Analysis of α-particle-induced chromosomal aberrations by chemically-induced PCC. Elaboration of dose-effect curves

Abstract

In a similar way to high-dose exposures to low-LET radiations, cells show difficulties reaching mitosis after high-LET radiation exposure. For this reason, techniques have been proposed that are able to analyze chromosome aberrations in interphase by prematurely condensing the chromosomes (PCC-techniques). Few dose-effect curves for high-LET radiation types have been reported, and none for α-particles. The aim of this study was to evaluate, by chemically-induced PCC, the chromosome aberrations induced by several doses of α-particles. Monolayers of peripheral lymphocytes were exposed to an α-source of Americium-241 with a mean energy entering the cells of 2.7 MeV. Lymphocytes were exposed to 10 doses, from 0–2.5 Gy, and then cultured for 48 h. Colcemid and Calyculin-A were added at 24 and 1 h before harvesting, respectively. During microscope analysis, chromosome rings and extra chromosome pieces were scored in G2/M-PCC and M cells, while dicentric chromosomes were only scored in M cells. As the dose increased, fewer cells were able to reach mitosis and the proportion of G2/M-PCC cells increased. Chromosome rings were hardly observed in M cells when compared to G2/M-PCC cells. Extra fragments were more frequent than rings in both G2/M-PCC and M cells, but with lower frequencies than in G2/M-PCC cells. The distribution of dicentrics and extra fragments showed a clear overdispersion; this was not so evident for rings. The dose-effect curves obtained fitted very well to a linear model. Damaged cells after α-particle irradiation show more difficulties in reaching mitosis than cells exposed to γ-rays. After α-particle irradiation the frequency of all the chromosome aberrations considered increased linearly with the dose, and α-particles clearly produced more dicentrics and extra chromosome pieces with respect to γ-rays. After α-particle exposure, the existence of extra chromosome fragments in PCC cells seems to be a good candidate for use as a biomarker for dose assessment. However, the observed frequencies of different types of chromosomal aberrations could be influenced by some methodological aspects; for this reason, and in order to avoid possible methodological bias, standardization of the technique will be desirable.

Keywords:

• Alpha emitters
• biological dosimetry
• chromosome abberations
• PCC

Introduction

Dose assessment after exposures to ionizing radiation (IR) has long been performed by the analysis of chromosome aberrations in mitotic cells. The analysis of dicentric chromosomes in peripheral blood lymphocyte metaphases has allowed the development of biological dosimetry and is the method of choice to be used in suspected overexposures to IR (International Atomic Energy Agency [IAEA] 2011). Peripheral blood lymphocytes are usually in the resting phase G0, and must be stimulated to reach metaphase. During this process, the cells must overcome different cell cycle checkpoints that control adequate progression. When cells show DNA damage, the DNA repair machinery is activated with other signaling pathways, inducing cell progression delay and triggering programmed cell death if necessary. Among the checkpoints, the G2/M one ensures cells do not enter mitosis before they have the chance to repair the damaged DNA (Jeggo and Löbrich 2006; Niida and Nakanishi 2006). Considering the difficulty highly damaged cells have in overcoming the G2/M checkpoint and hence in reaching mitosis, for biological dosimetry purposes it is difficult to assess exposures higher than 4–5 Gy (IAEA 2011), because at higher doses only a few cells can reach mitosis. A delay in cell cycle progression, mainly due to a prolonged G2-arrest, has also been described for high-LET radiation exposures (Lee et al. 2011). For this reason, the standard dicentric analysis would considerably underestimate the effect of a high-LET irradiation (Lee et al. 2011).

In cases where dose-assessment at higher doses of low-LET radiation or after high-LET radiation is needed, the evaluation of chromosome aberrations in interphase cells using Premature Chromosome Condensation (PCC) techniques has been proposed. PCC-techniques can be carried out by fusing unstimulated peripheral blood lymphocytes with mitotic CHO cells, allowing an estimation of the dose immediately after blood collection (Pantelias and Maillie 1983). Another way to obtain interphase chromosomes is the chemical induction of PCC using Calyculin-A or okadaik acid (Durante et al. 1998; Kanda et al. 1999), although in this case cells need to be stimulated to growth. Both methods, fusion and chemically induced PCC, are now accepted for biological dosimetry (IAEA 2011). Chemically PCC has been applied in biodosimetry in cases of accidental exposures to a high dose of IR (Hayata et al. 2001; Yao et al. 2013) and several dose-effect calibration curves for low-LET radiation have been published using different biomarkers such as chromosome rings or extra chromosome pieces (Lamadrid et al. 2007; Balakrishnan et al. 2010; Lindholm et al. 2010; Puig et al. 2013). For high-LET radiation few dose-effect curves based on PCC have been published, and none for α-particle irradiation (Lamadrid et al. 2007; 2011; Wang et al. 2007; Lee et al. 2011). The aim of the present study was to analyze, by chemically-induced PCC, the chromosome aberrations induced at several doses of α-particle irradiation and to construct dose-effect curves.

Materials and methods

Peripheral blood lymphocytes isolation and culture

Peripheral blood samples were obtained by venipuncture from a 27-year-old female with no history of exposure to ionizing radiation or clastogenic agents. This project was approved by the Animal and Human Experimentation Ethics Committee of the Universitat Autònoma de Barcelona (Reference: 2624). Lymphocyte isolation, culture and irradiation were performed as previously described (Schmid et al. 1996). Briefly, lymphocytes were isolated from whole blood using the Percoll (Biochrom, Berlin, Germany) gradient centrifugation technique. Isolated cells were washed twice in Roswell Park Memorial Institute (RMPI)-1640 medium (Biochrom) and resuspended in the same medium at a final concentration of 5 × 10⁶ lymphocytes per ml. A total of 0.5 ml of the lymphocyte suspension and 4.5 ml of RPMI-1640 medium, supplemented with 16% (v/v) of fetal calf serum (Biochrom), 27 μg of phytohaemagglutinin (PHA) (Biochrom) and (100 U.I. ml⁻¹ penicillin, 100 mg·ml⁻¹ streptomycin), were mixed and incubated for 3 h at 37 °C, in special dishes containing a stainless steel ring and a 2 μm thick Mylar foil (Goodfellow, Bad Nauheim, Germany), to allow lymphocyte attachment. Immediately before irradiation the residual medium was carefully drawn off, and monolayers of human peripheral lymphocytes were irradiated (Schmid et al. 1996; Regulla et al. 2002).

Irradiation procedure

The monolayers of attached lymphocytes were irradiated with α-particles from an Americium-241 source located at the Helmholtz Zentrum München (Neuherberg, Germany). The irradiation device was designed by Roos and Kellerer (1989) and has already been employed in studies of chromosome analysis in human lymphocytes using fluorescence plus Giemsa staining or fluorescence in situ hybridization techniques, resulting in a high degree of accuracy and reproducibility (Schmid et al. 1996; Barquinero et al. 2004; Schmid and Roos 2009). Details of the source are fully described in Schmid et al. (1996). During irradiations the most probable energy of α-particles entering the cells was 2.7 MeV with a dose-mean LET of 150 keV·μm⁻¹. Monolayers of human peripheral blood lymphocytes were exposed at 10 doses: 0, 0.05, 0.1, 0.2, 0.5, 0.7, 1, 1.5, 2 and 2.5 Gy of α-particles at a dose-rate of 0.1 Gy·min⁻¹. Although during the exposure the lymphocytes were kept in the chamber at 37 °C, remaining without medium during irradiation. Due to the dose-rate and the irradiation conditions it was not possible to obtain chromosome spreads at higher doses.

Cell culture and chromosome analysis

After irradiation, the lymphocytes were scraped off and cultured in supplemented RPMI-1640 medium for 48 h. To obtain metaphases and PCC spreads in the same sample, Colcemid (Boehringer Manheim, Manheim, Germany) was added after 24 h of culture at a final concentration of 0.1 μg·ml⁻¹, and 1 h before harvesting Calyculin-A (LC Laboratories, Interlabo, Spain) was added at a final concentration 50 nM. The cells then were treated with hypotonic solution (KCl, 0.075 M) (Sigma-Aldrich, Munich, Germany) for 20 min at 37 °C and fixed with Carnoy’s solution (methanol:acetic acid, 3:1, v/v) (Sigma-Aldrich). Once cells were dropped onto slides and air-dried, the slides were placed for 24 h at 37 °C. The staining was done with Leishman’s stain (Merck, Madrid, Spain).

Chromosome spreads were analyzed with a Zeiss Axio Imager.Z2 microscope (Izasa, Barcelona, Spain) coupled to a Metafer^® Slide Scanning System V3.8 (Izasa, Barcelona, Spain). Cytogenetic analysis was performed as previously described (Puig et al. 2013). In brief, each cell was classified as a G2/M-PCC cell or M cell according to the absence and presence of the centromeric constriction respectively, and it was assumed that G2/M-PCC cells were in G2 phase of the cell cycle and M cells in metaphase. Only cells with 46 or more pieces were considered. The number of chromosome rings and the number of extra chromosome fragments was recorded in both cell types. A chromosome ring was only considered when a corresponding fragment was present. Other fragments exceeding 46 and not related to a ring were considered as extra fragments. Dicentric chromosomes were only recorded in M cells.

Statistical analysis

To test if there were differences in the frequencies of chromosome aberrations between G2/M-PCC and M cells, the Wilcoxon paired test was used. To check the Poisson distribution of rings and extra fragments in the cells, the u-test was used. Values of u outside the interval ±1.96 indicated that the aberrations did not follow a Poisson distribution (Papworth 1975). Dose-effect curves were fitted to a linear model by means of the weighted least-squares approximation. For rings the inverse of the mean was used as the weighting factor because departures from the Poisson distribution were not observed. For extra fragments and dicentrics the reciprocal of the estimated variance was used as the weighting factor because overdispersion was present at all doses. Differences in α coefficients were tested by means of the standard difference test (Barquinero et al. 1999).

Results

To evaluate the cytogenetic effect of α-particle irradiation by PCC, a total of 11,327 cells were analyzed, including cells in G2 phase (G2/M-PCC cells) and cells in metaphase (M cells). As can be seen in Figure 1, the proportion of M cells in relation to total cells analyzed decreased as the dose increased, from 85% at 0 Gy to 1% at 2.5 Gy. Inversely, the proportion of G2/M-PCC cells increased with the dose, from 15% at 0 Gy to 99% at 2.5 Gy. At the lowest doses of irradiation, 0.05–0.1 Gy, percentages of M and G2/M-PCC cells were similar.

Figure 1. Percentages of M and G2/M-PCC cells at each dose. The error bars indicate the standard error of the mean.

For each dose and cell type the number of rings, and their distribution among cells, are shown in Table 1. At 0 and 0.05 Gy no rings were observed, and from 0.01–2.5 Gy the frequency of chromosome rings observed in G2/M-PCC cells was significantly higher than that observed in M cells (p = 0.0078). In fact in M cells rings were only observed after 0.5, 1 and 1.5 Gy. When G2/M-PCC cells and M cells were considered together as total cells, the frequency of rings increased with dose. Ring distribution among cells was in almost all cases in agreement with the Poisson distribution, and only at 1 and 2.5 Gy were u values indicative of a significant overdispersion. Dose-effect curves for ring chromosomes were established for G2/M-PCC and total cells (Figure 2). The linear relationship and the coefficients obtained are shown in Figure 2. Because ring chromosomes were hardly observed in M cells, the coefficients obtained for G2/M-PCC and total cells are almost the same. M cells did not show enough ring chromosomes to establish a dose-effect relationship.

Figure 2. Linear relationship for chromosome rings. Observed frequencies (±SE) in G2/M-PCC and total cells (G2/M-PCC and M cells together), error bars indicate the standard error of the mean. The values (±SE) of the coefficients are: y = 0.017(± 0.002)D for G2/M PCC cells, and y = 0.017(± 0.002)D for total cells.

Table 1. Cell distribution of chromosome rings for each dose and cell type.

			Cells with X chromosome rings
Dose (Gy)	Cell type	Scored cells	0	1	2	3	4	Total rings	y ± SE	σ^2	DI	u Value
0 Gy	M	977	977	0	0	0	0	0	–	–	–	–
	G2/M-PCC	167	167	0	0	0	0	0	–	–	–	–
	Total	1144	1144	0	0	0	0	0	–	–	–	–
0.05 Gy	M	839	829	0	0	0	0	0	–	–	–	–
	G2/M-PCC	1069	1069	0	0	0	0	0	–	–	–	–
	Total	1908	1898	0	0	0	0	0	–	–	–	–
0.1 Gy	M	724	724	0	0	0	0	0	–	–	–	–
	G2/M-PCC	895	893	2	0	0	0	2	0.002 (±0.002)	0.002	0.999	−0.033
	Total	1619	1617	2	0	0	0	2	0.001 (±0.001)	0.001	0.999	−0.025
0.2 Gy	M	258	258	0	0	0	0	0	–	–	–	–
	G2/M-PCC	1210	1205	5	0	0	0	5	0.004 (±0.002)	0.004	0.997	−0.091
	Total	1468	1463	5	0	0	0	5	0.003 (±0.002)	0.003	0.997	−0.083
0.5 Gy	M	386	385	1	0	0	0	1	0.003 (±0.003)	0.003	1	–
	G2/M-PCC	833	870	11	0	0	0	11	0.013 (±0.004)	0.013	0.988	−0.257
	Total	1219	1255	12	0	0	0	12	0.010 (±0.003)	0.010	0.991	−0.233
0.7 Gy	M	104	104	0	0	0	0	0	–	–	–	–
	G2/M-PCC	1001	989	12	0	0	0	12	0.012 (±0.003)	0.012	0.989	−0.257
	Total	1105	1093	12	0	0	0	12	0.011 (±0.003)	0.011	0.990	−0.245
1 Gy	M	166	165	1	0	0	0	1	0.006 (±0.006)	0.006	–	–
	G2/M-PCC	663	653	7	1	0	0	9	0.014 (±0.005)	0.016	1.210	4.062
	Total	829	818	8	1	0	0	10	0.012 (±0.004)	0.014	1.189	4.062
1.5 Gy	M	80	79	1	0	0	0	1	0.013 (±0.013)	0.013	1	–
	G2/M-PCC	848	827	20	1	0	0	22	0.026 (±0.006)	0.028	1.066	1.395
	Total	928	906	21	1	0	0	23	0.025 (±0.005)	0.026	1.063	1.394
2 Gy	M	24	24	0	0	0	0	0	–	–	–	–
	G2/M-PCC	617	596	21	0	0	0	21	0.034 (±0.007)	0.033	0.968	−0.584
	Total	641	620	21	0	0	0	21	0.033 (±0.007)	0.032	0.969	−0.573
2.5 Gy	M	3	3	0	0	0	0	0	–	–	–	–
	G2/M-PCC	463	447	14	2	0	0	18	0.039 (±0.010)	0.046	1.186	2.907
	Total	466	450	14	2	0	0	18	0.039 (±0.010)	0.046	1.186	2.921

Metaphase cells (M), prematurely condensed chromosome cells in G2/M (G2/M-PCC) and total cells (M plus G2/M-PCC cells), frequency ± standard error (y ± SE), variance (σ²), dispersion index DI (σ²/y), and u values (u) to test the agreement with the Poisson distribution. u values greater than +1.96 indicate overdispersion.

Extra fragments were scored in G2/M-PCC cells and M cells (Table 2). Their frequency clearly increased with dose. For all doses the extra fragments frequency was significantly higher in G2/M-PCC cells than in M cells (p = 0.002). As can be seen in Table 2, the distribution of extra fragments among cells for G2/M-PCC and M cells was overdispersed at all doses (u values > 1.96). For extra fragments, dose-effect curves were established for M cells, G2/M-PCC cells and total cells. The linear dose-effect relationships and the coefficients obtained are shown in Figure 3. The α-coefficients of the dose-effect curves for extra fragments for G2/M-PCC and for total cells did not differ statistically, but both were significantly higher than that for M cells (p < 0.001). Dicentric chromosomes were only recorded in M cells. As can be seen in Table 3, their frequency increased with the dose and their distribution showed a clear overdispersion. Due to the low number of M cells observed after 2 and 2.5 Gy irradiation, 24 and 3 respectively, these two doses were not considered in calculating the linear coefficient of the dose-effect curve Figure 4.

Figure 3. Linear relationships for extra fragments. Observed frequencies (±SE) of extra fragments in G2/M-PCC, M, and total cells (G2/M-PCC and M cells together), error bars indicate the standard error of the mean. The values (±SE) of the coefficients are: y = 0.865(± 0.018) + 1.527(± 0.025)D for G2/M-PCC cells, y = 0.018(± 0.03) + 0.2969(± 0.022)D for M cells, and y = 0.359(± 0.009) + 1.795(± 0.021)D for total cells.

Figure 4. Linear relationship for dicentric chromosomes. Observed frequencies (±SE) of dicentric chromosomes in M cells, error bars indicate the standard error of the mean. The values (±SE) of the coefficients are: y = 0.373(± 0.074)D.

Table 2. Cell distribution of chromosome fragments for each dose and cell type.

			Cells with X extra chromosome fragments
Dose (Gy)	Cell type	Scored cells	0	1	2	3	4	≥5	Total extra chromosome fragments	y ± SE	σ^2	DI	u Value
0 Gy	M	977	974	3	0	0	0	0	3	0.003 (±0.002)	0.003	0.998	−0.055
	G2/M-PCC	167	154	5	5	3	0	0	24	0.144 (±0.042)	0.292	2.035	9.633
	Total	1144	1128	8	5	3	0	0	27	0.024 (±0.006)	0.048	2.015	24.732
0.05 Gy	M	839	797	36	5	0	1	0	50	0.060 (±0.010)	0.082	1.382	7.900
	G2/M-PCC	1069	667	171	110	53	36	32	878	0.821 (±0.042)	1.924	2.343	31.043
	Total	1908	1464	207	115	53	37	32	928	0.486 (±0.026)	1.257	2.584	48.937
0.1 Gy	M	724	706	13	3	2	0	0	25	0.035 (±0.009)	0.058	1.688	13.347
	G2/M-PCC	895	666	87	59	45	18	20	529	0.591 (±0.042)	1.578	2.669	35.320
	Total	1619	1372	100	62	47	18	20	554	0.342 (±0.025)	0.974	2.847	52.590
0.2 Gy	M	258	225	18	5	6	2	2	64	0.248 (±0.049)	0.615	2.480	40.408
	G2/M-PCC	1210	401	263	191	149	84	122	2190	1.810 (±0.058)	4.020	2.221	30.031
	Total	1468	626	281	196	155	86	124	2254	1.535 (±0.051)	3.774	2.458	39.503
0.5 Gy	M	386	360	12	10	2	1	1	49	0.127 (±0.029)	0.334	2.635	22.920
	G2/M-PCC	833	446	108	109	78	41	51	1027	1.233 (±0.060)	3.044	2.469	29.980
	Total	1219	806	120	119	80	42	52	1076	0.883 (±0.045)	2.450	2.776	43.841
0.7 Gy	M	104	90	10	4	0	0	0	18	0.173 (±0.046)	0.222	1.284	2.095
	G2/M-PCC	1001	247	176	154	128	111	185	2499	2.497 (±0.075)	5.688	2.278	28.593
	Total	1105	337	186	158	128	111	185	2517	2.278 (±0.071)	5.634	2.473	34.622
1 Gy	M	166	151	8	5	1	0	1	26	0.157 (±0.046)	0.351	2.241	11.500
	G2/M-PCC	663	215	119	100	68	54	107	1456	2.196 (±0.096)	6.085	2.771	32.232
	Total	829	366	127	105	69	54	108	1482	1.788 (±0.082)	5.602	3.134	43.430
1.5 Gy	M	80	63	7	7	2	1	0	31	0.388 (±0.095)	0.721	1.862	5.504
	G2/M-PCC	848	119	106	139	133	107	244	2892	3.410 (±0.093)	7.328	2.149	23.647
	Total	928	182	113	146	135	108	244	2923	3.150 (±0.090)	7.478	2.374	29.589
2 Gy	M	24	12	3	5	3	1	0	26	1.083 (±0.262)	1.645	1.518	1.793
	G2/M-PCC	617	101	86	92	79	66	193	2158	3.498 (±0.126)	9.718	2.778	31.219
	Total	641	113	89	97	82	67	193	2184	3.407 (±0.123)	9.623	2.824	32.642
2.5 Gy	M	3	1	0	2	0	0	0	4	1.333 (±0.667)	1.333	1.000	0.000
	G2/M-PCC	463	66	51	48	55	52	191	1972	4.259 (±0.157)	11.963	2.809	27.498
	Total	466	67	51	50	55	52	191	1976	4.240 (±0.160)	11.946	2.817	27.717

Metaphase cells (M), prematurely condensed chromosome cells in G2/M (G2/M-PCC) and total cells (M plus G2/M-PCC cells), frequency ± standard error (y ± SE), variance (σ²), dispersion index DI (σ²/y), and u values (u) to test the agreement with the Poisson distribution. u values greater than +1.96 indicate overdispersion.

Table 3. Cell distribution of dicentric chromosomes among cells in metaphase.

		Cells with X dicentrics
Dose (Gy)	Scored cells	0	1	2	3	4	Total dicentrics	y± SE	σ^2	DI	u Value
0	977	977	0	0	0	0	0	–	–	–	–
0.05	839	811	24	4	0	0	32	0.038 (±0.07)	0.046	1.213	6.692
0.1	724	714	7	2	1	0	14	0.019 (±0.07)	0.033	1.697	20.582
0.2	258	235	16	6	1	0	31	0.120 (±0.026)	0.176	1.466	12.834
0.5	386	349	23	11	3	0	54	0.140 (±0.024)	0.224	1.604	15.052
0.7	104	86	8	9	1	0	29	0.279 (±0.065)	0.436	1.564	13.480
1	166	141	15	7	3	0	38	0.229 (±0.047)	0.372	1.623	12.846
1.5	80	51	16	8	4	1	48	0.600 (±0.106)	0.901	1.502	10.924
2	24	10	3	7	2	2	31	1.292 (±0.272)	1.781	1.379	6.886
2.5	3	1	0	1	1	0	5	1.667 (±0.882)	2.333	1.400	6.819

Observed frequency ± standard error of the mean (y ± SE), variance (σ²), dispersion index DI (σ²/y), and u values (u) to test the agreement with the Poisson distribution. Values greater than +1.96 indicate overdispersion.

Discussion

By chemically-induced PCC several dose-effect curves for low-LET radiation have been published to date (Lamadrid et al. 2007; Balakrishnan et al. 2010; Lindholm et al. 2010; Puig et al. 2013), the main purpose of these curves being the dose-assessment after high-dose exposures. For high-LET radiation-types few curves have been reported using PCC (Lamadrid et al. 2007; 2011; Wang et al. 2007; Lee et al. 2011) and none of them for α-particles. In this sense the main purpose of this paper was to evaluate by PCC the chromosome aberrations induced by α-particle exposures and to establish dose-effect relationships using different cytogenetic biomarkers. The first observation of this study was the clear reduction in the proportion of M cells in respect to the G2/M-PCC cells as the dose increased. This situation is in accordance with previous studies done with low LET radiation (Kanda et al. 1999; Puig et al. 2013). For γ-rays, using blood samples from the same donor and with a similar protocol to chemically-induce PCC, the observed percentage of M cells with regard to total cells was 51.6% ± 1.1% and 3.4% ± 0.4% after 1 and 3 Gy of γ-irradiation, respectively (Puig et al. 2013). After α-particle irradiation the percentage of M cells observed at similar doses is 20% ± 1.4% and 0.6% ± 0.3% after 1 and 2.5 Gy irradiation. This result indicates that, when compared with γ-irradiation, cells irradiated at similar doses with α-particles show more difficulties in reaching mitosis. This is in agreement with previous results where a major effectiveness of α-particles in reducing the mitotic index was described (Lücke-Huhle et al. 1983; Ritter et al. 2002). This delay seems to be linked to the presence of cells with complex aberrations (George et al. 2003a). In fact, after different heavy-ion exposures and applying FISH chromosome painting to distinguish between simple and complex exchange type aberrations, George et al. (2003b) reported higher frequencies of chromosome exchanges after Calyculin-A treatment with regard to those after Colcemid treatment. This result was also consistent with a previous study from the same group (George et al. 2001) indicating that chromosome exchanges, particularly complex type exchanges, were affected by the cell cycle delay. In our group we have seen that after α-particle irradiation complex aberrations are induced at low doses, whereas for low-LET radiation complex aberrations appeared from doses of 2–3 Gy (Barquinero et al. 1999, 2004; Rodríguez et al. 2009).

By the chemically-induced PCC technique, chromosome spreads show different morphologies according to the presence or absence of the centromeric constriction and chromatids alignment (Lamadrid et al. 2007; Puig et al. 2013). In the present study two cell morphologies were observed, G2/M-PCC cells and M cells. Another morphology described in a previous work with low-LET radiation (Puig et al. 2013) was the M/A-PCC cells, without the centromere constriction and with the two chromatids separated. This cell type was not observed in the present study. In the study with low-LET radiation, G2/M-PCC and M cells proportions were clearly dose-dependent, whereas the proportion of M/A-PCC was constant, indicating that their presence could result from a methodological aspect linked to the use of Calyculin-A (Puig et al. 2013). A possible explanation for an absence of M/A-PCC cells in the present study is again methodological. Although Calyculin-A treatment was also performed, here isolated lymphocytes were stimulated to growth to obtain a monolayer of lymphocytes attached to the Mylar foil for irradiation, whereas in the previous study whole peripheral blood was irradiated and then cultured.

Irrespective of the cell type, the frequency of chromosome aberrations considered showed a clear increase with the dose. Among the chromosome aberrations recorded, chromosome rings showed the lowest frequency, and their frequency in M cells was clearly lower than in G2/M-PCC cells. In fact, rings were hardly found in M cells, being unobservable at most doses. The poor induction of chromosome rings when compared with the other chromosome aberrations is widely described, even after α-particle irradiation (Schmid et al. 1996; Barquinero et al. 2004). With Calyculin-A induced PCC and after exposures to mixed gamma-neutron radiation, a saturation has been described in the induction of chromosome rings at doses around 5 Gy with a frequency of 0.33 per cell (Lamadrid et al. 2011). However, another possible explanation is the absence of M/A-PCC cells in the present study; as was observed in our previous study, chromosome rings are much easier to detect when both chromatids are separated (Puig et al. 2013). On the other hand, extra fragments were detected at all doses, but in a similar way to rings, the frequency in G2/M-PCC cells was always significantly higher than in M cells. Higher frequencies of extra fragments in G2/M-PCC cells to those in M-cells were also observed after X- or γ-ray irradiation (Kovacs et al. 1994; Kanda et al. 2004; Puig et al. 2013). Using pan-telomeric and pan-centromeric PNA probes it was reported that the G2/M checkpoint negatively selects cells with incomplete chromosome aberrations (Rodríguez et al. 2009). When the cell distribution of extra fragments was considered a clear overdispersion was observed. However, it is interesting to note that whereas for G2/M-PCC cells the u values are relatively constant, for M cells the u values increased from 7.9 at 0.05 Gy to 40.4 at 0.2 Gy, and then decreased. This result seems to indicate that although the induction of extra fragments by α-particle irradiation results in a non-dose dependent overdispersion, the suitability of different damaged cells in entering mitosis modifies the distribution of extra acentric fragments in M cells.

The distribution of dicentric chromosomes in M cells showed an initial increase in overdispersion (between 0.05 and 0.1 Gy), followed by a gradual decrease up to 2.5 Gy. A similar variation in the overdispersion was also reported by Schmid et al. (1996). The frequencies of dicentrics observed in the present study are similar to those reported previously using the same methodology of irradiation and the same source (Schmid et al. 1996; Barquinero et al. 2004; Schmid and Roos 2009), indicating the reproducibility of the method. This result indicates that the frequency of dicentrics observed after the conventional Colcemid treatment is similar to the one that can be observed in M cells after the co-treatment with Colcemid and Calyculin-A.

Regarding dose-effect relationships, the linear coefficient obtained for ring chromosomes (0.017 ± 0.02 Gy⁻¹, the same for G2/M-PCC and for total cells) was lower than one reported for neutrons (0.059 ± 0.02 Gy⁻¹) with a similar dose range (from 0–3.8 Gy) using the same methodology to induce PCC (Lamadrid et al. 2011). One explanation could be that α-particles are less efficient than 0.49 MeV neutrons in inducing chromosome rings. However, the linear coefficients obtained here are similar to that obtained by analyzing G2/M-PCC cells after γ-ray irradiation (0.019 ± 0.002 Gy⁻¹) and clearly lower for total cells after γ-rays (0.027 ± 0.002 Gy⁻¹) (Puig et al. 2013). If only G2/M-PCC cells are considered the low number of chromosome rings observed in the present study makes any comparison difficult. But when total cells are considered it is difficult to argue that α-particles are less efficient than γ-rays in producing chromosome rings. Probably the low linear coefficient obtained could be attributed to the absence of M/A-PCC cells. In this cell type chromosome rings are more easily detected (Puig et al. 2013). Another possible explanation for the lack of chromosome rings observed in the present study would be due to the irradiation conditions. In the present study lymphocytes were exposed to α-particles as monolayers of attached cells. It has been demonstrated that nucleus morphology in such situation is flattened in respect to settled lymphocytes that maintain a round shape morphology (Schmid et al. 2011). Taking into account that individual chromosomes occupy localized domains in the nucleus (Cremer and Cremer 2001), and that DSB occurring at the periphery of these chromosome domains will be more prone in producing chromosome interchanges than DSB occurring in the inner territory, which will tend to produce more chromosome intrachanges (Cremer et al. 1993). Irradiating flattened nucleus (i.e. enlarging chromosome territory surface) would result in a decrease of chromosome intrachanges such as chromosome rings. In fact, with the same device used here to irradiate with α-particles, Schmid et al. (2011) described that the ratio between dicentric and ring chromosomes, the F-ratio, was clearly lower after irradiating settled lymphocytes than after irradiating attached lymphocytes, indicating a lesser induction of chromosome rings in relation to the induction of dicentric chromosomes. Taking into account that the proportion of M/A-PCC cells as well as the relationship between different types of radiation-induced chromosome aberrations could depend on some methodological aspects, if dose-effect curves for ring chromosomes using chemically-induced PCC are starting to be produced among biological dosimetry laboratories a harmonized protocol seems desirable.

A major effectiveness of α-particles with respect to low-LET radiation types, like γ-rays, in producing chromosome damage is clearly observed when dicentrics or extra chromosome pieces are considered. The linear coefficient for dicentrics chromosomes in M cells is 0.373 ± 0.074 Gy⁻¹, slightly higher than the other two coefficients obtained using the same source and blocking division with a Colcemid treatment only (0.27 ± 0.02 Gy⁻¹ and 0.29 ± 0.02 Gy⁻¹ for Schmid et al. (1996) and Barquinero et al. (2004), respectively. The slight, but non-significant, increase could be due to the combined use of Colcemid and Calyculin A here used to obtain chromosome spreads. Other reported linear coefficients after α-particles irradiation are in the same range: 0.37 ± 0.02 Gy⁻¹ from Purrott et al. (1980), 0.29 ± 0.02 Gy⁻¹ from Edwards et al. (1980) and 0.24 ± 0.02 Gy⁻¹ from Moquet et al. (2001).

The linear coefficient for dicentrics is 17-fold higher than that obtained after γ-rays (0.021 + 0.005 Gy⁻¹) (Barquinero et al. 1995) and also clearly higher than that obtained using PCC after γ-rays (Puig et al. 2013), 0.072 ± 0.008 Gy⁻¹, where the dose range analyzed was from 0–20 Gy. Similarly, α-particles are also more efficient than γ-rays in producing extra chromosome fragments. For total cells the linear coefficient observed here (1.795 ± 0.021 Gy⁻¹) is 4- to 5-fold higher than the other two obtained for γ-rays (0.393 ± 0.029 Gy⁻¹ and 0.41 ± 0.03 Gy⁻¹ by Puig et al. (2013) and Balakrishnan et al. (2010), respectively. It is interesting to note that although α-particles are more efficient in producing extra chromosome fragments than γ-rays, the G2/M checkpoint seems to reduce their presence in M cells to a similar proportion. After α-particle irradiation the linear coefficient in G2/M cells is 1.53 ± 0.02 Gy⁻¹ 5-fold higher than that obtained in M cells 0.30 ± 0.02 Gy⁻¹. A similar reduction is observed after γ-rays from 0.41 ± 0.02 Gy⁻¹ in G2/M cells to 0.07 ± 0.01 Gy⁻¹ in M cells (Puig et al. 2013).

Considering that exchange type aberrations result from the misrepair of different DSB and that the initial damage after high-LET exposure, like the α-particles analyzed here, shows a higher degree of complexity (Brenner and Ward 1992) that results in an increase of the half-life of rejoining (Heilmann et al. 1993; Asaithamby et al. 2008), one would expect the presence of a quadratic component for some types of exchange type aberrations. A possible explanation for the absence of this quadratic component is the method used to analyze the aberrations. It it is not possible to distinguish between dicentrics or rings coming from simple or complex exchanges by analyzing solid stained chromosomes. Additionally, it has been described that after α-particle exposures, although the degree of complexity in the exchange type aberrations is higher than that observed after low-LET radiation types, the proportion of complex aberrations in respect to the total is constant with the dose (Barquinero et al. 2004), this would make the observation of a quadratic component difficult when all exchanges are considered together. It is interesting to note that in the study from George et al. (2003b) for all the particles evaluated, whereas simple aberrations fitted to linear curves, when complex exchanges alone were considered the obtained curves fitted to a linear quadratic model.

In conclusion, when the proportion of PCC and M cells was compared, the results indicated that damaged cells after α-particle irradiation have more difficulties entering mitosis than cells exposed to γ-rays. After α-particle irradiation the frequency of all chromosome aberrations considered increased linearly with the dose. However, ring chromosomes were not very informative due to the smaller amounts observed in the two cells types analyzed, G2/M-PCC and M cells; dicentric chromosomes observed in M cells gave similar results to those obtained by the classical method to obtain metaphase spreads blocking cell division with Colcemid; and the frequency of extra chromosome pieces was clearly reduced in M cells with regard to G2/M-PCC cells. The amount of extra chromosome pieces observed in all cell types analyzed suggests that this type of aberration should be a good candidate to be used as a biomarker of exposure after high-LET radiation types if chemically-induced PCC techniques are used. However, in order to avoid possible methodological bias, standardization of the technique will be desirable.

Acknowledgements

The authors express their appreciation to Sabine Wenzel for her technical support and Ulrike Kulka for her considerations. This work received financial support from the Consejo de Seguridad Nuclear. The sponsor had no role in the study design, data collection, analysis or interpretation. M.R.C., L.B. and J.F.B. belong to a consolidated research group of the Generalitat de Catalunya (2014 SGR 354).

Disclosure statement

The authors report no conflicts of interest. The authors alone are response for the content and writing of the paper.