John R K Savage
(MRC Radiation and Genome Stability Unit, Harwell, Didcot, OX11 0RD,
Visible changes to chromosome structure and morphology have played a very
important part as indicators of genetic damage in both clinical and cancer
Most of the changes encountered in clinical studies are "secondary" or
"derived" aberrations. This is true also in cancer studies, except that here,
there is an ongoing production of aberrations, so that in some cells, a mixture
of primary and secondary changes is present, and a continuously changing
karyotype (true chromosomal instability).
To appreciate these observed secondary changes we need to understand the
primary changes from which they are derived, and it is the purpose of this
article to provide a brief introduction to them.
Primary aberrations are those seen at the first post-induction division,
when all the parts are present and there has been no selection by passage
through mitosis, nor any modification by subsequent chromosome duplication
Most commonly, observation is made at metaphase, using "solid-staining" with
dyes which give high-contrast chromatin staining and negligible cytoplasmic
coloration. For more critical work, the chromosomes are banded in various ways,
which allows chromosome identification, detection of some forms invisible with
solid-staining, and offers more precise positioning of the lesion interaction
points (Savage, 1977).
Recently, resolution and classification of transmissible forms has been considerably
improved by the introduction of fluoresence in situ hybridisation (FISH)
chromosome "painting" (Lucas et al., 1992; Savage and Tucker, 1996; Tucker et
al., 1995) .
Classification of Primary changes
For purely pragmatic and diagrammatic purposes, we can regard the chromosomal
changes we see down the microscope as being the result of "breaks" followed
by "re-joins" of the chromosome thread. However, we must always remember that,
in reality, their origin is much more complicated (Savage, 1998; Savage and
Since the chromosome we see and score at metaphase has two (sister-)
chromatids, it is convenient (and conventional) to divide all aberrations into
two broad types:
Chromosome-type where the breaks and re-joins always affect
both sister-chromatids at any one locus. Examples in Figure 1.
Chromatid-type where the breaks and re-joins affect only
one of the sister-chromatids at any one locus (Fig 2).
The distinction is important. For some aberration-inducing agents, like
ionizing radiation, the type of aberration recovered at metaphase reflects the
duplication status of the chromosomes in the treated cell. But, for the
majority of chemical agents which can induce aberrations, for ultra-violet
light, and most probably all "spontaneous" (and de novo aberrations)
only primary chromatid-types are recovered. When, at
subsequent interphase, the chromatids duplicate, surviving aberrations (and
bits of aberrations) are converted into apparent chromosome-types, some
of which are then transmitted almost indefinitely to further cell generations.
These are the "derived" aberrations, and many are so modified that it is
impossible to deduce their primary origin.
Thus, following an "acute" treatment with any clastogen, surviving cells
in later generations carry only chromosome-type changes. The
presence in such cells of chromatid-type aberrations is, therefore, an
indicator of an ongoing production of primary structural changes, i.e. of some
form of chromosome instability.
Nearly all the aberrations we see with solid staining appear to
result from the interaction ("re-joining") of two breaks, so we can further
classify them on the basis of where these breaks are situated in relation to
the chromosome arms (Savage, 1976).
- If the breaks are situated in the arms of different (non-homologous or
homologous) chromosomes we have the category of INTERCHANGES.
- If the breaks are in the opposite arms of the same chromosome, we have
the category of INTER-ARM INTRACHANGES.
- If the two breaks are both in the same arm of a chromosome, we have the
category of INTRA-ARM INTRACHANGES.
These three categories are often referred to collectively as
- Finally, some aberrations appear to arise from a single, open break in
just one arm. This category we term "BREAKS" or "DISCONTINUITIES". Many
(perhaps all) of them are, in reality, intra-arm intrachanges where one end has
failed to join up properly, though the limitations of microscopical resolution
do not permit us to be certain that the re-joining is really incomplete.
The newer techniques, like FISH chromosome painting, are telling us that a
lot of the chromosome-type aberrations we see and score as "simple" two-break
interactions actually involve more than two breaks, and often more than two
chromosomes, i.e. they are COMPLEX EXCHANGES (Savage and Simpson, 1994; Savage
and Tucker, 1996; Savage et al., 1984; Simpson et al., 1995). These types have
always been known to be fairly frequent for chromatid-types , but until
the advent of FISH, were always considered to be rare for chromosome-types.
However, as most complexes will be non-transmissible, and therefore rarely
encountered in clinical studies (INSERTIONS, and CYCLICAL EXCHANGES involving
several chromosomes, are examples of transmissible complexes), we will not look
at them in any detail.
Interaction between the four ends of two breaks can obviously take place in
three ways :
- Join back to re-form the original chromosomes ("RESTITUTION") so that no
aberration is produced.
- Re-join in such a way that an acentric fragment is always formed
(ASYMMETRICAL RE-JOINING, A ). These forms are invariably visible with
solid staining. The fragment (which, if visible microscopically, will contain
many megabases of DNA) will be lost at anaphase, and, in the case of
INTERCHANGES and INTER-ARM INTRACHANGES, there will also be mechanical
separation problems producing "anaphase bridges". Asymmetrical aberrations are
therefore almost always cell-lethal, and so rapidly disappear from a population
of continuously dividing cells (Lea, 1946; Savage, 1989). Thus, they are rarely encountered in
clinical situations where there is not an ongoing induction of aberrations.
Instability in cancer cells, however, does lead to the occasional presence of
asymmetrical ("unstable") changes.
- Re-join in a way that never leads to an acentric fragment unless one of
the re-joins is incomplete (SYMMETRICAL RE-JOINING, S ). Many such symmetrical
chromosome-type exchanges are not visible with solid-staining, and their
accurate detection requires special techniques like banding or FISH-painting
(Lucas et al., 1992). In contrast, most symmetrical chromatid-type exchanges
are visible with solid staining because of the retention of sister-chromatid
adherence until metaphase. Because there is no loss of genetic material, and
no mechanical problems at mitosis, most symmetrical forms are transmissible
to future cell generations (Savage, 1976; Savage, 1995), hence they constitute
the bulk of the recovered "derived" aberrations encountered in clinical and
As mentioned, re-joining can sometimes be (apparently) INCOMPLETE. This is
much more frequent for chromatid-type aberrations (typically 30-50% of
interchanges) than it is for chromosome-type aberrations (difficult to
measure accurately, but probably around 3-5%). Incompleteness leads to genetic
loss, and so to increased cell lethality.
The four basic categories discussed above are seen in their simplest forms
for chromosome-type aberrations, as shown in Figure 1. Traditionally,
certain forms have specific names, as indicated. Symmetrical forms are seldom
visible with solid staining (probably less than 20% of reciprocal
translocations lead to an obvious change in chromosome morphology), so, in
the diagrams, these forms are shown in two colours, just as they are detected
when using FISH chromosome (or arm) painting. There is no reliable method as
yet for detecting paracentric inversions.
Because sister-chromatids tend to adhere, strongly, along their lengths,
many chromatid-type symmetrical forms remain visible without recourse
to special staining methods, as Figure 2 shows. Moreover, the presence of
sister-chromatids allows additional lesion interactions (inter-chromatid
intra-arm intrachanges) not possible with chromosome-type changes.
As a consequence, there is a much higher frequency of chromatid-type
changes, and a much greater variety of forms, compared with
chromosome-types, following a given treatment.
Moreover, the interactions within chromosome arms which are now possible
make chromatid-type aberrations a much more likely source of the
complicated duplications/deletions etc. encountered in clinical and cancer
Combinations of the various categories are frequent, especially for
chromatid-types (interchange/intra-arm intrachange particularly
so, giving rise to configurations like "triradials" (Savage and Harvey, 1994)), but these, of
course, constitute a type of complex exchange. Surviving remnants of such
events are responsible for some of the curious anomalies recorded in clinical
and cancer cell studies.
Relationship to the cell cycle
Conventionally, the period between successive mitoses ("INTERPHASE") is
sub-divided into three phases G1, S and G2 . For critical work, further
sub-division of S is possible (Savage et al., 1984). G1 is the pre-duplication period, when
the cell begins to prepare for DNA synthesis and the next mitosis. If the cell
is not going to divide again, it passes out of cycle during this phase into
another phase termed G0. From this phase it may, or may not, be possible to
call it back into a division cycle. Usually, however, cells pass on to
irreversible differentiation with their chromosomes unduplicated.
S-phase is a discrete period of interphase of a few hours duration during
which the chromosomal DNA and protein is duplicated, and the new chromatin
segregated into the sister-chromatids. Each chromosome has a precise programme
of replication, closely associated with its G-band pattern.
Pale G-bands always replicate early in S-phase, dark G-bands later, and
constitutive heterochromatin tends to be among the very last regions to
replicate (Aghamohammadi and Savage, 1990; Savage et al., 1984).
During G2 , the newly replicated chromosomes undergo a rapid programme of
condensation, packing and coiling to produce the familiar metaphase chromosomes
where we normally identify and score aberrations. These condensed chromosomes
facilitate transport of the genetic material to the daughter cells at mitosis.
This condensation and packing readily obscures, modifies and disguises
aberrations which are produced during interphase - a point that should always
be borne in mind when interpreting what we see down the microscope.
Most aberration-inducing agents can introduce lesions into the chromatin
at all stages of the cell cycle, but relatively few of them can produce
actual structural changes in G1,( and therefore give rise to primary
chromosome-type changes) or in S and G2 (producing primary
Ionising radiation, restriction endonucleases, and a few chemicals like
bleomycin and some antibiotics are amongst those that can.
Almost all remaining aberration producing agents are "S-dependent" ;
surviving unrepaired lesions from G1 or G2 have to pass through a scheduled
S-phase to convert them into exclusively chromatid-type
Any interference with or abnormality in the processes of chromatin
replication also leads to chromatid-type aberrations visible at next
mitosis. It is almost certain that the vast majority of "spontaneous" and
de novo aberrations arise in this way. Chromosome instability syndromes
also probably produce aberrations via defective S-phase pathways.
However they are produced, the resulting chromatid-type aberrations
are qualitatively (but not quantitatively) identical.
Meaningful quantitative work with chromatid-types is extremely
difficult because observed frequencies fluctuate with time of sample after
treatment, and are subject to dramatic modifications as the result of mitotic
perturbation and differential cell selection. This makes comparison between
different treatments, or the production of sensible dose-response curves,
virtually impossible (Savage and Papworth, 1991).
Aberration transmission and stability
Although there is an enormous range of primary aberration forms,
very few of them are transmissible to future cell generations long term, so
only a handful of secondary ( or "derived" ) forms are recovered
(Savage, 1976; Savage, 1995).
The following paragraphs list the kinds most likely to be encountered,
together with comments and a note about probable primary origin.
RECIPROCAL TRANSLOCATION :
Involves no mechanical separation problems at anaphase, and usually no
genetic loss or imbalance. Problems can occur at meiosis because of multivalent
formation, and degrees of sterility may arise.
At the molecular level, the re-joining points can disrupt important
genetic sequences, leading to inactivation, mutation or position effects (e.g.
the t(9;22) Ph1 chromosome of CML).
Derived directly from chromosome-type reciprocal translocations or
from one segregation sequence of symmetrical chromatid-type
interchanges. (Note that the alternative interchange segregation leads to
imbalance and cell lethality).
PERICENTRIC INVERSION :
Very similar properties to those for reciprocal translocations given above.
Large inversions lead to meiotic bridges, sterility and cell death.
Derived directly from chromosome-type or chromatid-type
PARACENTRIC INVERSION :
Very difficult to detect at the chromosome level unless they are very large
(many megabases of DNA). Again the re-joining points can disrupt important
genetic sequences, and reverse segments of the reading frame. Large inversions
will give problems at meiosis.
Derived directly from chromosome-type paracentric inversions, or
from one form of chromatid-type intra-chromatid intra-arm intrachange
(Revell-type 3 (Revell, 1959; Savage, 1976; Savage, 1989).
INTERSTITIAL DELETION :
The loss of small segments of a chromosome (usually in only one homologue)
is not uncommon. Many mutations that have been genetically sequenced have been
shown to be actually small deletions.
Very occasionally, the loss of quite large segments appears to be
compatible with cell survival.
Derived directly from chromosome-type interstitial deletions
("double minutes") and from the alternative form of chromatid-type
intra-chromatid intra-arm intrachange (Revell-type 2
(Revell, 1959; Savage, 1976; Savage, 1989)) to that which
produces paracentric inversions. Segregation products from some complex
chromatid-type interchanges can also carry deletions.
TERMINAL DELETION :
It is now questionable whether true stable terminal deletions actually
exist. All those that have been investigated using the new fluorescent
telomere probes are found to be "capped" by telomere sequences. This either
means that they are disguised interstitial deletions, where one re-join point
was almost terminal, or that survival has been rendered possible by de
novo telomere synthesis. The recent development of end-specific telomere
probes should be able to solve this question
(Boei and Natarajan, 1998; Boei et al., 1998).
Derivation, if genuine, from various forms of incomplete
chromosome-type or chromatid-type intrachanges and interchanges,
followed by telomerase activity to achieve capping.
INTERSTITIAL DUPLICATION :
Segments of a chromosome repeated in tandem, sometimes in reverse sequence.
This may not necessarily arise from a pre-existing structural aberration ;
segment amplification and re-duplication is a well attested phenomenon under
certain conditions (e.g. HSR regions following chronic methotrexate exposure).
Nevertheless, there are primary aberrations which can survive as segmental
Most likely derived from one form chromatid-type inter-chromatid
intra-arm intrachange (Revell-type 1
(Revell, 1959; Savage, 1976; Savage, 1989)). Some forms of complex
chromatid-type interchanges can segregate to give surviving chromosomes
with duplicated segments.
INTERSTITIAL INSERTION :
Deletion of a segment and its insertion into another chromosome within the
same cell is a fairly common transmitted aberration. Much less common is the
insertion of a segment additional to the two complete homologues within a
All insertions are derived from complex exchanges, since, by definition,
their production requires the interaction of a minimum of 3 lesions. Either
chromosome-type or chromatid-type complex interchanges may be
involved, the range of inter-intrachanges in the latter being particularly
productive of insertions.
Occasionally, a surviving dicentric may be found, usually without the
related acentric fragment. Very often, the two centromeres lie very close
together, because, under these circumstances, only one of the centromeres is
active, so anaphase bridges do not form. Likewise, an occasional centric-ring
may survive, again usually very small so that "fall-free" separation always
happens. Larger rings are very unstable with respect to size, and the positive
selection pressure towards very small rings soon eliminates the big
Most of the above comments apply to the situation in normal individuals
and cells. When we turn to cancer-derived cells, or to transformed cell lines
growing in culture, the situation is somewhat different. These cells are
inherently chromosomally unstable. There is a continuous production of
structural change so that new primary changes are superimposed on the
already existing background of secondary aberrations, and these new
ones, in their turn, become secondary.
Moreover, some of the new changes are being produced in already abnormal
chromosomes, so the observed aberrations are often very complicated and
On top of this, it is clear that most cancer cells are very tolerant of
chromosomal loss, or gain, as is evidenced by considerable numerical variations
and multiple chromosome copies. These facts make cancer cytogenetics a very
difficult and uncertain field for investigation, and considerable credit goes
to those workers whose careful and painstaking efforts have produced meaningful