British Journal of Radiology (2007) 80, S63-S67
© 2007 British Institute of Radiology
doi: 10.1259/bjr/24766640
Reconstruction of doses from ionizing radiation using fluorescence in situ hybridization techniques
A A Edwards, MSc
M Sz
ui
ska, PhD
and
D C Lloyd, PhD
Health Protection Agency, Radiation Protection Division, Chilton, Didcot OX11 0RQ, UK
Correspondence: David Lloyd, Health Protection Agency, Radiation Protection Division, Chilton, Didcot OX11 0RQ, UK. E-mail: david.lloyd{at}hpa.org.uk
 |
Abstract
|
|---|
This paper reviews cytogenetic methods of biological dosimetry. The most reliable indicator of exposure to ionizing radiation is the observation of dicentrics in human peripheral lymphocytes. The major disadvantage is that dicentrics cannot be used for exposures that occur many years prior to blood sampling. In such cases, translocations are the aberrations of choice, and recent developments in their measurement using fluorescence in situ hybridization techniques are highlighted.
|
Introduction
|
|---|
For many years, ionizing radiation has been known to damage cells, and one well-documented effect is alterations in the structure of chromosomes. The particular rearrangements formed depend on the stage of the cell cycle that is irradiated. If the irradiation takes place during G2 or S phase, then chromatid-type aberrations are seen at the following metaphase. In the case of G1 irradiation, when repair is complete before the cell enters the S phase, chromosome-type aberrations are seen. The latter may appear as dicentrics, centric rings, acentric fragments, translocations, insertions and several other forms. These aberrations can be looked upon as products from exchanges of material between chromosomes or within the same chromosome. The exchanges are considered simple if the aberrations are apparently the result of only two breaks, and complex if more than two are involved.
The purpose of this paper is to review how the observation of these aberrations may be used for the purpose of biological dosimetry and, therefore, be of help in radiological protection.
|
Methods
|
|---|
For biological dosimetry purposes, a cell is required that is mature, has a long residence time in the body and can be sampled with as little discomfort as possible. The T lymphocyte fulfils all the above criteria, as the cell is distributed all around the body, is in the G0 stage of the cell cycle and undergoes no further divisions when mature. After an exposure to ionizing radiation, a blood specimen of about 10 ml is taken. Cell culture techniques that enable the chromosomes to be observed have been described in detail in [1]. Briefly, phytohaemagglutinin is added to stimulate the cells to divide and, after 45 h of culture at 37°C, Colcemid is added to stop the cells at metaphase. At about 48 h, metaphase spreads are made on slides according to standard techniques. At the beginning of culture, bromodeoxyuridine (BrdU) is added so that cells in their second division may be distinguished from those in their first division by the "harlequin" appearance of the chromosomes when stained with the fluorescence plus Giemsa method. For biological dosimetry, several hundreds or even thousands of cells need to be scored, and precise details of what is regarded as a scorable cell and how to score the aberrations depend on the staining technique. When the blood sample is taken up to about 1 year following exposure, the aberrations of interest are dicentrics, which are easily seen using solid staining of the chromosomes with Giemsa. The time limitation arises because cells containing dicentrics are removed from the peripheral pool of lymphocytes with a half-time of approximately 3 years. The advantage of using dicentrics is that the spontaneous yield is low, of the order of 1 in 1000 or 2000 cells, and a dose of 0.1 Gy of 
-rays would produce about two dicentrics in 1000 cells [2]. It is important to score cells only in their first metaphase because early research demonstrated that, when cells pass into the next cycle, those carrying chromosomal damage are more likely to be eliminated and so the aberration frequency is lowered [3]. The normal practice is to score only cells with the full chromosome complement visible, i.e. there must be exactly 46 visible centromeres. In addition, any aberrations seen must be complete, that is a dicentric or a centric ring must be accompanied by a fragment. Some laboratories incorporate centric rings in their calibration curves but, because it is generally believed that they occur rarely (one ring for every 10–20 dicentrics), their inclusion makes little practical difference. However, there are occasional reports of one centric ring for every three dicentrics [4], but the origin of this is presently unclear. This observation could be related to exposures occurring a long time before blood sampling or to long-term, low dose rate exposures because a centric ring may be more able to negotiate cell division in vivo than a dicentric. Thus, for biological dosimetry purposes, it would be more reasonable to restrict analysis to dicentrics alone.
|
Statistical analysis and interpretation
|
|---|
Calibrations are made by irradiating blood samples with a range of doses and, provided fresh blood is used and held at 37°C, the dicentric yield per unit dose is the same in vitro as it is for in vivo irradiations. The in vitro irradiation conditions must ensure that exposure of blood cells is uniform. Dose build-up and absorption in the blood sample, scattering of photons and the inverse square law are all relevant to the design of the irradiation procedure [1].
Dose–effect relationships fit well the following equation:
where Y is the mean yield of dicentrics per cell, D is the absorbed dose and C, 
and 
are fitted coefficients. The spontaneous level is C, the initial slope at low doses is , and describes the curvature at higher doses. Typical values of the coefficients for dicentric yield, derived from both measurements and interpolation between measurements, are given in Table 1
for a variety of radiation qualities [2]. The coefficients and both vary with radiation quality, and values are for acute exposure, i.e. when the duration of exposure is less than 0.5 h. For continuous exposures lasting a day or more, remains as in Table 1, but is conventionally set to zero. For intermediate times, an adjustment needs to be made based on Lea's formula [5]:
|
View this table:
[in this window]
[in a new window]
|
Table 1. Suggested calibration coefficients for dicentric yield for common radiations with estimates of standard error. The coefficient is to be used for acute exposure only (see text). Based on reference [2]
|
|
where x = t/T0, t is the time of continuous exposure and T0 = 2 h. Then, the coefficient , shown in Table 1
, is multiplied by a factor G given by Equation (2), which is dependent on the time over which the exposure occurs.
In most high-dose accidents, victims approach close to the source, producing non-uniform exposure of the body. In these situations, which are often accompanied by skin burns, the measurement of dicentrics can detect the non-homogeneity by analysis of the distribution of dicentrics among cells. For uniform exposure, the distribution should be Poisson, while a wider distribution suggests a non-uniform irradiation. The extent of non-uniformity of exposure can be inferred from the difference between the Poisson and the wider distributions. Details of methods of analysis appear in [1, 6].
|
Doses from past exposures
|
|---|
Two methods of chromosome aberration dosimetry may be used when doses are received more than about 1 year prior to blood sampling. The first is to measure dicentric yield and correct for the loss of cells containing dicentrics when the date of exposure is well defined, assuming a disappearance half-life of 3 years, although there is interindividual variation depending on factors such as infections. In particular, when high doses sufficient to cause significant reductions in white cell count are involved, the disappearance rate of dicentrics is shortened with half-lives lower than 1 year being reported [7]. The second is to measure yields of translocations in lymphocytes, which persist through cell division. For exposures that have occurred more than about 10 years prior to sampling, the cells scored are predominantly daughter cells of stem cells thought to exist in the bone marrow at the time of irradiation. For times between 1 year and 10 years, a mixture of stem cell descendants and long-lived mature lymphocytes is assayed. The latter could, for example, be isoforms of CD45 memory T cells [8]. Any variability in sensitivity between lymphocytes and bone marrow stem cells, as well as any loss of aberrations during the passage through an unknown number of cell divisions, are possible confounding factors. This would pose a problem for calibration, as only cells able to pass through multiple cell division are generally observed. The failure to pass through cell division is a more serious calibration problem following neutron exposure and, for that reason, neutrons are excluded from the following discussion.
The advent of fluorescence in situ hybridization (FISH) or "chromosome painting" has simplified the scoring of translocations compared with the earlier chromosome banding techniques, allowing the transition of translocations through cell division to be investigated. Owing to the observation of complex rearrangements, a characteristic of high doses of low linear energy transfer (LET) radiation and of high LET, additional difficulties arise with respect to the interpretation of data for retrospective biological dosimetry. A special issue of a journal [9] gives preliminary views of the laboratories attempting to solve these problems. More considered views, taking into account subsequent information, have been published and represent the views of European laboratories [10, 11]. These are summarized in the following sections.
What to paint
To date, there is broad agreement that the centromeres of each chromosome should be clearly visible to distinguish dicentrics from translocations. In the view of some laboratories, the centromere can be identified because the two chromatids of each chromosome and the single indentation ("waist") are clearly visible, while in others, the extra brightness of the centromere using the 4'6-diamidino-2-phenylindole•2HCl (DAPI) filter is sufficient for its location. Alternatively, the centromere can be specifically FISH painted. There is evidence to show that it does not matter which of the chromosomes are painted because each one takes part in exchanges with a probability approximately proportional to its DNA content [12]. Therefore, it is common practice to paint three of the larger chromosomes in groups 1–12, with about 20% of the genome being highlighted, which means that about 30% of all translocations are seen. Thus, the yield in the full genome can be calculated and is often expressed as translocations per genome-equivalent cell [13]. Hence the pooling of data from several experiments is possible [11].
What to score
There is now a consensus that translocations should be scored in stable cells only. The judgement of stable cells relies upon the absence of dicentrics, acentrics or centric rings anywhere in the metaphase. As about 3000 cells would be required in order to obtain a reasonable estimate of spontaneous yield in control subjects, it would be impracticable to select cells and score them using the same criteria as for dicentrics. Clearly, a more rapid and, therefore, less rigorous approach needs to be adopted, and the following criteria have been developed. At low microscope magnification, a scorable cell should, at a glance, contain about 46 chromosomes in each metaphase, and all the painted chromosomes (usually six) should be present. At high magnification, any apparently shortened painted chromosomes should be accompanied by the deleted fragment somewhere in the cell (i.e. the total length of each painted chromosome is present somewhere in the cell). If cells are included where painted material is absent, there is an increased likelihood that they also contain translocations and, as these are also a feature of control subjects, they are unlikely to be related to radiation exposure and should thus be excluded from the analysis [14].
For recording and communication purposes, a modification of the PAINT system of nomenclature [15] is used, as opposed to a more complex alternative known as S&S [16]. For the PAINT system, aberrations are described in terms of letters, with A or a referring to counterstained chromosomes and B or b to a painted chromosome. A capital letter indicates that the piece has a centromere and a lower case shows it does not. The abbreviations, dic, t and ace refer to dicentrics, translocations and acentrics, respectively. Thus, dic(AB) means a dicentric between a painted and an unpainted chromosome, and it is usually accompanied by ace(ab). The corresponding reciprocal translocation is described as [t(Ab) + t(Ba)]. Initially, translocations were described as reciprocal and terminal or complete and incomplete, which all have mechanistic connotations. For this reason, the terms two-way and one-way have evolved, and their use is intended to be purely visual without any implied mechanism. It is now believed that many radiation-induced translocations seen as one-way exchanges are in reality reciprocal [17].
Persistence of translocations
For retrospective biological dosimetry by FISH, the lymphocytes examined at the time of blood sampling were stem cells at the time of irradiation and, therefore, had gone through many cell divisions. Hence cells that had unstable aberrations disappeared because the daughter cells were not viable. If translocations appear preferentially in unstable cells, their yield would reduce (i.e. persistence would be compromised). Alternatively, if the production of aberrations in bone marrow cells differs from that in mature lymphocytes, there would then also be a change in yield with time. Experiments that followed cells through several cycles suggest that translocations are very stable following low LET irradiation, particularly the two-way variety [18]. There is some evidence that the one-way types decrease because they are either associated with complex rearrangements, most of which are unstable, or they are incomplete exchanges with associated fragments and so the cell as a whole is unstable. A follow up of the victims of the Goianian accident shows a large reduction (about a factor of 3) in translocation yield compared with dicentric yield measured shortly after the accident [19]; however, no attempt was made to distinguish one-way and two-way translocations. The observation applied to about 20 subjects, exposed non-uniformly, who had estimated average whole body doses from about 0.5 Gy to 4 Gy. It should be noted that there is in vitro evidence that, in simulated part body exposures, both types of translocation yield reduce because translocations are confined to the exposed group of cells, which are more likely to be unstable by virtue of also containing dicentrics and fragments [20]. A follow up of three people following an accident in Estonia shows that translocations decrease initially with time and then level off [21]. However, when translocations are measured in stable cells only, there is no appreciable loss of translocation yield with time [22]. Additionally, a follow up of a tritiated water intake by a worker has shown no decrease in translocation yield with respect to the original dicentric yield for either one- or two-way translocations over a period of 11 years [23]. Therefore, it seems that the translocation yield is persistent with time provided it is measured in genomically stable cells.
Most radiation accidents, especially where high doses are received, should be appreciated promptly, well within the time-frame for using the dicentric assay. The main application for FISH is thus envisaged as assessing long-term cumulative exposure at lower dose rate. It therefore becomes a useful tool in dose reconstruction to facilitate, for example, epidemiological studies such as those currently under way around Chernobyl and in the Urals.
Control levels
Control levels of translocations have been measured and reviewed [11, 24]. In Figure 1, the combined data from eight laboratories for the dependence of control translocation levels on age measured with FISH are shown. While it is clear that the spontaneous frequency for translocations is confounded markedly by age, it is not so for dicentrics, as the rate of dicentric formation is counterbalanced by the rate of their removal. Therefore, age needs to be accounted for in FISH dosimetry. However, some extra variation in translocation yield between unexposed individuals exists, even after correction for age [11, 24]. Despite conflicting published data on the effects of smoking [9], some authors [11] have concluded that smoking habits are not the cause.
View larger version (20K):
[in this window]
[in a new window]
|
Figure 1. Dependence of the background level of stable translocations on age measured with fluorescencein situ hybridization (FISH) [11].
|
|
Calibration
As has been pointed out previously, there is no guarantee that the induced yield of translocations per unit dose in stem cells is the same as in mature lymphocytes, but more recent experiments have indicated that translocation yields are stable in vivo provided they are measured in stable cells. This gives confidence that there is no loss of yield due to cell division or from a sensitivity mismatch between stem cells and mature lymphocytes. Therefore, mature lymphocytes may provisionally be used for calibration purposes [25]. Retrospective dosimetry usually involves chronic exposure, for which the quadratic term in Equation (1) is essentially zero. Therefore, the main difficulty is to score enough cells at low doses to define the linear term in Equation (1), which is the most important for retrospective dosimetry. The data available so far are too scarce to determine the linear term in this way with sufficient accuracy. In addition, the stability of the scored cells was generally not recorded, as much of the published data predate the current criteria and so the requirement to score stable cells may not be fulfilled. This being the case, the linear term for translocations may be provisionally derived from the relatively well-known linear term for dicentrics because the two aberration types are stable and unstable homologues induced initially with a ratio very close to 1.0.
Sensitivity
Based on knowledge to date and the assumptions about calibration, an estimate of the lowest dose that can be measured by translocations can be made. Converted to genome equivalents, the translocation yield (one-way and two-way) produced by 1 Gy of chronic irradiation from cobalt-60 is about 15 per 1000 cells. For a 60 year old, the control level is about 10 per 1000 cells. This means that, by scoring 1000 genome-equivalent cells, a dose of about 0.5 Gy can be detected. The pre-exposure control level in any one individual is of course not known, and so it has to be inferred from a generic database. For a young person, less than 20 years old in whom the assumed control yield is around three translocations per 1000 genome-equivalent cells, the detection limit reduces to about 0.3 Gy.
|
Conclusions
|
|---|
There are decades of experience to show that the measurement of dicentrics in human lymphocytes produces a reasonable estimate of absorbed dose to a person. The lower limit of detection is about 0.1 Gy of -rays, and the highest dose that can be measured is well into the personal lethal range. At the higher doses, partial body exposures can be detected, and some correction can be made to account for time since exposure. For exposures that occur some years prior to blood sampling and will usually be delivered over a long period of time, the dicentric assay is inadequate. In this case, the scoring of translocations proves useful, and the system using FISH has been developed into a routine method. Translocation frequencies persist with time provided they are measured in stable cells. More data, particularly from the follow up of accidental exposure to humans, are required to confirm this. There is still outstanding variation in translocation yields in control subjects even after the correction for age, so that more work on other potential confounders is warranted.
Received for publication May 12, 2006.
Revision received August 29, 2006.
Accepted for publication September 28, 2006.
|
References
|
|---|
- International Atomic Energy Agency. Cytogenetic analysis for radiation dose assessment. A manual. Technical Report Series no. 405. Vienna: IAEA, 2001
- Edwards AA. The use of chromosomal aberrations in human lymphocytes for biological dosimetry. Radiat Res 1997;148:S39–S44.[Medline]
- Wagner R, Schmid E, Bauchinger M. Application of conventional and FPG staining for the analysis of chromosome aberrations induced by low levels of dose in human lymphocytes. Mutat Res 1983;109:65–71.[Medline]
- Sevankaev AV, Lloyd DC, Poletnya OI, Zhloba AA, Moiseenko VV, Edwards AA. Chromosomal aberrations in lymphocytes of residents of area contaminated by radioactive discharges from the Chernobyl accident. Radiat Prot Dosim 1995;58:247–54.[Abstract]
- Lea DE, Catcheside DG. The mechanism of the induction by radiation of chromosome aberrations in Tradescantia. J Genet 1942;44:168–74.
- Szuiska M, Edwards AA, Lloyd DC. Statistical methods for biological dosimetry. HPA document. HPA-RPD-011. Chislton: Health Protection Agency, 2006
- Buckton KE. Chromosome aberrations in patients treated with x-irradiation for ankylosing spondylitis. In: Ishihara T, Sasaki MS, editors. Radiation induced damage in man. New York, NY: Alan R Liss, 1983: 491–511
- Mitchie CA, MacLean A. Lymphocytes longevity: a coincidence of approaches. Int J Radiat Biol 1994;66:229[CrossRef][Medline]
- Edwards AA. (editor) Fluorescence in situ hybridisation (FISH) biological dosimetry. Radiat Prot Dosim 2000;88(1):
- Edwards AA, Lindholm C, Darroudi F, Stephan G, Romm H, Barquinero J, et al. Review of translocations detected by FISH for retrospective biological dosimetry applications. Radiat Prot Dosim 2005;113:396–402.[Abstract/Free Full Text]
- Whitehouse CA, Edwards AA, Tawn EJ, Stephan G, Oestreicher U, Moquet JE, et al. Translocation yields in peripheral blood lymphocytes from control populations. Int J Radiat Biol 2005;81:139–45.[CrossRef][Medline]
- Cigarran S, Barrios L, Barquinero JF, Caballin MR, Ribas M, Egozcue J. Relationship between the DNA content of human chromosomes and their involvement in radiation-induced structural aberrations, analysed by painting. Int J Radiat Biol 1998;74:449–55.
- Lucas JN, Deng W. Views on issues in radiation biodosimetry based on chromosome aberrations measured by FISH. Radiat Prot Dosim 2000;88:77–86.[Abstract]
- Edwards AA, Maznik N, Moquet J, Hone P, Vinnikov V, Lloyd DC, et al. Choosing metaphases for biological dosimetry by fluorescence in situ hybridization (FISH). Radiat Res 2002;157:469–71.[CrossRef][Medline]
- Tucker JD, Morgan WF, Awa AA, Bauchinger M, Blakey D, Cornforth MN, et al. Proposed system for scoring structural aberrations detected by chromosome painting. Cytogenet Cell Genet 1995;68:211–21.[Medline]
- Savage JRK, Simpson PJ. FISH "painting" patterns resulting from complex exchanges. Mutat Res 1994;312:51–60.[Medline]
- Boei JJWA, Vermeulen S, Fomina J, Natarajan AT. Detection of incomplete exchanges and interstitial fragments in X-irradiated human lymphocytes using a telomeric PNA probe. Int J Radiat Biol 1998;73:599–603.[CrossRef][Medline]
- Darroudi F. The use of FISH translocations analyses for retrospective biological dosimetry: How stable are stable chromosome aberrations? Radiat Prot Dosim 2000;88:101–9.[Abstract]
- Natarajan AT, Santos SJ, Darroudi F, Hadjidikova V, Vermeulen S, Chatterjee S, et al. 137Cesium-induced chromosome aberrations analyzed by fluorescence in situ hybridization: eight years follow-up of the Goiania radiation accident victims. Mutat Res 1998;400:299–312.[Medline]
- Guerrero-Carbajal YC, Moquet JE, Edwards AA, Lloyd DC. The persistence of FISH translocations for retrospective biological dosimetry after simulated whole or partial body irradiation. Radiat Prot Dosim 1998;76:159–68.[Abstract]
- Lindholm C, Romm H, Stephan G, Schmids E, Moquet J, Edwards A. Intercomparison of translocations and dicentric frequencies between laboratories on a follow-up of the radiological accident in Estonia. Int J Radiat Biol 2002;78:883–90.[CrossRef][Medline]
- Lindholm C, Edwards A. Long-term persistence of translocations in stable lymphocytes from victims of a radiological accident. Int J Radiat Biol 2004;80:559–66.[CrossRef][Medline]
- Lloyd DC, Moquet JE, Oram S, Edwards AA, Lucas JN. Accidental intake of tritiated water: a cytogenetic follow-up case on translocation stability and dose reconstruction. Int J Radiat Biol 1998;73:543–7.[CrossRef][Medline]
- Sorokine-Durm I, Whitehouse C, Edwards AA. The variability of translocation yields amongst control populations. Radiat Prot Dosim 2000;88:93–9.[Abstract]
- Savage JRK, Papworth DG, Bauchinger M, Natarajan AT, Pantelias GE, Griffin CS, et al. Constructing a 2B calibration curve for retrospective dose reconstruction. Radiat Prot Dosim 2000;88:69–76.[Abstract]
Top
Abstract
Introduction
