conformations and structural
located in clusters of four on the terminal hexaarabinofura- organization of the mycolyl-AG complex.25 An inner leaflet
noside through 1,5 linkages. However, only two-thirds of molecular model of the M. tuberculosis cell wall was
the terminal arabinose residues are mycolated. The linker constructed. (See the summary given in the Methods section.)
disaccharidephosphate connects the galactan region of AG The mycolate hydrocarbon chains were determined to be
to a peptidoglycan. Figure 2 provides an overview of the tightly packed and perpendicular to the “plane” formed by
structure and organization of the mycolyl-AG peptidoglycan the oxygen atoms of the 5-hydroxyl groups of the terminal
complex. arabinose residues to which the mycolic acids bind. The
In the physical organization model for the mycobacterial average packing distance between mycolic acids of M.
cell wall proposed by Minnikin, the mycobacterial cell wall tuberculosis was estimated to be approximately 7.3 Å. This
is composed of an asymmetric lipid bilayer (Figure 3).11 The previous computational study provides support for the
inner leaflet contains mycolic acids covalently linked to AG. Minnikin model and also presents a way to probe the
In this leaflet, mycolic acid hydrocarbon chains are tightly mechanism of low permeability of the cell wall and, in turn,
packed in a parallel fashion, and predominantly oriented in the intrinsic drug resistance of M. tuberculosis.
a direction perpendicular to the cell wall surface. The outer To investigate M. tuberculosis cell wall permeability to
leaflet contains the extractable lipids that accommodate the organic solutes in a quantitive manner and to gain more
uneven surface of the inner leaflet caused by the two insight into the nature of this barrier, a variety of drugs were
branches of mycolic acid that are not equal in length. It has modeled as solutes undergoing transport across the M.
been shown that the outer leaflet is moderately fluid, while tuberculosis cell wall. The properties of the solute-
the inner leaflet has very low fluidity.24 This low fluidity of membrane complexes were studied by the means of MDS,
the inner leaflet may be the result of the unique chemical especially the diffusion coefficients of the solute molecules
structures of mycolic acids and their tight packing, which, inside the cell wall. For the purpose of comparing the
in turn, could account for the low permeability of the overall permeability of the M. tuberculosis cell wall to those of
cell wall. The cell wall permeability barrier prevents drugs bacterial and animal membranes, models for the complexes
1068 Biomacromolecules, Vol. 5, No. 3, 2004 Hong and Hopfinger
Figure 2. Overview of the mycolyl-AG peptidoglycan complex (with permission from the Annual Review of Biochemistry, Volume 64, copyright
1995 by Annual Reviews www.annualreviews.org; minor modifications made).
Figure 4. Chemical structure of a PMA molecule. Carbon atoms located at position 13 on both the R and the mero chains are pointed out.
residues that are close to the termini of the arabinan chain and fourth columns of the PMs. The primary (largest) inertial
and covalently linked to the mycolic acids was chosen to axis of each solute was aligned parallel to the Z axis of the
represent the structural features of the AG complex (Figure mycolates (i.e., the direction of the mycolate hydrocarbon
2). Exhaustive random conformational sampling was per- chains). The solute was then moved along the Z axis until
formed on this structure. The conformations generated were its center of mass was in the “plane” formed by the carbon-
then ranked according to their conformational potential 13 atoms of the mycolates’ mero chains; see Figure 4.
energies. The low energy conformations were chosen to form The 13 cell wall-solute complexes generated were each
complexes with the PMAs. Energy minimization was per- subjected to extensive MDS. The MOLSIM package,29 with
formed on the resultant PMA-AG complexes, and the an extended MM2 force field,30,31 was used to perform the
complex with the lowest conformational potential energy was MDS. The simulation temperature was set at 311 K and held
used as a starting point in extensive MDS. The average constant during the MDS by coupling the system to an
distance between the oxygen atoms of the 5-OH groups of external fixed temperature bath.32 The initial periodic bound-
the terminal arabinose residues, where the mycolic acids bind, ary conditions were set at 36.5 Å for the X and Y dimensions,
was calculated on the basis of the MD trajectories taken after 80 Å for the Z dimension, and 90° for the surface orientation
the complex had reached equilibrium. Because the mycolic angle, γ. The simulation sampling time was 140 ps with
acids exist as mycolates in the inner leaflet, a pseudo- intervals of 0.001 ps for a total sampling of 140 000
mycolate (PM) was generated by esterifying the carboxyl conformations for each of the complexes. Geometries gener-
group of a PMA molecule with methanol. A PM monolayer ated during the MDS were recorded every 0.1 ps. Three
was then constructed using the average packing distance different random seeds were used to initiate the MDS. The
between the oxygens of the 5-OH groups that, in turn, MDS routine of the MOLSIM software package, as described
represents the inner leaflet of the M. tuberculosis cell wall. above, was applied throughout this study unless stated
2. Construction of the 13 Solutes. Thirteen known drugs otherwise. Methyl ester carbons and the terminal carbons of
were selected as solutes to study the permeability of the M. the PM mero chains were each assigned heavy masses of
tuberculosis cell wall. These drugs include seven first- and 1000 amu. The purpose of the heavy mass assignment is to
second-line anti-TB drugs, five general antimycobacterial take the influence of additional bonded structure into
drugs, and one compound that has no antimycobacterial consideration. For example, restrictions in the movement of
activity. The structures of these 13 “solutes” are given in the terminal carbons would occur in a MDS if the mero
Figure 5. The HyperChem 6.01 software26 was used to build chains had not been artificially shortened.
the three-dimensional structures of the solutes. The MM+ The displacement, ∆di, between two adjacent time steps
molecular mechanics force field, also implemented in Hy- was determined for each atom of each solute from its MD
perChem, was utilized in the solute structure optimizations. trajectory, and the individual diffusion coefficient of every
Each structure was subject to conjugate gradient energy solute atom was calculated using the mean-square displace-
minimization using the Polak-Ribiere first derivative ment method.33 The mean-square displacement is given as
method,27 until a derivative convergence criterion of 0.1 kcal/ n
(mol Å) was reached or until a maximum number of
iterations were performed. The HyperChem software sets the ∑
i)1
∆di2
maximum number of iterations to be 15 times the number χ2 ) (1)
of atoms in the system. After each structure was energy n
minimized in the MM+ force field, it was further energy where n is the total number of time steps in MDS. From the
minimized by the AM1 semiempirical method.28 Partial Einstein diffusion equation, the diffusion coefficient, D, is
atomic charges were estimated by performing a single point given by
AM1 calculation on the minimized structure.
3. Construction and MDS of the Complexes of the M. χ2
tuberculosis Cell Wall Model with the 13 Solutes. On the D) (2)
2∆t
basis of the estimated average packing distance between
mycolic acids from our previous study, the unit cell where ∆t is the size of one time step.34 The diffusion
parameters of the PM monlayer were selected to be a ) 7.3 coefficient of the complete solute molecule is then calculated
Å, b ) 7.3 Å, and γ ) 90°.25 Each solute was inserted into as the average of the diffusion coefficients of all atoms of
the M. tuberculosis cell wall, modeled by the PM monolayer, the solute.
as is illustrated in Figure 6. The solutes were located in the 4. Construction and MDS of the Complexes of the
region defined by the second and third rows and the third DPPE Monolayer Model with the Solutes. A DPPE
1070 Biomacromolecules, Vol. 5, No. 3, 2004 Hong and Hopfinger
monolayer model was previously built25 (see Figure 7A for M. tuberculosis cell wall exists in a relatively highly ordered
the structure of DPPE). The complexes of the DPPE state. Hence, the solutes were inserted into an intact PM
monolayer with the thirteen solutes were constructed using monolayer model to mimic this dense and highly ordered
the MI-QSAR software.35 To prevent unfavorable van der state.
Waals interaction between the solute molecule and the Each solute molecule was placed between the regions of
membrane DPPE, one of the “center” DPPE molecules was head groups and aliphatic chains of the monolayer with the
removed from the monolayer model and the solute molecule most polar group of the solute molecule “facing” toward the
was inserted in the space created by the missing DPPE head group region of the monolayer. MDSs were performed
molecule. This procedure was not applied in the construction on the monolayer-solute complex using MOLSIM. The
of the solute-PM complex model. In the previous study,25 periodic boundary conditions, which were used to build the
it was found that the average packing distance of mycolic DPPE monolayer model, were employed (a ) 37.5 Å, b )
acids in a PM monolayer is correspondingly less than that 37.5 Å, c ) 80 Å, and γ ) 92°).25 The monolayer-solute
of DPPEs in a DPPE monolayer, and the inner leaflet of the complexes were simulated for 70 ps using time intervals of
Mycobacterium tuberculosis Cell Wall Permeability Biomacromolecules, Vol. 5, No. 3, 2004 1071
Results
Table 4. Components of the Diffusion Coefficients with Respect to the Individual Axes (10-8 cm2/s) for the Solute Molecules in the M.
tuberculosis Cell Walla
〈Dma〉 Dma (1) Dma (2) Dma (3)
solute 〈Dx〉 〈Dy〉 〈Dz〉 Dx Dy Dz Dx Dy Dz Dx Dy Dz
isoniazid 0.4 0.4 0.3 0.4 0.4 0.3 0.3 0.4 0.3 0.4 0.4 0.3
PAS 0.4 0.4 0.3 0.4 0.4 0.4 0.4 0.4 0.3 0.4 0.4 0.3
ethionamide 0.3 0.4 0.3 0.4 0.3 0.3 0.3 0.4 0.3 0.3 0.4 0.3
ethambutol 0.5 0.5 0.3 0.5 0.5 0.3 0.5 0.4 0.4 0.5 0.5 0.4
ciprofloxacin 0.3 0.3 0.2 0.3 0.3 0.3 0.3 0.2 0.2 0.3 0.3 0.2
ofloxacin 0.3 0.3 0.2 0.3 0.3 0.2 0.3 0.3 0.2 0.2 0.3 0.2
clofazimine 0.3 0.3 0.2 0.3 0.3 0.3 0.2 0.2 0.2 0.3 0.3 0.2
amithiazone 0.3 0.4 0.3 0.3 0.3 0.2 0.3 0.4 0.3 0.3 0.4 0.3
dapsone 0.5 0.5 0.4 0.6 0.5 0.6 0.4 0.4 0.4 0.4 0.4 0.3
hydnocarpic acid 0.4 0.5 0.2 0.4 0.5 0.2 0.4 0.5 0.2 0.5 0.5 0.2
chaulmoogric acid 0.5 0.5 0.2 0.5 0.4 0.2 0.4 0.6 0.2 0.4 0.5 0.2
thiocarlide 0.3 0.2 0.2 0.3 0.3 0.2 0.3 0.2 0.2 0.2 0.2 0.2
phencycline 0.3 0.3 0.2 0.3 0.3 0.2 0.3 0.3 0.2 0.2 0.3 0.2
a The indices, (1-3), represent the MDS experiments initiated by three different random seeds. The diffusion coefficients reported in the second to
fourth columns are the average values from the three MDS experiments. The diffusion coefficients of hydnocarpic acid and chaulmoogric acid, which
exhibit significant lateral diffusion, are shown in bold.
acid are similar, having a two-dimensional molecular simi- implies identical similarity.
larity measure of 0.906 and the AMS P+-AMS and HBD- It is a core working hypothesis in medicinal chemistry that
AMS have measured similarities of 1.000. structurally similar molecules are likely to express similar
Mycobacterium tuberculosis Cell Wall Permeability Biomacromolecules, Vol. 5, No. 3, 2004 1075
Table 6. Diffusion Coefficients (10-8 cm2/s) of the Solute Molecules in the DPPE and DMPC Monolayers (DDPPE, DDMPC)
solute 〈DDPPE〉 DDPPE (1) DDPPE (2) DDPPE (3) 〈DDMPC〉 DDMPC (1) DDMPC (2) DDMPC (3)
isoniazid 1.5 1.6 1.5 1.5 1.3 1.4 1.2 1.3
PAS 1.6 1.7 1.5 1.5 1.2 1.3 1.3 1.1
ethionamide 1.5 1.6 1.5 1.5 1.3 1.3 1.3 1.4
ethambutol 1.7 1.8 1.6 1.6 1.6 1.6 1.6 1.7
ciprofloxacin 1.2 1.2 1.2 1.1 1.0 1.0 1.0 1.0
ofloxacin 1.0 1.0 1.1 0.9 0.9 0.9 0.9 1.0
clofazimine 1.0 1.0 0.9 1.0 0.8 0.8 0.9 0.8
amithiazone 1.7 1.8 1.5 1.7 1.4 1.3 1.4 1.3
dapsone 1.7 1.7 1.8 1.7 1.4 1.5 1.4 1.4
hydnocarpic acid 1.6 1.5 1.6 1.6 1.4 1.4 1.4 1.3
chaulmoogric acid 1.4 1.4 1.4 1.5 1.5 1.4 1.5 1.5
thiocarlide 1.3 1.3 1.4 1.3 1.2 1.1 1.2 1.2
phencycline 1.2 1.2 1.2 1.2 0.9 0.9 0.9 1.0
biological behavior profiles including absorption distribution, and the corresponding absorption of solute inside the cell
metabolism, elimination, and toxicity (ADMET) properties. are both low even though the diffusion coefficient may be
In this case, it would appear that similar compounds have large.
similar permeation behavior within a common cell wall Lateral diffusion may also affect cell wall function. In fact,
environment. This conclusion is supported by several ex- lipid analysis of M. Vaccae grown in the presence of
amples, such as chaulmoogric acid and hydnocarpic acid, chaulmoogric acid has demonstrated that chaulmoogric acid
isoniazid and PAS, amithiazone and dapsone, and the is taken up by the organism and incorporated into cell wall
spherical and bulky solutes. components.44 It is also observed that cell growth is retarded
2. Comparison of Diffusion Coefficients from the DPPE by the addition of chaulmoogric acid to the growth medium.44
and the DMPC Monolayer Models. There is no significant Therefore, it is interesting to postulate that the antimyco-
difference between the DDPPE and the DDMPC values for the bacterial properties of chaulmoogric acid result from its
solute molecules studied (Table 6). But, as expected, solutes uptake into or distortion of the cell wall structure. In the
in both phospholipid membrane models have correspondingly bacterial and animal membranes modeled by DPPE and
larger diffusion coefficients than the Dma of the M. tuber- DMPC monolayers, respectively, chaulmoogric acid under-
culosis cell wall model. Most solutes retain three-dimensional goes transverse diffusion across the membrane equally well
movement patterns in both the DPPE and DMPC monolayers compared to lateral diffusion. Thus, this compound is not
such as those observed in the M. tuberculosis cell wall model, readily retained inside the phospholipid membranes and its
except for chaulmoogric acid. This molecule moves unifor- cellular absorption and distribution properties are likely to
mally with respect to the three reference axes in the DPPE reflect its diffusion coefficients in these membranes. Al-
and DMPC monolayers. Only hydnocarpic acid from the though chaulmoogric acid is traditionally used to treat
solute dataset retains its principal movement within, as leprosy, this molecule might also provide a promising anti-
opposed to through, the membrane in the DPPE and DMPC TB mechanism: the uptake into or disruption of the M.
monolayers. That is, hydnocarpic acid undergoes significant tuberculosis cell wall structure. Thus, chaulmoogric acid
lateral diffusion in both DPPE and DMPC membranes. could be an important lead for future anti-TB drug develop-
ment. Hydnocarpic acid, similar to chaulmoogric acid, is also
Discussion used to treat leprosy, and its antimycobacterial effects may
also be due to the uptake into or distortion of the mycobac-
The molecular shape of a solute is found to be an important terial cell wall structure.
factor for the permeation behavior of a solute through the The resolution of solute diffusion into transverse and lateral
M. tuberculosis cell wall. For example, clofazimine and components in membranes might be very important in
thiocarlide have similar MWs. However, bulky-shaped studying ADMET properties of drug candidates. It is shown
clofazimine has a Dma value significantly lower than that of that lateral diffusion plays a significant role in the transport
the flat-shaped thiocarlide. Moreover, phencycline and clo- of molecules across tissues, including the blood-brain
fazimine have the same Dma values even though the MW of barrier, corneal membrane, and intestinal membranes (Figure
phencycline is only half that of clofazimine. 12).45 However, to get inside a target cell and exert its action,
A major finding of this MDS study is the observation for the drug candidate has to have good transverse diffusion
some solutes of predominant lateral diffusion within, as properties. Among the solute molecules currently studied,
opposed to transverse diffusion across, the M. tuberculosis hydnocarpic acid is the only molecule that displays pre-
cell wall model. Chaulmoogric acid is one of the solutes that dominant lateral diffusion in the bacterial and animal
prefers to move “parallel” to the cell surface inside the cell membrane models, while the rest of the solutes exhibit,
wall structure (lateral diffusion) as opposed to moving along essentially, equal lateral and transverse diffusion. Considering
the hydrocarbon chains of mycolates (transverse diffusion). that these solute molecules are actual drugs and have
As a result of preferred lateral diffusion, such solutes remain clinically acceptable ADMET properties, it is not clear from
inside the cell wall. Overall, the apparent penetration rate the results of this study what the delicate balance between
1076 Biomacromolecules, Vol. 5, No. 3, 2004 Hong and Hopfinger
Chem21 Group, Inc., and X.H. acknowledges a University (22) McNeil, M.; Daffe, M.; Brennan, P. J. Evidence for the nature of
Fellowship from UIC. We also thank Mr. Jianzhong Liu of the link between the arabinogalactan and peptidoglycan of myco-
bacterial cell walls. J. Biol. Chem. 1991, 265, 18200-18206.
our research group and Professor Scott G. Franzblau of the (23) McNeil, M.; Daffe, M.; Brennan, P. J. Location of the mycolyl ester
Institute for Tuberculosis Research at UIC for their helpful substituents in the cell walls of mycobacteria. J. Biol. Chem. 1991,
comments and discussions over the course of this work. 266, 13217-13223.
(24) Liu, J.; Rosenberg, E. Y.; Nikaido, H. Fluidity of the lipid domain
References and Notes of cell wall from Mycobacterium chelonae. Proc. Natl. Acad. Sci.
U.S.A. 1995, 92, 11254-11258.
(1) WHO Tuberculosis Fact Sheet. http://www.who.int/mediacentre/
factsheets/who104/en/index.html (accessed May 2003). (25) Hong, X.; Hopfinger, A. J. Construction, Molecular Modeling, and
(2) WHO report 2003. http://www.who.int/gtb/publications/globrep/ Simulation of Mycobacterium tuberculosis Cell Walls. Biomacro-
index.html (accessed June 2003). molecules, 2004, 5, 1052-1065.
(3) Dolin, P. J.; Raviglione, M. C.; Kochi, A. Global tuberculosis (26) HyperChem Program Release 6.01 for Windows; Hypercube, Inc.:
incidence and mortality during 1990-2000. Bull. W. H. O. 1994, Gainesville, FL, 2000.
72, 213-220. (27) Polak, E. Computational Methods in Optimization; Academic Press:
(4) Jarlier, V.; Nikaido, H. Mycobacterial cell wall: structure and role New York, 1971.
in natural resistance to antibiotics. FEMS Microbiol. Lett. 1994, 123, (28) Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. J. P.
11-18. AM1: a new general purpose quantum mechanical model. J. Am.
(5) Brennan, P. J.; Nikaido, H. The envelope of mycobacteria. Annu.
Chem. Soc. 1985, 107, 3902-3909.
ReV. Biochem. 1995, 64, 29-63.
(6) Barry, C. E., III; Mdluli, K. Drug sensitivity and environmental (29) Doherty, D. C. MOLSIM User’s Guide; The Chem21 Group, Inc.;
adaptation of mycobacterial cell wall components. Trends Microbiol. Lake Forest, IL, 1997.
1996, 4, 275-281. (30) Allinger, N. L. Conformational analysis. 130. MM2. A hydrocarbon
(7) Jarlier, V.; Nikaido, H. Permeability barrier to hydrophilic solutes force field utilizing V1 and V2 torsional terms. J. Am. Chem. Soc.
in Mycobacterium chelonei. J. Bacteriol. 1990, 172, 1418-1423. 1977, 99, 8127-8134.
(8) Trias, J.; Benz, R. Permeability of the cell wall of Mycobacterium (31) Hopfinger, A. J.; Pearlstein, R. A. Molecular mechanics force-field
smegmatis. Mol. Microbiol. 1994, 14, 283-290. parametrization precedures. J. Comput. Chem. 1984, 5, 486-492.
(9) Chambers, H. F.; Moreau, D.; Yajko, D.; Miick, C.; Wagner, C.; (32) Berendsen, H. J. C.; Postman, J. P. M.; Gunsteren, W. F. V.; Nola,
Hackbarth, C.; Kocagoz, S.; Rosenberg, E.; Hadley, W. K.; Nikaido, A. D.; Haak, J. R. Molecular dynamics with coupling to an external
H. Can penicillins and other beta-lactam antibiotics be used to treat
bath. J. Chem. Phys. 1984, 81, 3684-3690.
tuberculosis? Antimicrob. Agents Chemother. 1995, 39, 2620-2624.
(10) Jarlier, V.; Gutmann, L.; Nikaido, H. Interplay of cell wall barrier (33) Tinoco, I.; Sauer, K.; Wang, J. Physical Chemistry. Prentice-Hall:
and beta-lactamase activity determines high resistance to beta-lactam New York, 1985.
antibiotics in Mycobacterium chelonae. Antimicrob. Agents Chemoth- (34) Einstein, A. InVestigations on the theory of the Brownian moVement;
er. 1991, 35, 1937-1939. Methuen & Co., Ltd.: London, 1926.
(11) Minnikin, D. E. In The biology of the Mycobacteria; Ratledge, C., (35) MI-QSAR User’s Manual, version 2.0; The Chem21 Group, Inc.; Lake
Standford, J. L., Eds.; Academic: London, 1982; Vol. 1, pp 95- Forest, IL, 1997.
184. (36) Kulkarni, A.; Hopfinger, A. J. Membrane-interaction QSAR analy-
(12) Minnikin, D. E.; Goodfellow, M. In Microbiological classification sis: application to the estimation of eye irritation by organic
and identification; Goodfellow, M., Board, R. G., Eds.; Academic:
compounds. Pharm. Res. 1999, 16 (8), 1245-1253.
London, 1980; pp 189-256.
(13) Dobson, G.; Minnikin, D. E.; Minnikin, S. M.; Parlett, J. H.; (37) Kulkarni, A.; Han, Y.; Hopfinger, A. J. Predicting caco-2 cell
Goodfellow, M. In Chemical methods in bacterial systematics; permeation coefficients of organic molecules using membrane-
Goodfellow, M., Minnikin, D. E., Eds.; Academic: London, 1985; interaction QSAR analysis. J. Chem. Inf. Comput. Sci. 2002, 42, 331-
pp 237-265. 342.
(14) Brennan, P. J. In Microbial lipids; Ratledge, C., Wilkinson, S. G., (38) Iyer, M.; Mishru, R.; Han, Y.; Hopfinger, A. J. Predicting blood-
Eds.; Academic: London, 1988; Vol. 1, pp 203-298. brain barrier partitioning of organic molecules using membrane-
(15) Brennan, P. J. Structure of mycobacteria: recent developments in interaction QSAR analysis. Pharm. Res. 2002, 19 (11), 1611-1621.
defining cell wall carbohydrates and proteins. ReV. Infect. Dis. 1989, (39) Hopfinger, A. J.; Wang, S.; Tokarski, J. S.; Jin, B.; Albuquerque,
11 (Suppl.), 420-430. M.; Madhav, P. J.; Duraiswami, C. Construction of 3D-QSAR models
(16) Hunter, S. W.; Murphy, R. C.; Clay, K.; Goren, M. B.; Brennan, P.
using the 4D-QSAR analysis formalism. J. Am. Chem. Soc. 1997,
J. Trehalose-containing lipooligosaccharides. A new class of species-
specific antigens from mycobacterium. J. Biol. Chem. 1983, 258, 119, 10509-10524.
10481-10487. (40) Duca, J. S.; Hopfinger, A. J. Estimation of molecular similarity based
(17) Hunter, S. W.; Gaylord, H.; Brennan, P. J. Structure and antigenicity on 4D-QSAR analysis: formalism and validation. J. Chem. Inf.
of the phosphorylated lipopolysaccharide antigens from the leprosy Comput. Sci. 2001, 41, 1367-1387.
and tubercle bacilli. J. Biol. Chem. 1986, 261, 12345-12351. (41) Elson, E. L. Membrane dynamics studied by fluorescence correlation
(18) Liu, J.; Barry, C. E., III; Besra, G. S.; Nikaido, H. Mycolic acid spectroscopy and photobleaching recovery. Soc. Gen. Physiol. Ser.
structure determines the fluidity of the mycobacterial cell wall. J. 1986, 40, 367-383.
Biol. Chem. 1996, 271 (47), 29545-29551. (42) Frye, C. D.; Edidin, M. The rapid intermixing of cell surface antigens
(19) Yuan, Y.; Crane, D. C.; Musser, J. M.; Sreevatsan, S. Barry, C. E., after formation of mouse-human heterokaryons. J. Cell Sci. 1970, 7,
III. MMAS-1, the branch point between cis- and trans-cyclopropane-
319-335.
containing oxygenated mycolates in Mycobacterium tuberculosis. J.
Biol. Chem. 1997, 272 (15), 10041-10049. (43) Moore, P. B.; Lopez, C. F.; Klein, M. L. Dynamical properties of a
(20) Barry, C. E., III; Lee, R. E.; Mdluli, K.; Sampson, A. E.; Schroeder, hydrated lipid bilayer from a multinanosecond molecular dynamics
B. G.; Slayden, R. A.; Yuan, Y. Mycolic acids: structure, biosynthesis simulation. Biophys. J. 2001, 81, 2484-2494.
and physiological functions. Prog. Lipid Res. 1998, 37, 143-179. (44) Cabot, M. C.; Goucher, C. R. Chaulmoogric acid: assimilation into
(21) Daffe, M.; Brennan, P. J.; McNeil, M. Predominant structural features the complex lipids of mycobacteria. Lipids 1981, 16 (2), 146-148.
of the cell wall arabinogalactan of Mycobacterium tuberculosis as (45) Edwards, A.; Prausnitz, M. R. Predicted permeability of the cornea
revealed through characterization of oligoglycosyl alditol fragments to topical drugs. Pharm. Res. 2001, 18 (11), 1497-1508.
by gas chromatography/mass spectrometry and by 1H and 13C NMR
analysis. J. Biol. Chem. 1990, 265, 6734-6743. BM0345155