White Collar–1, a DNA Binding Transcription Factor and a Light Sensor Qiyang He, Ping Cheng, Yuhong Yang, Lixing Wang, Kevin H. Gardner, Yi Liu

Strains, culture conditions and molecular analysis: The bd, a (wild-type clock) strain was used as the wild-type strain in this study. A wc-1 null (wc-1^-) strain was created by repeat induced point mutation (RIP) (S1) and used as the host strain for wc1-2 (S2) and wc-1.lov constructs. Sequencing of the mutant wc-1 gene in the wc-1^- strain revealed the presence of many RIP-induced G/C-A/T mutations resulting multiple premature stop codons in the WC-1 open reading frame. The wc-1.lov construct contained a deletion of aa 391–507 of wc-1 ORF generated by site-specific mutagenesis and it was transformed into wc-1^- strain at the his-3 locus. 241-23 (his-3, bd, wc-2^ko) is the wc-2^- strain (S3) used as the host strain for the Myc-His-WC-2 construct, which was created by inserting an 6-histidine tag into the Nde1 site of pWC2-1.Myc (S2). Liquid culture conditions and race tube assays were the same as previously described (S4, S5). Protein extraction, Western blot analysis, immunoprecipitation assays and Northern blot analysis were as previously described (S4, S6, S7). For cultures subjected to light treatment, ~1000 lux of light was used. All molecular experiments were repeated 2–3 times and similar results were obtained. The race tubes shown are representative samples from at least six replicate tubes. Amido black staining of the Western blot membranes was used to ensure equal loading of proteins. For Northern blot experiments, ribosomal RNA was used as the loading control.

Purification of the WC complex from Neurospora: To facilitate purification of WC-1 from Neurospora, we designed an affinity scheme that took advantage of the in vivo WC-1/WC-2 association. A construct was made in which the WC-2 protein was tagged with two different epitope tags, 5-c-Myc and 6-His tags in the N and C terminus of the protein, respectively. This construct was then transformed into a wc-2 knock-out strain (S3) where it was able to rescue the circadian conidiation phenotype of the wc-2^- strain after a single LD transition (Fig. 4A) and other light defects of the wc-2^- mutant (data not shown). Therefore, this tagged WC-2 protein functions as the wild-type WC-2 protein. 12–15 liters of wc-2^-,Myc-His-WC-2 cells were harvested at DD30 and extracted in lysis buffer (50mM Tris, pH 7.4, 20mM NaCl and 10% glycerol plus protease inhibitors) (typical yield: 1–1.5g of protein) (S7). These total protein lysates were applied on a Q-Sepharose column (80 ml bed volume) equilibrated with buffer A (20 mM Tris, pH 7.4). After washing with buffer A, column-bound materials were eluted with a 800 ml linear gradient from 20 mM NaCl to 500 mM NaCl in buffer A. After Western blot analysis, fractions containing WC-1 were pooled (elution at ~260mM NaCl), and supplied with imidazole to a final concentration of 10 mM, and loaded onto a 5-ml nickel column equilibrated with buffer C (20mM Tris, pH 7.4, 300 mM NaCl, 10 mM imidazole). After washing with buffer C containing 20mM imidazole, the WC complex was eluted with buffer C containing 80 mM imidazole. The eluted protein fractions were combined and concentrated to ~ 2 ml and immunoprecipitated with 15 l c-Myc monoclonal antibody coupled agarose beads (9E10, Santa Cruz Technology). After incubating at 4°C for 3 hrs, the beads were washed alternately in 20mM Tris, pH 7.4, 50mM NaCl and 20mM Tris, pH 7.4, 1M NaCl for four times before washed twice in water (pH 5.6 – 6.0).

To perform fluorescence spectroscopic analyses, the purified WCC from two separate purification products were combined. Beads were boiled in 100 l of water for 5 min to release the bound chromophore and centrifuged to remove denatured proteins. During all of these, protein samples were exposed to minimal amounts of light. Fluorescence excitation spectrum in Fig. 4C was obtained by monitoring the emission at 530 nm. Emission spectra were measured by using an excitation wavelength of 390 nm (Fig. 4D) or 450 nm (Fig. 4E), respectively. No significant fluorescent signals were detected either using 10 g of the total proteins before immunoprecipitation or using purification products from extracts of a wild-type strain lacking tagged WC-2 (data not shown), establishing the specificity of the fluorescent signals observed in Fig. 4 C & D. To distinguish which flavin species was associated with WCC, emission spectra were obtained using samples prepared under neutral and acidic conditions (pH 2.2). Comparisons between these spectra were expected to discriminate between FAD and FMN based on prior observations that the fluorescence signal of FMN decreases in acidic conditions while the signal of FAD significantly increases (S8). To perform TLC analysis, the boiled chromophore solution was concentrated to 2–3 l before applied onto chromatography paper. TLC was performed as described (S9) using butanol-acetic acid-water (12:3:5) as solvent. To determine the FAD/protein stoichiometry, WC-1 protein concentration was estimated by SDS-PAGE using BSA as standard, and the amount of FAD associated with WCC was calculated by fluorescence emission spectra at 520 nm (excitation wavelength at 390 nm) using a series of FAD standards.

Supplemental Text:

FAD differs from FMN by the presence of an adenosine monophosphate that is linked to FMN via a pyrophosphate linkage. The binding of FAD to WC-1 suggests that additional protein motifs may be present to accommodate the larger size of this molecule compared to FMN. A sequence alignment of the LOV domains from WC-1 and several other proteins (Fig. S3A) shows that all of these have a reasonably high sequence conservation, especially among the group of eleven FMN-interacting residues identified in the crystal structure of a LOV domain from the PHY3 phototropin from Adiantum (S10). This implies that the WC-1 LOV domain will adopt the same general PAS domain fold as PHY3 and bind the FMN moiety of FAD in a comparable manner. This arrangement is essential to the mechanism of the proper photocycling of these proteins, which proceed through a covalent adduct between the C(4a) position on the flavin and a highly conserved cysteine residue located in the ’A helix.

This sequence alignment also identifies a critical difference between the WC-1 and phototropin LOV domains in the form of a significant extension in the loop connecting the ’A and C helices of these proteins. These two helices provide most of the residues that directly contact FMN in the PHY3 LOV domain, wedging the isoalloxazine ring in a scissors-grip type arrangement. The presence of the longer linker in WC-1 could allow a greater degree of flexibility in the orientation of these two helices, potentially creating additional space within the domain to facilitate the binding of significant portions of the ribose and/or AMP moieties of FAD. Additionally, these linkers are rich in basic residues that could provide ionic interactions with the pyrophosphate linker unique to FAD. Experimental support for the importance of this linker in FAD binding is provided by studies of point mutants in the FAD-binding PAS domain of Aer, an E. coli protein responsible for aerotactic behavior (S11, S12). These works identified several point mutations in the ’A/C linker, including both missense mutations and single residue insertions/deletions, that interfered with FAD binding and aerotaxis. While this linker is of slightly different composition than that observed in WC-1, this suggests that this region is important for FAD binding within a PAS domain.

Supplemental Figure 1. In the wc-1.lov strain, frq mRNA and FRQ protein levels do not respond to the light/dark transition. (A) Northern blot analysis showing the lack of circadian rhythms of the levels of frq and ccg-1 mRNA in the wc-1.lov strain in DD after a LD transition. Note the rapid disappearance of frq after the LD transition in the wild-type and the similar frq levels after the LD transition in the wc-1.lov strain. (B) Western blot analysis showing that the levels and the phosphorylation states of FRQ were arrhythmic in the wc-1.lov strain in DD after a LD transition.

Medium version | Full size version

Supplemental Figure 2. Circadian rhythms in the wc-1.lov strain could be entrained by temperature. (A) Western blot analysis showing that the circadian cycling of FRQ protein in the wc-1.lov strain after a temperature treatment. Cultures were first kept in DD at 19°C before they were transferred to 25°C. (B) Northern blot analysis showing the circadian rhythm of ccg-1 in the wc-1.lov strain after the temperature treatment.

Medium version | Full size version

Supplemental Figure 3. (A) Amino acid sequence alignment of LOV domains from phototropins, WC-1, VVD, ZTL and FKF1. This alignment was modified from (S10) and uses the same nomenclature for structural elements. Both LOV domains of the Adiantum PHY3 and Arabidopsis NPH1 were used in the alignment. Black residues are identical in all LOV domains and gray residues are the identical in most LOV domains shown here. Asterisks mark the 11 FMN-interacting residues found in the LOV2 domain of Adiantum PHY3. For simplicity, the first residue of each LOV domain was numbered as the first residue. (B) Phylogenetic tree with bootstrap values (GeneBee, cluster algorithm) showing that the LOV domains of ZTL and FKF1 are more closely related to those of WC-1 and VVD than to phototropin LOV domains.

Medium version | Full size version

References

S1. E. U. Selker, Annu. Rev. Genet. 24, 579 (1990).

S2. P. Cheng, Y. Yang, K. H. Gardner, Y. Liu, Mol. Cell. Biol. 22, 517 (2002).

S3. M. A. Collett, N. Garceau, J. C. Dunlap, J. J. Loros, Genetics 160, 149 (2002).

S4. B. Aronson, K. Johnson, J. J. Loros, J. C. Dunlap, Science 263, 1578 (1994).

S5. Y. Liu, M. M. Merrow, J. J. Loros, J. C. Dunlap, Science 281, 825 (1998).

S6. P. Cheng, Y. Yang, C. Heintzen, Y. Liu, EMBO J. 20, 101 (2001).

S7. N. Garceau, Y. Liu, J. J. Loros, J. C. Dunlap, Cell 89, 469 (1997).

S8. E. J. Faeder, L. M. Siegel, Anal. Biochem. 53, 332 (1973).

S9. F. M. Huennekens, S. P. Felton, Methods Enzymol. 3, 950 (1957).

S10. S. Crosson, K. Moffat, Proc. Natl. Acad. Sci. 98, 2995 (2001).

S11. S. I. Bibikov, L. A. Barnes, Y. Gitin, J. S. Parkinson, Proc. Natl. Acad. Sci. 97, 5830 (2000).

S12. A. Repik, et al., Mol. Micro. 36, 806 (2000).