The [Fe-II(CYS)(4)](2-) site of rubredoxin from Clostridium pasteurianum (Rd(red)) has been studied by Mossbauer spectroscopy in both purified protein and whole cells of Escherichia coli overproducing it. Excellent fits were obtained to an S = 2 spin Hamiltonian for D = 5.7(3) cm(-1), E/D = 0.25(2), delta = 0.70(3) mm/s, DeltaE(Q) = -3.25(2) mm/s, eta = 0.75(5), A(x) = -20.1(7) MHz, A(y) = -11.3(2) MHz, and A(z) = -33.4(14) MHz. These parameters were analyzed with crystal-field theory for the D-5 manifold of iron(II), revealing a d(z(2)) orbital ground state that is admixed by similar to0.21 d(x(2) -y(2)). The spin-Hamiltonian parameters are consistent within the D-5 theory, apart from the zero-field splitting parameter, D. This problem was solved by extending the crystal-field treatment with spin-orbit coupling to spin-triplet d-d excited states of the iron. Theoretical estimates are given for the spin-triplet (D-T) and spin-quintet contributions (D-Q) to D based on excitation energies derived from time-dependent density functional theory, TD-DFT. The computational results were interpreted in terms of crystal-field theory, yielding the Racah parameters B = 682 cm(-1) and C = 2583 cm(-1). The theoretical analysis gives the relative magnitudes D-Q:D-T:D-ss = 51%: 42%:7% (D-ss originates from spin-spin interaction). The DFT analysis corroborates the pivotal role of the torsion angles (omega(j)) of the C-S-i bonds in shaping the electronic structure of the iron(II) site. Rd(red) in overexpressing whole cells accounts for 60% of the Mossbauer absorption. The Rd(red) spectra from whole cells are virtually identical to those of the purified protein. By using the theoretical omega dependence of the spin Hamiltonian parameters, the torsions for Rd(red) in whole cells and purified protein samples are estimated to be the same within 2degrees. These findings establish Mossbauer spectroscopy as a structural tool for investigating iron sites in whole cells.