Exposure to anticancer drugs can result in transgenerational genomic instability in mice

Abstract

The genetic effects of human exposure to anticancer drugs remain poorly understood. To establish whether exposure to anticancer drugs can result not only in mutation induction in the germ line of treated animals, but also in altered mutation rates in their offspring, we evaluated mutation rates in the offspring of male mice treated with three commonly used chemotherapeutic agents: cyclophosphamide, mitomycin C, and procarbazine. The doses of paternal exposure were approximately equivalent to those used clinically. Using single-molecule PCR, the frequency of mutation at the mouse expanded simple tandem repeat locus Ms6-hm was established in DNA samples extracted from sperm and bone marrow of the offspring of treated males. After paternal exposure to any one of these three drugs, expanded simple tandem repeat mutation frequencies were significantly elevated in the germ line (sperm) and bone marrow of their offspring. This observed transgenerational instability was attributed to elevated mutation rates at the alleles derived from both the exposed fathers and from the nonexposed mothers, thus implying a genome-wide destabilization. Our results suggest that paternal exposure to a wide variety of mutagens can result in transgenerational instability manifesting in their offspring. Our data also raise important issues concerning delayed transgenerational effects in the children of survivors of anticancer therapy.

In recent decades, improvements in treatment have dramatically increased survival rates among cancer patients (1). However, along with surgery, the mainstays of cancer treatment are radiotherapy and chemotherapy, both of which are potentially genotoxic and mutagenic. This can lead to the development of secondary, treatment-related conditions, including further cancers, as well as predispose survivors to various nonmalignant diseases (2). Given that the results of numerous publications clearly show that exposure to ionizing radiation and numerous commonly used drugs results in mutation induction in the mouse germ line (3, 4), there is growing concern regarding the genetic risk of anticancer therapy, particularly given the steady increase in long-term survival of childhood cancer survivors (5). It should be stressed that the hereditary effects of human exposure to mutagens, including anticancer drugs, remain poorly understood and merit thorough analysis.

Recent mouse studies have shown that the genetic effects of exposure to ionizing radiation and some chemical mutagens might not be restricted to the directly affected parental germ line, but also can manifest in their offspring (6–10). One of the key aspects of transgenerational effects is destabilization of the offspring's genomes, detectable across at least two generations. Given that the development of cancer is a multistep process in which somatic cells acquire mutations (11), ongoing genomic instability may predispose the offspring of irradiated parents to cancer. However, data on transgenerational instability in mice so far have been obtained by profiling the offspring of males exposed to high doses of acute whole-body irradiation (6–9), whereas cancer patients undergoing radiotherapy receive substantially smaller doses to noncancerous tissues (1). Therefore, whether low-dose parental irradiation can destabilize their offspring remains to be established.

In contrast to radiotherapy, in which normal tissues are shielded, chemotherapy treatments are systemic and provide little if any protection of nontumorous tissues against the damaging effects of anticancer drugs. Given the high doses of acute exposure to mutagenic anticancer drugs, it would appear that the genetic effects of chemotherapy might exceed those after radiotherapy. Although the mutagenic potential of a majority of anticancer drugs has been established in mice (4), little is known about whether parental exposure to these drugs can also result in transgenerational instability (12). Here we present the results of a systematic study that analyzed the effects of paternal exposure to widely used anticancer drugs in terms of the manifestation transgenerational instability in mice.

Results

Experimental Design.

Mutation frequencies in sperm and bone marrow were measured at the Ms6-hm expanded simple tandem repeat (ESTR) locus in the male F₁ offspring of two reciprocal crosses: ♂CBA/Ca × ♀BALB/c and ♂BALB/c × ♀CBA/Ca (Fig. 1A). Male mice were given a single dose of one of three anticancer drugs. At 8 wk after exposure (to allow mature sperm to develop from the exposed spermatogonial stem cells), these F₀ males were mated with nonexposed females, CBA/Ca males with BALB/c females and vice versa. We evaluated mutations at both the smaller (∼2.5 kb) BALB/c-derived allele and the larger (∼3 kb) CBA/Ca allele (Fig. 1B), which allowed us to differentiate between the maternally and paternally derived alleles and measure mutation frequencies for each. DNA (typically more than 100 PCR-amplifiable molecules per animal for each of the two tissues) was analyzed from six animals in the control group (three offspring from different litters of each reciprocal cross) and from four animals in each of the three drug test groups (two offspring from different litters of each reciprocal cross).

Fig. 1.

Design of transgenerational study (A) and mutation detection at the Ms6-hm ESTR locus by SM-PCR (B). ESTR PCR products from one offspring are shown. Mutations at the CBA/Ca and BALB/c alleles are indicated by arrows.

Treatment.

Male mice were given a single dose of cyclophosphamide (CPP), mitomycin C (MMC), or procarbazine (PCH). Doses were calculated using Food and Drug Administration guidelines recommending multiplying human doses by 12.3 per kilogram of body weight to yield equivalent doses for mice (13).

CPP is a widely used anticancer drug and immunosuppressive agent that after metabolic activation forms DNA adducts, mostly N⁷-substituted guanine derivatives and phosphotriester adducts (14). We previously showed that a single dose as low as 40 mg/kg can increase mutation frequency at the Ms6hm ESTR locus in the germ line of directly exposed male mice (15). Single doses for anticancer treatment in humans are typically between 400 and 1,200 mg/m² (16); the equivalent mouse dose is 130–400 mg/kg. To evaluate the transgenerational effects of paternal exposure to CPP, we analyzed the offspring of male mice given a single clinically relevant dose of 150 mg/kg.

The streptomycin-derived antibiotic MMC is a widely used anticancer drug. After metabolic activation, MMC forms a number of sequence-specific DNA adducts, including alkylated monoadducts and cross-linked adducts (17). Single doses of 2.5 or 5 mg/kg lead to dose-dependent statistically significant increases in ESTR mutation frequency in the germ line of directly exposed male mice (15). Single doses in humans are typically between 10 and 20 mg/m² (16), with equivalent mouse doses of 3–6 mg/kg. To evaluate the transgenerational effects of paternal exposure to this MMC, male mice were given a single clinically relevant dose of 5 mg/kg.

PCH is used to treat a number of cancers, including Hodgkin's lymphoma. Metabolites of PCH inhibit DNA polymerase and react directly with DNA (18). A dose of 50 mg/kg is sufficient to more than double the ESTR mutation frequency in the germ line of directly exposed male mice (15). Single doses in humans do not exceed 150 mg/m² (16), and the equivalent mouse dose of 50 mg/kg was used.

ESTR Mutation Frequencies.

We evaluated the frequency of ESTR mutation in DNA samples extracted from sperm and bone marrow using single-molecule PCR (SM-PCR). This approach involves diluting bulk genomic DNA and amplifying multiple samples of DNA, each containing approximately one amplifiable ESTR molecule (Fig. 1B).

The frequency of ESTR mutation in the offspring of two reciprocal crosses, ♂CBA/Ca × ♀BALB/c and ♂BALB/c × ♀CBA/Ca, did not differ significantly within each group (Table S1); thus, we combined the data for these crosses. Table 1 summarizes the ESTR mutation data. We found that paternal exposure to all three anticancer drugs resulted in highly significant, approximate twofold increases in ESTR mutation frequency in the germ line and bone marrow of their offspring. In most cases, ESTR mutation frequencies were significantly elevated at the F₁ alleles derived from the exposed F₀ fathers and from the nonexposed mothers (Fig. 2A and Table 1). Thus, we conclude that paternal exposure to anticancer drugs results in transgenerational instability, which affects alleles derived from both parents.

Table 1.

Summary of ESTR mutation data

Tissue, group	Mutations, n	Progenitors, n	Frequency ± SE.	Ratio*	t†	P†
Sperm
Control
Paternal allele	32	409	0.0782 ± 0.0145	—	—	—
Maternal allele	19	437	0.0435 ± 0.0103	—	—	—
Total	51	846	0.0603 ± 0.0088	—	—	—
CPP
Paternal allele	44	291	0.1510 ± 0.0248	1.93	2.53	0.0116
Maternal allele	24	258	0.0929 ± 0.0200	2.14	2.20	0.0281
Total	68	550	0.1237 ± 0.0161	2.05	3.46	0.0006
MMC
Paternal allele	24	205	0.1168 ± 0.0254	1.49	1.32	0.1873
Maternal allele	32	268	0.1192 ± 0.0226	2.74	3.06	0.0023
Total	56	474	0.1182 ± 0.0169	1.96	3.04	0.0024
PCH
Paternal allele	22	182	0.1206 ± 0.0274	1.54	1.36	0.1101
Maternal allele	22	218	0.1011 ± 0.0228	2.32	2.30	0.0218
Total	44	400	0.1100 ± 0.0176	1.82	2.53	0.0115
Bone marrow
Control
Paternal allele	25	403	0.0621 ± 0.0129	—	—	—
Maternal allele	26	483	0.0538 ± 0.0109	—	—	—
Total	51	886	0.0576 ± 0.0083	—	—	—
CPP
Paternal allele	48	405	0.1185 ± 0.0186	1.91	2.49	0.0130
Maternal allele	39	404	0.0966 ± 0.0165	1.79	2.16	0.0310
Total	87	809	0.1076 ± 0.0124	1.87	3.34	0.0009
MMC
Paternal allele	37	324	0.1141 ± 0.0200	1.84	2.18	0.0296
Maternal allele	34	351	0.0968 ± 0.0176	1.80	2.08	0.0378
Total	71	675	0.1051 ± 0.0133	1.83	3.03	0.0025
PCH
Paternal allele	34	311	0.1095 ± 0.0201	1.76	1.99	0.0470
Maternal allele	32	321	0.0996 ± 0.0187	1.85	2.11	0.0352
Total	66	632	0.1045 ± 0.0137	1.81	2.92	0.0036

*Ratio of the corresponding ESTR mutation frequency in control.

^†Student test and probability of difference from the control group.

Fig. 2.

The effects of paternal exposure on ESTR mutation frequencies in the germ-line and somatic tissues in their offspring. (A) The frequency of ESTR mutation at the alleles derived from the exposed fathers and the nonexposed mothers. (B) Transgenerational instability in the offspring of male mice exposed to ionizing radiation (8), ethylnitrosourea (10) and three anticancer drugs. SEs are shown.

Discussion

Our study was specifically designed to establish the transgenerational effects of paternal exposure to anticancer drugs. Our analysis of ESTR mutation frequencies in the offspring of male mice exposed to clinically relevant doses of three widely used anticancer drugs revealed significantly elevated F₁ germ-line and somatic mutation rates. These results provide important clues about the mechanism of transgenerational instability and may have far-reaching implications for the evaluation of long-term effects of human exposure to anticancer drugs.

Our analysis of ESTR mutation frequencies in the F₁ offspring of male mice exposed to a variety of mutagens demonstrates several striking similarities among the transgenerational effects of paternal exposure to ionizing radiation, chemical mutagens, and anticancer drugs. First, paternal exposure to all mutagens analyzed to date can destabilize F₁ genomes (Fig. 2B). Thus, our data support our previously reported notion that that transgenerational instability is not attributable to specific subset of DNA lesions, such as double-strand DNA breaks, but rather is triggered by generalized DNA damage in male germ cells (10). Indeed, irradiation results in a wide spectrum of DNA lesions, ranging from base damage to double-strand DNA breaks (19), whereas exposure to ethylnitrosourea mainly causes alkylation of DNA (20). This also holds true for exposure to the anticancer drugs analyzed in the present study. Phosphotriester adducts constitute the most frequent alterations by CPP (21), alkylated monoadducts and cross-linked adducts dominate in the spectrum of MMC-damaged sites (22), and PCH metabolites inhibit DNA polymerase and also react directly with DNA (23). Thus, it appears that relatively high-dose acute exposure to various germ-line mutagens can result in transgenerational instability manifesting in the offspring.

Second, according to our data, the F₁ mutation rate is equally elevated in two tissues (sperm and bone marrow), as well as at the alleles derived from the exposed fathers and the nonexposed mothers. These data are in line with the results of previous studies on the transgenerational effects of paternal irradiation showing significantly increased ESTR mutation frequencies in the germ-line and somatic tissues in the offspring of irradiated male mice (8, 9, 24); the same tissue-wide destabilization of the F₁ genome has been demonstrated in the present study. Moreover, other recent studies have shown a similarly elevated mutation rate in the offspring of irradiated male mice at the alleles derived from exposed and nonexposed parents (7, 8, 25); the same holds for the offspring of anticancer drug-treated males. Thus, we conclude that the transgenerational effects are attributed to a genome-wide destabilization manifesting in many, possibly all, tissues.

Finally, according to the results of our previous studies, transgenerational instability in the offspring of irradiated male mice is an epigenetic phenomenon (7–9). This conclusion is based mainly on the fact that vast majority of the F₁ offspring show similar levels of instability in numerous tissues (8). Given the very low likelihood that even a few offspring may carry de novo mutations affecting the same set of genes involved in maintaining genome stability, this observation rules out the possibility that transgenerational effects are due to Mendelian segregation of radiation-induced mutations arising in the germ line of exposed males, thus suggesting that the radiation-induced signal leading to genomic instability is inherited in an epigenetic fashion. Our current data also show that ESTR mutation frequencies are equally elevated in all of the offspring of male mice exposed to anticancer drugs, thus implying that similar epigenetic mechanisms may underlie the transgenerational effects after paternal exposure to ionizing radiation and chemotherapeutic agents. Of note in this regard are the results of a recent study showing that exposure to chemotherapy regimens used for testicular cancer treatment can alter the pattern of DNA methylation in spermatozoa of treated rats (26), which may potentially survive the reprogramming during early development and alter the F₁ epigenetic landscape.

Our data showing significantly elevated F₁ germ-line and somatic mutation rates at a noncoding tandem repeat DNA locus raises the important question of whether transgenerational changes can also destabilize protein-coding genes and thereby affect health-related traits in the offspring of exposed parents. Although this remains to be established, some previous studies suggest that paternal exposure to anticancer drugs may lead to a genome-wide destabilization. First, a significantly increased rate of dominant lethal mutations has been reported in the F₁ offspring of male rats exposed to CPP (11). Second, the results of numerous animal studies provide strong evidence of health-related transgenerational effects of paternal irradiation, including elevated frequencies of mutation at protein-coding genes (8, 25) and chromosome aberrations (reviewed in ref. 27). Given the aforementioned similarities between the transgenerational effects of ionizing radiation and anticancer drugs, it appears plausible that paternal exposure to anticancer drugs also might result in a genome-wide destabilization.

In conclusion, our present and previous findings (15) are important to increasing our understanding of the genetic mechanisms underlying the long-term effects of human exposure to anticancer drugs. Here, as well in a previous study (15), we analyzed the effects of paternal exposure to the clinically relevant doses of anticancer drugs on mutation induction and transgenerational instability in mice. The results of the two studies clearly show that exposure to a number of widely used anticancer drugs results in mutation induction in the germ line of directly exposed males and can destabilize the genomes of their offspring. These data raise the important issue of delayed transgenerational effects in the children of cancer survivors. The genetic effects of human exposure to anticancer drugs remain poorly understood (28, 29), and our data underscore the importance of further studies aimed at establishing the magnitude of mutation induction in the germ line of anticancer therapy survivors, as well as transgenerational instability in their children.

Materials and Methods

Animals.

Male 6-wk-old CBA/Ca and BALB/c inbred mice were obtained from Harlan. Each animal received a single dose of the test compound, dissolved in PBS and delivered via i.p. injection, at age 8 wk. Then, 8 wk later, these male mice were mated with nonexposed 8-wk-old females. Tissues, including caudal epididimy and bone marrow, were subsequently collected from the 8-wk-old male hybrid F₁ offspring of exposed and nonexposed control mice. All animal procedures were carried out under Home Office Project License PPL 80/2267.

Chemicals.

CPP and MMC were obtained from Sigma-Aldrich. PCH was obtained from Sequoia Research Products.

DNA Preparation.

Sperm DNA and bone marrow samples were prepared in a laminar flow hood as described previously (8, 15). Approximately 5 μg of each DNA sample was digested with 20 U of MseI for 2 h at 37 °C. (MseI cleaves outside the ESTR array and distal to the PCR primer sites used for PCR amplification and was used to render genomic DNA fully soluble before dilution.) Each digested DNA sample was diluted to ∼10 ng mL⁻¹ before mutation analysis.

Mutation Detection.

The frequency of ESTR mutation was evaluated with SM-PCR (8, 15, 30). The Ms6-hm ESTR locus was amplified in 10-μL reactions using 0.4 μM flanking primers HM1.1F (5′-AGA GTT TCT AGT TGC TGT GA-3′) and HM1.1R (5′-GAG AGT CAG TTC TAA GGC AT-3′). The Roche Expand High-Fidelity PCR System was used (0.035 U μL⁻¹) with 1 M betaine and 200 μM of each dNTP. Amplification was carried out with 30 cycles of 96 °C for 20 s, 58 °C for 30 s, and 68 °C for 8 min on a PTC-225 Tetrad DNA engine (MJ Research). To increase the robustness of the estimates of individual ESTR mutation frequencies, an average of 145 amplifiable molecules were analyzed for each tissue in each male mouse.

PCR products were resolved on a 40-cm-long agarose gel and detected by Southern blot hybridization as described previously (31). The frequencies of ESTR mutation were estimated using an approach proposed by Chakraborty (29) with modifications.