X-band EPR investigations of an altered nitrogenase MoFe protein for which the alpha-subunit His(195) residue has been substituted by Gln(alpha-Gln(195) MoFe protein) revealed that it exhibits three new S = 1/2 EPR signals when incubated under turnover conditions in the presence of acetylene (C2H2). These three signals are designated S-EPR1, S-EPR2, and S-EPR3. We now report Q-band EPR and C-13 and H-1 ENDOR of the alpha-Gln(195) MoFe protein when incubated under turnover conditions in either H2O or D2O buffers with (C2H2)-C-12, (C2H2)-C-13, Or C2D2 as the substrate. ENDOR measurements from S-EPR1 prepared with (C2H2)-C-13 reveal interactions with three distinct C-13 nuclei, indicating that at least two C2H2-derived species are bound to the cofactor of the alpha-Gln(195) MoFe protein under turnover conditions. Although distinct, two of these species have approximately isotropic hyperfine tensors, with hyperfine splittings of A(C1,C2) similar to 2.4 MHz; the third has a smaller hyperfine splitting, A(C3) less than or equal to 0.5 MHz at g(1).H-1 ENDOR measurements further show strongly coupled proton signals (A similar to 12 MHz) that are associated with bound C2Hx, The observation of this signal from the C2H2/D2O sample indicates that this proton is not exchangeable with solvent in this cluster-bound state. Conversely, the absence of a signal in the C2D2/H2O sample indicates that there is no strongly coupled proton derived from solvent. We propose that we are monitoring a C2H2 species that is bound to the FeMo-cofactor by bridging two Fe ions of a 4Fe4S "face", thereby stabilizing the S = 1/2 cluster state. Q-band EPR also resolves rhombic features in the spectrum of SEPR2, giving g = [2.007, 2.000, 1.992], but ENDOR showed no C-13 signals with enriched substrate, confirming an earlier suggestion that this signal is not derived from C2H2.