Rapid DNA fiber technique
for measuring linear and circular DNA molecules/probes
DNA size measurements are an integral part of other laboratory procedures, such as physical mapping, subcloning or sequencing. The large variety of vectors in use today, allows cloning and propagation of DNA inserts of various sizes: 0.1-10 kb in plasmids, 5-25 kb in phages, 30-45kb in cosmids, 80-300 kb in P1 clones, P1 artificial chromosomes (PACs) or bacterial artificial chromosomes (BACs) and up to 2000kb in yeast artificial chromosomes (YACs). Some of these recombinant DNA molecules are circular (plasmids, cosmids, PACs, BACs) whereas others are linear (phages, YACs). With the exception of YACs (propagated in yeast) all other vectors are propagated by transforming bacterial hosts, usually strains of E. coli. Short of direct sequencing, two main approaches are in use today for measuring the size of DNA probes: (1) Restriction enzyme digestion, followed by gel-electrophoresis and sizing of DNA fragments by comparisons with known molecular weight markers (Fig 1B), and (2) DNA fiber procedure(s). In fluorescence in situ hybridization (FISH) on DNA fibers, a DNA template is stretched on a slide , then a labeled DNA probe is hybridized onto it and the physical length of the hybridization signals measured. The hybridization signals appear as an array of closely spaced dots, resembling beads on a string (Fig 1H). Sizing is done by measuring the length (pixels, microns) of the signal array and comparing it with a known standard (micrometer). However, if the stretching technique involves subjecting whole cells to chemical reagents to release the DNA directly on the slide, individual fibers are stretched to different degrees, depending on how much residual protein is still attached to particular DNA strands. Fiber FISH is a valuable tool for mapping DNA probes relative to one another and for measuring the size of small microdeletions. DNA of any size can be resuspended in a solution at appropriate concentration and pH, and stretched on chemically pretreated, positively charged slides using one of the procedures collectively known as molecular combing. The DNA is stained with a green fluorescent dye (YOYO®-1, Molecular Probes) and visualized using a fluorescent microscope (Fig. 1). Various DNA fiber techniques were previously published, but the stretched DNA was usually subjected to hybridizations or was enzymatically cut to yield restriction maps (optical mapping). Here, we describe a molecular combing protocol, which allows DNA molecules to be stretched on silanized glass slides in seconds, and can be used with DNA ranging from less than 10kb to 300 kb or longer in size. To our knowledge, this protocol appears to be the fastest DNA stretching procedure described to date. We show that the DNA can be stretched with various degrees on a chemically treated glass surface, especially when the DNA molecule is circular. The procedure requires only minute amounts of DNA (10-20ng) and can be performed in about fifteen minutes, in any laboratory with access to a fluorescence microscope. Visual inspection at the microscope is very simple, as the DNA fibers are bright and very visible, especially with a 63x or a 100x objective. DNA fiber measurements (Fig 1D-F) were compared with results from restriction digest gel electrophoresis (Fig 1B) and direct sequencing (Fig 1K) information. For example, BamHI-digested DNA of three different, but partially overlapping PACs, was separated on an agarose gel and was stained with ethidium bromide (Fig 1B). With that enzyme, PAC#1 was 137kb, PAC#2 was 139 kb and PAC#3 was 91kb. Several other single or double restriction digests (XhoI, SfiI, EcoRI, HindIII, NotI) of the same PACs yielded length values which could vary by as much as 10-15kb (about 10%) for the same probe. The same DNA was then stretched and measured on glass slides. Based on these and other similar comparisons, we derived simple computations, using correction factors that allow the conversion of size measurements in DNA length from microns to kilobases. DNA probes are prepared according to standard laboratory protocols (usually alkaline lysis). Vortexing should be avoided as it will shear larger DNA molecules. A small aliquot of isolated DNA is diluted to 2-5ng/ul in 10mM aminomethyl propenediol (AMP, Sigma) buffer, pH 8.2-8.5. After gentle mixing, 7-8 µl of DNA solution is pipetted close to the free (not frosted) end of a silanized, positively charged slide (Sigma). One end of a 24x40 mm coverslip is positioned so that it touches the edge (free end) of the slide. The angle between the coverslip and slide will be gradually decreased, by tilting the coverslip progressively, until it comes in contact with the drop of DNA solution. At that moment, the drop will spread laterally on the slide, along the coverslip edge. Then, holding the touching edges of the slide and coverslip between the thumb and index finger, the coverslip is gradually pressed downward, toward the slide, until it covers the slide. This move will spread the liquid between the slide and coverslip, from the free end towards the frosted end of the slide. The liquid flow between the slide and coverslip is the force that stretches the DNA. Continuous pressure on the slide/coverslip assembly, between the two fingers, should last several seconds, until the liquid spreads completely. Pressure should be further applied for another 10-15 seconds, to make sure that all excess liquid is squeezed out from between the slide and coverslip. The coverslip is lifted with a sharp forceps and removed, and the slide is air-dried. The DNA can be stained in several ways: a) 10-7 YOYO®-1 dye (Molecular Probes) can be added to the DNA solution prior to stretching. After DNA stretching, the slide is air-dried, mounted with antifade solution and examined. b) after stretching, 50-100 µl of a 10-7 M YOYO®-1/20 mM beta-mercaptoethanol solution can be placed on the slide, covered with a coverslip, incubated for 5 minutes at room temperature to stain the DNA, then visualized at the microscope. Alternatively, the coverslip can be removed, the slide rinsed in water for a few seconds, air-dried and mounted with antifade. c) antifade solution containing 10-7 M YOYO®-1 can be used to mount the slide prior to microscopy. After 3-5 minutes, slides are examined at the fluorescence microscope using a common FITC filter (for YOYO). Although we use mostly the first procedure, all three staining protocols yield the same results, and it is up to the researcher to choose which one to use. DNA fiber size is measured using any software that calculates the number of pixels between two points of a digital image (Segmented Ruler®, Adobe® Photoshop®). The pixel-to-kilobase conversion is obtained by imaging a microscopic ruler using the same objective/magnification (Fig 1G). In our case, 10 µm = 73.5 pixels, when using the 100x objective. At this magnification, the theoretical size of a 100 kb (34 µm) long DNA molecule should be 250 pixels, if the DNA molecule is stretched to its full length. In our experiments, we measured DNA molecules between 8kb (Fig 1J) and 220 kb (Fig 1K). It is conceivable that the procedure would be capable of handling larger DNA molecules, around 300kb or even longer. To find the optimal relationship between the degree of DNA stretching and the actual size in kb, we performed numerous measurements using DNA probes of known sizes (by sequencing or restriction digest). After stretching, all circular DNA probes tested yielded DNA fibers with a variety of geometrical shapes (Fig 1C). Depending on the number of kinks in a molecule, the absolute size can vary from one measured molecule to another. These kinks are produced by the direction of the liquid flow during stretching and the number of attachment points of the DNA molecule to the glass. For the purposes of this work, four classes of shapes are more important: linear fibers, closed circles, collapsed circles and knotted molecules. Molecular distribution among these classes depends primarily on the age of the DNA preparation is and the care with which long DNA molecules are handled. In a fresh DNA preparation, the majority of molecules (60-70%) stretch on the slide like circles or knots. In an older preparation, one that was subjected to repeated centrifugations or cycles of freezing and thawing, large molecules such as PACs and BACs break. Such DNA will stretch like linear fibers of multiple sizes with very few circles. 1) Linear DNA fiber. They vary greatly in size for the same DNA probe, because the stretching force of the liquid flow often breaks the DNA molecules in fragments of various sizes. To assess the degree of stretching of a linear molecule, we used as control a DNA of known size (lambda phage DNA). Theoretically, this 48.5 kb molecule should be 16.5 µm long, and should span about 122 pixels. However, when lambda DNA was stretched onto silanized slides, fibers of various sizes were visible after YOYO®-1 staining. This suggests that the DNA breaks, either in solution or during stretching or both. To calculate the average size of a full length lambda molecule (48.5 kb), we measured all reasonably long fibers (132-161 pixels) in five microscopic fields, in different areas of the slide (Fig 2). Because of random breakages in the DNA molecules, only the longest of these fibers were considered when calculating the average size of a molecule (in this case, only fibers between 156-161 pixels). In Figure 2, the average size of the 15 seemingly intact DNA molecules was 159 pixels (see also Fig 1A). The slight differences may be due to the way the DNA attaches to the slide. Sometimes, one or both ends of a molecule show an increase in thickness, similar to the aspect when the molecule breaks. This suggests that the "thicker" areas close to a free end in the DNA molecule may not be fully stretched. This data indicates that, at least in our conditions, linear DNA molecules are over-stretched at 130% of the theoretical DNA size. Similar value was obtained with molecules as long as 190 kb and 220 kb. Therefore, to correct for over-stretching, the measured size (pixels or microns) of a linear fiber needs to be multiplied by 0.77 to convert it to actual length. Because of this factor, the real correspondence between physical measurement and size of a linear fiber is: 100 kb = 44.2 µm = 325 pixels. Therefore, the measured length (pixels) of a linear fiber should be divided by 3.25 (= 130/100 x2.5) to convert the DNA size in kb. Especially for large probes (>180-200 kb), full-length linear fibers may be more difficult to use for calculations, as molecules break easily. Restriction digest of a circular DNA probe prior to stretching would linearize all molecules, but would eliminate the various geometrical forms described below, which can provide valuable size information. For linear molecules, the longest fibers found on the slide should be measured. 2) Closed circles and ovals. A circular DNA is rarely stretched smoothly, instead the fiber makes numerous, irregular little turns. Measurements of DNA probes of known size show that circular DNA is stretched closest to or at it's theoretical size. (100 kb = 34 nm = 250 pixels). Therefore, in order to convert the size of a circle (in pixels) into actual DNA size (kb), the pixel number of a circular fiber is divided by 2.5 (= 250/100). 3) Collapsed circles. In this case, the DNA fiber is short and looks thicker compared to other circular or linear fibers on the slide. This increased thickness is due to the twisting, to some degree, of the "sides" of the circular molecule around each other. The collapsed circles are very convenient to use for sizing DNA probes, as they are easier to measure, especially in the case of large DNA molecules (200-300 kb). Our measurements indicated that DNA in collapsed circles is somewhat over-stretched, to about 104% of the theoretical values. In order to convert the measured size (in pixels) of a collapsed circle into theoretical size (in pixels), the measured value is multiplied by 1.925 (=2.5x0.77). The 2.5 value is the factor required to convert the size of the collapsed circle to the measured size of the linear DNA fiber, if the same molecule were stretched completely on the slide. To convert the size of a collapsed circle from pixels directly into kb, the measured value is multiplied by 0.77. 4) Knotted DNA molecules. Whether originating from circular or linear molecules, they appear like small knots of DNA (Fig 1C, 1E, 1I) and are not usable for sizing. Fiber DNA measurements were performed using two different microscope settings (Leica and Olympus), in two different laboratories, with identical results. Our data indicate that this fast and simple procedure allows DNA size measurements with an accuracy similar to or better than gel electrophoresis (5-10%). Questions,
comments, ideas?
Send an e-mail to:
Octavian
Henegariu
|