Observation of single-stranded DNA on mica and highly oriented pyrolytic graphite by atomic force microscopy

(1)

Observation of single-stranded DNA on mica and highly

oriented pyrolytic graphite by atomic force microscopy

Jozef Adamcik

a,b

, Dmitry V. Klinov

c

, Guillaume Witz

a

, Sergey K. Sekatskii

a

, Giovanni Dietler

a,* a_{Laboratoire de Physique de la Matie`re Vivante, Ecole Polytechnique Fe´de´rale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland}

b_{Department of Biochemistry, Institute of Chemistry, Faculty of Sciences, P. J. Sˇafa´rik University, 04154 Kosˇice, Slovakia} c

Shemyakin-Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences, 117871 Moscow, Russia Received 25 July 2006; revised 4 September 2006; accepted 8 September 2006

Available online 18 September 2006 Edited by Lev Kisselev

Abstract Atomic force microscopy was used to image single-stranded DNA (ssDNA) adsorbed on mica modified by Mg2+, by 3-aminopropyltriethoxysilane or on modified highly oriented pyrolytic graphite (HOPG). ssDNA molecules on mica have compact structures with lumps, loops and super twisting, while on modified HOPG graphite ssDNA molecules adopt a confor-mation without secondary structures. We have shown that the immobilization of ssDNA under standard conditions on modified HOPG eliminates intramolecular base-pairing, thus this method could be important for studying certain processes involving ssDNA in more details.

Keywords:Atomic force microscopy; Single-stranded DNA; Double-stranded DNA; Mica; Highly oriented pyrolytic graphite

1. Introduction

Atomic force microscopy (AFM) is extensively used for studying biological molecules, in particular DNA[1]. One of the characteristics of AFM is its two-dimensional representa-tion of biomolecules, because the molecules are immobilized on appropriate surfaces. Mica is a commonly used substrate for biological samples, due to its very flat surface that can be modified for the adsorption of the biomolecules. In general, two modifications of the bare mica surface are used for DNA immobilization. In one case, DNA is immobilized on mica by deposition from a solution containing divalent ions such as Mg2+ [2]. It is well known that divalent ions act as bridges between DNA and the mica substrates, leading to a weak adsorption of DNA. It is expected that DNA molecules equilibrate on the surface[3]and assume a 2D conformation. A second modification of mica consists to treat the mica surface with 3-aminopropyltriethoxysilane (APTES-modified mica)[4,5], leading to a strong adsorption of the DNA mole-cules. This strong adsorption is akin to a projection from three- to two-dimensions. The strong interaction between

DNA and surface irreversibly adsorbs the DNA molecules from solution, permitting to recover their three-dimensional properties[6].

Highly oriented pyrolytic graphite (HOPG) has also an atom-ically ﬂat surface, which can be used for DNA imaging[7–9], but unlike mica, its surface is hydrophobic and the DNA immobilization is not very easy. Nevertheless, there are HOPG modiﬁcations allowing DNA adsorption on HOPG[7,8].

Already various properties of double-stranded DNA (dsDNA), DNA-small molecules and DNA–protein interac-tions have been investigated by AFM [6,10–16]. However, the investigation of single-stranded DNA (ssDNA) molecules by AFM is not straightforward, because ssDNA has the ten-dency to form inter-strand base pairing leading to a wealth of secondary structures. Often ssDNA has been coated with RecA for its visualization by AFM[17]. In the present work, we introduce a simple way for immobilization and visualiza-tion of ssDNA on HOPG, which slightly denaturates ssDNA and prevents the formation of intra-strand base-pairing. We compare the method with the commonly used protocols for adsorption of DNA on mica surface. The present method is particularly suitable when the length of the ssDNA part of the molecules is aimed, instead of the secondary structure: the former case would be of interest for example when en-zymes/proteins generating, manipulating or binding to ssDNA are in the focus of the study.

2. Material and methods

2.1. Sample preparation for AFM imaging

Plasmid pBR322, Lambda DNA (Lambda Mix Marker 19) and restriction enzyme EcoRI were purchased from Fermentas, while ssDNA M13mpl8 was purchased from Sigma. Plasmid pBR322 was treated with EcoRI to obtain a homogenous linear DNA. All DNA preparations were diluted in 1 mM Tris–HCl buffer to a final DNA concentration of 1lg/ml. MgCl2was added in the DNA solution to a final concentration of 5 mM before deposition on freshly cleaved mica. APTES mica was prepared by the evaporation method described by Lyubchenko et al.[4]. A 10ll aliquot of the DNA solution was deposited onto the appropriate surface (freshly cleaved mica and APTES-mica, respectively) and incubated for 10 min at room temper-ature. The sample was then rinsed with nanopure (Ultra High Quality) water (USF Elga, High Wycombe, England) and dried with air. The same procedure was applied for deposition of heated DNA. Lambda DNA solution was heated at 75°C for 30 s before deposition to yield ssDNA molecules.

For the adsorption on HOPG (ZYA quality from MikroMasch), 30ll of graphite modiﬁer ‘‘GM’’ (stock solution without dilution) obtained from Nanotuning (Nanotuning, Chernogolovka, Russia; Abbreviations:AFM, atomic force microscopy; ssDNA, single

stran-ded DNA; HOPG, highly oriented pyrolytic graphite *

Corresponding author. Fax: +41 21 693 04 22. E-mail address:Giovanni.Dietler@epﬂ.ch(G. Dietler).

(2)

http://www.nanotuning.com) was deposited on a freshly cleaved HOPG surface. After 10 min the droplet was removed from HOPG and 30ll of DNA solution was deposited on the modiﬁed HOPG and incubated for 10 min. The droplet was removed and the sample surface was dried with air.

2.2. AFM imaging

Images were collected using a Nanoscope IIIa (Veeco Inc., Wood-bury, NY, USA) operated in tapping mode in air. Ultrasharp non-contact silicon cantilevers (NT-MDT Co., Zelenograd, Moscow, Russia) with a nominal tip radius of <10 nm were used and were driven at oscillation frequencies in the range of 150–300 kHz. During imag-ing, the surface was scanned at a rate of one line per second. Images were simply ﬂattened using the Nanoscope III software and no further image processing was done. The contour length of DNA molecules was measured with Ellipse[18].

3. Results and discussion

3.1. Imaging of dsDNA

First, we present results for dsDNA on the three diﬀerent substrates.Fig. 1shows typical AFM images of linear dsDNA deposited (a) on mica from a solution containing divalent ions Mg2+; (b) on mica treated by APTES; and (c) on modiﬁed HOPG. InFig. 1A are imaged DNA molecules deposited in the presence of Mg2+_{, which show few crossings, because this}

deposition method permits a 2D equilibration of dsDNA[3]. InFig. 1B, the DNA molecules have many crossings, because the strong adsorption on APTES-mica is akin to a geometrical projection from 3D conformation to a 2D plane. Valle et al.[6] have studied the behavior of linear dsDNA on APTES modi-fied mica, when molecules have many crossing. They could show that the 3D properties could be recovered even if ad-sorbed on a 2D surface. InFig. 1C images of dsDNA mole-cules on ‘‘GM’’ modified HOPG are shown: the dsDNA conformation resembles to that of DNA molecules on mica in the presence of Mg2+ (Fig. 1A) with few crossings. Thus, dsDNA on ‘‘GM’’ modified HOPG has a certain degree of 2D relaxation.

3.2. Imaging of ssDNA

It is not possible to evaluate native ssDNA on the usual deposition substrates for dsDNA, because ssDNA forms many secondary structures. However, Woolley and Kelly have dem-onstrated a method for studying ssDNA molecules on poly-lysine, which yielded extended ssDNA molecules that were imaged by AFM[19].

Here, we study ssDNA deposited on modiﬁed mica or on modiﬁed HOPG and we demonstrate a very easy way to visu-alize it. To obtain ssDNA, Lambda DNA was heated at 75°C for 30 s before deposition. Alternatively, commercially avail-able ssDNA M13mpl8 was deposited as a comparison for imaging ssDNA.

Adsorption in the presence of Mg2+;Fig. 2A and B display AFM images of ssDNA M13mpl8 and heated linear DNA, respectively. In the latter case, the images include ssDNA and rests of non-denaturated dsDNA. In the presence of Mg2+, ssDNA and dsDNA molecules are well adsorbed and can be easily identiﬁed. However, the ssDNA parts of the mol-ecules are not elongated and have very compact structures with many kinks and nodes due to intra-strand base-pairing, while the dsDNA parts are in almost 2D equilibrium conformation (Fig. 2B). Additionally, Mg2+ions can link two single strands.

Our results are in a good agreement with previously ﬁndings [7]. Moreover, it was not possible to measure the height and width of ssDNA.

Fig. 2C and D represent images of ssDNA M13mpl8 and heated linear DNA on APTES modiﬁed mica, respectively.

Fig. 1. AFM images of Lambda dsDNA adsorbed on the surface of: (A) mica in the presence of 5 mM Mg2+_{; (B) mica modiﬁed by APTES;} and (C) ‘‘GM’’ modiﬁed HOPG. The bars represent 250 nm on all images.

(3)

The binding of ssDNA on APTES mica is very strong and the conformation of ssDNA is diﬀerent in comparison to the dsDNA conformation (seeFig. 1B and Ref. [6]). We can ob-serve that dsDNA part of the molecules has many crossings due to its strong absorption on the APTES surface (Fig. 2D). At the extremities of the double-stranded parts, it is possible to observe the beginning of separation of the double helix (Fig. 2D). The ssDNA parts are compact with many loops and intra-strand base-pairing (Fig. 2C). The height of the paired ssDNA strands is not exactly the same as for dsDNA, probably indicating that these parts do not have the typical double helix structure. ssDNA structure on APTES mica reminds RNA molecules with many loops and base-pair-ings. In spite of the compact structure of ssDNA, it was possi-ble to estimate its height and width in the loops. For dsDNA the height was 0.7 ± 0.1 nm and the width 12 ± 3 nm; for ssDNA the height was 0.3 ± 0.1 nm and the width 7 ± 2 nm.

Deposition of DNA on HOPG is not very easy due to its hydrophobic surface. Brett and Chiorcea have shown that dur-ing deposition of ssDNA and dsDNA on HOPG, DNA was condensed forming complex networks[8]. For ssDNA immo-bilization on HOPG a ﬁlm of graphite modiﬁer ‘‘GM’’ was spread on the substrate. ‘‘GM’’ is an aqueous solution of poly-meric molecules consisting of a hydrophobic part

(hydrocar-bons) and of a hydrophilic polypeptide part, terminated by an amine group. At deposition on a HOPG surface, it forms an uniform film (with 0.5–0.7 nm thickness and 0.1–0.2 nm roughness) with the hydrophobic part of the polymeric mole-cules adhering to a hydrophobic HOPG surface and the hydro-philic part with the amine oriented towards the water at the water-graphite interface. Thus, the surface gains a positive charge and the ability to adsorb negative biomolecules. In comparison with ssDNA on modified mica, we have obtained very interesting results. ssDNA molecules do not have second-ary structures as in the case of mica deposition. It was possible to observe the individual ssDNA molecules without intra-strand base-pairing (Fig. 3) and the ssDNA molecules are not as extended as in the case of poly-lysine [19] (Fig. 3A and B). It was possible to estimate the height and width of ssDNA molecules. For dsDNA the height was 0.9 ± 0.1 nm and width 12 ± 3 nm; for ssDNA the height was 0.35 ± 0.05 nm and width 7 ± 2 nm, similarly to the results on APTES-mica. The advantage of ‘‘GM’’ modifier is that ssDNA molecules are deposited under standard conditions during adsorption, while in the case of poly-lysine special con-ditions are needed[19]. The contour length of dsDNA linear pBR322 (4361 bp) molecules was also compared for APTES and ‘‘GM’’ adsorption methods [18]. In the case of APTES

Fig. 2. AFM images of: (A) ssDNA M13mp18; (B) heated Lambda dsDNA adsorbed on the surface of mica in the presence of 5 mM Mg2+; (C) ssDNA M13mp18; and (D) heated Lambda dsDNA on mica modiﬁed by APTES. The bars represent 250 nm on all images.

(4)

mica the contour length was 1480 ± 9 nm (average over 50 molecules) and corresponded with the theoretical value of 1483 nm. In the case of ‘‘GM’’ modifier the contour length was 1525 ± 15 nm, which was 3% longer than in the case of APTES mica. Thus, dsDNA was slightly stretched on ‘‘GM’’. In case of ssDNA the contour length could only be measured for ssDNA adsorbed on ‘‘GM’’ modified HOPG. The contour length of ssDNA M13mpl8 on HOPG was 2590 ± 30 nm (average over 25 molecules). It is 5% longer in comparison with theoretical value of dsDNA but 10% shorter in comparison with the contour length of extended molecules of ssDNA M13mpl8 in the case of poly-lysine[19]. It means that this deposition technique for ssDNA on modified HOPG does not put ssDNA under significant stress. The contour length of ssDNA M13mpl8 corresponds to 0.36 nm/base while in case of poly-LL-lysine it was 0.39 nm/base. It correlates well with recent transient electric birefringence measurements that indicate a spacing of 0.32–0.52 nm/base for ssDNA, depending on sequence[20]. InFig. 3C and D we repeated the deposition of heated Lambda dsDNA similarly as it was done forFig. 2B and D. By comparing these two sets of images, one can easily note that in the case of ‘‘GM’’ modified HOPG (Fig. 3C and D), the dsDNA as well as the ssDNA are well visible and in a 2D relaxed state. Also on these images, it is possible to

ob-serve two ssDNA strands emanating from a dsDNA strand as indicated by the circles inFig. 3C and D.

In summary, we have described here a straightforward meth-od for observing ssDNA on HOPG and found that ssDNA can be deposited on the substrate without forming secondary structures. This protocol permits to determine the contour length and molecular weight of isolated ssDNA molecules as well as of ssDNA segments together with dsDNA segments. This method could open the possibility to investigate processes where ssDNA is involved, like DNA replication and transcrip-tion, the mechanism of enzyme action on DNA or DNA hybridization. This protocol is complementary to other meth-ods, like electron microscopy or gel electrophoresis.

Acknowledgment: This study was partially supported by the Swiss National Science Foundation through the Grant 200021-101851.

References

[1] Hansma, H.G. (2001) Surface biology of DNA by atomic force microscopy. Annu. Rev. Phys. Chem. 52, 71–92.

[2] Vesenka, J., Guthold, M., Tang, C.L., Keller, D., Delaine, E. and Bustamante, C. (1992) Substrate preparation for reliable imaging of DNA molecules with the scanning force microscope. Ultrami-croscopy 42-44, 1243–1249.

Fig. 3. AFM images of: (A, B) ssDNA M13mpl8 and (C, D) heated Lambda dsDNA adsorbed on modiﬁed HOPG. The bars represent 250 nm on all images.

(5)

[3] Rivetti, C., Guthold, M. and Bustamante, C. (1996) Scanning force microscopy of DNA deposited onto mica: equilibration versus kinetic trapping studied by statistical polymer chain analysis. J. Mol. Biol. 264, 919–932.

[4] Lyubchenko, Y., Shlyakhtenko, L., Harrington, R., Oden, P. and Lindsay, S. (1993) Atomic force microscopy of long DNA: imaging in air and under water. Proc. Natl. Acad. Sci. USA 90, 2137–2140.

[5] Lyubchenko, Y.L. and Shlyakhtenko, L.S. (1997) Visualization of supercoiled DNA with atomic force microscopy in situ. Proc. Natl. Acad. Sci. USA 94, 496–501.

[6] Valle, F., Favre, M., De Los Rios, P., Rosa, A. and Dietler, G. (2005) Scaling exponents and probability distributions of DNA end-to-end distance. Phys. Rev. Lett. 95, 158105.

[7] Klinov, D.V., Martynkina, L.P., Iurchenko, V.Iu., Demin, V.V., Strel’tsov, S.A., Gerasimov, Iu.A. and Vengerov, Iu.Iu. (2003) Eﬀect of supporting substrates on the structure of DNA and DNA–trivaline complexes studied by atomic force microscopy. Bioorg. Khim. 29, 397–402.

[8] Brett, A.M.O. and Chiorcea, A.M. (2003) Atomic force micros-copy of DNA immobilized onto a highly oriented pyrolytic graphite electrode surface. Langmuir 19, 3830–3839.

[9] Severin, N., Barner, J., Kalachev, A. and Rabe, J.P. (2004) Manipulation and overstretching of genes on solid substrates. NanoLetters 4, 577–579.

[10] Hansma, H.G., Revenko, I., Kim, K. and Laney, D.E. (1996) Atomic force microscopy of long and short double-stranded, single-stranded and triple-stranded nucleic acids. Nucleic Acid Res. 24, 713–720.

[11] Hansma, H.G., Bezanilla, M., Zenhausern, F., Adrian, M. and Sinsheimer, R.L. (1993) Atomic force microscopy of DNA in aqueous solutions. Nucleic Acid Res. 21, 505–512.

[12] Adamcik, J., Viglasky, V., Valle, F., Antalik, M., Podhradsky, D. and Dietler, G. (2002) Eﬀect of bacteria growth temperature on the distribution of supercoiled DNA and its thermal stability. Electrophoresis 23, 3300–3309.

[13] Viglasky, V., Valle, F., Adamcik, J., Joab, I., Podhradsky, D. and Dietler, G. (2003) Anthracycline-dependent heat-induced transi-tion from positive to negative supercoiled DNA. Electrophoresis 24, 1703–1711.

[14] Edwardson, J.M. and Henderson, R.M. (2004) Atomic force microscopy and drug discovery. Drug Discov. Today 9, 64–71. [15] Bustamante, C. and Rivetti, C. (1996) Visualizing protein–nucleic

acid interactions on a large scale with the scanning force microscope. Annu. Rev. Biophys. Biomol. Struct. 25, 395–429. [16] Herbert, A., Schade, M., Lowenhaupt, K., Alfken, J., Schwartz,

T., Shlyakhtenko, L.S., Lyubchenko, Y.L. and Rich, A. (1998) The Zalpha domain from human ADAR1 binds to the Z-DNA conformer of many diﬀerent sequences. Nucleic Acid Res. 26, 3486–3493.

[17] Hansma, H.G., Laney, D.E., Bezanilla, M., Sinsheimer, R.L. and Hansma, P.K. (1995) Applications for atomic force microscopy of DNA. Biophys. J. 68, 1672–1677.

[18] Marek, J., Demjenova, E., Tomori, Z., Janacek, J., Zolotova, I., Valle, F., Favre, M. and Dietler, G. (2005) Interactive measure-ment and characterization of DNA molecules by analysis of AFM images. Cytometry A 63, 87–93.

[19] Woolley, A.T. and Kelly, R.T. (2001) Deposition and character-ization of extended single-stranded DNA molecules on surfaces. NanoLetters 1, 345–348.

[20] Mills, J.B., Vacano, E. and Hagerman, P.J. (1999) Flexibility of single-stranded DNA: use of gapped duplex helices to determine the persistence lengths of poly(dT) and poly(dA). J. Mol. Biol. 285, 245–257.