Spontaneous and bleomycin-induced chromosome damage in non cancer thyroid patients


Dr Roberto Scarpato, Dipartimento di Biologia, University of Pisa, Via Derna 1 56100 Pisa, Italy. Tel.: 39-050-2211509; fax: 39-050-2211527; e-mail: rscarpato@biologia.unipi.it


Background  Presence of chromosome damage in lymphocytes of patients affected by several diseases, including cancer, was detected by the micronucleus (MN) assay. Individual susceptibility to DNA damage, considered as a risk factor for cancer, can be also evaluated using the bleomycin (BLM) sensitivity test.

Materials and methods  We aimed to evaluate spontaneous or BLM-induced MN frequencies in autoimmune (AI, n = 19) and non autoimmune (NAI, n = 11) thyroid patients, not receiving 131I radiometabolic therapy with respect to a control group of 18 healthy subjects. According to thyroid function, patients were also divided into hypothyroid (n = 10), euthyroid (n = 13) or hyperthyroid (n = 7) subjects.

Results  Spontaneous MN frequencies of AI and NAI patients did not differ from those of controls. Hypothyroid patients had more elevated MN basal levels (9·00 ± 1·71‰) than hyperthyroid (3·75 ± 1·17‰, < 0·05) and euthyroid (5·38 ± 0·97‰, < 0·01) patients or healthy subjects (4·17 ± 0·63‰, < 0·01). In particular, the hypothyroid AI group showed the highest value (9·79 ± 2·26‰, < 0·01). All thyroid patients responded differently to BLM than controls (39·90 ± 2·48‰ vs. 31·08 ± 2·51‰, = 0·0377). The NAI group had BLM-induced MN levels (45·00 ± 2·56‰) significantly higher (= 0·0215) than AI patients (36·95 ± 3·49‰) or healthy subjects (31·08 ± 2·51‰). No significant difference was seen when patients were stratified according to autoimmunity.

Conclusions  We report that hypothyroid patients exhibit a moderate increase in the level of spontaneous genome damage, and that AI thyroid patients resulted to be less sensitive than NAI patients to the mutagen sensitivity test. In prospective, it may be of interest to reinvestigate hypothyroid patients when correction of their dysfunction is achieved.


Disorders of the thyroid gland that are characterized by the presence of inflammation associated with normal or impaired thyroid function can be mainly autoimmune or with a minor incidence, also infectious in origin. In fact, alteration of the immune system, especially towards endocrine organs, can result in the development of some pathological conditions in 5–10% of the human population. Among these, Hashimoto’s thyroiditis and Graves’ hyperthyroidism represent the most common autoimmune disorders targeting the thyroid gland [1].

Several studies have reported the presence of chromosomal anomalies (e.g. micronuclei) in proliferating peripheral lymphocytes, from people affected by pathological conditions, hypothesizing that the observed damage may either cause or be a manifestation of the disease [2–6]. It has also been suggested that an excess of genomically instable cells may predispose healthy subjects to undergo carcinogenesis [7]. Susceptibility to endogenous and/or exogenous DNA damage, which depends on the correct functioning of the cellular machinery devoted to maintain the genome integrity, may also play an important role in the pathogenesis of cancer. A methodology to indirectly assay, is the mutagen sensitivity test, which estimates the DNA repair capability of peripheral lymphocytes on the basis of the number of double strand breaks (DSB) induced in vitro by the anticancer drug bleomycin [8]. In this view, an increased level of DSB has been found in the lymphocytes of subjects affected by either different types of cancer [9] or precancerous pathologies such as the Barrett’s oesophagus and endometriosis [10,11].

To date, little or nothing is known about the extent of spontaneous genome damage in circulating cells of non cancer thyroid patients, with the exception of those studies aiming to elucidate the induction and persistence of genotoxic effect by 131I radiometabolic therapy in Graves’ hyperthyroidism [12,13]. With this background, using the micronucleus (MN) assay in peripheral lymphocytes, we evaluated spontaneous and bleomycin-induced chromosome damage in patients affected by autoimmune and non autoimmune thyroid disorders with no history of therapeutic or diagnostic exposure to ionizing radiation.

Materials and methods

Study population

The study population consisted of 30 thyroid patients referred to the Metabolism Unit of the Department of Internal Medicine, and a group of 18 healthy controls recruited among the medical personnel employed at the E. Lotti Hospital of Pontedera (Pisa) who had, at the time of blood sampling, no past or present history of thyroid alteration. Each participant enrolled in the study was physically examined and diagnosis of thyroid pathology was confirmed (for patients) or excluded (for healthy controls) on the basis of medical inspection and laboratory findings as described below. Thyroid patients were subdivided into an autoimmune (AI, n = 19) and a non autoimmune (NAI, n = 11) group on the basis of the specific disease diagnosed, and into hypothyroid (n = 10), euthyroid (n = 13) or hyperthyroid (n = 7) subjects according to the functioning of the thyroid. Each subject also signed an informed consent form and filled in a detailed questionnaire on personal data and life styles (tobacco smoking, alcohol consumption and drug use). Intake of chemotherapeutic drugs and occurrence of viral infections in addition to ionizing radiation exposure for radiodiagnostic or radiotherapeutic purposes in the last 6 months were criteria to exclude subjects from the study. The study was reviewed and approved by the ethical committee of the ASL 5 of Pisa.

Laboratory evaluation

Laboratory evaluation included measurement of serum concentrations of TSH, free T3 (FT3) and free T4 (FT4), as well as serum titres of antithyroglobulin (TgAb), antithyroperoxidase (TPOAb) and, when necessary, anti-TSH receptor (TRAb) antibodies. Circulating T3 and T4 were measured by commercial RIA kits (AMERLEX-MAB FT3/FT4 Kit; Amersham, Milano, Italy). Serum TSH (DiaSorin, Stillwater, MN, USA), TgAb, TPOAb and TRAb (ICN Pharmaceuticals, Costa Mesa, CA, USA) were evaluated by IRMA method. Normal ranges were: TSH, 0·4–4·0 mIU L−1; FT4, 7·0–17·0 pg mL−1; FT3, 1·8–4·8 pg mL−1; TgAb, < 100 IU mL−1; TPOAb, < 50 IU mL−1 and TRAb < 1 IU L−1.

MN assay and mutagen sensitivity test

Venous blood samples were obtained by venipuncture and collected into heparinized tubes. For each subject, four independent cultures were set up and incubated at 37 °C for 72 h by adding 0·3 mL of whole blood to 4·7 mL RPMI-1640 medium (Invitrogen, Milano, Italy) supplemented with 20% foetal bovine serum (Invitrogen), 1·5% phytohaemagglutinin (Invitrogen) and 1% antibiotic/antimycotic (Invitrogen). After 24 h, cultures for the mutagen sensitivity test received bleomycin (BLM, 15 000 IU from Nippon Kayaku, Sanofi-aventis, Milano, Italy) at 5·0 μM final concentration.

Cytochalasin B (6·0 μg mL−1 final concentration; Sigma-Aldrich, Milano, Italy) was added at 44 h to all cultures to block cell cytokinesis. Cell harvesting was carried out according to the standard procedures as described elsewhere [14]; after hypotonic treatment with 0·075 M KCl, lymphocytes were prefixed in acetic acid:methanol 5 : 3, fixed in 100% methanol, washed twice in methanol:acetic acid (7 : 1) and then dropped onto clean glass slides. The air-dried slides were then stained in 5% Giemsa. According to standard criteria, 2000 binucleated cells (1000 cells per culture) on coded slides from each subject were scored for the presence of MN by an experienced microscopist using an optical microscope equipped with a 40× objective (400× final magnification). A micronucleus can be formed in dividing cells by either chromosome acentric fragments or whole chromosomes in late migration during anaphase that are unable to be included in the two daughter nuclei [15]. MN frequency was expressed as the number of micronucleated binucleates per 1000 binucleated scored cells.

In the case of the BLM sensitivity test, the actual induction level was also calculated for each subject, subtracting to the MN mean value obtained from the corresponding basal level of BLM-treated cultures. Data were statistically elaborated by means of multifactor anova to assess the effect of the thyroid alterations on spontaneous or BLM-induced MN frequencies adjusted for confounding factors (gender, smoking habit and age). All factors were included in the statistical analyses as categorical variables with the exception of age, which was used as a covariate. When required, a multiple range test was performed (Bonferroni method). All statistical calculations were carried out using the Statgraphics Plus version 5·1 (Statistical Graphics Corporation, 2001, Rockville, MD, USA) software package.


The demography and clinical features of the subjects enrolled in the study are summarized in Table 1 Women are more prevalently affected than men in both AI (n = 19) and NAI (n = 11) patients; a similar ratio is thus maintained in the control group (n = 18). Both mean age and the proportion of smokers to non smokers do not differ significantly among the subjects grouped according to the disease status.

Table 1.   Demographic and clinical characteristics of the study population
SubjectsNo.Age (years)TSH (mIU L−1)*FT4 (pg mL−1)FT3 (pg mL−1)TgAb (IU mL−1)TPOAb (IU mL−1)*
  1. Values, if not differently reported, are expressed as mean ± SD.

  2. AI, autoimmune thyroid pathology; NAI, non autoimmune thyroid pathology.

  3. *For each subject group, TSH and TPOAb values are expressed as median. Minimum and maximum values are also reported in brackets.

  4. Hyperthyroid patients: Graves’ hyperthyroidism (n = 5); toxic nodular goitre (n = 2). Hypothyroid patients: Hashimoto’s thyroiditis (n = 7); iodine deficiency (n = 3).

 AI1947·40 ± 11·101·595 (0·002–12·27)10·46 ± 3·342·83 ± 0·74216·66 ± 141·49304·6 (5·00–3200·00)
 NAI1145·70 ± 8·400·883 (0·002–27·17)11·25 ± 4·672·78 ± 1·6190·61 ± 29·296 (2·20–107·30)
Thyroid function
 Hyperthyroid745·43 ± 8·160·267 (0·002–12·27)11·91 ± 6·573·19 ± 2·00146·56 ± 97·81102·6 (3·90–3200·00)
 Hypothyroid1042·90 ± 12·003·366 (0·20–27·17)10·74 ± 3·472·83 ± 0·83219·71 ± 163·41130·45 (2·2–3153·0)
 Euthyroid1350·50 ± 8·700·941 (0·341–3·95)10·13 ± 1·872·60 ± 0·58145·64 ± 112·777·8 (3·00–2162·00)
 Controls1845·10 ± 10·101·485 (0·94–3·50)11·81 ± 2·372·58 ± 0·5044·64 ± 21·114 (1·50–8·20)
 Male735·14 ± 13·00     
 Female4148·05 ± 8·21     
Smoking habit
 Non smoker3045·83 ± 10·16     
 Smoker1846·72 ± 10·00     

The results of spontaneous chromosome damage in peripheral cells of the study population are given in Table 2 Thyroid patients as a whole showed MN frequencies slightly, but not significantly, higher than those of healthy subjects. Also when the patients were classified according to autoimmunity, we did not observe any significant increase, even though both AI and NAI subjects showed higher frequencies than controls. At variance, the different functioning of the thyroid gland was able to significantly affect MN spontaneous frequencies (= 0·0047), and hypothyroid patients exhibited statistically more elevated MN levels (9·00 ± 1·71‰) than either patients with hyperthyroidism (3·75 ± 1·17‰, < 0·05) and normal function (5·38 ± 0·97‰, < 0·01) or healthy controls (4·17 ± 0·63‰, < 0·01). In addition, the hypothyroid AI patients showed higher MN levels (9·79 ± 2·26‰, < 0·01) as compared with the other AI groups or healthy controls, whereas no difference was observed among the NAI patients (Fig. 1). Gender and age did not cause variation in the MN baseline level.

Table 2.   Multifactor anova for baseline MN frequencies in the study population stratified by health status, thyroid disease or thyroid function, and gender, smoking habit or age
Source of variationMean ± SEF-ratioP-value
  1. AI, autoimmune thyroid pathology; NAI, non autoimmune thyroid pathology.

  2. *Significantly different from MN frequencies of hyperthyroid patients (< 0·05, Bonferroni multiple range test).

  3. Significantly different from MN frequencies of euthyroid patients and controls (< 0·01, Bonferroni multiple range test).

Health status
 Thyroid patients6·13 ± 0·84 0·0743
 Controls4·16 ± 0·63 
Thyroid disease
 NAI5·77 ± 1·081·630·2074
 AI6·34 ± 1·18
 Controls4·16 ± 0·63
Thyroid function
 Euthyroidism5·38 ± 0·975·010·0047
 Hypothyroidism9·00 ± 1·71*,†
 Hyperthyroidism3·75 ± 1·17
 Controls4·16 ± 0·63
 Female5·83 ± 0·651·970·1685
 Male2·86 ± 0·66
 Non smokers5·25 ± 0·780·500·4847
 Smokers5·64 ± 0·89
 Age 0·000·9700
Figure 1.

 Spontaneous MN frequencies in controls and thyroid patients stratified according to both thyroid function and autoimmunity. **Significantly different (< 0·01, multiple range test) from euthyroid, hyperthyroid or healthy controls. Bars represent the average values ± SE of each subject group.

Table 3 and Fig. 2 report the results of the mutagen sensitivity test. All together, thyroid patients responded differently to BLM treatment than healthy controls (39·90 ± 2·48‰ vs. 31·08 ± 2·51‰, = 0·0377). Moreover, the NAI group had BLM-induced MN levels (45·00 ± 2·56‰) significantly higher (= 0·0215) as compared with those of either the AI patients (36·95 ± 3·49‰) or controls (31·08 ± 2·51‰). Conversely, we did not observe any significant difference when the patients were classified according to the thyroid gland functioning. Figure 2 shows the distribution curves of the subjects clustered according to the net number of BLM-induced micronucleated cells. About 45% of both healthy controls and the autoimmune patients did not exceed 24%MN, whereas more than 50% of the NAI subjects showed 40% induction.

Table 3.   Multifactor anova for BLM-induced MN frequencies in the study population stratified by health status, thyroid disease or thyroid function, and gender, smoking habit or age
Source of variationMean ± SEF-ratioP-value
  1. AI, autoimmune thyroid pathology; NAI, non autoimmune thyroid pathology.

  2. *Significantly different from MN frequencies of controls (< 0·01, Bonferroni multiple range test).

  3. Significantly different from MN frequencies of autoimmune patients (< 0·05, Bonferroni multiple range test).

Health status
 Thyroid patients39·90 ± 2·484·600·0377
 Controls31·08 ± 2·51
Thyroid disease
 NAI45·00 ± 2·56*,†3·900·0279
 AI36·94 ± 3·49
 Controls31·08 ± 2·51
Thyroid function
 Euthyroidism40·08 ± 4·091·330·2764
 Hypothyroidism40·20 ± 3·91
 Hyperthyroidism33·57 ± 4·75
 Controls31·08 ± 2·51
 Female36·95 ± 1·980·120·7310
 Male34·50 ± 6·29
 Non-smokers36·28 ± 2·620·010·9194
 Smokers37·11 ± 2·66
 Age 3·050·0883
Figure 2.

 Distribution curves of the percentage of the study population grouped according to the net number of BLM-induced MN. AI, autoimmune patients; NAI, non autoimmune patients.

There was no correlation between MN frequencies (either spontaneous or induced) and TSH, TgAb or TPOAb levels (data not shown).


Several studies have investigated the level of genome damage in thyroid cancer patients environmentally exposed to radiation after the Chernobyll fallout [16,17], and in Graves’ disease patients or in children affected by thyroid cancer who underwent radiotherapy [13,18], as marker of the risk for occurrence of primary or secondary tumours. Higher spontaneous levels of chromosomally altered lymphocytes were also observed in people affected by thyroid nodules following occupational exposure to ionizing radiation [19].

Our study reports, for the first time, the presence of spontaneous chromosome damage in peripheral blood lymphocytes from non cancer thyroid patients without any history of therapeutic ionizing radiation exposure. The results indicate that, irrespective of thyroid functioning, the level of damage detected in the autoimmune group is only slightly higher than that observed in non autoimmune patients. Rather, when we looked at the thyroid gland function irrespective of autoimmunity, hypothyroid patients showed moderately, but significantly increased MN frequencies as compared with hyperthyroid or euthyroid subjects, in addition to healthy controls.

High levels of chromosome breakage were shown in peripheral cells of scleroderma patients, an autoimmune disorder of the connective tissue [2,20]. The observed damage was seen to correlate to the presence of plasmatic inflammatory molecules constituting the so-called clastogenic factor [21], and to the seroimmunological profile of the patients (i.e. presence of anticentromere antibodies) [3]. Other chronic inflammation-based disorders such as atherosclerosis, Chron’s disease or ulcerative colitis and pneumoconiosis, induced genome damage at peripheral level [4,22,23]. Thus, the failure to detect chromosome damage in autoimmune thyroid disease per se, also relies on its lower grade and persistence of inflammation with respect to the other inflammatory disorders. Moreover, we propose that the increased frequency of spontaneous genome altered cells observed in hypothyroid patients is the consequence of oxidative damage. Oxidative stress, which is because of impairment of the physiologic equilibrium between serum or intracellular antioxidant activities and the production of reactive oxygen species (ROS), is generally associated to hyperthyroidism. However, increased rate of ROS generation and lipidic peroxidation together with alteration of the antioxidant systems have been recently reported in hypothyroid patients as well as in animal models of hypothyroidism [24–28]. It is now well established that thyroid hormones can modulate transcription and/or expression of several genes involved in growth, development and differentiation, especially of the central nervous system [29]. As compared with the euthyroid condition, in fact, hypothyroidism was proven to increase the extent and the duration of apoptosis in rats’ down-regulating or over-expressing the anti-apoptotic Bcl-2 or the pro-apoptotic Bax gene, respectively [30]. Thus, alteration of the antioxidant/pro-oxidant systems would allow endogenous ROS to induce DNA lesions such as single-strand breaks (SSB) or, if not properly repaired, DSB which may originate, in turn, chromosome deletions and MN. Previous studies showed that patients affected by Parkinson’s disease and treated with levodopa, with respect to de novo and untreated patients, exhibited in peripheral lymphocytes either more elevated levels of caspase-3 activity or reduced anti-apoptotic BCL-2 protein, or an increased DNA fragmentation and MN [31,32]. The pivotal role of oxidative stress in inducing genome damage at peripheral level has been documented also for endocrine-metabolic syndromes such as type 2 diabetes mellitus and polycystic ovary syndrome [5,6].

Notably, the highest MN basal level was observed among the hypothyroid autoimmune patients. This would suggest that the expression of spontaneous chromosome damage might be, at least partially, modulated by autoimmunity, confirming also that the two aspects are linked to each other. It is well known, in fact, that variation of thyroid function in autoimmune diseases, from hypo to hyperthyroidism and vice versa, actuates, at least at early stage, through the apoptosis pathway, which is started by anomalous production of cytokines. Moreover, it is in our opinion that the moderate increase in the baseline MN level of our patients can spontaneously regress as soon as the euthyroid condition has been restored by an appropriate therapy. Despite the fact that these patients could be reinvestigated at this time, previous studies would indicate that a drug- or dietary supplement-based intervention strategy suppresses or attenuates the occurrence of peripheral chromosome damage in diseases characterized by oxidative damage [4,33,34].

Interesting findings were also obtained from the mutagen sensitivity test. The induction of genome damage was detected in all subjects from the study population, but thyroid patients as a whole responded to BLM treatment at a higher extent than healthy controls. Furthermore, it appeared evident that, whereas both AI and controls shared lower level of induced MN, the NAI patients were the most susceptible group towards DNA damage induced by the mutagenic anticancer drug. It is worth noting the fact that, even when baseline MN frequencies were subtracted to the BLM induced levels, the significance of the finding maintained exactly unaltered. The ability of thyroid hormones to interfere with transcription or gene expression would suggest a modulating activity on the cell systems involved in the activation/detoxification of mutagens and carcinogens and/or in the DNA damage response. Some studies, in fact, highlighted as BLM sensitivity can be affected, at least in part, by allelic variants in genes encoding for metabolism and DNA repair enzymes [35,36], or controlling cell growth and/or maintenance and regulation of cell cycle [37].

The relationship between the incidence of cancer and thyroid diseases appears still rather controversial, even though evidence exists of a higher prevalence of thyroid autoimmune or non autoimmune disorders in patients affected by breast cancer, haematological neoplasias, multiple myeloma or papillary carcinoma of the thyroid gland [38–41]. On the other hand, growing literature evidences indicate that regulatory T-cell subtypes, in addition to prevent autoimmunity, are also able to accumulate in the tumour microenvironment causing cancer progression. The mechanisms by which they act remain largely undefined, but several auto-antibodies, including TgAb opposing to regulatory T-cell activity, are recently shown to be protective against different types of cancer [42–44]. Also the results of the mutagen sensitivity test would seem to indicate that thyroid autoimmunity may afford a protective role against cancer, at least with respect to NAI patients.

In conclusion, our results indicate the presence of either spontaneous or induced genome damage in peripheral lymphocytes of patients who were, at the time of blood sampling for cytogenetic analysis, in a hypothyroid condition or affected by non autoimmune disease, respectively. Further studies should confirm and possibly clarify the molecular basis of the observed effects, especially those concerning cancer predisposition in thyroid patients, for which the two most important factors remain the presence of thyroid nodules and administration of radioiodine therapy.

Conflict of interest

The authors have nothing to declare.


Dipartimento di Biologia, University of Pisa, Via Derna 1, 56100 Pisa, Italy (R. Scarpato, I. Tusa, I. Sbrana); Department of Internal Medicine, University of Pisa, Pisa, Italy (A. Antonelli, P. Fallhai).