Materials and Methods FtsK50C construct design. …

Materials and Methods

FtsK50C construct design.

       PredictProtein1 was used to create a multiple sequence alignment, which

predicted conserved regions, secondary structure elements, and unstructured regions.

3D-PSSM threading2 was used to predict tertiary folds adopted by the -domain.

FtsK50C construct cloning.

       Plasmid pTM123, which contained the gene for FtsK50C was inverse PCR

amplified using AccuTaq LA (Sigma) and the primers (1) 5-Phos-

TGATGAGGTGCGGGTGGTTTCGATGGCGCT and (2) 5-Phos-

ACCTTCGCTTTCGCTGTCGGA, to insert a UGA stop codon after residue 487 in

the FtsK50C gene (which corresponds to residue 1254 of full-length FtsK from E. coli).

PCR products were blunt-ended with Vent DNA Polymerase (NEB), gel purified, and

circularized using T4 DNA ligase (NEB). Clones were propagated in E. coli strain

DH5 and verified by sequencing.



FtsK50C and FtsK50C expression and purification.

       FtsK50C and FtsK50C were expressed and purified as described3. The purest

fractions from each purification (>95% pure), as determined by SDS-PAGE and

coomassie blue staining, were pooled and quantified using the Bradford Method with

BSA (NEB) as a standard. Stocks were stored at 4°C without detectable loss of

activity over the course of the experiments. FtsK50C displayed comparable DNA-

dependent ATP hydrolysis characteristics to FtsK50C (data not shown).
Triplex Displacement Assays.

       Triplex substrates were prepared as described4. Triplex displacement assays

were carried out as described4 with 235nM FtsK50C, 470 nM FtsK50C, or dialysis

buffer. Background was subtracted as described4.



Optical Tweezers Substrates Preparation.

 DNA single molecule substrates were prepared as described3. Briefly, biotin- and

digoxygenin-modified PCR products were digested with KpnI or XbaI (NEB), to

create linker molecules. Lambda DNA (NEB) was circularized using T4 DNA ligase

(NEB), digested with KpnI and XbaI, and the 41 kbp fragment was gel purifed and

ligated to biotin- and digoxygenin-modified linker molecules to create DNA tethers.



Optical Tweezers Assay.

       FtsK50C and FtsK50C were used at 150nM. At these concentrations, actively

translocating particles of aggregated enzyme were observed and used to track the

position of a single active FtsK4. This optical tweezers particle assay has been used

extensively used to detect FtsK directionality (Refs. 1, 4). We believe that the FtsK

particle motion observed in these experiments is due to a single active complex within

the particle, as we previously showed (Ref. 4). The reasons are: (i) We never observe

DNA reeling in from both sides of the FtsK particle. If multiple motors were active

due to stochastic loading of the motors on the DNA, then we would expect to see

some examples of the DNA being reeled in from both directions. (ii) We measured

similar velocities at low FtsK concentrations where we see no aggregation and at

concentrations >100 nM which promote aggregation. (iii) The change in direction of

movement at non-FRS sequences is extremely rapid. The transition between the
forward and backward movements by the FtsK particle occurs within a 30th of a

second, the temporal resolution of our assay. If multiple motors were active at the

same time, then we would expect a slower transition. (iv) If there were multiple

motors binding stochastically, there should occasionally be more than one motor

acting on a DNA. However, the standard deviation of the rate is relatively small (4.0 ±

1 kb/sec). (v) Translocation rates and other properties of the FtsK motor remained the

same independently of whether we observed FtsK particles in the optical tweezers or

single FtsK complexes in the magnetic tweezers.



       Reactions were conducted at room temperature in 50mM Tris, pH7.5/ 5mM

MgCl2/ 3mM ATP/ 50mM NaCl/ 1mM DTT/ 0.1mg/ml BSA. DNA tethers were

bound between 2.8 µm diameter anti-digoxygenin bead held in an optical trap and

2.2µm streptavidin-coated bead immobilized on a micropipette. Data was captured on

digital video at 30Hz using a SONY Digital8 video walkman and analyzed with NIH

Image and Matlab. Particles were tracked by following the movement of the centroid

of FtsK particles in the direction parallel to the DNA tether as a function of time. A

translocation event is defined as the whole period of activity in which a single FtsK

particle moves for at least 0.5 µm at a minimum speed of 0.25 µm/sec.

       These data were employed to calculate the number of FtsK particles traveling

in the permissive n+(d) or non-permissive n-(d) orientation as a function of their

position d on  DNA (Fig. 1C). The probability of FtsK traveling in the permissive

(non-permissive) direction was calculated by integrating n+(d) (n-(d)) for all values of

d in the interval [0, 44] kbp and normalizing by the total number of events. The net

distance translocated by an FtsK particle was measured by subtracting the initial

position of the FtsK particle on  DNA from that at the end of the translocation event.
       A burst was defined as the region of a translocation event in which the FtsK

particle moves in one direction without pausing or turning around. The burst size was

thus defined as the total distance translocated in a burst. Bursts were identified using a

burst-finding algorithm with data averaged using a four point rolling average with a

lower threshold velocity of 0.1µm/s. The sign of the burst size was defined as positive

when the particle traveled in the permissive orientation and negative when it traveled

in the non-permissive one. The distribution of burst sizes was calculated from a total

of 441 bursts for FtsK50C and 262 bursts for FtsK50C.

       Optical tweezers were operated under constant feedback mode, and the tension

on the DNA was always >20 pN. Although FtsK-induced loop formation could have

occurred at the lowest forces used (20 pN), loop sizes and frequency were minimized

by the immediate increase in force generated when looping displaced the bead from

the optical trap. At forces >30 pN, FtsK is unable to efficiently loop DNA as we have

previously shown (Refs. 1, 4). FtsK translocation velocities were measured by

tracking the position of the centroid of FtsK particles as a function of time. Looping

of DNA would not have changed the position of FtsK particles but rather the distance

between the beads. Therefore, our velocity measurements directly reflect on the

translocation of FtsK. The mean velocities measured with this assay (4 ± 1.5 kbp/s)

were similar to those reported in Ref. 2 (~ 5.5 kbp/s at 3 mM ATP concentration), our

previous optical tweezers experiments (5 ± 1 kbp/s, Ref. 4), and our magnetic

tweezers experiments (~ 4.5 kbp/s).



Magnetic Tweezers Substrates Preparation.

       A 12.2 kbp region of the E. coli K-12 genome (nucleotides 3178400-

3190613) devoid of any FRS or RAG family motifs4,5 was PCR amplified using
AccuTaq LA DNA polymerase and primers

AGTAGTCTAGAGCGTGGAATCCAGGGCGCA and

GTGCAACCGGTACCTCTTCTCGTTC. PCR products were digested with XbaI,

KpnI, and DpnI and gel purified. Products were ligated into XbaI and KpnI sites of

pBS-SK(+). Clones were propagated in E. coli strain DH5 and verified by

restriction digestion. To insert FRS sequences into these clones in the desired

location, inverse PCR with AccuTaq LA DNA polymerase and the primers 1)

5'ATGATCCATGGGCAGGGCAGGGCAGGGCCCGTCCATATTCAACGGCT

and 2) 5'ATGATCCATGGTGGATAGCGCACGCGCCCTT to introduce the FRS

sequences and a unique NcoI site for circularization. PCR products were digested

with NcoI and DpnI (NEB) and gel purified. T4 DNA ligase (NEB) was used to

circularize products. Clones were propagated as described above and verified by

sequencing. To create DNA tether molecules, the plasmid described above was

digested with either XbaI and KpnI or XbaI and SalI to create inserts of 12.2 kbp or

8.8 kbp with the FRS repeat in the center or 2.7 kbp from one end, respectively.

These molecules were ligated to biotin- and digoxygenin-modified PCR products to

create tether molecules.



Magnetic Tweezers Experiments

       Magnetic tweezers experiments were performed by using an FtsK

concentration ranging from 10 to 15 nM. At these concentrations, we have not

observed FtsK aggregates. In this assay, loop formation by FtsK was the signal used

to monitor FtsK activity. These experiments were conducted on a high-power

magnetic tweezers instrument described elsewhere 6 7 (Hong S.C, Stone M.D,

Humphries D., N.M., C.B., and N.R.C., manuscript in preparation). Briefly, the
optical setup is a home-made inverted microscope, equipped with high power hybrid

permanent magnets that can be translated to control precisely the degree of tension

applied to single DNA molecules. The magnets enabled the application of stretching

forces up to ~10 pN using 1 micron-diameter paramagnetic beads. Dynamic changes

in DNA extension were measured in real time at 10-30 Hz by comparing the bead

diffraction ring pattern with a previously calibrated set of images taken at known

focal displacements. To ensure accuracy of our extension measurements, each

experimental setup was individually calibrated by an automated routine. Activity

assays were conducted in FtsK reaction buffer containing: 50 mM Tris-HCl pH 7.5,

3mM ATP, 10-15nM FtsK50C or FtsK50C , and 0.1 mg/ml BSA. The distance

between the magnets and the sample was held constant throughout each experiment,

resulting in constant force (~ 5-8 pN) during measurements. Modified DNA

molecules possessing multiply-labeled biotinylated and digoxigenated ends were

oriented between an -digoxigenin antibody coated glass surface (Roche) and a 1 µm

diameter streptavidin bead (Dynal, My One Beads). FtsK activity was observed as a

change in the DNA extension resulting from the formation or release of a DNA loop.

In this assay, we have observed both spontaneous reverse translocation and abrupt

loop release by the motor, as previously described in Ref. 2. In the range of forces

used, however, we mostly observe translocation reversals.

       In measuring translocation rates, waiting times prior to the onset of activity

were required to be five times longer than the total burst time. The translocation rate

was then measured in each burst by using a piecewise linear fitting algorithm. For the

determination of mean translocation rates, 142 events on 6 tethers were considered for

FtsK50C and 186 events on 7 tethers for FtsK50C. Occupancy time distributions for

FtsK50C and FtsK50C were measured by calculating the histogram of DNA
extensions during periods of activity and pauses for nine experiments on three and

five tethers, respectively. Extensions corresponding to zero or the full length of the

tether were not considered for clarity. The occupancy time distribution for FtsK50C in

Fig. 2D (blue) is higher at higher DNA extensions (after passing through the FRS

repeat). Before translocation begins, the DNA tether has full extension and FtsK

translocation by looping shortens the DNA extension. As shown in Fig. 2B, the

recognition of FRS by FtsK leads to a translocation reversal, which implies that most

of the time the DNA tether is longer than the position of FRS in the tether. Only in a

few cases FtsK fails to recognize FRS and thus translocates the full length of the

DNA tether. For this reason, most of the time the DNA length is higher than that

corresponding to the position of FRS, and so the distribution for FtsK50C is

asymmetric with respect to this position.

The occupancy time distribution for FtsK50C (Fig. 2d, red) is, on the other hand,

homogeneous. We have shown in the optical tweezers experiments that the mean

processivity of the FtsK motor (FtsK50C) is ~ 9 kbp. The tether used in the magnetic

tweezers experiments was 12 kbp long, and so most often the activity of FtsK led to

the full tether translocation (see Fig. 2C). For this reason, this distribution is

homogeneous.



Supplementary References

1.      Rost, B., Yachdav, G. & Liu, J. The PredictProtein Server. Nucleic Acids
        Research 32, W321-W326 (2003).
2.      Bates, P.A., Kelley, L.A., MacCallum, R.M. & Sternberg, M.J. Enhancement
        of protein modeling by human intervention in applying the automatic
        programs 3D-JIGSAW and 3D-PSSM. Proteins Suppl 5, 39-46 (2001).
3.      Pease, P.J. et al. Sequence-directed DNA translocation by purified FtsK.
        Science 307, 586-90 (2005).
4.      Levy, O. et al. Identification of oligonucleotide sequences that direct the
        movement of the Escherichia coli FtsK translocase. Proc Natl Acad Sci U S A
        102, 17618-23 (2005).
5.   Bigot, S. et al. KOPS: DNA motifs that control E. coli chromosome
     segregation by orienting the FtsK translocase. Embo J 24, 3770-80 (2005).
6.   Strick, T. Experimental Setup for Magnetic Tweezers (Strick et al). in
     Biophysical Journal (1998).
7.   Strick, T.R., Allemand, J.F., Bensimon, D. & Croquette, V. Behavior of
     supercoiled DNA. Biophys J 74, 2016-28 (1998).