Skip to main content
. 2012 Jun 19;40(16):7728–7738. doi: 10.1093/nar/gks473

Figure 3.

Figure 3.

Looping probability as a function of loop length at constant repressor concentration. Surprisingly, the sequence dependence of Figure 2E for the 94 bp construct is absent at other loop lengths. However, the bottom panels show data for constructs where 36 bp of either E8 or TA nearest O1 has been replaced with the lacUV5 promoter sequence (and for technical reasons O1 has been replaced with O2, which should not affect our measurements of J-factors as demonstrated by the data in Figure 2D). The presence of this promoter ‘restores’ a sequence dependence to looping across several helical periods. (A) Total looping probability (that is, both looped states summed) for the constructs Oid-E8-O1 and Oid-TA-O1, at 100 pM repressor. The red hatched region represents a prediction for where the TA data should fall, assuming the TA sequence has a J-factor anywhere from 5 to 30 times larger than the J-factor for the E8 sequence [a range based on the cyclization J-factors of (29)]. The lengths used in earlier cyclization assays (29) are a subset of those shown in this figure. (B) Looping probabilities for the two looped states separately (labeled ‘bottom’ (‘B’) and ‘middle’ (‘M’) as in Figure 2C and F) for the constructs in (A). The two states alternate in likelihood: the bottom state predominates near 89 and 100 bp, but the middle state near 94 and 106 bp. It is more clear in this panel than in (A) that E8 and TA are in phase with each other, with a period close to the canonical period of 10 bp, everywhere except near 94 bp, where TA has a maximum that is instead at 95–96 bp for E8. Therefore, a simple offset in phase between the two sequences cannot account for the behavior at 94 bp. (C) Looping J-factors for the constructs shown in (A). The J-factors for both E8 and TA span at least an order of magnitude as a function of loop length, and the J-factors for the two looped states (see Figure 4 and Supplementary Figure S8B) can also differ by an order of magnitude at a given loop length. However, as shown in Figure 4, this degree of modulation by operator phasing is less than might be predicted, depending on the assumptions made about Lac repressor conformation and flexibility. (D) Looping probabilities for constructs where part of the looping sequence of the constructs in (A) has been replaced with the 36-bp lacUV5 promoter. The red hatched region is the same kind of cyclization-based prediction as in (A). In sharp contrast to the data in (A), with the promoter sequence in the loop, TA loops as much or more than E8 at all lengths measured, as would be expected from cyclization and nucleosome formation assays with the pure E8 and TA sequences. Note that because of the replacement of O1 by O2, the looping probabilities for these constructs will not necessarily match those of (A) even when the J-factors for the loops, plotted in (F), are the same. (E) As in (B), here the two looped states have been separated out for the constructs in (D). With the promoter in the loop, the two sequences have the same phasing even at 94 bp [and in fact share the same phasing as the pure E8 constructs in (A)]. Interestingly, the preferred looped state with the promoter is almost exclusively the middle state at all lengths—note for example that at 107 bp without the promoter, the two looped states are comparable in likelihood (see also Figure 2F), but with the promoter at 107 bp only the middle state contributes to looping (see also Supplementary Figure S8D and E). (F) J-factors for the constructs in (D) (open circles), overlaid on the J-factors for the no-promoter E8 construct shown in (C) (grayed-out closed circles). The addition of the promoter to the loop does not appreciably change the J-factors for E8-containing loops, only those the TA-containing loops. See Supplementary Figure S8C for the J-factors of the two states of (E). Solid, dashed and dotted lines in (A), (B), (D), and (E) are guides to the eye only, not theoretical predictions or fits. Their purpose is to highlight general trends. Example bead motion-versus-time trajectories for these constructs can be found in Section S7 in the Supplementary Data, and the effective tether lengths of the two looped states as a function of the loop length, with and without the promoter, are presented in Section S6 in the Supplementary Data.