Combining combing and secondary ion mass spectrometry to study DNA on chips using ¹³C and ¹⁵N labeling

Preparation of DNA

In principle, DNA could be extracted from any organism to perform CIS. DNA generated in vitro can also be used. For the experiment reported here, for example, we extracted DNA from the 168 (trpC2) strain of Bacillus subtilis and a derivative (MT119, trpC2 leuB6 r- m-) harboring a 21.1 kb long, chloramphenicol-resistant plasmid (pHV1431 plus insert of 10.3 kb)³. On modified silicon surfaces at concentrations of 0.2 µg/mL, most of the DNA is combed as single fibers. Higher concentrations can lead to fibers running together whilst lower concentrations result in an insufficient number of fibers on the images. Combing works well with buffers containing NaCl at ionic strengths of around 0.2M.

Bacterial culture medium: 14g/L of K₂HPO₄, 6g/L of KH₂PO₄, 1g/L of sodium citrate, 10mM MgSO₄, 0.01% (w/v) of tryptophan, 0.005% (w/v) of leucine, 0.2%(w/v) of ¹³C-glucose (Isotec) and 0.8g/L of ¹⁵NH₄Cl (Isotec). Lysis buffer: 50mM of Tris-HCl (pH8.0), 10mM of EDTA (pH8.0), 150mM of NaCl and 5mg/ml of lysozyme. Proteinase K: Final concentration of 0.2mg/mL (Roche). Sarcosyl Buffer: Final concentration of 1.2%. Phenol/chloroform: Solution PCI: phenol-chloroform-isoamylalcohol (25:24:1 v/v) (the phenol is first saturated in NaCl 150mM and buffered at pH of about 7); Solution CI: chloroform-isoamylalcohol (24:1 v/v). RNase: Final concentration of 20µg/mL (Roche).

Other materials required are: 100% cold ethanol; 70% (v/v) ethanol; NaCl solution: 0.2M to 1M; PureYield^TM plasmid midiprep system (Promega); restriction enzyme; 0.3M Potassium acetate pH7.0; TaKaRa EX Taq^TM system (TaKaRa); Expand Long Template PCR System (Roche Applied Science); QIAquick PCR Purification kit (QIAGEN).

DNA preparation from bacterial cultures. DNA can be prepared from any cells using a procedure that removes contaminating peptide and RNA molecules. In the case of the labeling and extraction of B. subtilis DNA, B. subtilis cells were cultivated at 30–37°C for ~20 generations in bacterial culture medium containing stable isotopes (here, ¹³C and ¹⁵N, see above) to saturation. Then, to prepare chromosomal DNA free of peptides and RNA, one milliliter of a freshly saturated culture was centrifuged at 15000 g for 2 min at 4°C. The supernatant was discarded. Pelleted cells were resuspended in 0.5 mL of lysis buffer. The cell resuspension was incubated for 20 minutes at 37°C. Proteinase K was added to a final concentration of 0.2 mg/mL in addition to 20µL of sarcosyl buffer 30% (1.2%, final concentration) and incubated for 20 min at 65°C. To remove peptides and cell fragments by a phenol/chloroform treatment, we lowered the temperature of the sample on ice for 3 min and then added 500µL of solution PCI (see Preparation of DNA) at 4°C and vortexed strongly for 30 s. The mixture was centrifuged at 15000 g for 15 min at room temperature in a benchtop centrifuge. The aqueous solution was recovered and 500 µL of solution CI was added to it (see Preparation of DNA). This mixture was vortexed strongly and centrifuged as above. RNase was added to a final concentration of 20 µg/mL to the aqueous, nucleic acid-containing phase and incubated for 10 min at 37°C. DNA was purified by a second phenol/chloroform extraction as above. We added 2.2 volume of 100% -20°C ethanol to the aqueous solution and centrifuged at 15000 g for 20 min at 4°C in a benchtop centrifuge. The supernatant was discarded and the DNA pellet was washed in 70% cold ethanol. This mixture was centrifuged again for 10 min at 4°C at 15000 g and the supernatant was discarded. The DNA pellet was dried for 10 min under vacuum and resuspended carefully in pure water for at least 12 h at room temperature (note that incomplete solubilization of ethanol-precipitated DNA severely perturbs combing). DNA concentration was measured by absorbance at 260/280 nm using a Nanodrop 2000 (Thermo Scientific). Depending on the fragment length, DNA concentration should be adjusted to within the range 0.2 to 2 µg/mL by dilution in NaCl solutions that may range up to to 1M; here, we used DNA at 0.2µg/mL in 0.2M NaCl.

In this paper, we only report CIS in the case of chromosomal DNA. CIS can also be used to analyse plasmid DNA (not shown). In the case, for example, of a 20 kb plasmid from B. subtilis, plasmid DNA may be extracted using the PureYieldTM plasmid midiprep system (Promega Corporation, Madison, WI, USA), followed by linearization of the plasmid by a single cutter restriction enzyme according to suppliers’ instructions (e.g. New England BioLabs, Inc., Hitchin, UK) to give fragments of the appropriate size; this linearization is needed before combing because of the difficulty of combing supercoiled circular molecules. After linearization, the DNA can be purified by the phenol/chloroform procedure and recovered by ethanol precipitation in the presence of 0.3 M potassium acetate pH 7.

DNA preparation by PCR (protocol used to prepare PCR-generated DNA fragments 1-20 kb long). DNA fragments generated by PCR can be analyzed by CIS³. We routinely produce fragments <5 kb using the TaKaRa Ex Taq^TM system (Takara Shuzo Co., Ltd, Shiga, Japan). For products of 5-20 kb, we use the Expand Long Template PCR System from Roche Applied Science (Mannheim, Germany). The PCR products can be purified with the QIAquick PCR Purification kit (QIAGEN GmbH, Hilden, Germany). PCR products can then be diluted for CIS as described above.

Preparation of silicon surfaces

This experiment required an efficient fume hood. The materials used were: P-type 100 crystalline silicon wafers (Siltronix) with a resistivity > 1 Ω cm (typically 10 mm × 10 mm × 380 μm); Piranha solution obtained by mixing 3 volumes of 96% of sulfuric acid [VLSI (very large-scale integration)-grade] and 1 volume of 30% hydrogen peroxide (VLSI-grade), to be used immediately; 50% hydrofluoric acid (VLSI-grade); acetone; isopropyl alcohol (iPrOH); chloroform (CHCl₃); ethanol (EtOH); 1-tetradecene; a photochemical reactor for UV irradiation (λ=312 nm) (our home-made reactor has 8 tubes in a circle around a Schlenk tube with a fan at the bottom to limit any increase in temperature); a goniometer system (DIGIDROP, GBX, France).

Preparation of hydrogenated-terminated silicon surfaces. In the following preparation, it should be noted that the piranha solution is a powerful oxidant that reacts violently with organic materials; it can cause serious skin burns and must be handled with great care in a well-ventilated fume hood while wearing appropriate chemical safety protection. Moreover, HF is a hazardous acid that can cause severe damage to tissues if burns are not treated properly. Etching of silicon should therefore be carried out in a fume hood with the right safety measures, which include a face shield and double-layered nitrile gloves.

The silicon surface was cleaned in an ultrasonic bath for 5 min periods in acetone and isopropylic alcohol. It was then rinsed extensively with ultrapure water and immersed for 20 min in a piranha solution to remove organic contaminants on the surface. This was followed by the immersion of the clean surface in an aqueous solution of HF (50%, as provided by the supplier) to generate a hydrogen-terminated surface (Si-H)⁹. This surface was washed extensively with water and then dried by blowing with nitrogen. The resulting surface typically had a contact angle of 85° for a 1 µL water droplet. Note that freshly prepared Si-H surfaces were used immediately for DNA combing or for grafting hydrocarbon chains (see below). Before combing, coated silicon surfaces were protected from dust and dried because a thin film of water greatly reduces DNA adsorption.

Formation of organic monolayers on hydrogen-terminated silicon surfaces. To obtain highly hydrophobic surfaces, alkenes with C₁₄ alkyl chains were grafted onto freshly prepared Si-H surfaces using the hydrosilylation reaction of 1-tetradecene with the hydrogen-terminated silicon surface⁹. The freshly hydrogen-terminated silicon surface (see above) was immersed in a Schlenk tube containing 10 ml of previously deoxygenated, neat 1-tetradecene under nitrogen bubbling. This was then irradiated at 312 nm in a photochemical reactor for 3 h. We removed unreacted and physisorbed 1-tetradecene by rinsing with chloroform and ethanol at room temperature. The silicon substrate was dried under a stream of nitrogen. We verified that water contact angles (using deionized water in the ambient atmosphere at room temperature) are around 104° as measured to an accuracy of ± 2° with a remote, computer-controlled, goniometer system (N.B., protect the surface from dust, see above).

Combing on silicon. The 'drop' and 'lift' methods^10,11 can both be used to comb DNA on the Si-C₁₄ surface. Both 'drop' and 'lift' methods entail the DNA becoming attached to the surface and then drawn out and aligned perpendicular to the triple line or the meniscus, respectively. The ‘drop’ method entails depositing 10 µL of prepared DNA solution on the silicon surface, incubating it for 10 min on the bench, tilting the surface with tweezers to 45° to cause the drop to roll off the surface, and washing the surface by immersion in water and rapidly air-drying; the quality of combing via this method can be adversely affected by vibrations occurring during the movement of the drop. Here, we used the ‘lift’ method, which entails immersing a part of the silicon surface in a DNA solution (see above); after 5 min incubation, the silicon surface was pulled out of the solution at a constant speed (600 µm/min) and rapidly air-dried. In both ‘drop’ and ‘lift’ methods, washing the surfaces after combing helps avoid perturbation of D-SIMS analysis by saline crystals. The tightly controlled, motor-driven ‘lift’ method gives more reproducible, better quality results but takes longer and requires more DNA than the manual ‘drop’ method. The ‘lift’ method is better than the drop method for fragments less than 5 kb; the ‘drop’ method is satisfactory with fragments equal to or greater than 5 kb.

D-SIMS methods

Cs flooding method. To avoid a premature destruction of thin samples before SIMS detection¹², a cesium flooding system is recommended to deposit neutral cesium on samples^13,14. We used a UHV Cs evaporation system which is available for purchase from the Luxembourg Institute of Science and Technology (LIST) . It comprises a neutral cesium evaporator and an independent stand-alone UHV chamber. A suitcase under UHV (10^-8 to 10^-10 Torr) was used to transfer samples to the NanoSIMS50 thereby avoiding an immediate reaction between the neutral cesium deposit and air. To analyze combed DNA (as performed here), we flooded the surface with Cs⁰ at 1.5 Å/s for 1800 s before SIMS imaging. An alternative to cesium flooding is to coat samples with Au; a Cressington Sputter Coater can be used to deposit a layer of Au, approximately 60 nm thick, on wafers with combed DNA but the quality of the images subsequently obtained is often poor.

NanoSIMS 50 analyses. In SIMS analyses of samples labeled with carbon and nitrogen, there are several factors that should be borne in mind. First, carbon can be detected as C^- and in multi-clusters that include CN^- and C₂^- (N.B. CN is easier to ionize and therefore detect than C and C₂ because the electron affinity is 3.8eV for CN^-, 1.26eV for C^- and 3.27eV for C₂^-). Second, nitrogen can be detected in the multi-cluster CN (note that nitrogen forms fewer types of multi-clusters thant carbon even though it can form to a very limited degree multi-clusters that include NO, N₂H, and NS). Third, the probability of formation of the CN multi-cluster in the mixing-recombination under the primary beam is related to the distance between the carbon and nitrogen atoms¹⁵; this means that differentially labeled macromolecules can be colocalized if they are within 2 nm of one another; it also means that the carbon and nitrogen in the DNA are more likely to recombine with one another than with carbon and nitrogen that are further away. Note that even a trace amount of contaminants is an important problem for the highly sensitive CIS method and strenuous efforts should therefore be taken to protect samples from contamination by, for example, an atmosphere containing C and N. Silicon surfaces with combed DNA were analyzed in the multi-collection image mode of the NanoSIMS 50 (Cameca, Gennevilliers, France). The NanoSIMS 50 was used in the negative secondary-ion mode with the Cs⁺ primary ion beam. In this case high spatial resolution images were obtained using a primary beam around 0.5–1.0 pA in intensity and an impact energy of 16 keV. The surface was rastered by the cesium beam on a surface from 5 × 5 to 15 × 15 µm². The instrument was tuned to limit the dispersion in aperture and energy in order to obtain a minimal mass resolution of M/ΔM = 5000 (10% height peak measurement).

Aphelion 3.2 was used to analyze the results pixel by pixel or line by line, whilst ImageJ plus the OPEN MIMS plug-in (which opens the .im format¹⁶ and is available from http://www.nrims.harvard.edu) was used to analyze the data in an ROI (Region Of Interest). The data were summed and the images colored using WinImage 2. The images of the ¹²C (m = 12.000000), and/or ¹³C (m = 13.00335484), ¹²C¹⁴N (m = 26.003074), ¹²C¹⁵N (m = 27.00010898) or ¹³C¹⁴N (m = 27.00642885), ¹³C¹⁵N (m = 28.00346382) were acquired simultaneously in 256 × 256 pixels with a dwell time of 10 ms per pixel. Note that ¹²C¹⁵N and ¹³C¹⁴N cannot be detected simultaneously in the same ‘sputter section’. The sulfur ³²S (m = 31.9720718) distribution can be also acquired to control the quality of the preparation.