Chromosome Fragile Sites in Mentally Retarded Males: Increased Incidence with Seizures and Diphenylhydantoin Therapy

Abstract

Chromosome fragile site or lesion data were examined in 154 institutionalized mentally retarded males with or without seizures or treated with anti-seizure medication. Blood lymphocytes were cultured using three different cell culture conditions and the incidence of specific chromosome fragile sites (10q25, 16q22, and 12q23) or lesions determined. Increased fragile sites were seen in mentally retarded males with seizures compared to those without seizures in cells grown in folate-deplete Medium 199. Those with seizures and treated with diphenylhydantoin had a higher incidence of induced fragile sites (p < 0.001) relative to similar patients treated with anti-seizure medication other than diphenylhydantoin. These results suggest that a cohort of patients with mental retardation and seizures are more likely to have induced cytogenetic changes when treated with diphenylhydantoin than mentally retarded individuals without seizures.

Introduction

Fragile sites represent areas of chromosomes that are prone to breakage. They occur spontaneously or may be induced by certain culture conditions. Approximately 107 chromosome fragile sites have been described and classified into two general categories: common (83 of the 107 fragile sites) and rare (remaining 24 of the 107) fragile sites.¹ Expression of common fragile sites are induced by sub-stances such as 5-azacytidine and 5-bro-modeoxyuridine (BrdU), whereas inhibitors of thymidylate synthetase, such as 2′-deoxy-5-fluorouridine (FUdR) induces rare fragile sites.

Specific chromosomal abnormalities have been described in patients with seizures; however, whether or not this relationship is the result of medication or seizures is not clearly defined. Kitsiou-Tzeli et al,² in an attempt to evaluate whether seizures induced chromosomal abnormalities, studied 8 untreated children with seizures and showed no significant increase in the incidence of chromosomal breakages compared to normal controls.

In addition, a number of therapeutic drugs including diphenylhydantoin have been shown to be mutagenic and induce chromosomal abnormalities in vitro, but chromosomal abnormalities associated with anti-seizure medication in patient studies have shown conflicting results. Two small series of epileptic patients on long-term anticonvulsant therapy failed to show an association with cytogenetic abnormalities or sister chromatid exchanges.^3,4 Kitsiou-Tzeli et al² reported 36 epileptic children undergoing long-term anticonvulsant therapy showed no significant cytogenetic effects. On the other hand, other investigations using similar patient populations and sample size did show cytogenetic changes with anticonvulsant drugs.^5,6 In addition, Musumeci⁷ and personal communication reported that cells grown in BrdU culture conditions showed chromosome lesions at chromosome bands 10q25, 16q22, and 12q23 in 8 to 25 percent of mentally retarded patients with epilepsy.

Therefore, to define more clearly the relationship of seizures, anti-seizure medication, and genetic alterations, the incidence of fragile sites or lesions induced by several culture conditions in a relatively large number of mentally retarded individuals with and without seizures was compared with individuals with seizures treated with either diphenylhydantoin or other anti-seizure medication.

Materials and Methods

Patient Population

154 institutionalized mentally retarded males (age range from 8 to 76 years, average age of 40 years) previously screened for fragile X syndrome, and with no known cause of mental retardation were identified. The protocol for this study was approved by the Vanderbilt University Medical Center (VUMC) Institutional Review Board for Human Research and the Cloverbottom Developmental Center Human Review Committee. Medical records were assessed for age, race, history of seizures, and anti-seizure medications. In order to analyze a similar number of cells grown in different culture conditions, a subset of the 154 mentally retarded males was used to compare fragile sites induced bv folate-deficient culture condition (Medium 199) or folate replete Roswell Park Memorial Institute (RPMI) 1640 culture medium supplemented with FUdR or BrdU following established protocols.

Cytogenetic Studies

Peripheral blood lymphocytes were cultured in folate-deficient Medium 199^* supplemented with 5 percent fetal bovine serum^* inactivated at 65°C, for 30 minutes, phytohemagglutinin (M-form^*) and penicillin/streptomycin^* with the pH adjusted to 7.8 at the time of culture preparation. The cultures were incubated at 37°C for 96 hours.

Peripheral blood cultures were also established in folate-replete RPMI 1640 medium supplemented with 5 percent fetal bovine serum, phytohemagglutin (M-form) and antibiotics (penicillin/streptomycin). Cultures were incubated at 37°C for 96 hours, and exposed to FUdR (10⁻⁷ M) 24 hours prior to harvest. Samples cultured in RPMI 1640 medium supplemented with 5 percent fetal bovine serum, phytohemagglutin (M-form) and antibiotics (penicillin/streptomycin) were also incubated at 37°C for 72 hours and exposed to 30 mg/L of BrdU^† 6 hours prior to harvest. Colcemid (0.1 μg/ml)^* was added to each of the cultures 45 minutes prior to cell harvest. Cells were harvested following treatment with 0.56 percent KCl solution at 37°C for 8 minutes, fixed with 3:1 methanol/acetic acid solution and washed four times.

Chromosome slides were prepared and stained with Giemsa using conventional methods. A minimum of 525 cells (i.e., 975 from Medium 199, 525 from RPMI 1640 with FUdR, and 700 from RPMI 1640 with BrdU) were analyzed by the same observer from each subject for chromosome fragile sites or lesions. The location of the fragile sites or lesions were identified after destaining and GTG banding by comparing photographic images of the uniformly stained chromosomes with the same chromosomes after banding at the microscopic level.⁸ Fragile sites or lesions were scored as gaps (width of a chromatid) or breaks (greater than the width of a chromatid or out of the plane) involving one or both chromatids. The location and frequency of all fragile sites were recorded on chromosome idiograms, and specific sites (e.g., 10q25, 16q22, and 12q23) were compared. Statistical analysis was performed using Chi-square test with Yates’ correction.

Results

Patient Population

154 mentally retarded individuals were identified with an age range of 8 to 76 years, mean of 40 years (table I) and 86 (56 percent) of the patients had seizures while the remaining 68 (44 percent) did not. All were male, and the majority were Caucasians; 45 percent were successfully treated with diphenylhydantoin monotherapy, and 55 percent were treated with other anti-seizure medication.

TABLE I.

Mentally Retarded Male Patients: Clinical Features

	Total	With Seizures (Percent)	Without Seizures (Percent)
Number of patients	154	86 (56)	68 (44)
Mean ages (years)	40	40	41
Range of ages (years)	8–76	9–76	8–75
Number of blacks	33	15 (45)	18 (55)
Number of Caucasians	120	70 (58)	50 (42)
Number of hispanics	1	1 (100)	0 (0)

Absolute numbers are shown.

Cytogenetic Studies

Comparison of fragile sites in patients with seizures versus without seizures

Of the 86 patients with seizures, 47, 75, and 74 were studied using BrdU, FUdR, and M199 culture conditions, respectively (table II). Exactly 10.6 percent of the cells cultured in BrdU showed fragile sites with an average of 5.3 fragile sites per patient. 30.1 percent of the cells cultured in FUdR showed fragile sites with an average of 8.2 fragile sites per patient, and 8.5 percent of the cells cultured in M199 showed fragile sites with an average of 4.3 fragile sites per patient.

TABLE II.

Incidence of Cells with Fragile Sites in Patients with and without Seizures

CC	Patients with Seizures			Patients without Seizures			p–valueb
	Cases	FS(%)a	Ave # FS/Pt	Cases	FS(%)a	Ave # FS/Pt
BrdU	47	10.6	5.3	54	11.3	5.6	p > 0.05
FUdR	75	30.1	8.2	66	30.4	7.5	p > 0.05
M199	74	8.5	4.3	68	5.9	3.1	p < 0.05

CC = Culture condition.

FS = Fragile sites.

Pt = Patient.

^aPercent of cells with fragile sites were determined by dividing the number of cells with fragile sites by the total of cells observed in that culture condition.

^bp–value comparison of fragile sites percentage between patients with or without seizures (chi–square test with Yates’ correction).

Of the 68 patients without seizures, 54, 66, and 68 were studied using BrdU, FUdR, and M199 culture conditions, respectively. 11.3 percent of the cells cultured in BrdU showed fragile sites with an average of 5.6 fragile sites per patient. 30.4 percent of the cells cultured in FUdR showed fragile sites with an average of 7.5 fragile sites per patient, and 5.9 percent of cells cultured in M199 showed fragile sites with an average of 3.1 fragile sites per patient.

Significantly more cells from mentally retarded males with seizures showed fragile sites compared to cells from patients without seizures when cultured in folate-deplete Medium 199 (p < 0.05; chi-square test). Significant differences in the number of fragile sites or lesions in cells between mentally retarded patients with or without seizures cultured in folate-replete RPMI exposed to BrdU or FUdR were not observed.

Comparison of fragile sites in patients with seizures treated with diphenylhydantoin versus other anti-seizure medication

Of the 26 patients with seizures treated with diphenylhydantoin, 16, 25 and 25 were studied using BrdU, FUdR and M199 culture conditions, respectively (table III). 15.1 percent of the cells cultured in BrdU showed fragile sites with an average of 7.6 fragile sites per patient. 32.4 percent of the cells cultured in FUdR showed fragile sites with an average of 8.8 fragile sites per patient, and 9.2 percent of the cells cultured in M199 showed fragile sites with an average of 4.8 fragile sites per cell.

TABLE III.

Incidence of Cells with Fragile Sites in Patients with Seizures and Treated with Diphenyhydantoin Compared to Other Anti–seizure Medication

CC	Patients Treated with DPH			Patients Treated with Other Medication			P–value
	Cases	FS(%)a	Ave # FS/Pt	Cases	FS(%)a	Ave # FS/Pt
BrdU	16	15.1	7.6	14	6.9	3.4	p < 0.001
FUdR	25	32.4	8.8	21	38.0	9.5	p > 0.05
M199	25	9.2	4.8	19	9.9	5.1	p > 0.05

DPH = Diphenylhydantoin.

CC = Culture condition.

FS = Fragile sites.

Pt = Patient.

^aPercent of cells with fragile sites were determined by dividing the number of cells with fragile sites by the total number of cells observed in that culture condition.

^bp–value comparison of fragile sites percentage between patients treated with diphenylhydantoin or other medication (chi–square test with Yates’ correction).

Of the 25 patients with seizures treated with medication other than diphenylhydantoin, 14, 21, and 19 were studied using BrdU, FUdR, and M199 culture conditions, respectively. 6.9 percent of the cells cultured in BrdU showed fragile sites with an average of 3.4 fragile sites per patient. 38 percent of the cells cultured in FUdR showed fragile sites with an average of 9.5 fragile sites per patient. 9.9 percent of the cells cultured in M199 showed fragile site with an average of 5.1 fragile sites per cell.

Significantly more cells from mentally retarded males with seizures and treated with diphenylhydantoin showed fragile sites compared to cells from patients treated with medications other than diphenylhydantoin when cultured in folate-replete RPMI 1640 exposed to BrdU (15.1 percent vs 6.9 percent, p < 0.001). Significant differences were not seen in the number of fragile sites or lesions in cells grown in Medium 199 or RPMI with FUdR from patients with seizures on diphenylhydantoin compared to patients with seizures on medication other than diphenylhydantoin.

Comparison of fragile sites at 10q25, 16q22 or 12q23 in patients with seizures or without seizures

To determine if mentally retarded patients with seizures and treated with anti-seizure medications had a different percentage of fragile sites at 10q25, 16q22, or 12q23 compared to mentally retarded patients without seizures as reported by Musumeci⁷ and personal communication, 86 patients with seizures were studied and 48 were analyzed in BrdU culture conditions for 10q25, 16q22, and 12q23 fragile sites. 8.6 percent of the cells cultured in BrdU showed fragile sites at 10q25, 16q22, or 12q23. In addition, of the 68 patients without seizures 54 were studied using BrdU culture conditions for 10q25, 16q22, or 12q23 fragile sites. 18.3 percent of the cells cultured in BrdU showed fragile sites at 10q25, 16q22, or 12q23 (table IV).

TABLE IV.

Incidence of Cells with Fragile Sites at 10q25, 16q22 or 12q23 in BrdU Culture Conditions in Patients with and without S\eizures

Sites	48 Patients with Seizures			54 Patients without Seizures			P–valueb
	Total #FS/Total # Cells Observed	# FS	FS(%)a	Total #FS/Total # Cells Observed	# FS	FS(%)a
	255/2397			301/2669
10q25, 10q22 or 12q23		22	8.6		55	18.3	p<0.01
10q25		7	2.7		23	7.6
16q11		14	5.4		32	10.6
12q23		1	0.4		0	0

FS = Fragile sites.

^aPercent of cells with fragile sites were determined by dividing the number of cells with fragile sites at each site(s) by the total number of cells with fragile sites in BrdU culture conditions.

^bp–value comparison of fragile sites percentage between patients with siezures or without seizures (chi–square test with Yates’ correction).

Significantly more cells from mentally retarded males without seizures showed induced fragile sites at 10q25, 16q22, or 12q23 compared to cells from patients with seizures when cultured in folate-replete RPMI 1640 exposed to BrdU (p < 0.01).

Discussion

The frequency of fragile sites or lesions in individuals with seizures has not been previously documented, although there have been a number of cytogenetic studies correlating behavioral findings with chromosomal fragility including schizophrenia,^9,10 fragile X syndrome,^11,12 Prader-Willi syndrome,^13,14 Rett syndrome,¹⁵ and Tourette syndrome.¹⁶ Our data indicate that cells from mentally retarded males with seizures, when exposed to Medium 199 with reduced folate levels, had significantly more fragile site expression than those from mentally retarded patients without seizures. No difference in fragile site expression was found in lymphocytes grown in folate-replete conditions containing BrdU or FUdR, agents known to induce chromosome fragility. In addition, it was found that a small but significant increase in number of total BrdU chromosome breaks in folate-replete culture condition in mentally retarded males with seizures on diphenylhydantoin compared to mentally retarded patients with seizures and treated with medications other than diphenylhydantoin.

Clinical differences among the mentally retarded males with increased fragile site expression and mentally retarded males without increased fragile site expression were not apparent. Therefore, the small difference between groups may indicate that a cohort of seizure patients or patients treated with diphenylhydantoin are at risk for increased fragile sites, or alternatively all seizure patients with mental retardation are at a slightly increased risk. In addition, Musumeci⁷ and personal communication reported that cells grown in BrdU culture conditions showed 10q25, 16q22, and 12q23 sites in 8 to 25 percent of mentally retarded patients with epilepsy, while our data showed mentally retarded males without seizures had a significantly higher incidence of fragile sites at 10q25, 16q22, and 12q23 (p < 0.01) relative to similar patients with seizures. Whether this discrepancy is due to phenotypic differences of the patients or differences in treatment is not known.

Previous studies have been somewhat contradictory, and investigators have had difficulty showing an association between chromosomal abnormalities, seizures, and medication; although anticonvulsants including primidone, diphenylhydantoin, and carbamazepine have been implicated as genotoxic and therefore may cause chromosome abnormalities.¹⁷ Investigations undertaken by Esser et al¹⁸ in epileptic children during long-term therapy with phenytoin and primidone showed no increase of structural or numerical chromosome abnormalities.

In addition, Goyle et al⁴ were unable to demonstrate significant differences in sister chromatid exchange frequency, a sensitive indicator of early chromosome changes, in epileptic patients before and after anticonvulsant therapy. These findings suggest that chromosome fragility may be a more sensitive measure of chromosome changes than classical karyotype analysis or sister chromatid exchange frequency.

Furthermore, the specific induced chromosome abnormality may be drug dependent as Kitsiou-Tzeli et al² did not find significant differences in chromatid or chromosome breakage rate or sister chromatid exchange frequency in children treated with anticonvulsant therapy of phenobarbitol, valproic acid, or carbamazepine. Unfortunately, the latter study did not evaluate diphenylhydantoin as in our study. Habedank et al,⁵ however, reported increased sister chromatid exchanges in epileptic children during long-term therapy with phenytoin.

The increased chromosome fragility or lesions seen in our study culture conditions may mimic a physiologic pathway leading to chromosome breakage in treated patients. Diphenylhydantoin is known to decrease plasma or serum folate levels. Therefore, increased chromosome breakage in patients may occur through a pathway that in part involves the effective reduction of folate metabolism in patients on diphenylhydantoin.

In summary, our study does support increased chromosome fragility of mentally-retarded individuals with seizures and those treated with diphenylhydantoin. The use of folate deficient culture condition was also important in detecting these effects. These findings may have important clinical implications in the treatment of seizure patients with diphenylhydantoin.