DETECTION OF DNA FRAGMENTATION

DNA from most mammalian cells undergoing apoptosis displays a characteristic series of bands (a so-called nucleosomal ladder, Figure 3.1C) after agarose gel electrophoresis.¹³ This fragmentation pattern results from preferential cleavage of DNA in the linker regions between nucleosomes compared with the DNA that is wrapped around histone octamers.¹⁴ Accordingly, this cleavage pattern is not specific for any particular endonuclease but, instead, reflects the structure of chromatin.

Because any nuclease that introduces DNA double-strand breaks will randomly yield this pattern when intact chromatin is digested,^15-17 there has been considerable debate about the nature and identity of the nuclease responsible for the apoptotic DNA fragmentation. Early indirect evidence implicated a variety of endonucleases, including DNase I,^18,19 DNase II,^20,21 and DNase γ.^22,23 Subsequent studies demonstrated that caspase-activated deoxyribonuclease (CAD) plays a critical, nonredundant role in apoptotic internucleosomal cleavage in a number of different cell types.^24-27 According to current understanding, this endonuclease is complexed with inhibitor of CAD (ICAD), which acts as both a chaperone (to permit folding of the nascent CAD polypeptide) and an inhibitor.²⁴ The constitutively expressed CAD/ICAD complex resides in nuclei,²⁸ where cleavage of ICAD by caspase-3, a protease activated during apoptosis (see below), frees CAD^29,30 and allows it to begin digesting chromatin.

Even though the discovery of CAD and ICAD provides a plausible link between other apoptotic events and endonuclease activation, other nucleases might also contribute to apoptotic internucleo- somal DNA degradation in certain cell types. Targeted disruption of the ICAD²⁶ or CAD²⁷ genes abolishes apoptosis-associated internucleosomal cleavage in some cell types^26,27 but not others.³¹ In the latter cells, endonuclease G released from mitochondria during apoptosis has been implicated in internucleosomal cleavage.³¹

Regardless of which endonuclease performs the cleavage, the demonstration that nucleases are activated during apoptosis has led to the development of a variety of techniques for detecting apoptotic cells.

Conventional Agarose Gel Electrophoresis

The original description of internucleosomal DNA fragmentation during apoptosis relied on agarose gel electrophoresis.¹³ For this technique, DNA was prepared from cells or tissues using any of a number of protocols. One commonly used approach involved solubilization of cellular contents in sodium dodecyl sulfate (SDS), digestion of polypeptides with protease K, and extraction with phenol to remove peptide fragments. Application of the resulting total cellular DNA to an agarose gel with suitable separation properties (e.g., 1 to 2% [w/v] agarose) readily demonstrated a ladder of oligonucleosomal fragments.

The advantages of detecting DNA fragmentation by this approach are its relatively low cost and simplicity. There are, however, several disadvantages as well. First, the technique is relatively insensitive. Because the nucleosomal fragments are spread throughout the length of the gel, a substantial fraction of the total DNA must be cleaved before fragments are detectable using ethidium bromide. Second, the technique remains inherently qualitative rather than quantitative. Although agarose gel electrophoresis distinguishes between internucleosomal DNA degradation and the ran­dom DNA fragmentation that accompanies necrosis,¹¹ it is difficult to determine whether 10% or 25% of the total DNA in a particular lane is present in the nucleosomal ladder. Merely photographing agarose gels and digitizing the images does not yield valid quantification unless appropriate controls to assure linearity are performed. Finally, even if the amount of cleaved DNA is precisely quantitated, the interpretation of the result is open to question. Conversion of 25% of the total DNA to nucleo- somal fragments might indicate cleavage of a quarter of the DNA in 100% of the cells or cleavage of 100% of the DNA in a quarter of the cells. For these reasons, assays that assess apoptosis on a cell-by-cell basis (see the sections entitled “Flow Cytometry,” “Terminal Deoxyribonucleotidyl Transferase-Mediated dUTP Nick-End Labeling (TUNEL),” and “Comet Assay”) are preferred.

Extraction of Small Fragments

Alternative methods of quantitating DNA fragmentation have also been developed. The first of these is based on the observation that double-stranded DNA fragments below 10 to 20 kilobases are extracted from nuclei in buffers containing magnesium chelators.³²’³³ To extract these fragments, cells undergoing apoptosis are lysed in buffer consisting of 20 mM Tris (pH 7 to 8), ethylenedi­aminetetraacetic acid (EDTA), and a neutral detergent. After a 10- to 30-minute incubation at 4°C, the intact chromatin is pelleted, which leaves the nucleosomal fragments in the supernatant. DNA in the two fractions can then be assayed using colorimetric or fluorimetric techniques. Alternatively, if DNA in the cells is uniformly radiolabeled (i.e., greater than one cell cycle) before application of the apoptotic stimulus, DNA in the various fractions can be quantified by scintillation counting.

This approach is simple and readily applicable to large numbers of samples. Moreover, it provides relatively precise quantitation of how much DNA has been cleaved to oligomers of nucleosome-sized fragments. This information can be very useful in time-course studies of nuclease activation during apoptosis. By using this technique alone, however, it is impossible to determine whether partial DNA fragmentation results from complete degradation in a subset of the cells or partial degradation of the DNA in all of the cells. Moreover, because small DNA fragments resulting from necrosis will also be extracted under the conditions described above, this technique cannot be used to distinguish apoptosis from necrosis.

Flow Cytometry

Solubilization of oligonucleosomal fragments in buffers lacking divalent cations also plays a role in the detection of apoptotic cells by flow cytometry after staining with propidium iodide.³⁴ When this approach is applied, cells treated with an apoptotic stimulus are fixed with ethanol, washed with an aqueous buffer lacking divalent cations (e.g., 0.1% sodium citrate), treated with RNase A, reacted with propidium iodide, and subjected to flow cytometry.³⁵ Alternatively, unfixed cells are lysed in buffer containing neutral detergent and EDTA, stained with propidium iodide, and subjected to flow cytometry (Figure 3.1D).³⁶ In either case, low-molecular-weight fragments are extracted from the cells, decreasing their apparent DNA content.

This technique is simple, relies on widely available equipment, and is amenable to high- throughput analysis. Moreover, because thousands of particles can be analyzed in each sample, a precise determination of the percentage of subdiploid events is possible. Nonetheless, this approach has two notable limitations. Because the low-molecular-weight DNA fragments generated during the course of necrosis should also be extractable in cation-free aqueous buffers, this technique does not distinguish between apoptotic and necrotic cells. In addition, because one cell can give rise to multiple apoptotic bodies, counting the number of particles that contain strand breaks observed after γ-irradiation have been widely studied using the comet assay.^55,56 In addition, it has been reported that necrotic cells yield a comet when this approach is applied.⁵³ Nonetheless, when the occurrence of apoptosis has been documented by other techniques, the comet assay can be useful for studying the process on a cell- by-cell basis.