Chemical equilibrium modeling techniques for the analysis of high-resolution bacterial metal sorption data

An ion selective electrode was used to monitor binding of Cd2+ on two bacteria, Bacillus subtilis (Gram+) and Escherichia coli (Gram-), as a function of increasing pH. A competitive Langmuir sorption isotherm was used in conjunction with a linear programming method (LPM) or FITEQL to fit experimental data. Results obtained with simulated data showed that LPM is less sensitive than FITEQL to variations in sorption data. Application of the LPM to experimental data found three discrete metal binding sites on B. subtilis and E. coli with -log equilibrium constant (pK(S)) values of -0.80 +/- 0.20, 0.63 +/- 0.09, and 2.35 +/- 0.10, and -0.60 +/- 0.10, 0.25 +/- 0.19, and 1.93 +/- 0.17, respectively, at a constant ionic strength, I = 0.1 M (KNO3)The corresponding site densities were 0.09 +/- 0.01, 0.07 +/- 0.01, and 0.07 +/- 0.01, and 0.01 +/- 0.00(2), 0.02 +/- 0.01, and 0.04 +/- 0.01 mu mol of Cd2+/mg of B. subtilis or E. coli. From FITEQL, pK(S) values of -1.18 +/- 0.15, 0.40 +/- 0.11, and 2.31 +/- 0.32 for B. subtilis and -1.46 +/- 0.34, 0.20 +/- 0.12, and 1.87 +/- 0.12 for E. coli were recovered with site densities of 0.10 +/- 0.07, 0.07 +/- 0.06, and 0.06 +/- 0.02, and 0.02 +/- 0.005, 0.02 +/- 000(4), and 0.04 +/- 0.04 mu mol of Cd2+/Mg of B. subtilis or E. coli, respectively. Total site densities of 0.22 +/- 0.02 and 0.06 0.01 mu mol/mg were obtained by LPM for B. subtilis and E. coli, whereas FITEQL yielded values of 0.23 +/- 0.02 and 0.08 +/- 0.07 mu mol/mg. Both LPM and FITEQL produced feasible results, but LPM was less sensitive to error and did not require an a priori assumption of the number of binding sites. (C) 2001 Academic Press.