Electron microscopic techniques

Electron microscopy was the original technique used by Kerr et al. (1) in their seminal publication on apoptosis. It provides the most detailed information for the assessment of cell morphology, and hence the most accurate determination of apoptosis in tissues. However, it requires more expensive equipment and takes longer than other methods. Protocol 11 describes a method for transmission electron microscopic study of cell lines and isolated cells from tissues. Prior to starting Protocol 11, cultured cells must be harvested by

• Acridine orange: 2 mg/ml stock in PBS, pH7.4. Prior to use dilute 1:400 in PBS (Sigma)

• Propidium iodide: 2 mg/ml stock in PBS, pH 7.4. Prior to use, dilute 1:1500 in PBS (Sigma P4170).

conventional means. Cells from normal tissues and tumours may be isolated by a variety of methods, including scraping and mechanical disaggregation, in combination with mild enzymatic treatment or chelating agents. Readers are advised to seek more specialized texts for the optimal preparation of specific tissues. Particular care needs to be taken when preparing samples for electron microscopy that are also going to be used for immunolabelling of proteins and RNAs. Here, the use of low-temperature embedding procedures is recommended in order to preserve antigenicity (9,10).

Protocol 11. Preparation of cells for transmission electron microscopy (TEM)


• 2% paraformaldehyde

• glutaraldehyde


11. Resuspend the cells at a density of ~2 X 107/ml in 0.1 M sodium cacodylate buffer. For this protocol a 1.5 ml plastic centrifuge tube is a useful size to carry out the procedures in.

12. After 5 min, pellet the cells by centrifugation, and resuspend in 2% paraformaldehyde/2% glutaraldehyde in 0.1 M sodium cacodylate buffer (Tooze fixative). Leave the cells in the Tooze fix for 1 h.

13. Pellet the cells by centrifugation, and resuspend in 0.1 M sodium cacodylate buffer.

14. Pellet the cells again after a further 5 min, and resuspend in 1% osmium tetroxide, in 0.1 M sodium cacodylate buffer. Leave for 1 h.

15. After 1 h, pellet and resuspend the cells in 2% magnesium uranyl acetate (in 70% ethanol). Leave overnight, in the dark.

16. The following day, dehydrate the cells in a graded series of alcohols (3 x 70%, 90%, 95%), allowing 15 min in each.

17. Carry out three further changes in 100% ethanol, allowing 1 h between each change.

18. After dehydration, allow 1-2 h for infiltration of absolute alcohol/ Spurr's resin mixture (1:1).

19. Change the alcohol/resin mixture for 100% Spurr's resin and allow further infiltration overnight.

10. The next day carry out a further three changes of resin, and finally polymerize the block overnight at 60°C.

• 2% magnesium uranyl acetate

• Spurr's resin mixture

11. Cut the sections (at 30-50 nm thickness) on to water using an ultra microtome,

12. Expand the sections with chloroform, prior to transferring to copper mesh support grids. Allow the sections to dry.

13. Stain the sections with 2% uranyl acetate (in 70% ethanol) for 20 min.

14. Wash the sections three times with deionized water, prior to staining tn 0.3% lead citrate (lead nitrate and lead acetate are also used routinely) for 3-4 min.

15. Wash three times in deionized water and allow to air-dry.

16. Examine the sections using a transmission electron microscope.

Electron microscopic (EM) evidence of apoptosis is illustarted in Figure

4. Quantitation of apoptotic events 4.1 Methods

This is a very labour-intensive task, requiring much time being spent on the microscopic examination of prepared samples. Cells in suspension, stained with fluorescent nuclear dyes, are loaded on a haemocytometer and viable

Figure 3. (a) Transmission electron micrograph of several apoptotic bodies which have been phagocytosed by a neighbouring cell. One of the phagocytosed bodies appears to be a cell which itself has previously phagocytosed another cell. The micrograph is of the base of a small intestinal crypt. The elecron-opaque bodies marked G are Paneth cell granules. To the bottom right of the picture is the nucleus of the phagocytosing cell (N). We are grateful to Dr T.D. Allen for the production of this electron micrograph, (b) Budding apoptotic bodies and crescents in HL60 cells treated with 20 jxM etoposide for 6 h (supplied by G.P. Studzinski).

Figure 3. (a) Transmission electron micrograph of several apoptotic bodies which have been phagocytosed by a neighbouring cell. One of the phagocytosed bodies appears to be a cell which itself has previously phagocytosed another cell. The micrograph is of the base of a small intestinal crypt. The elecron-opaque bodies marked G are Paneth cell granules. To the bottom right of the picture is the nucleus of the phagocytosing cell (N). We are grateful to Dr T.D. Allen for the production of this electron micrograph, (b) Budding apoptotic bodies and crescents in HL60 cells treated with 20 jxM etoposide for 6 h (supplied by G.P. Studzinski).

cells and apoptotic cells are logged simply by use of a tally counter. Cytospun slides, cell monolayers, and tissue sections may be scored in a similar way, except that a graticule can be placed in the eye piece of the microscope to substitute for the haemocytometer.

Methods do exist to make counting easier and to reduce the time spent staring down a microscope, which include flow cytometry (see Chapter 4). In our laboratory we use a Zeiss Axiohome system. This consists of a microscope linked to a computer, which can monitor the co-ordinate position of the microscope stage and allows the user to overlay a projected image of tally symbols on the counted cells. The microscope automatically keeps a record of the total number of each symbol used. This system is particularly useful when scoring tissue sections with little or no structural organization, i.e. tumours and hair follicles. When scoring, it is the general practice to record observations of 1000 cells from 10 random fields. Video capture of slide images is one method which allows the observer more independence from the microscope. Digital images can then be analysed on screen; however, this does not allow the observer to focus up and down on an object in the slide that may be slightly out of plane (focus), which can be critical for assessing the number of small apoptotic bodies. The use of serial images from confocal microscopy is desirable, since it eliminates this problem.

For structurally organized tissue such as the intestine, it is possible to score the cells on a positional basis. This technique has revealed important biological differences between cells and allowed a number of issues to be addressed. One critical procedure to carry out this scoring is the optimal preparation and orientation of tissue sections. The method used in our laboratory is outlined in Protocol 12.

Protocol 12. Preparation of tissue for the positional scoring of apoptotic events in gastrointestinal epithelia of mice


1. Excise the small and large intestine and flush with ice-cold PBS.

2. Separate the intestine into different regions, i.e. duodenum, jejunum, upper ileum, lower ileum, caecum, mid-colon/rectum.

3. Cut into 1 cm lengths.

4. 'Bundle' together up to 10 pieces of intestine:

(a) cut an 8-10 cm length of micropore tape and form a loop by sticking the two ends together,

(b) lay the pieces of intestine parallel to each other within the loop and bind together with micropore tape,

(c) cut a 4-5 mm thick cross-sectional disc from the taped bundle.

5. Process these smaller bundles for histology, as described in Protocol 3.

6. Cut sections through the bundles and process by one of the protocols described in Section 2.2.

7. Examine the slides.

If the tissue is prepared as described in Protocol 12, slides with good transverse sections of the intestine and good longitudinal crypt sections will be obtained. Cells can be numbered from the base of the crypt, starting at cell position one, with each side of a crypt (half-crypt) being scored separately. Fifty half-crypts are commonly scored for each region of the intestine; ideally from 4-5 bundles containing 5-10 individual cross-sections. The average length of a small intestinal half-crypt is 20 cells, i.e. 50 half-crypts is equivalent to 1000 cells. The data are recorded on a laptop computer, using custom-written software for data acquisition and analysis (PC Crypts, Steve Roberts, Bio-statistics and Computing Department, PICR). The recording of data in this way allows the frequency distributions of events (i.e. apoptosis, mitosis) to be plotted for each cell position. This is illustrated in Figure 4. All positions can be related to cell lineage position and, thus, stem cell properties can be studied (11,12).

4.2 Problems in scoring apoptotic events

Quantitation of apoptotic events in cellular/tissue systems is always a topic that raises much debate. The major problem is whether the methods used accurately determine the frequency of apoptotic events, i.e. cell deaths, and how to interpret such numbers (11,12). A number of factors can influence the frequency of apoptotic events, or 'apoptotic index'. These include:

(a) Half-life. The half-life of apoptotic cells differs between cell systems/ tissues. In a tissue/cell line where apoptotic cells have a long half-life, samples at different times over a given period will show a higher frequency of events, compared with tissue where apoptotic cells have a short half-life, even though the absolute number of cell deaths may be the same.

Half-life itself is influenced by other factors. These include the rate at which cells are cleared by phagocytosis, migration of cells in a hierarchical system such as the intestinal tract, and the rate at which degradative changes take place. We have previously published data that suggest that the half-life of apoptotic cells may vary according to the cytotoxic agent used to induce apoptosis (see ref. 12 and references therein).

(b) Section thickness. Apoptotic cells are smaller than neighbouring, viable cells. They also fragment into yet smaller apoptotic bodies. Therefore, if very thin sections are cut through a tissue block, the apoptotic nuclei may

Viable Cell = 1 Apoptoiic Cell = 2

Half-Crypts #1:121211211111121112121111111 #2:121212121111111111111 etc...

Cell Position

Figure 4. Scoring of apoptosis or a cell positional basts in the intestinal crypts. Small intestinal crypts from a mouse exposed to 7-radiation are shown in (a). Individual cells can be assigned a numerical value, according to their histological appearance. This is recorded on a laptop computer as a string of numbers, as demonstrated in <b). Pooling of data from 50 half-crypts can then be used to construct frequency distributions (c): the data shown are for irradiated crypts 4 h after exposure to 16 Gy ^-radiation (solid line) and unirradiated crypts (broken line). In addition to apoptosis, we commonly use this technique to analyse mitotic events, incorporation of tritiated thymidine, and the expression of nuclear antigens, be 'missed', resulting in an underestimation of apoptotie frequency. The use of thicker sections may allow better visualization of apoptoiic events, but overall the structural organization of the tissue may become less clear, as is the case with the intestinal tract. Whole-mount preparation of crypts can give better estimates of the frequency of apoptotie events (13, 14); however, their use is not suitable for routine examination of a large number of tissue samples,

(c) Criteria selected for scoring. It is important for observers to be consistent in their application of selection criteria for apoptotie cells. Most workers would tend to err on the side of caution and only score those cells that are unambiguously apoptotie. Apart from recognizing apoptotie cells, there is the question of whether one should count every apoptotie body or try to assess how many apoptotie events are represented by a group of apoptotie bodies. This situation can be illustartcd by a comparison of the scoring of small intestinal epithelium and of proliferating hair follicles. The well-ordered structure of the small intestinal epithelium may allow the determination of whether given apoptotic bodies are the result of a single apoptotic event. In contrast, the hair follicle does not visibly demonstrate a highly ordered structure (although there is a cellular hierarchy), making this kind of assessment impossible. In this case, it is better to score all apoptotic bodies. As with the half-life of apoptotic cells/bodies, the number of apoptotic fragments arising from a single cell death may vary according to the cytotoxic insult.

(d) Circadian rhythm. This is certainly of relevance in the gastrointestinal tract, and probably also in other tissues, although diurnal variation in some tissues is less well characterized. We have shown that the rate of cell proliferation and cell migration in the intestinal crypts varies according to the time of day at which observations are made (15). When making comparisons between different treatments/experiments, therefore, it is essential that observations that are being contrasted are made at the same time of day.

(e) Cellular hierarchies. In hierarchical tissues, such as the intestinal epithelium, the sensitivity of any given cell to an apoptotic stimulus is determined by its position in that hierarchy. It is therefore useful to represent the frequency of apoptotic events in relation to this position within the hierarchy. This information can be very useful in determining whether effects are observed in selected cell types.

(f) The denominator of the apoptotic index. The apoptotic index (AI) is the number of apoptotic cells/bodies divided by the total number of cells. Within the well-ordered intestinal epithelium, the denominator can be determined reasonably accurately. For tissues with less visible organization, i.e. tumours and hair follicles, cell counts will be less accurate. Also, in heterogeneous tissue such as tumours, there can be variation in the rates of cell proliferation and cell death in different regions of the tissue. Consequently, it can be useful to determine the ratio of the apoptotic index in relation to the proliferation index (cell in S phase, measured by bromodeoxyuridine or tritiated thymidine incorporation) or the mitotic index (MI). The AI:MI ratio effectively eliminates the denominator and any uncertainties associated with it. It also provides useful information on any net changes in cell number that are occurring within a tissue.

Was this article helpful?

0 0

Post a comment