Fig. 1. Overview of parallel expression approach.

Fig. 2. Utility of E. coli cell lysate screening for determining domain boundaries. Fragments of human PAS kinase were expressed in E. coli BL21(DE3) as C-terminal fusions to GB1 using a T7 RNA polymerase expression vector (Amezcua et al., 2002). Each spectrum was generated from a lysate generated by a 1-liter culture grown in uniformly ¹⁵N-labeled M9 media, which was concentrated to 1 ml before acquiring 2D ¹⁵N¹H HSQC spectra (40 min, 37°). (A–C) An overlay of spectra from GB1–PAS kinase fusions (black contours; PAS kinase residues are listed on each panel) with a reference from an isolated GB1 sample (red).

Fig. 3. Discovery of novel secondary structural elements by comparison of multiple constructs in the photosensory LOV2 domain of Avena sativa phototropin 1 (AsLOV2). Comparisons of ¹⁵N¹H HSQC peaks originating from residues within the predicted PAS domain showed significant chemical shift changes among constructs with differing C-termini, as demonstrated in spectra from AsLOV2 (black) and the shorter AsLOV2ΔJα (red). The data strongly suggested that the C-terminal region interacted with the PAS domain in the dark. When these experiments were repeated with blue light illumination, these ¹⁵N¹H HSQC peaks became independent of the construct length, providing evidence that light-induced conformational changes led to the displacement of the C-terminal Jα helix. Adapted with permission from Harper et al. (2003).