Topographic characterization and electrostatic response ofM-DNA studied by atomic force microscopy

1. Introduction

In recent years, the scientific community has focused on the search for an appropriate molecular wire for the incoming nanotechnology. Among other molecules, DNA has experienced intense research trying to characterize its electrical properties [1–8]. To date, the electric behaviour of B-DNA is still under debate. Metallic engineered DNA, M-DNA, has also been proposed as a molecular wire [9, 10]. M-DNA has attracted interest because some unusual electrical properties have been reported [9, 10]. However, to date the effect of the metal ions on the structure of the regular B-DNA is not clear and no surface characterization has been reported. If future experiments confirm the possibilities of thismodifiedmolecule in the field of nanotechnology, topological characterization of these molecules should be of great importance. M-DNA has been described as a complex of B-DNA with the divalent metal ions Co2+, Ni2+ and Zn2+. Present knowledge concerning M-DNA properties mainly derives from studies with bacterial and synthetic DNAs [11] These studies have characterized, in different detail, several properties.

(i) M-DNA does not bind ethidium bromide, thus a 'fluorescence assay' in the presence of ethidium bromide can be used to monitor M-DNA formation.
(ii) The mobility of M-DNA in agarose gels has been postulated to be similar to that of B-DNA; this observation suggests that metal ions do not cause condensation of the DNA.
(iii) It was suggested that the imino protons of T and G were replaced in M-DNA by the metal ion during the B-DNA to M-DNA transition.

We have used atomic force microscopy (AFM), also called scanning force microscopy (SFM), to characterize the surface of the metal modified DNA. The AFM was invented in 1986 [12]. Since then, it has become a routine technique of surface characterization. The technique has experienced an enormous growth, specially in the field of biophysics, and in particular in DNA imaging [13–23]. In this paper we describe a simple protocol to produce MDNA. Cobalt and nickel modified DNA results in a reduction of its longitudinal dimension as well as an increment of the molecular height. We deduce that the presence of these metal ions in the DNA condenses the molecule. Release of the DNA molecules fromthe M-DNAstatewas achieved experimentally
by removal of the divalent cations through EDTA chelation.

In order to investigate whether the electrical transport properties of DNA are affected by the intercalation of metal ions we have used AFM. Firstly, we have carried out a topographic characterization of the M-DNA molecules and secondly, we have used AFM to study the electrical properties of the modified DNA molecules. This has been accomplished by performing two types of electrostatic experiment on MDNA following [8, 24]. These two experiments show no evidence for molecular wire behaviour of M-DNA as reported on non-modified DNA [8].

2. Materials and methods

2.1. M-DNA sample preparation

To generate DNA linear fragments of different lengths, the pUC18 and YEp13 plasmids were linearized at unique BamHI sites. Two linear fragments of 2690 bp and 10 667 bp were obtained. These DNA fragments were used to prepare DNA modified molecules with the divalent ions Ni2+ and Co2+. The DNA fragments were dissolved in 90 mM tris borate buffer, pH 8.5, at a concentration of approximately 0.1 mg ml−1. 5 μl DNA aliquots were added to 20 μl of a buffer containing 90 mM tris borate and 10 mM NiCl2 or 1 mM CoCl2, pH8.5. The pH of the resulting assay mixture was maintained at approximately 8.5 with 1MKOH. Typical reaction times were 40–45 min at room temperature. Because the M-DNA does not bind ethidium bromide [9], M-DNA formation was followed by agarose electrophoresis measuring the fluorescence of intercalated ethidium bromide. To localize M-DNA in the agarose electrophoresis the DNA samples were labelled with [α-32P]dCTP by filling in the protruding 5 ends in the presence of Klenow fragments.

2.2. AFM imaging

AFM images were obtained with a commercial AFM [25]. We have used non-contact dynamic SFM for all the images shown. We used soft cantilevers of nominal force constant 1 N m−1, resonance frequency 75–80 kHz and tip radius 15–20 nm [26]. The cantilever is oscillated at its resonance frequency. The amplitude and the relative phase of the oscillation are measured through the normal force signal. The feedback is performed on the amplitude signal channel. In this manner tip–sample distance is kept constant. The tip is at, typically, a few nanometres above the surface and the amplitude of the oscillation at which the images are taken is about 25 nm. For more details see [27, 28].

2.3. AFM sample preparation

Mica is an appropriate substrate for imaging biomoleculeswith AFM [29–31]. Mica exposes a negatively charged surface when cleaved, therefore, divalent cations like Zn2+ or Mg2+ are employed to adsorb negatively charged molecules like DNA [17]. We have found that the absence of this divalent ion inhibits the adsorption of the molecules to the mica. The standard DNA sample preparation protocol for AFM consists of a final mixture of 1–10 mM MgCl2 and 10–100 ng of DNA diluted in tris EDTA buffer placed on a freshly cleaved mica disc and extensively washed with double-deionized ultrapure water. Finally, the sample is dried with a gentle stream of air or dry nitrogen. Using an AFM, DNA molecules so prepared can be easily imaged [21, 32]. For M-DNA, no extra addition of divalent cations was necessary. 5μl of the M-DNA reaction were placed on a mica disc and then we followed the washing and drying part as described above. The concentration of MDNA molecules was checked by AFM inspection.

For the first electrostatic experiment on M-DNA, a 30 nm thick film of gold was evaporated on the samples partially covering the molecules as described in [1]. The electrostatic method employed here is described in detail in [8]. For the second electrostatic experiment no additional evaporation was needed. AFM tips employed in the electrostatic experiments were covered with 20 nm of titanium and 20 nm of gold.

3. Results and discussion

Nickel and cobalt M-DNAsampleswere prepared as described in section 2. It is known that ethidium bromide (EtBr) does not bind M-DNA [9]. We have used this property of M-DNA to design a 'fluorescence assay', in the presence of EtBr, to monitor M-DNA formation. 5 μl DNA aliquots were added to 20 μl of a buffer containing 90 mM tris borate and from 0.1 to 100 mM NiCl2 or from 10 μM to 50 mM CoCl2 , pH 8.5, using the assay conditions described above. After 45 min incubation at room temperature the samples were subject to electrophoresis in a horizontal 0.8% agarose gel at 10 V cm−1 of gel for 45 min in 0.5 × TB (45 mM tris borate, pH 8.5) containing 5 μg EtBr. The minimum cation concentration that allows no binding of EtBr to DNA was selected for Ni2+ (10mM) and Co2+ (1 mM) ions (figure 1(A)). As can be seen in figure 1 (lanes 2 and 3), the fluorescence of the modified DNA disappears because M-DNA does not bind ethidium bromide.

The unmodified B-DNA remains fluorescent (lane 1). This simple and easy testwas used routinely to check the production of M-DNA. In order to know the exact position of the M-DNA in the gel, the DNA was radioactively labelled with [α-32P]dCTP by filling in the protruding 5 ends in the presence of Klenow fragment DNA polymerase. After agarose gel electrophoresis of the DNA samples, gels were autoradiographed at −70 ◦C with an intensifying screen (figure 1(B)). One interesting find is that the M-DNA does not enter the gel matrix. This can be explained in two manners: either M-DNA does not have any type of net charge and in consequence does not run into the agarose gel or the modified DNA cannot enter the gel because of sterical factors. The pore size of a 0.8% agarose gel is about 50 nm [33]. If the modified DNA does not enter the gel because it cannot fit the pore, it must have an structure of tens of nanometres. This result is in contradiction with the mobility in agarose gels of M-DNA reported before [9]. To know more about the physical appearance of the M-DNA, we have used the ability of the AFM to image insulating samples at the nanometre scale.

 

M-DNA samples for AFM experiments were bound to a mica surface for imaging. Interestingly, itwas not necessary to add any extra ions to the M-DNA samples to assist binding to the mica substrate. This indicates that the modified DNA has some kind of charge. We should, then, discard the hypothesis of the lack of charge as the reason why the M-DNA does not enter an agarose gel. We think that the excess of divalent ions (Ni2+ or Co2+) binds to the surface and charges it. The positively charged mica is then appropriate for binding the M-DNA. It has been reported that the use of other divalent ions than Mg2+ also promotes adhesion of DNA to the mica surface [34, 35].

Atomic force microscopy images of Ni2+ and Co2+ MDNA are shown in figure 2. A set of images at different magnifications is shown. No major differences between Ni2+ or Co2+ modified species of DNA were found from the AFM images. Figure 2(A) is a 5 μm × 5 μm image where MDNA molecules can be seen. The B-DNA, in this case, is a linearized plasmid of 10 667 bp (3.6 μm). One characteristic of the M-DNA molecules is that they are much shorter than the regular B-DNA. In figure 2(A) the longest molecule found has a length of 600–700 nm. There is a typical fivefold shortening in the length. Another interesting feature of the sample is the o-ring shapes detected everywhere. This can be explained because the Yep13 plasmid was linearized with BamHI. This restriction enzyme leaves a four-base overhang. During the MDNA production process the DNA could be annealed forming a circular shape. The interesting feature of theDNAshortening during the B-DNA to M-DNA transition is clearly visible in figure 2(B) (a 3 μm × 3 μm image) and figures 2(C) and (D) (details of 1 μm size).

4. Conclusion

We have presented one of the first AFM measurements of metallic engineered DNA samples (M-DNA). We have shown an easy and quick method to obtain M-DNA and a way to test the production through a 'fluorescence assay' in the presence of ethidium bromide. Our results demonstrate that M-DNA cannot enter an 0.8% agarose gel because the size of the modified structures is bigger than the pore size.

AFM measurements on M-DNA samples shows a fivefold prove that the reaction of the M-DNA formation is reversible upon the addition of EDTA. The AFM images demonstrate that the M-DNA molecules treated with EDTA recover their nominal length and height typical for the DNA in B-form. The electrostatic experiments performed on M-DNA molecules show no evidence for molecular wire behaviour since no topographic modification on the molecules is observed when
a bias voltage is applied between the metallized tip and the electrode that covers them. In addition, no change in the resonance frequency of the cantilever is detected when the biased tip passes over the M-DNA molecules, confirming the insulating nature of the M-DNA.shortening in the length of the DNA as well as an increment in the height of almost one order of magnitude.

This kind of condensation could be important to explain previous results obtained on this kind of molecule regarding the electrical properties of the modified DNA.We have also used theAFMto prove that the reaction of the M-DNA formation is reversible upon the addition of EDTA. The AFM images demonstrate that the M-DNA molecules treated with EDTA recover their nominal length and height typical for the DNA in B-form.

The electrostatic experiments performed on M-DNA molecules show no evidence for molecular wire behaviour since no topographic modification on the molecules is observed when a bias voltage is applied between the metallized tip and the electrode that covers them. In addition, no change in the resonance frequency of the cantilever is detected when the biased tip passes over the M-DNA molecules, confirming the insulating nature of the M-DNA.

---

Invite your mail contacts to join your friends list with Windows Live Spaces. It's easy! Try it!