Forensic Science International

108 (2000) 187203

www.elsevier.com / locate / forsciint

Post-mortem analysis of apoptotic changes associated

with human skin bruises
Toshiko Sawaguchi

a,b ,

*, Bharat Jasani c , Makio Kobayashi d ,

Bernard Knight a

Wales Institute of Forensic Medicine, Department of Pathology, University of Wales College of Medicine,
Cardiff, UK
b
Department of Legal Medicine, Tokyo Women s Medical University, Tokyo, Japan
c
Immunocytochemistry and Molecular Pathology Unit, Department of Pathology, University of Wales
College of Medicine, Cardiff, UK
d
Department of Pathology, Tokyo Women s Medical University, Tokyo, Japan
Received 7 October 1998; received in revised form 22 October 1999; accepted 23 November 1999

Abstract
The estimation of the age of skin bruises is of importance in forensic medicine, especially in
child abuse cases. Time-dependent changes in bruise colour and / or associated histological features
have been used with a limited degree of success. An increased rate of apoptosis in the injured
tissue has been considered as a novel time-dependent marker of cell death, by injury inflicted in a
rat model. The object of the present study was to apply the TUNEL method of DNA end labelling
to identify and enumerate apoptotic cells in bruised and normal skin in order to study the
relationship of apoptotic cell density with the age of the bruise. A commercially available DNA
end labelling kit, TUNEL method, was standardised, validated and used for this purpose. Twenty
unselected post-mortem cases with bruises due to a variety of causes were studied. The apoptotic
cells stained with TUNEL reaction were counted in 10 high power fields in the epidermis, as well
as in the dermis of formaldehyde-fixed paraffin-embedded skin specimens. The mean positive cell
densities (61 S.E.) were compared with respect to the age of the bruise. In the epidermis, the
mean apoptotic cell count was statistically significantly greater in the bruised skin compared to
normal skin in 2- to 6-day-old bruises; whilst in the dermis the same was true in 3- to 8-day-old
bruises. The overall findings suggest that there is a quiescent period prior to the increase in the
apoptotic cell activity that is seen following skin bruising. This is so provided the post-mortem
*Corresponding author. Department of Legal Medicine, Tokyo Womens Medical University, 8-1 Kawadacho, Shinjuku-ku, Tokyo 162, Japan. Tel. / fax: 181-3-5269-7300.
E-mail address: [email protected] (T. Sawaguchi)
0379-0738 / 00 / $ see front matter 2000 Elsevier Science Ireland Ltd. All rights reserved.
PII: S0379-0738( 99 )00210-8

188

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

skin samples were collected within a lapse of 6 days or less between the time of death and
formalin fixation and paraffin embedding to avoid the bias made by the difference of length of
post-mortem interval. 2000 Elsevier Science Ireland Ltd. All rights reserved.
Keywords: Apoptosis; Human skin bruises; TUNEL method

1. Introduction
Mechanical trauma to the skin may lead to injury referred to as a bruise. Although
often combined with abrasions or lacerations, a pure bruise lies beneath an intact
epidermis and consists of an extravascular collection of blood that has leaked from blood
vessels damaged by mechanical impact. An extravasation of such blood that should be
arbitrarily larger than a few millimeters in diameter, is usually termed a bruise or
contusion in forensic medicine [1].
Attempts to estimate of the age of skin bruises is of considerable importance in
forensic pathology, particularly in child abuse cases. Though much research has been
carried out on the age estimation of open wounds, only a few studies have been reported
on the study of bruises [25]. In three of these studies the age of the bruise was
estimated using time-dependent changes in bruise colour [2,3] or light-microscope
histology [4]. Using this approach a bruise with a yellow colour was estimated to be
more than 18 h old provided it related to subjects aged ,65 years [2]. A red coloration
was seen in up to 50% of bruises less than 1 week old whilst a yellow coloration was
seen in up to 25% of bruises over 1 day old in children [3]. The light-microscopical
changes included: neutrophil leukocyte exudate peaking between 24 to 48 h; macrophage leukocyte exudate was maximal between 12 to 36 h; fibroplasia with the highest
incidence between 48 to 60 h; and haemosiderin deposits in macrophages rising from 48
h and reaching the highest density at 72 h. The relation lack of specificity and variability
of these macroscopic and histological criteria has promoted a search for more reliable
alternative markers.
Apoptosis is a specific form of cell death with highly characteristic morphological and
biochemical features: condensation of chromatin and cytoplasm and fragmentation of
these into the so-called apoptotic bodies, followed by a specific pattern of DNA
breakdown through activation of certain endonucleases. Apoptosis is an active, genetically programmed process which has been studied recently as a relatively accurate
post-mortem marker of brain injury associated with closed head trauma inflicted in a rat
model [6]. The data showed that severe head injury is associated with a significant
increase in brain cell death by apoptosis up to 10 h after the trauma. Another study also
conducted on rat brain has shown that prolonged formaldehyde fixation and post-mortem
delay up to 2 days have little or no influence on the rate of apoptosis in brain cells
induced by unilateral intrastriate injection of N-methyl-D-aspartic acid (NMDA) as
measured by in situ end labelling (ISEL) of fragmented DNA [7]. The present study
was therefore set up to examine the usefulness of DNA end labelling for analysis of
apoptotic cell density in relation to the age of bruises in post-mortem human skin
specimens.

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

189

2. Materials and methods

2.1. Tissue specimens

Samples were taken from 20 unselected corpses (clinical details listed in Table 1) with
,1- to 7-day-old abdominal skin bruises (site distribution and possible cause of bruise as
shown in Table 2). The test samples were taken from a part of the bruised area and the
control samples were taken from the contra-lateral normal-appearing skin sites. Tissues
were fixed in 4% paraformaldehyde overnight (for 1824 h), embedded in paraffin wax
in a routine manner and serial (5 mm thick) sections prepared. Two blocks from bruised
skin and one block from normal skin were examined from each individual.

2.2. Histological examination

Sections stained with routine haematoxylin and eosin (H&E) stain were used to
confirm the presence of extravasated red cells in the dermis as a characteristic of the
bruise.
Table 1
Clinical details of each case
Case
no.

Sex

Age
(years)

PMI
(day)

Cause of death

75

65

3
4
5
6
7
8
9
10
11
12

M
M
M
M
F
M
F
F
F
M

73
46

71
63

1
1
2
7
2
2
5
0
3
4

13

80

14
15
16

F
F
F

17
18
19
20

M
M
M
M

1a: Myocardial infarction; 1b: coronary atherosclerosis;

2: ruptured atheromatous aneurysm of abdominal aorta (operated)
1a: Peritonitis; 1b: intestinal fistula and abdominal abscess;
1c: atheromatous aneurysm of abdominal aorta (operated)
1a: Cerebral lacerations and haemorrhage with skull fracture
1a: Myocardial infarction; 1b: coronary atherosclerosis
Cause not known
1a: Cardiogenic shock; 1b: leaking abdominal atherosclerotic aneurysm
Cause not known
1a: Bronchopneumonia; 1b: fatty degeneration of liver and heart
1a: Ruptured abdominal atherosclerotic aortic aneurysm
Cause not known
Cause not known
1a: Myocardial infarction; 1b: coronary atherosclerosis;
2: chronic obstructive airways disease
1a: Bronchopneumonia, renal failure; 1b: ruptured atheromatous
aneurysm of abdominal aorta; 2: calcific aortic and mitral valve disease
Cause not known
Cause not known
1a: Adult respiratory distress syndrome with lobar pneumonia;
1b: ruptured spleen (operated)
1a: Blunt injury and chest and belly
1a: Multiple injuries including fractures to the skull, spine, ribs and pelvis
1a: Multiple injury
Cause not known

77
57
83

3
83

44
21
78

3
2
2
3

190

1.
2.
3.
4.
5.

6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
a

4% Paraformaldehyde fixation for 1824 h

Embedding in paraffin
5-mm sections on Super Frost plus slides
Preheating at 608C for 30 min in oven
Deparaffinisation
in 100% xylene five times each for 2 min
in 96% ethanol four times each for 2 min
in PBS once for 2 min
Wash in running tap water for 20 min followed by microwave treatment
with high power for 5 min at 800 W in 0.01 M citric acid buffer (pH6.0)
Wash in PBS twice each for 10 min
Intrinsic POD blocking: 3% H 2 O 2 in methanol for 30 min at room temperature
Wash in PBS twice each for 10 min
TUNEL reaction mixture (75 ml per 1 sample) at 34548C for 60 min in oven,
applied under Gene Flame
Wash in PBS four times each for 10 min
Converter-POD (70 ml per section) at 378C for 30 min in oven with Gene Frame
Wash in PBS four times each for 10 min
0.5 mg / ml DAB with 17 ml H 2 O 2 per 50 ml for 10 min at room temperature
Wash in PBS four times each for 10 min
Counter stain in methyl green (in tap water, methyl green solution at 608C for 5 min,
in deionised water for 1 min, in tap water 510 times in acetone containing 0.05% acetic acid)
Dehydration in a series of 96% alcohol baths followed by xylene baths
Permanent mounting
TUNEL mixture including TdT conjugated with fluorescein. Convertor POD including anti-fluorescein antibody conjugated with POD.

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

Table 2
Protocol of standard TUNEL method a

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

191

2.3. TUNEL method

To detect the apoptotic cells the TUNEL method was carried out using In Situ Cell
Death Detection Kit, POD (Boehringer Mannheim, Mannheim, Germany) with minor
modifications. The details of the method are listed in Table 2.

2.4. Positive and negative controls

Follicular centres and interfollicular zone cells of formalin-fixed paraffin-embedded
human tonsil sections were used for checking sensitivity and specificity of the TUNEL
reaction associated with each batch of analysis.
Negative controls were also set up through omission of the TdT enzymes in the
TUNEL reaction mixture, or through pretreatment of sections with DNase-I.
The overall reliability (specificity, sensitivity and reproducibility) of the method was
further evaluated and validated using human endometrium at different stages of the
menstrual cycle, as well as tissue-culture cell blocks prepared from cytotoxic drugrelated cells exhibiting low, moderate and high apoptotic cell death rates, using the
following approach.
Endometrial biopsy blocks from surgical pathology archives considered to be from
different stages of the normal menstrual cycle as confirmed by dates and morphological
criteria, were reviewed. These included a set of four pairs of blocks, each pair
representative of either early proliferative, mid-secretory, late secretory, or perimenstrual
endometrium curettings. Sections from these were stained for apoptotic cells using the
standardised TUNEL method, and graded for relative apoptotic activity on a single-blind
basis by an assessor unaware of the biopsy staging data.
A human follicular B-lymphoma cell line, DOHH2 (a gift from Dr. J.C. KluinNelemans, Leiden, The Netherlands), with details as originally described by KluinNelemans et al. [8], routinely maintained in RPMI 1640 supplement with 5% fetal calf
serum (FCS), 100 units / ml each of penicillin and streptomycin, and 2 mM glutamine,
was passaged twice weekly at an initial density of 5310 4 cells / ml at 378C in a humid
atmosphere of 5% CO 2 / 95% air. Apoptosis was induced with etoposide (VP-16;
Vepesid, Bristol Myers, Syracuse, NY, USA), and approximately, 1310 7 cells / ml of
untreated and treated cells, respectively, and an even (50:50) mixture of the two types of
cells, untreated and treated by etoposide, were pelleted by centrifugation at 1250 g for
10 min at 48C. The pellets were washed once with phosphate-buffered saline (PBS) in
suspension, repelleted and fixed in 5 ml of 4% paraformaldehyde for 18 h, and
embedded in paraffin wax in a routine manner. Sections cut from these blocks were
stained for apoptotic cells using the standardised TUNEL procedure, and the apoptotic
cell rate calculated for each of them.

2.5. Counting of apoptotic cells in skin samples

Apoptotic nuclei recognisable from the positive reaction of the TUNEL method were
counted in contiguous high power fields (objective 340) along the whole length of the
epidermis; the count for each field was added and the sum was divided by the number of

192

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

fields. For the dermis, two columns perpendicular to the length of the epidermis were
visualised in the mid portion of the dermis. The contiguous high power field (objective
340) was moved along these column(s) towards the deeper end of the biopsy, the
apoptotic cell count for each field was added and the sum divided by the number of
fields counted.

2.6. Statistical analysis

The coefficient of variation (Cv5standard error / mean) relating to the interblock
differences in the apoptotic rate of bruised skin specimens taken from the same
individual was calculated for cases with complete data. The linear regression analysis
relationship between the Cv and mean (see Fig. 1), was used to calculate the RSD for all
cases with or without complete data using the 95% confidence limits.

3. Results
The negative controls based either on omission of the TdT enzyme or inclusion
and / or exclusion of DNase pretreatment (Fig. 5), consistently produced uniformly
negative results on tonsil, uterine endometrium, and normal as well as bruised skin
specimens. DNase pretreatment increases the nonspecific nicks and those nicks lead the
nonspecific weak nuclear staining as nonspecific positive staining. In this sense, DNase
pretreatment is recognised as a positive control. At the same time, DNase treatment
digest the chromatin condensation which is specific in apoptotic cells and decrease

Fig. 1. Linear regression analysis relationship between Cv and mean with 95% confidence limits (mean5mean
of apoptotic rate by age of bruise, Cv5standard error / mean) (A) in epidermis; RSD50.6320.553mean, (B) in
dermis; Cv50.7220.833mean.

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

193

Table 3
Patterns of staining reaction obtained with the TUNEL method
Grade

Situation of staining

0
1
1a
1b
2
3
3a
3b
3c

No staining
Cytoplasmic staining only
Weak cytoplasmic staining only
Strong cytoplasmic staining only
Weak nuclear staining plus cytoplasmic staining
Nuclear staining only
Weak nuclear staining only
Moderately strong nuclear staining only
Strong nuclear staining only

specific apoptotic strong nuclear reactions. In this sense, DNase pretreatment is

recognised as a negative control in TUNEL reaction. Hence, DNase pretreatment is
recognised as a paradoxical control in the TUNEL reaction.
Amongst the positive control and test specimen results regardless of the tissue source,
the background staining associated with the TUNEL reaction was found to be either
completely absent, or occurred as weak or strong cytoplasmic staining, or as varying
intensity of nuclear staining. The frequencies with which these patterns of background
reactions occurred are tabulated in Table 3. A few of these patterns are illustrated in Fig.
2. Only the TUNEL reactions associated with uniform, relatively weak background
nuclear staining without any accompanying cytoplasmic staining (pattern 3a) were found
to reliably reproduce the expected frequency of apoptotic cells in the selected external
positive control specimens (Fig. 2). This was quantitatively validated to be the case for
the cell line preparation as indicated in Table 5. Hence, only the test results showing
pattern 3a type background nuclear staining were selected for quantitative analysis of the
apoptotic activity. Unfortunately, using this criterion only about one-fifth of all the skin
studies undertaken proved trustworthy for analysis. The optimal reactions selected on the
basis of this criterion pertaining to the endometrium, the cell line preparation, as well as
the epidermis and dermis from bruised and normal skin specimens are illustrated in Figs.
3 to 4 (see Fig. 5 for results either on omission of TdT or inclusion and / or exclusion of
DNase pretreatment for tonsil) and Figs. 6 to 7, respectively.
Analysis based on optimal stained sections, showed the variation in mean apoptotic
cell rate in the normal epidermis and dermis to be limited to a range between 0 and 0.5
apoptotic cells / HPF up to 6 days post-mortem interval before fixation of the tissue. The
rate was found to rise precipitously to 2 apoptotic cell / HPF in epidermis fixed following
a post-mortem interval of 7 days or greater. Hence, only skin specimens fixed within less
than 6-day post-mortem interval were used for formal analysis. The relationships
between the age of bruise in days and the mean apoptotic cell rates observed for the
selected epidermis and dermis are graphically represented in Figs. 8 and 9.
Using the standard error estimates extrapolated from Fig. 1, the overall bruised shin
apoptotic rate (BSAR) observed for the epidermis associated with bruises of age
between 2 to 6 days (illustrated in Fig. 8), was found to be statistically significantly
higher than that of the corresponding normal skin apoptotic rate (NSAR) (P,0.05;

194

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

Fig. 2. Apoptotic cell of fibroblast by TUNEL reaction and patterns of stained reactions (see Table 3). (A)
Uniform weak nuclear staining in tonsil (grade 3a), (B) uniform weak nuclear staining in epidermis (grade 3a),
(C) uniform strong nuclear staining (grade 3c), (D) no staining (grade 0).

Fig. 3. Apoptotic cells stained by TUNEL reaction in uterine endometrium (A) perimenstrual, (B) late
secretory, (C) secretory, (D) proliferative.

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

195

Fig. 4. Validation of the TUNEL procedure using nontreated cell line. (A) Untreated cells, (B) untreated
etoposide-treated cells (50:50) mixture, (C) etoposide-treated cells, see Table 5.

196

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

Fig. 5. Controls of TUNEL reaction and standard of TUNEL reaction using sections of human tonsil (A)
without TUNEL reaction mixture including TdT, (B) pretreatment of DNase, (C) standard TUNEL reaction.

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

Fig. 6. Apoptotic cells in normal epidermis. (A) Normal epidermis, (B) bruised epidermis.

197

198

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

Fig. 7. Apoptotic cells in normal dermis. (A) Normal dermis, (B) bruised dermis.

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

199

Table 4
Frequency of various patterns of TUNEL reactions in the epidermis in bruised and normal tissue blocks
Pattern of
staining reaction
0
1a
1b
2
3a
3b
3c

Type of tissue
Bruise (n5317)

Normal (n5533)

77 (24.3%)
25 (7.9%)
8 (2.5%)
13 (4.1%)
68 (21.5%)
32 (10.1%)
94 (29.7%)

83 (24.9%)
22 (6.6%)
1 (0.3%)
9 (2.7%)
75 (22.5%)
38 (11.4%)
105 (31.5%)

Student t-test).Similarly, the overall BSAR observed for fibroblasts in the dermis
associated with bruises from 3 to 8 days old (as illustrated in Fig. 8) was statistically
significantly higher than the corresponding NSAR (P,0.05; Student t-test). Furthermore, the epidermal apoptotic cells associated with non-bruised skin were always found
to be restricted to the stratum granulosum, whilst those associated with bruised skin were
found both in the stratum granulosum as well as in the stratum spinosum.
The validation of TUNEL method using cell lines is shown in Table 5.
Linear regression analysis relationship between Cv (standard error / mean) and mean
(mean of apoptotic rate by age of bruise) with 95% confidence limits were carried out in
epidermis and in dermis.

4. Discussion
Apoptosis activity found in the post-mortem skin is most probably analogous to the
shrinkage necrosis associated apoptosis observed in liver cells following various
types of injury [9]. In general, the apoptotic process is associated with condensation of
cytoplasm followed by phagocytosis and digestion by surrounding cells, the steady state
mass of a tissue being related to the balance between cell formation (mitosis) and cell
destruction (via apoptosis) [10]. In epidermal cells, cytoplasmic condensation takes the
form of abnormal and premature keratinisation or dyskeratosis [10]. The light-microTable 5
Validation of the TUNEL method using cell line

Observed apoptotic cells /

total cells ratio (%)
Predicted ratio (%)

A
untreated
cells

B
untreated
cellsetoposide
treated cells
(50:50 mixture)

C
Etoposide-treated
cells

12.10

24.10

31.03

21.6

200

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

Fig. 8. Relationship between aging of bruises and apoptosis in epidermis.

scopic criteria for the identification of apoptotic keratinocytes have been published and
combined with the distinctive cytological features of keratinocytes [11]. It has been
suggested that DNA fragmentation is initiated in the granular keratinocyte layer and is
identical in pattern to that seen in other examples of programmed cell death [12]. Due to
abundance of tonofilaments in keratinocytes, usually rigid and large apoptotic bodies are
formed [13] that remain visible for several hours [14] even though shrinkage and
budding of apoptotic cells occur within a few minutes [15].
As to the relationship between apoptosis and physical injury in skin, a number of
papers on acute UV-induced dermatitis [10,13] and on changes after radiation [16,17]
has been published confirming UV-induced and radiation-induced apoptosis in the skin is
an active, genetically programmed phenomenon [13].
Apoptotic changes detected following somatic death may only relate to the effect of
the post-mortem interval (PMI) prior to fixation [18]. In this study, the epidermis was
however, relatively free of apoptotic activity up to 6 days post-mortem, thus making it
feasible to meaningfully use apoptosis as a marker in post-mortem forensic investigations. The result of this study based on skin specimens, is similar to the study

Fig. 9. Relationship between aging of bruises and apoptosis of fibroblasts in dermis.

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

201

conducted on post-mortem brain tissue [7], in that no influence of PMI was noted on the
low numbers of apoptotic cell nuclei seen up to at least 2 days post-mortem.
The slight time-difference observed between the apoptotic rate in the epidermal layer
and in the dermal layer maybe explained on basis of the following points:
(1) The direct impact effect which has caused bruises on epidermis and on dermis is
different.
(2) The cell cycle of fibroblasts and epidermal cells are different.
(3) The fibroblasts in the dermal layer are directly affected by cytokines released by
infiltrating macrophages but the epidermal cells are only indirectly or not affected by
those factors.
(4) In the dermis, fibroplasia becomes maximal after 4860 h after bruising [4].
If the increase in apoptotic cell count is under the influence of cytokines that are
released by infiltrating macrophages, the increase in the frequency of apoptotic cells may
start between 12 to 36 h following the injury, a period during which macrophages are
expected to infiltrate the site [4]. According to this hypothesis it is expected that the
apoptotic cell count will increase at an earlier period in the dermis than in the epidermis.
However, we observed the reverse of this hypothesis: the percentage increase in
apoptotic cell count was found to increase slightly earlier in the epidermis than in the
dermis.
The increase of fibroblasts, the so-called dermal fibroplasia, takes place between 24
to 66 h following skin injury [4]. However, the percentage increase of fibroblast-like
associated apoptotic cells in the dermis occurred between 48 and 216 h. Following skin
injury, mesenchymal cells increase the amount to effect tissue replacement and
reorganization. It is possible that during this recognisational process, any excess of cells
undergoing proliferation are probably eliminated through apoptosis.
The apoptotic cell rates in the stratum granulosum were almost the same between the
bruised and normal skins. On the other hand, the apoptotic activity in the stratum
spinosum in bruised skin appeared to be consistently higher than that in the normal skin.
As a result, any excess of epidermal apoptotic cells were almost entirely confined to the
stratum spinosum in bruised skin.
An additional question raised by this study is how apoptotic cells survived during the
post-mortem interval. Apoptotic cells are normally phagocytosed by neighbouring cells
almost immediately. However, phagocytosis may not occur adequately after somatic
death and since a new round of apoptosis is not likely to occur after somatic death, the
presence of cellular features suggestive of apoptosis may reflect a process that took place
prior to somatic death. In severe closed head injury, DNA fragmentation was found
significantly increase, with the maximum level occurring at 10 h after trauma, and the
DNA and nuclei yields decreased 10 h after injury and remained at a reduced level at all
subsequent sampling intervals [6].
Somatic death is an event and cellular death is a process rather than event [19]. The
interval between somatic death and cellular death may be called the supra-vital
period. Apoptosis is recognised as one of cellular vital signs during supra-vital period,
because apoptosis is originally an antemortem phenomenon [18,20].
The TUNEL method is a relatively recent development [21]. Problems of reproducibility of this method have been widely acknowledged. When the tissues from

202

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

post-mortem bodies are used for the TUNEL method, the reproducibility is expected to
be further compromised because of the often unavoidable variation in the fixation time
and the post-mortem interval. In this study the fixation time was limited to between 16
and 24 h, and the post-mortem time interval to less than 6 days.
Dependable quality TUNEL reaction (type 3a pattern with weak nuclear staining only;
Table 4) was found to be obtainable in only around 20% studies. We believe that this
low reproducibility is due to factors inherent to the TUNEL technique, in particular, the
uncontrollability of the microwave pretreatment, and the temperature of terminal
deoxynucleotidyl transferase (TdT) enzyme reaction, respectively. Though some studies
have recommended the use of the microwave pretreatment to obtain good quality results,
others have reported that this pretreatment compromises the reproducibility of the results
[22]. The reproducibility of the ISEL method based on the use of E. coli DNA
polymerase I seems to be higher than that of the TUNEL technique which uses TdT
instead. Further work is needed to optimise and standardise these factors in order to
improve the overall efficiency and reliability of the TUNEL method.

Acknowledgements
The authors are grateful to Dr. S. Leadbetter, Dr. S. Claydon, Dr. R. James for
providing tissue samples and also to Professor G. Williams and Professor P. Smith,
Department of Pathology, University of Wales College of Medicine for providing the
tissue and cell samples for the validation of the TUNEL method. A part of this study
was carried out by the grant of Uehara Life Science Foundation.

References
[1] B. Knight, Contusions or bruises, in: Forensic Pathology, 2nd Edition, Edward Arnold, London, 1996, p.
140.
[2] N.E. Langlois, G.A. Gresham, The ageing of bruises: a review and study of the colour changes with time,
Forensic Sci. Int. 50 (1991) 227238.
[3] T. Stephenson, Y. Bialas, Estimation of the age of bruising, Arch. Dis. Child. 74 (1996) 5355.
[4] R.N. Thornton, R.D. Jolly, The objective interpretation of histopathological data: an application to the
ageing of ovine bruises, Forensic Sci. Int. 31 (1986) 225239.
[5] K. Kibayashi, K. Hamada, K. Honjyo, S. Tsunenari, Differentiation between bruises and putrefactive
discolorations of the skin by immunological analysis of glycophorin A, Forensic Sci. Int. 61 (1993)
111117.
[6] S.V. Pravdenkova, A.G. Basnakian, S.J. James, B.J. Andersen, DNA fragmentation and nuclear
endonuclease activity in rat brain after severe closed head injury, Brain Res. 729 (1996) 151155.
[7] P.J. Lucassen, W.C.J. Chung, J.P. Vermenlen et al., Microwave-enhanced in situ end labeling of
fragmented DNA: parametric studies in relation to postmortem delay and fixation of rat and human brain,
J. Histochem. Cytochem. 43 (1995) 11631171.
[8] H.C. Klwin-Nelemans, J. Limpens, J. Meerabux et al., A new non-Hodgkins B-cell line (DOHHZ) with
a chromosomal translocation t (14, 18)(q32,q21), Leukaemia 5 (1991) 221224.
[9] J.F.R. Kerr, Shrinkage necrosis: a distinct mode of cellular death, J. Pathol. 105 (1971) 1320.
[10] P.L. Olson, M.A. Everell, Epidermal apoptosis: cell deletion by phagocytosis, J. Cutaneous Pathol. 2
(1975) 5357.

T. Sawaguchi et al. / Forensic Science International 108 (2000) 187 203

203

[11] J.G. Lovas, Apoptosis in human epidermis: a postmortem study by light and electron microscopy, Aust. J.
Derm. 27 (1986) 15.
[12] C.A. McCall, J.J. Cohen, Programmed cell death in terminally differentiating keratinocytes: role of
endogenous endonuclease, J. Invest. Dermatol. 97 (1991) 111114.
[13] D. Weedon, J. Searle, J.F. Kerr, Apoptosis. Its nature and implications for dermatopathology, Am. J.
Dermatopathol. 1 (1979) 133144.
[14] R. Paus, T. Rosenbuch, N. Haas, B.M. Czarnetzki, Patterns of cell death: the significance of apoptosis for
dermatology, Exp. Dermatol. 2 (1993) 311.
[15] J.F. Kerr, Patterns of cell death, Methods Arch. Exp. Pathol. 12 (1988) 1854.
[16] T.P. Fitzpatrick, A.Z. Eisen, K. Wolff, I.M. Freedberg, K.F. Austen, in: Dermatology in General
Medicine, 3rd Edition, McGraw Hill, New York, 1987, p. 52.
[17] L. Geng, C.S. Potten, Changes after irradiation in the number of mitotic cells and apoptotic fragments in
growing mouse hair follicles and in the width of their hairs, Radiat. Res. 123 (1990) 7581.
[18] T. Sawaguchi, B. Jasani, B. Knight, Identification of apoptotic cell nuclei in postmortem skin biopsies
using the TUNEL reaction: effect of post-mortem time interval, Acta Crim. Japan 63 (1997) 3943.
[19] B. Knight, The pathophysiology of death, in: B. Knight (Ed.), Forensic Pathology, 2nd Edition, Edward
Arnold, London, 1996, pp. 5152.
[20] Y. Nakanuma, What histology is normal and abnormal in the liver?, Jpn. J. Leg. Med. 46 (1992)
361365.
[21] Y. Gavrieli, Y. Sherman, S.A. Ben-Sasson, Identification of programmed cell death in situ via specific
labeling of nuclear DNA fragmentation, J. Cell Biol. 119 (1992) 493501.
[22] J. Strater, A.R. Gunthert, S. Bruderlein, P. Moller, Microwave irradiation of paraffin-embedded tissue
sensitizes the TUNEL method for in situ detection of apoptotic cells, Histochem. Cell Biol. 103 (1995)
157160.