1. A Human Nuclear-Localized Chaperone that Regulates Dimerization, DNA Binding, and Transcriptional Activity of bZIP Proteins 
Ching-Man A Virbasius, Susanne Wagner, Michael R Green 
Molecular Cell 
Volume 4, Issue 2, Pages (August 1999)
DOI: /S (00)80369-X

2. Figure 1
Detection, Purification, and Identification of a Nuclear Factor that Stimulates DNA Binding of GCN4
(A) Detection of a stimulatory activity in crude and heat-treated HeLa nuclear extract. DNA binding by a bacterially expressed GCN4 bZIP polypeptide was assayed in the presence of a purified His-tagged HTLV-I Tax protein (lanes 1 and 2), HeLa nuclear extract (lanes 3 and 4), or heat-treated HeLa nuclear extract (lanes 5 and 6). The endogenous AP-1/DNA complex and GCN4 bZIP/DNA complex are indicated.
(B) Purification and identification of the stimulatory activity. Mobility shift assays using eluates from S-Sepharose column are shown in the upper panel. Silver stain of the SDS-PAGE of the S-Sepharose fractions used in the mobility shift assays are shown in the middle panel. Renaturation of BEF activity from the S-Sepharose fraction 7, as monitored by mobility shift assays, are shown in the lower panel.
(C) BEF specifically stimulates DNA binding of bZIP proteins. Bacterially expressed fusion proteins were purified by affinity chromatography and assayed for DNA binding. Binding reactions were performed in the absence (lane 1) or presence of increasing amounts of purified BEF (lanes 2 and 3). Endogenous AP-1 DNA binding activity from HeLa nuclear extract was titered to low binding activity (lanes 1 and 3) and stimulated by addition of Tax (lane 2) or BEF (lane 4).
(D) Stimulation of GCN4 binding by BEF is concentration dependent. Increasing amounts of GCN4 peptide were assayed for DNA binding in the presence of BSA or BEF.
(E) BEF stimulates dimerization of GCN4. The GCN4 bZIP peptide (0.16 μM) was incubated with buffer (lanes 1 and 2), HTLV-I Tax (lanes 3 and 4), or purified BEF (lanes 5 and 6) in the presence (+) or absence (−) of glutaraldehyde. The positions of cross-linked dimer and free monomer are indicated.
Molecular Cell 1999 4, DOI: ( /S (00)80369-X)

3. Figure 2 BEF Recognizes the Leucine Zipper Subdomain of the bZIP
(A) The leucine zipper supports BEF function. DNA binding of a chimeric fusion protein λ-ZIP containing the GCN4 leucine zipper and the DNA binding domain of the λ repressor was assayed in the presence of Tax (lane 2) or BEF (lane 3).
(B) A nonconsensus basic region does not impair BEF function. Mobility shift assays using GST–ATF4 were performed in the presence of increasing amounts of Tax or BEF as indicated (upper panel). Control reactions using GCN4 are shown in the middle panel. A sequence comparison between the consensus bZIP and ATF-4 is shown below the autoradiogram.
(C) Mutations in the leucine residues abolish BEF function. GST–JUN fusion proteins containing mutations in the L1 or L3 position of the leucine zipper were analyzed for DNA binding and responsiveness to Tax and BEF. The control reaction contained GCN4. Positions of the mutated residues are indicated in the diagram below the autoradiogram.
Molecular Cell 1999 4, DOI: ( /S (00)80369-X)

4. Figure 3 BEF Has Properties of a Molecular Chaperone
(A) GroEL stimulates GST–ATF2(350–415) DNA binding. Upper panel, GST–ATF2(350–415) DNA binding was analyzed in the presence of increasing amounts of GroEL either in the absence (lanes 1–6) or presence (lanes 7–11) of ATP. The molar ratios of GST–ATF2(350–415) to GroEL are indicated. Lower panel, GST–ATF2(350–415) DNA binding activity was monitored in the absence or presence of GroES. ATP was added in all cases except for lane 1. The molar ratios of GST–ATF2(350–415), GroEL, and GroES are indicated.
(B) BEF prevents rhodanese aggregation. Aggregation of guanidinum hydrochloride–denatured rhodanese was measured as an increase in photometric absorbance. Upper panel, denatured rhodanese was diluted into buffer with no addition (diamonds) or in the presence of GroEL (circles) or DnaK (triangles). The absorbance of diluted, native rhodanese is shown (squares). Lower panel, aggregation assays were performed as above, except in the presence of dialysis buffer (squares) or BEF (diamonds).
Molecular Cell 1999 4, DOI: ( /S (00)80369-X)

5. Figure 4
BEF Synergizes with HTLV-I Tax to Promote bZIP Dimerization and DNA Binding
(A) BEF and Tax stimulate bZIP DNA binding synergistically. Binding assays using GCN4 were performed in the presence of limiting Tax (lane 2), BEF (lane 4), or both Tax and BEF (lane 3).
(B) BEF and Tax stimulate GCN4 dimerization synergistically. GCN4 peptide was cross-linked with glutaraldehyde in the presence of BEF (lane 2), Tax (lane 4), or both BEF and Tax (lane 3). The positions of the cross-linked dimer and free monomer are indicated.
Molecular Cell 1999 4, DOI: ( /S (00)80369-X)

6. Figure 5 Isolation and Characterization of a BEF cDNA Clone
(A) Nucleotide and predicted amino acid sequence of HeLa BEF. Amino acid residues matching those of the tryptic peptides derived from purified BEF are in bold.
(B) Cellular localization of BEF. Immunofluorescence staining of HeLa cells using an αBEF polyclonal antibody (left) or DAPI (right).
(C) DNA binding stimulatory activity of recombinant BEF. In the upper left and middle panels, mobility shift assays using various bacterially expressed proteins in the absence (lane 1) or presence of increasing recombinant BEF (lanes 2 and 3). In the upper right panel, mobility shift assays of GST–ATF4 or λ-ZIP were performed in the presence of BEF or HLTV-I Tax as indicated.
(D) Recombinant BEF stimulates bZIP dimerization. GST–ATF2 was incubated with buffer (lane 1), His-tagged BEF (lane 2), or HLTV-I Tax (lane 3) in the presence of glutaraldehyde. The positions of the cross-linked dimer and free monomer are indicated.
Molecular Cell 1999 4, DOI: ( /S (00)80369-X)

7. Figure 6 Chaperone Activities of Recombinant BEF
(A) Sucrose Gradient Analysis of BEF. His-tagged BEF was fractionated on a 10%–40% sucrose gradient. Fractions containing molecular weight markers and His-tagged BEF were quantitated by laser densitometry and plotted against fraction number as shown. Peak fractions containing the molecular weight markers are indicated. Inset, peak fractions containing the molecular weight markers were plotted against native molecular weight on a logarithmic scale from which the apparent molecular mass of BEF was determined.
(B) Recombinant BEF prevents rhodanese aggregation. Aggregation of denatured rhodanese was measured as an increase in photometric absorbance and expressed as a percentage of the absorbance of diluted, native protein. Denatured rhodanese (920 nM) was diluted in the presence of 217 nM His-tagged BEF (diamonds), 326 nM His-tagged BEF (circles), 434 nM His-tagged BEF (triangles), or dialysis buffer (squares).
(C) Stoichiometry of BEF-enhanced bZIP DNA binding. GCN4 bZIP polypeptide (3.8 nM) was incubated with increasing amounts of His-tagged BEF and GCN4–DNA complexes detected in a mobility shift assay. The results were quantitated by phosphoimager analysis, and the molar ratio of BEF to GCN4 dimer was plotted.
(D) Recombinant BEF enhances renaturation of ATF-1. Reactivation of denatured ATF-1 was monitored in a mobility shift assay. Guanidinium hydrochloride–denatured recombinant ATF-1 was allowed to renature in the presence of dialysis buffer (lane 1), increasing recombinant BEF (lanes 2–5), increasing BSA (lanes 7–9), or increasing Tax (lanes 10–11).
(E) Direct interaction between BEF and bZIPs. GST (lane 1), GAL–VP16 (lane 2), GST–CREB (lane 3), or GST–ATF1 (lane 4) were fractionated on a 10% SDS-PAGE, transferred onto immobilon-P membrane, and incubated with 35S-labeled BEF (left) or visualized by Coomassie blue staining (right).
Molecular Cell 1999 4, DOI: ( /S (00)80369-X)

8. Figure 7
Antisense BEF Impairs Transcriptional Activation by bZIP Proteins In Vivo
HeLa (lanes 1–6) or 293T (lanes 7–10) cells were cotransfected with E4CAT (lanes 1–3), G5E1bCAT and GAL4–VP16 (lanes 4–6), or 3XTRECAT (lanes 7–10) in the presence of empty expression vector alone (lanes 1, 4, 7, and 9) or the BEF antisense expression vector (lanes 2, 3, 5, 6, 9, and 10) as indicated. Schematic diagrams of the reporters are shown below the autoradiograms. TPA was added as indicated. Each antisense experiment was repeated four to six times in duplicate with comparable results. In each transfection mixture, total DNA was normalized by using the empty expression vector.
Molecular Cell 1999 4, DOI: ( /S (00)80369-X)