Oligonucleotide 11 carrying one single amino group at the apex of the topological loops was prepared as shown in Scheme 1. Amino groups were selected due to their special reactivity as nucleophiles. This reactivity may be used for further functionalization with compounds carrying carboxyl groups. A phosphoramidite derivative of 2′-deoxyuracil carrying an amino alkyl group at position 5 of the uracil nucleobase was used to introduce the amino group at one specific site of oligonucleotide B1 (Table 1). The site of modification was one of thymidines forming the unpaired loop at the apex of one of the topological loops of tile B*. First a 20-base oligonucleotide containing 5 amino-dT residues was prepared and the stability of this analogue in oligonucleotide synthesis conditions was confirmed by mass spectrometry analysis (Supplementary data). Then, oligonucleotide sequence 11 was synthesized and purified by denaturing gel electrophoresis. The purified oligonucleotide was hybridized with the rest of the oligonucleotides forming tile B* (B2–B4). DNA tile B* carrying one single amino group at the unpaired loop position was able to form correct B* DNA tile as observed by native gel electrophoresis analysis.

Next, oligonucleotide sequence 11 was hybridized with the rest of the oligonucleotides A1-A5 (1– 5) and B2-B4 (7–10) yielding the desired DNA array that was then deposited on mica substrates. As can be seen from the AFM image in Figure 2a the DNA lattice carrying an amino group at the topological loop was observed on mica. The spacing of the decorated columns was 32 ± 1 nm, which was the same value observed in the bibliography for unmodified DNA lattices [6].

Recently, the chemical functionalization of DNA arrays with thiols has been demonstrated [16]. The introduction of the thiol groups in the B1 and B5 DNA oligonucleotides was done using the strategy described above for the introduction of amino groups (Scheme 1). We used cytosine phosphoramidite derivatives for the introduction of the thiol groups protected with the t-butylthio group (Scheme 2). The use of the protected form of the thiol groups avoids the formation of undesired disulfide bonds. We modified tile B* by introducing one N⁴-[(t-butyldithio)ethyl]-dC or one N⁴-[(t-butyldithio)ethyl]-5- methyl-dC residue to replace one thymidine residue at the unpaired loop positions of the oligodeoxynucleotides (Scheme 2). This involved preparing the two longer oligonucleotides of tile B* (69, 70 bases), which carry one single one N⁴-[(t-butyldithio)ethyl]-dC or one t-butyldithio-ethyl-5- methyl-dC residue in the middle of the sequence. The appropriate phosphoramidites were synthesized as described [17,18]. Oligonucleotide sequences carrying thiol groups protected with the t-butylthio group (12, 13, Table 1) were purified by denaturing polyacrylamide gel electrophoresis (PAGE). The formation of the correct thiol-modified DNA arrays on mica have already been described [16].

Next, we prepared oligonucleotides carrying fluorescein, c-myc peptide and nanogold at the apex of the topological loops (Scheme 2). Fluorescein and the c-myc peptide sequence were selected because they are model antigenic compounds that can be recognized by monoclonal antibodies. Nanogold was selected as model gold nanoparticle. Maleimido derivatives of fluorescein and nanogold are commercially available. Maleimido- c-myc peptide was synthesized as described [17,18].

The synthesis of oligonucleotide conjugates carrying fluorescein (14) or c-myc peptide (15) started with the synthesis of thiolated oligonucleotide B5 (13, Table 1). This oligonucleotide was treated with tris(carboxyethyl)phosphine to remove the t-butylthio protecting group. The corresponding free thiol oligonucleotide was reacted with the corresponding maleimido derivatives as described [17–20] (Scheme 2) to yield the desired conjugates (14–15). Fluorescein and peptide conjugates 14 and 15 were purified by gel electrophoresis (supplementary information). As oligonucleotides 14 and 15 are too long for mass spectrometry analysis, we set up the conjugation reactions with a shorter DNA sequence and the corresponding fluorescein and c-myc conjugates were obtained in good yields. The desired conjugates were characterized by mass spectrometry (supplementary information).

The preparation of the nanogold conjugate 16 was slightly different. First thiolated oligonucleotide B1 (12, Table 1) was synthesized and purified by gel electrophoresis. The purified oligonucleotide 12 was treated with 0.1 M dithiothreitol (DTT) overnight at 50 °C to remove the t-butylthio protecting group and the resulting thiolated oligonucleotide was incubated with maleimido-Nanogold. The resulting nanogold conjugate 16 was used directly from the conjugation reaction without further purification. The functionalization of the Nanogold with the desired oligonucleotide was confirmed by agarose gel electrophoresis [21].

DNA tiles B* formed with these conjugates were able to form correct tiles as observed by gel electrophoresis analysis of hybridized equimolar mixtures of the corresponding B1-B5 oligonucleotides (supplementary information). Figure 2b shows the DNA lattices carrying fluorescein formed on mica. Figure 3 shows the arrays formed using nanogold-oligonucleotide conjugate 16. Nanogold was not observed as the size (around 1 nm) is similar to the size of double-stranded DNA. The observation of DNA arrays using peptide-oligonucleotide conjugate 15 on mica did not yield positive results in spite of the good formation of B* tile observed by native gel electrophoresis. At present we do not know if the presence of the peptide sequence prevents the deposition of the array on mica or prevents the formation of large DNA arrays.

Oligodeoxynucleotide sequences are shown in Table 1. The syntheses were performed on the Applied Biosystems Model 3400 DNA synthesizer using a scale of 0.2 μmol and standard 2-cyanoethyl phosphoramidites as monomers. The N⁴-[(t-butyldithio)ethyl]-2′-deoxycytidine and N⁴- [(t-butyldithio)ethyl]-5-methyl-2′-deoxycytidine phosphoramidites were prepared as described elsewhere [17,18]. After the addition of the N⁴-[(t-butyldithio)ethyl]-5-methyl-2′-deoxycytidine phosphoramidite, the oxidation solution used was a solution of tert-butyl hydroperoxide 10% in acetonitrile instead of the commercially available solution of iodine 0.02 M to avoid the oxidation of the thiol group described in reference [18]. The 5-amino-T-phosphoramidite was obtained from commercial sources. After the assembly of sequences, ammonia deprotection was performed overnight at 55 °C. Oligonucleotides were purified by polyacrylamide gel electrophoresis (PAGE) (see below). Purification of oligonucleotides by HPLC using DMT-on protocols yielded oligonucleotides that were not pure enough to generate complete DNA arrays (Supplementary section).

Oligonucleotide sequence 13 was used to perform the reactions with fluorescein diacetate 5-maleimide, and maleimido c-myc peptide prepared as described [17]. For the preparation of each conjugate 50 OD₂₆₀ units of crude oligonucleotide 13 were used. To cleave the disulfide bond, oligonucleotide 13 was dissolved in 1 mL of 0.1M triethylammonium acetate solution (pH = 7). Afterwards, 40 μL of a 0.5 M tris(2-carboxyethyl)phosphine hydrochloride (TCEP. HCl) solution were added to the solution and allow to react at 55 °C overnight. Under these conditions, the tert-butylthiol group was completely removed [18,19]. The resulting product was purified with Sephadex G-25 (NAP-10 column). The oligonucleotide carrying the free thiol group was eluted with 1.5 mL of sterile water.

Conjugation with fluorescein: To the resulting solution, 150 μL of a 1M triethylammonium acetate solution were added. Then, 100 μL of a solution 33 mM fluorescein diacetate 5-maleimide in DMF were added and allowed to react at room temperature overnight. The crude oligonucleotide was concentrated to dryness and then dissolved in 400 μL of a 0.2 M sodium hydrogen carbonate solution (pH = 9). The solution was heated at 55 °C for 1 h. Then 600 μL of sterile water were added and the excess of reagents and salts were removed by using a NAP-10 column. The oligonucleotide was eluted with 1.5 mL of sterile water and was purified by polyacrylamide gel electrophoresis (PAGE) (see below).

Conjugation with c-myc-peptide: To the resulting solution, 200 μL of a 1M triethylammonium acetate solution were added. Then, 500 μL of a solution 24 mM of the maleimido peptide [17,18,20] in sterile water were added. The pH was adjusted to 6 and the mixture was allowed to react at room temperature overnight. The crude oligonucleotide was concentrated to dryness and the excess of reagents and salts were removed by using a NAP-10 column. The oligonucleotide was eluted with 1.5 mL of sterile water and purified by polyacrylamide gel electrophoresis (PAGE) (see below).

Conjugation with maleimido-nanogold: Oligonucleotide sequence 12 was used to perform the conjugation with maleimido-Nanogold. First, thiolated oligonucleotide 12 was purified by polyacrylamide gel electrophoresis (PAGE) (see below). Then, the purified oligonucleotide was treated with 0.1 M dithiothreitol (DTT) overnight at 50 °C to remove the t-butylthio protecting group. The excess of DTT was removed by filtration over a NAP-10 column (Sephadex G-25). The desired thiol-oligonucleotide was eluted in 1 mL solution that was concentrated to 0.3 mL. The resulting thiolated oligonucleotide was incubated with commercially available maleimido-Nanogold. The resulting nanogold conjugate 16 was used directly from the conjugation reaction without further purification. The functionalization of the Nanogold with the desired oligonucleotide was confirmed by the presence of a brown spot in 3% agarose gel electrophoresis as described in reference [21].

Oligodeoxynucleotides were purified using denaturing gel electrophoresis. The gels contained 20% acrylamide (19:1, acrylamide/bisacrylamide) and 8.3 M urea, and were run at 55 °C on a Hoefer SE 600 electrophoresis unit. The running buffer comprised 89 mM Tris base, 89 mM boric acid, and 2 mM EDTA at pH 8.0. The sample buffer contained 10 mM NaOH, 1 mM EDTA, and a trace amount of Xylene Cyanol FF tracking dye. Gels were stained with ethidium bromide and the target band was excised and eluted in a solution containing 500 mM ammonium acetate, 10 mM magnesium acetate, and 1 mM EDTA. The eluates were extracted with n-butanol, which removes the ethidium bromide, followed by ethanol precipitation.

Complexes were formed by mixing a stoichiometric quantity of each strand, which was estimated by measuring the optical density at 260 nm. All 10 strands (A1-A5 and B1-B5) were mixed in 10 mM HEPES (pH 7.8), 12 mM MgCl₂, and 2 mM EDTA. The final concentration of DNA was 0.2–0.4 μM. The final volume was 50 μL. Mixtures were annealed from 90 °C to room temperature for 40 h in a 2 litre water bath insulated in a styrofoam box. Previous to the array formation the correct assembly of tile A and modified tile B* was checked by nature PAGE.

A sample of between 2–7 μL was spotted on freshly cleaved mica (Ted Pella, Inc.). The arrays were imaged in tapping mode in a buffer. The sample was deposited for 1–3 min and an additional 35 μL of fresh buffer was added to the liquid cell. All AFM imaging was performed on a NanoScope III (Digital Instruments) or Asylum MFP3D using commercial cantilevers with Si₃N₄ tips.