Simulations of DNA permeation through nanopores

In the second unit, you will learn how to manipulate DNA molecules and simulate their permeation through a synthetic nanopore.

Manipulating DNA

1

Enter cd ../5_manipulate_dna/ to start this section.

2

In the Tk Console type

mol load psf dsDnaAmber.psf pdb dsDnaAmber.pdb

In the Graphical Representations window, set the Drawing Method to Licorice and the Coloring Method to ResName.

3

You should now see an 8-basepair molecule of double-stranded DNA (dsDNA), colored by the residue names ADE, CYT, GUA, and THY; which correspond respectively to the bases adenine, cytosine, guanine, and thymine. (See Fig. 5. Try setting Selected Atoms in the Graphical Representations window to resname ADE, resname CYT, resname GUA, and resname THY in turn. Which colors correspond to which bases?

Figure 5: Double-stranded DNA colored by base type.

4

To determine the base sequence for the first strand, type the following in the Tk Console:

set a [atomselect top "segname ADNA and name C1'"]
puts [$a get {resid resname}]
$a delete

What is the sequence of the first strand (segment name ADNA)? What is the sequence of its complementary strand (segment name BDNA)?

5

There are several sets of parameters available for molecular modeling of DNA. We'll use the AMBER topology given in cornell.rtf and the interaction parameters given in cornell.prm. Another popular model of DNA uses the Charmm topology and parameter set. The script convertDnaToCharmm.tcl can produce a Charmm model from our AMBER model. The script applies patches using psfgen to change the topology from that of the AMBER model to that of the Charmm model using the Charmm topology file top_all27_prot_na.inp. Execute this script by typing source convertDnaToCharmm.tcl in the Tk Console.

6

In the Tk Console, enter
mol load psf dsDnaCharmm.psf pdb dsDnaCharmm.pdb. Set the Drawing Method to Licorice, the Coloring Method to Molecule, and Selected Atoms to all for both the AMBER DNA that we loaded earlier and the Charmm DNA. No difference in structure between the AMBER model and the Charmm model should be apparent. In the Tk Console, enter mol delete all.

7

Now we would like to produce single-stranded DNA (ssDNA) from dsDnaAmber.psf and dsDnaAmber.pdb. The script removeResidues.tcl deletes the residues of all atoms in a given selection. Open the script in your text editor. The first and second DNA strands have the segment names ADNA and BDNA, respectively. Set the value of selText in line 6 to segname BDNA so that the script will delete the second DNA strand. Save your changes and execute the script.

8

Let's check that we produced the ssDNA correctly. Enter mol load psf ssDna.psf pdb ssDna.pdb in the Tk Console. After examining your 8-mer ssDNA, type mol delete all.

ssDNA is much more flexible than dsDNA and easily bends into various conformations. The details of these conformations can be important for applications of bionanotechnology. For example, if ssDNA is to pass through a nanopore device, such as is proposed for a means of fast sequencing, it must be aligned somewhat along the axis of the pore. Molecules lying in the plane of the membrane or contorted in certain ways can make translocation more difficult or impossible. For this reason, we want the ability to easily generate any desired DNA conformation in silico.

Figure 6: Shaping single-stranded DNA. (a) The DNA begins in a straight conformation. (b), (c) Bending the DNA with Sculptor using two different paths as described in the text.

9

Here we will use the VMD script sculptor.tcl to shape ssDNA to our will. In the Tk Console, enter the following lines to load a 110-mer ssDNA molecule and open Sculptor:

mol load psf ssDnaLong.psf pdb ssDnaLong.pdb
source sculptor.tcl
sculptorGui

The Sculptor window should open. The script will map any long molecule aligned along the $z$-axis to a cubic spline whose form is given by the points in Path. If we are careful, the cubic spline allows us to bend the ssDNA smoothly, leading to conformations, that with some equilibration, could occur in nature. However, using Sculptor on structures that are not relatively straight along the $z$-axis, applying a tortuous path, or pressing the Sculpt button more than once without undoing the last operation will result in highly distorted and unphysical conformations. If this happens, simply reload the molecule.

10

Let's start by bending ssDNA into an L-shape. Delete the contents of Path, and type {0 0 1} {0 0 0} {1 0 0}. Press Sculpt. Rotate the molecule a bit and then press Undo. Your result should look like Fig. 6(b).

11

Now we'll bend the ssDNA in a U-shape. Delete the contents of Path and type {0 1 2} {0 1 0} {0 -1 0} {0 -1 2}. Imagine the positions of these coordinates in space. You should see that they form three sides of a rectangle. Production of a cubic spline from these control points will yield a U-shape as shown in Fig. 6(c). Press Sculpt. Undo this and then produce a few conformations of your own. Close Sculptor when you are finished. Then enter mol delete all.

Combining DNA and the synthetic nanopore

1

We now will combine our 8-mer ssDNA molecule with the nanopore. Execute the script combine.tcl, which will create pore+dna.psf and pore+dna.pdb. As shown below, the script combines the pore we created in Section 1.2 with the ssDNA using psfgen. The script is rather general, but can run into problems if segment names are duplicated between the scripts.

combine.tcl

# Input:
set psf0 ../1_build/pore.psf
set pdb0 ../1_build/pore.pdb
set psf1 ssDna.psf
set pdb1 ssDna.pdb
# Output:
set finalPsf pore+dna.psf
set finalPdb pore+dna.pdb

# Load the topology and coordinates.
package require psfgen
resetpsf
readpsf $psf0
coordpdb $pdb0
readpsf $psf1
coordpdb $pdb1

# Write the combination.
writepdb $finalPdb
writepsf $finalPsf

2

We've added the ssDNA without regard for the position of the pore. We now need to adjust the position of the molecule so that it is in a reasonable position for our translocation simulation. What is the charge of DNA? Which way will it move in an electric field pointing along the $z$-axis? Enter mol load psf pore+dna.psf pdb pore+dna.pdb in the Tk Console. Examine the system in the VDW representation. Using selection text like segname ADNA and within 4.0 of resname SIN allows us to see where the DNA has been placed too close to the .

Type the following commands into the Tk Console:

set sel [atomselect top "segname ADNA"]
$sel moveby {4 1 7}
set all [atomselect top all]
$all writepdb pore+dna.pdb
$sel delete
$all delete

VMD will not automatically update a selection defined by within commands after the ssDNA has been moved. To see the changes, simply change one letter in the Selected Atoms box, change it back, and press Enter. When you are convinced that the ssDNA is not too close to the , enter mol delete all.

Measuring ionic current with DNA

1

We've been running a lot scripts in our VMD session, some of which may have large global variables. This might be a good time to exit VMD and start a new VMD session to free any memory in these variables.

2

Enter cd ../6_current_dna/ in the Tk Console. Execute the solvation scripts addWater.tcl, cutWaterHex.tcl, and addIons.tcl in sequence.

3

To save the time it takes to equilibrate the system, we've included an equilibrium system (sample*) with which you can continue.

4

Calculate the value of eField necessary to apply 20 V along the $-z$-axis of the system with data from sample.xsc as you did in Section 1.6. Place this value in the configuration file run0.namd and execute NAMD with this file.

5

Execute the script electricCurrentZ.tcl to determine the ionic current. How does it compare with what you measured with no DNA in the system?

Simulating DNA translocation

1

Enter cd ../7_translocate/. In production simulations, translocation would be performed in solution. However, due to time constraints, we'll perform the translocation simulation in vacuum and then analyze the provided trajectory for a similar simulation in solution. We will also be using only short-range electrostatics (with a 12 Å cutoff) instead of PME electrostatics because the vacuum system has a nonzero charge. Electrostatic cutoffs are not recommended for most production simulations.

2

Execute constrainSilicon.tcl.

3

Run the NAMD configuration scripts shown in the table below sequentially. If you generated the pore with InorganicBuilder, you need to change cellBasisVector1 and cellBasisVector2 in eq0.namd to those you recorded. The final simulation may take several minutes to run, so if you are short on time you may want to skip this step and the one that follows.

NAMD script	steps	description
eq0.namd	201	minimization
eq1.namd	500	raise temperature from 0 to 295 K at constant $V$
eq2.namd	1000	constant $p$ and Langevin thermostat
run0.namd	8000	electric field 150 kcal/(mol Å e) at constant $V$

4

View the resulting trajectory in VMD by entering mol delete all and mol load psf pore+dna.psf dcd run0.dcd. Change the representation to VDW. Does the ssDNA translocate from one side of the pore to another?

5

Since your simulation was performed in a vacuum, we cannot analyze the ionic current. For this reason, the trajectory translocate.dcd along with the structure translocate.psf and extended system translocate.xsc has been provided. The data is from a 6 V translocation simulation of dsDNA. Execute electricCurrentZFrame.tcl to calculate the ionic current for this trajectory. Unlike the script of a similar name we used previously, electricCurrentZFrame.tcl records the time in DCD frames, instead of nanoseconds, to facilitate comparision with the trajectory. The results are placed in the file curr_6V.dat.

6

The script trackPositionZ.tcl operates in much the same way as
electricCurrentZFrame.tcl except that it determines the center of mass of the DNA relative to the center of the pore instead of the current. Enter source trackPositionZ.tcl. The $z$-position of the center of mass is stored as a function of frame number in pos_6V.dat.

7

Open the trajectory in VMD with the following commands:

mol delete all
mol load psf translocate.psf dcd translocate.dcd

In the Graphical Representations window, change Selected Atoms to resname SIN and y > 0. Change the Drawing Method to Beads. Create a new representation with the selection segname ADNA BDNA and the drawing method VDW.

8

Now plot current versus frame (curr_6V.dat) and center-of-mass position versus frame (pos_6V.dat) and compare it with the events that take place in the trajectory. How does the passage of the DNA change the current?