Enrichment for POU5F1 (Oct4) binding sites. (A) Perfect octamer, ATGCAAAT, containing oligonucleotide from the immunoglobulin heavy chain promoter (“wt” lanes 1,2) and its octamer scrambled control (“mut” lanes 3,4) were analyzed by EMSA with recombinant POU5F1(even number lanes) or no protein control (odd lanes). Mobility associated with singly bound oligonucleotides marked with an arrow; multiply bound oligonucleotides have been marked with a feathered arrow. Round 1 (lanes 5,6) of the POU5F1 enrichment was performed with the synthetic oligonucleotide pool as a probe. The singly shifted fraction was excised, reamplified, and used as a probe in lanes 7,8. The singly shifted fraction from round 2 was used as a probe in lanes 9,10. (B) EMSA performed using J1ES cell extract. Extract was preincubated with POU5F1 (Oct4) antibody (lanes 3,7,9) or with an antibody against a closely related POU2F1 (OCT1). Bands responsive to anti-POU5F1 antibodies are indicated with arrows. Lanes 6–9 display the oligonucleotide pool being shifted by ES extract. The starting oligonucleotide pool was used as probes for lanes 6,7, while lanes 8,9 use selected sequences from Round 2 (A, lane 8) of SELEX using the recombinant POU5F1. (C) Wild-type (wt), mutant, and the multiply shifted fraction of the oligonucleotide pool, excised (undetectable) from panel A, lane 8 (feathered arrow) were reamplified and used as a probe in an EMSA using recombinant POU5F1. (D) EMSA analysis performed using an increasing concentration gradient of recombinant POU5F1 protein on wild type and the multiply shifted fraction of the oligonucleotide pool.

MEGAshift tracks for the UCSC Genome Browser. Two of the 19 genomic regions corresponding to REST (A) and GADD45G (B) have been displayed in custom UCSC genome browser tracks. Annotation is stacked vertically along the chromosomal coordinate axis (X-axis). Starting from the top and proceeding down, the mouse (red) sequences are annotated for the following molecules: oligonucleotides cloned out of POU5F1 selected fraction(short, stacked bars). ChIP-PET regions (wide bars), normalized enrichment scores in grayscale for each duplicate probe pair for each microarray experiment (multiply bound, round 3, round 2, round 1). Enriched oligonucleotides are shaded darkly. Human (blue) is identical save ChIPped material is analyzed by microarray (ChIP-chip). Predicted OCT binding sites annotated below. Conservation determined by eight vertebrate BLASTZ alignments.

De novo motif identification. (A) For each oligo, singly bound (Y-axis) enrichment values were plotted against multiply bound enrichment (X-axis) values. POU5F1 contains two POU domains that recognize a bipartite signal as diagrammed in B. Half sites (ATGC, GCAT, AAAT, and ATTT) are counted in the entire set of oligonucleotidess, the set biased toward the singly bound state (above green line in A), and the set biased toward multiply bound (below the red line in A). Each permutation of half sites with more than twofold relative risk of being found in the multibound state versus the entire set is graphed. Red histogram bars mark relative risk (RR) of particular combination occurring in the multiply shifted fraction. Green bars for singly shifted fraction. Both measures are relative to the entire set and the blue dashed line marks zero enrichment (RR = 1).

MEGAshift protocol. (Step 1) All orthologous genomic regions enriched in both human and mouse POU5F1 (Oct4) ChIP experiments were aligned and resynthesized (step 2) as a tiled contig of 35-mers flanked by universal primer binding sites. The human genomic region was extended to cover the union of this overlap. (Step 3) This pool was amplified with labeled primers migrates as a single band and was then used in an EMSA activity with recombinant POU5F1. The shifted band was excised, reamplified, and either reshifted or analyzed by cloning or microarray. (Step 4) Microarray analysis. Shifted and unselected fraction were reamplified and the T7 containing template used to generate Cy3 (shifted) or Cy5 (unselected) targets for the custom oligonucleotidearray.

Changes in oligonucleotide enrichment throughout an POU5F1 SELEX experiment. (A) Enrichment scores for each round of SELEX were binned and graphed as a histogram. (B) Average enrichment scores were ranked with relevant oligonucleotides marked on the percentile bar. Gel shift assay was repeated for isolates cloned from selected (C) and unselected (D) fractions. (E) Microarray images corresponding to pool/pool and pool/round1 are drawn below the gel lane for each oligonucleotide.

Comparison of octamer site prediction and POU5F1 binding. (A) POU5F1 sites were scored for each oligonucleotide as the log probability that a random sequence would fit the octamer binding model better than the highest scoring window (Y-axis) in the oligonucleotide and plotted against enrichment (X-axis). Vertical line represents mean enrichment for each experiment. (B) De novo motif identification was performed using Gibbs sampling trials with varying amounts of input that were ranked according to enrichment in round 1 (red) or the multiply bound fraction (blue). Successful trials that converged on motifs with the POU5F1 consensus (ATGCAAAT) were recorded on the Y-axis. (C) Using the top 3% of enriched oligonucleotides, the effect of motif length was examined.