Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
GENOTOXICITY OF FLUORIDE Fluoride has been tested extensively for its genotoxicity. There are a number of published reports on the genotoxicity of fluoride in microbes, cultured mammalian cells, and animals. These data are summarized in Tables 6-! through 6-6. IN VITRO GENOTOXICITY TEST SYSTEMS Microbes NaF has been tested extensively for its ability to induce gene mutations in Ames Salmonella typhimurium reverse mutation assay by standard plate and preincubation tests and in other microbial systems, with and without metabolic activation at concentrations ranging from 0. ~ to 4,421 ,ug/plate Fable 6-~. The results were negative (Litton Bionetics, 1975; Martin et al., 1979; Gocke et al., 1981; Haworth et al., 1983; ArIauskas et al., 1985; Li et al., 1987a; Tong et al., 19X8~. However, in a suspen- sion assay (a modification of the standard Ames plate test), Nikiforova (19X2) reported positive findings in S. typhimurium strains TAl535 and TA98 at concentrations of I,000-l,500 ~g/plate. The reported increases in histidine-revertants (mutations) in S. typhimurium strains TAl535 and TA98 appear to be artifactual results of increased cell killing, which 91
92 C) a, - U) a, Do ox ~oo oo ~ ~ Ins °° °\ A c) ox ~ ~ o' _1 00 - ~ '1 ~ ~ ~ S ~ ~^ 5.~=a=~ S ~ ~ Ha . ; to -, ~ ~ ~ S ~ 5 o p Dais ' =°t 1~ ~ ~ =O ~ s a) - o ;^ · C) .g 01 o ._ - m [~) ~ 1 1 1 1 1 1 |.a a ~ a a ~ 8 3 ~ ~ ~ =~= ~ - ~ ~ ~ ~ ~ o o o o o - ~ an a At - Do _ an .= O CQ ~ ~o o i. i^~ t~ ~it o ~ c) - ~ ·-o - - - a] - c) .> - ~
93 . ._ ca a) ._ ._ ._ 4 - _ _ ._ ._ ._ In ~ A o o o al .~ of _ ~ 04 to ~ ~ ~ 8 o ~ ~ ~ _. oo Go o ~ .~§ C' ~ _ ~ C, ~ ._ ~ o ~ _ ~ A ~ ID d] V) - 3 ~ CQ Q ·e ~ ._ U) O _ ;~` ~ C) ._ ca or · 5 ~ ~ A ~ ~ .5 11 11 11 PA Bear 11 11
94 Health Elects of Ingested Fluoride might have increased histidine and generated small background colonies. NaF was not mutagenic to Saccharomyces cerevisiae strain D4 Litton Bionetics, 1975~. Table 6-l summarizes mutagenicity data of fluoride in microbial organisms and mammalian cells in culture. NaF was found to be negative in the Bacillus subtilis rec assay (Matsui, 1980), a test that measures DNA damage. However, Kanemat- su (1985) reporter! that both NaF and stannous fluoride (SnF2) were positive. The differing results could be due to the differing protocols used by the two investigators. NaF failed to induce gene conversion and aneuploidy in S. cerevisiae Litton Bionetics, 1975; Martin et al., 1979~; similar results were re- ported for potassium fluoride (KF) when tested in Neurospora (Griff~ths, 1981~. Mammalian Cells Fluoride has been tested in in vitro mammalian cell cultures for its ability to induce mutations, chromosomal aberrations, sister chromatic exchanges, DNA damage and repair, and cell transformation. Gene Mutations Fluoride has been tested for its mutagenicity in several in vitro mam- malian cell systems with ant} without metabolic activation Table 6-~. NaF and KF were strongly mutagenic in the mouse lymphoma L5178Y TK+~- test with and without S9 at concentrations ranging from 10 to 600 ,ug/mL (Cole et al., 1986; Caspary et al., 1987~; the authors speculated that the induced mutant colonies resulted from chromosomal damage ratherthan point (gene) mutations. This hypothesis was supported by the absence of induction of ouabain-resistant mutants in the same cell type (Cole et al., 1986~. Caspary et al. (1988) reported that NaF was muta- genie at the tk locus in human lymphoblastoid cells treated with NaF at 200-600 ~g/mL. for 4 hours. Recently, Crespi et al. (1990) reported that NaF was mutagenic at the tk and hgprt loci in human lymphoblastoid cells treated with NaF at concentrations of 200-600 ~g/mL for 28 hours and at the tk locus in ceils treated with 65 ~g/mL for 20 clays. Howev
Genotoxicity of Fluoride 95 er, a statistically significant response was observed only at concentrations that resulted in substantial cell death. In contrast, no mutagenicity was observed at the hgprt locus in rat liver epithelial cells treated with NaF at concentrations of 2-40 ~g/mL for 72 hours (Tong et al., 1988) or at the 6-tg locus in Chinese hamster ovary V79 cells treated with NaF at 10-400 ~g/mL for 24 hours (Slamenova et al., 1992~. Chromosomal Aberrations and Sister Chromatid Exchanges NaF has been shown to induce chromosomal damage in several in vitro mammalian cell systems (Table 6-2~. The reported chromosomal effects were primarily chromatic deletions or achromatic lesions (gaps); the definition and scoring of the latter events is not standardized, and their significance is unknown, and, in fact, questionable. These effects were not always clearly demonstrated and appeared to be protocol-clepen- dent Hi et al., 1988; Aardema et al., 1989~. Animal Cells Chromosomal aberrations were increaser! by 3 hours of exposure to NaF at concentrations of 25-100 ~g/mL in Chinese ham- ster ovary cells (but not at concentrations of 0. I-10 ~g/mL) (Aardema et al., 1989), at 50-200 ~g/mL in Syrian hamster embryo cells (Tsutsui et al., 1984a), and at 12.6-126 ~g/ml~ in red muntjac cells Tie et al., 1983~. Aarclema et al. (1989) concluded that the G2 stage of the cell cycle is a sensitive stage for NaF-induced chromosomal damage in Chi- nese hamster ovary cells. Chromosomal abnormalities also were reported in Chinese hamster ovary Don cells treated with NaF at 25-75 ~g/mL for 12-36 hours (Bale and Mathew, 1987~. The National Toxicology Pro- gram studies (NTP, 1990) in Chinese hamster ovary cells shower! no induction of aberrations in one laboratory where NaF was tested at con- centrations up to 200 ~g/mL without S9 ant! harvester] after 20/ hours, but a positive response was reported in a second laboratory at concentra- tions of 400-600 ~g/mL with a shorter (13 hours) harvest time. No chromosomal aberrations were induced with metabolic activation at concentrations up to 1,600 ,ug/mL at either harvest time (NTP, 1990~. No chromosomal aberrations were induced in Chinese hamster lung cells by NaF at concentrations up to 500,ug/mL (Ishidate, 19884.
96 A - .= as - a, as 2 ~ LL] · ~ o LL1 - - o a lo Do Cal _ Cal u' Ha o o ._ ._ ~_. ID \0 _ _ 01 1 Go 1. . oCal ~ in _ _ In Cal Cal - o o ._ ._ ~ 3 o .~ A C) .~ ~e u ~ ' ~ ' g ~ i 8 ;~ 5" ~ ~ 3o^ o = c~ _ ~ oo td~ - ~ O O _ ^ ^ ~ ~ ~·= f ~ ~ ~ ~ ~ 0 . ~. ~. ~. ~ . ~ . ~. ~ . ~ 0 0 0 0 g g g g ~ P" 0 0 0 0 _ 1 - o 0 0 ~ 0 0 0 ~ _ 1 1 LD c: 0 0 ._ ._ _ ~ c~ e ' a;' ~ CQ u ~ ~ o u) ^, CQ V] o ~ ~ o o ,r, ~D · - · -
97 ~ oo Go ox ~ ~ ~ ~ ~ ~ ^ ~ ~ y ~ S ,~ ~ ~ ~ ~ ~ ~ ~" ~a) an ~ a, At · ~ ~ ~ . ~ . ~ _ ~) 04 be 04 to 01 0 ~a, ~ a, ~ ~ Z ~ Z Z ~ Z o _. ~ Go ~ ~ ~ ~ o , ~ ~ Go o ~ ~ ~ ~ ~Aca ooo .~._._ CalCalCal a) ID ~Go Go cat A A o o o ._ ._ ._ _ ~ a' ~c. At ~ ~ ~ ~D U. CQ CQ ~q CQ CQ ~q ·_, o C) a~ U' · _. 11 U.
98 Health Effects of Ingested Fluoride Sister chromatic exchanges were not induced by NaF in Chinese hamster ovary cells treated with 250 ~g/mL for 24 hours Hi et al., 1987b) or with 160 ~g/mL for 24 hours (Tong et al., 1988~. However, sister chromatic exchanges were induced in Syrian hamster embryo cells treated with 20-80 ~g/mL without S9 (T'sutsui et al., 1984a), and in red muntjac cells treated with 126 ,ug/mL (He et al., 1983~. In the NTP studies, the incidence of sister chromatic exchanges was increased in Chinese hamster ovary cells treated with NaF at concentrations up to 1,600 ,ug/mlL with S9. Human Cells Several investigators have reported chromosomal aberrations in cultured human lymphocytes and fibroblasts at NaF con- centrations ranging from 20 to 40 ,ug/mL~ (Tsutsui, et al., 1984b; Al- banese, 19X7; Scott and Roberts, 1987~. Chromosomal aberrations were also observed in human leukocytes at concentrations ranging from ~ to 132 ~g/mL Qachimczak and Skotarczak, 1978~. Sato et al. (1989) reported the induction of chromosomal gaps, but not breaks or rearrange- ments, in human lymphocytes treated with fluoride at concentrations up to 44,ug/mL. However, other investigators did not observe chromosomal aberrations in human lymphocytes and leukocytes exposed in vitro to NaF at concentrations up to 125 ,ug/mL (Voroshilin et al., 1975; Kraiisz and Szymaniak, 1978; Matsuda, 1980; Gebhart et al., 1984~. Sister chrom- atid exchanges were also not observed in human lymphocytes exposed to NaF at concentrations up to 420 ,ug/mL or KF at concentrations up to 580 ~g/mL Thomson et al., 1985; Tong et al., 19881. DNA Damage and Repair Several investigators have studied the induction of DNA repair synthe- sis and unscheduled DNA synthesis in various in vitro mammalian cell systems (Table 6-3~. NaF has been shown to inhibit protein and DNA synthesis in cultured mammalian cells Holland, 1979a,b; Li et al., 19XX); the inhibition of DNA synthesis might be a secondary effect of the inhibition of protein synthesis or a direct inhibition of DNA polymerase (Skare et al., 1986a; Holland 1979a,b; Imai et al., 19831. NaF failed to induce DNA repair synthesis in primary rat hepatocytes at concentrations up to 160 ~g/mL (Tong et al., 1988~. Skare et al. (1986a) did not
Ger~otoxicity of Fluoride 99 TABLE 6-3 Effect of Fluoride Exposure on Induction of DNA Repair Synthesis and Unscheduled DNA Synthesis 8-500 8-120 10-40 System DNA repair synthesis 18 in rat hepatocytes UDS in human 1.5-8 fibroblasts UDS in rat 18-20 hepatocytes UDS in Syrian 12 hamster embryo cells UDS in human oral keratinocytes Exposure Flounde, Time, hr ,ug/mL 2-160 Result Negative Reference Tong et al., 1988 Skare et al., 1986a Skare et al., 1986a Tsutsui et al., 1984a Negative Negative Positive 4 100-300 Positive Tsutsui et al., 1984c UDS = Unscheduled DNA synthesis. observe unscheduled DNA synthesis in human fibroblasts treated with NaF at 8-500 ~g/mL. NaF at concentrations of 10-300 ~g/mL inclucect unscheduled DNA synthesis in Syrian hamster cells (Tsutsui et al., 1984a) and in human keratinocytes (Tsutsui et al., 1984c), but those results were not confirmed by other investigators who used concentrations that clid not induce high levels of toxicity (Skare et al., 1986a; Tong et al., 1988~. Transformation Transformation is a process that changes a normal cell into one that is capable of forming a tumor that might or might not be malignant. The main event that initiates transformation is a change in the genetic materi- al. Several investigators have studied the ability of NaF to transform cells in culture (see Table 6-4~. Dose-related increases in the frequencies of transformed colonies were observed in Syrian hamster embryo cells at NaF concentrations of 10-125 ,ug/mL (Tsutsui et al., 1984a; Jones et al., 19X8a,b; Lasne et al., 1988~. Morphological transformation was not induced in BALB/3T3 cells treated with NaF for 72 hours at concentra . /
100 Health Effects of Ingested Fluoride TABLE 6 - Effect of Fluoride on Cell Transformation . Exposure Fluoride, Time, d ~g/mL 75-125 S stem End Point y Syrian hamster Morphological 1 embryo cells transformation Syrian hamster Neoplastic embryo cells transformation Syrian hamster embryo cells Syrian hamster embryo cells Syrian hamster embryo cells BALB/3T3 cells Morphological 4 transformation Morphological transformation Morphological transformation Foci Sin standard assay. bin Promotion-type assay. 75-100 10-50 7 7 3 25-125 75-125 25-50 Result Positive Positive Positive Positive Positive Reference_ Tsutsui et al., 1984a Tsutsui et al., 1984a Jones et al., 1988a Jones et al., 1988b Lasne et al., 1988 Negativea Lasne et Positiveb al., 1988 tions of 25-50 ~g/mL (Lasne et al., 1988~. Syrian hamster embryo cells transformed by NaF at 75 or 100 ~g/mL and injected into newborn Syrian hamsters produced tumors at the site of injection after 141-320 days (Tsutsui et al., 1984a). Histological examination of the tumors formed in vivo revealed that the tumors were anaplastic fibrosarcomas. It should be noted that hamster embryo cells are unusually sensitive to the induction of transformation and are not considered to be representa- tive of the transformation susceptibility of other cell types. Thus, the overall significance of the fluoride transformation data are subject to question. Discussion If there is a threshold concentration of fluoride for genotoxicity in in vitro tests, how does it compare with concentrations found in humans? The results of mammalian-cell culture studies show a lowest effective dose at approximately 10 ,ug/mL (NaF at 10 fig is equivalent to fluoride ion at 4.5 ~g). That concentration should be compared with the steady- state concentration of fluoride ion at 0.02~.06 ~g/mL in human plasma
Genotoxicity of Fluoride 101 that exists even in areas where water supplies are fluoridated. If it is accepted that the sensitivity of human fibroblasts in vitro is a fair repre- sentation of cellular sensitivity in vivo and if a true threshold does exist, then clearly there is a large safety margin (Scott and Roberts, 19X7~. IN Vno GENOTO]UCITY TEST SYSTEMS NaF and other fluoride salts have been tested for their genotoxicity in Drosophila and rodents. As shown above in the in vitro test results, the in vivo test results are also mixed. The published data are generally weak and the descriptions of experimental protocols often fad] to provide dose-selection criteria and toxicity information, thus precluding accurate assessment of the adequacy of the test concentrations. Drosophila Several reports on the genotoxicity of NaF and other fluoride-contain- ing compounds in Drosophila have been published Gable 6-5~. Most of the studies used inadequate controls, thereby preventing assessment ofthe effect of NaF alone. However, there are a few studies that allow critical analyses of the mutagenicity of fluoride. Gerdes (1971) exposed Dro- sophilla males to hydrogen fluoride (HF) at 2.9 ppm and 4.3 ppm by inhalation and observed dose-related increases in sex-linked recessive lethal mutations. Similar results were obtained by Mitchell and Gerdes (1973) when NaF and SnF2 were administered to Drosophila by feeding in a glucose solution. Vogel (1973) reported that NaF in diet induced whole chromosomal loss and partial chromosomal loss, an indication of chromosomal breakage in postmeiotic germ cells of males. However, Gocke et al. (1981) reported that feeding of SnF2 or sodium monofluoro- phosphate (Na2FPO3) did not induce sex-linked recessive lethal mutations in Drosophila. Other studies have reported no induction of sex-linked recessive lethal mutations (Mukherjee and Sobels, 1968; MacDonald and Luker, 1980) or dominant lethality and sex-chromosome loss (Buchi and Burki, 1975), but the genetic systems used or the numbers of flies treated Or examined probably were inadequate.
102 ·~ to ._ o - o V' ·4 - o · _ a' CQ o U' off ~ ·- ,c i U) a, ^ for' ~ ~:t' ~ o, ~ o ~ in ~ Bill ,,~ O. ~ at I,, ~ i, ~ it .t it 5 ~ in,> ~ ~ ~a ~ lo,, oc~' ~ ~ G ~Cal. DO all ~C~1 on ~CM O ~ Cal ~.Cal Do ~0\ 1 Cal d. ~ sit t_ ~1 I ~ ~ ~a ~ · - · - · - · - · - · - · - · - ~ a ao ~oo Ct ~ ~ ° >~ g O ~ ;. O ~ ~ .~ 2 2 2 ~g I c ~2 ~- 2 9 2 s~ C. _ .n 11
Genotoxicity of Fluoride 103 Rodents Fluoride has been tested for its in vivo genotoxicity in mice, rats, and hamsters. The in vivo studies in rodents include tests for mutations, chromosomal aberrations, sister chromatic exchanges, DNA damage, and related genotoxic effects in germ cells (Table 6~. Somatic CeUs Induction of sister chromatic exchanges, chromosomal aberrations, and micronuclei was reported in the bone-marrow cells of mice aciministereci NaF at 10-40 mg/kg of body weight by gavage and by intraperitoneal or subcutaneous injection (Ma et al., 1986; Pati anti Bhunya, 19871. How- ever, the study of Ma et al. (1986) is presented in abstract form and no data are available; Pati and Bhunya (1987) inclucled chromatic) gaps anti breaks in their analysis of aberrations and relied largely on the gaps for their conclusions that NaF was cIastogenic. The results were dose- anti time-dependent but showed no route sensitivity. Fractionated dosing yielded weaker genotoxic response. The significance of gaps is not unclerstoocI, and they are not normally user] in aberration analysis. In contrast, no cIastogenic effects were seen in bone marrow of Swiss- Webster mice aciministerec] NaF at 50 mg/kg in feed for seven genera- tions (Kram et al., 1978~. Martin et al. (1979) also tiic! not observe increased frequency of chromosomal aberrations in bone marrow or testis cells of Swiss-Webster mice that receiver! fluoride at either 50 mg/L for at least five generations or I-100 mg/L for 6 months as compared with animals that received clistilled water. Mohamecl and Chandler (1982) a(iministerecl drinking water containing fluoride at 0, I, 5, 10, 50, 100, or 200 mg/L to BALB/c mice for 3-6 weeks. Cytological studies on bone-marrow-cell chromosomes showeci that NaF at I-200 mg/L induceci chromosomal changes in a dose-clepen- dent manner. The frequency of the inducer! chromosomal damage was significantly higher at fluoride concentrations as little as ~ mg/L. How- ever, because of the very high frequency of aberrant cells in control animals ant] uncertainty regarding the nature of aberrations scored, the validity of these findings is questionable. Li et al. (1989) conflicted a
104 o o on m Cal Do ~rig 0\ _ _ 00 ~ ~ _^ ~ ~ I'] ~ ~ '5 _ _ o a 3 · o" o Cal tm i~ 04 on ~ o o ~ ° o ~V 3 in · .n C) m m at ~ 4) ~ ~ ~ ._ a' ~ ~ ~ _ Pi ·! ·t t4 t4 Z Z 0£ 00 o ~O O ~ .~.-m.~m .w ~ O O O O Z ==P~ ~4 a 04~- 00 ees $$$ $ · 000 0 O ~q cn o 3 .~ ~ ~ ~ 3 3 ca u' ~i; - t3] j: j~} ~Q U~ ~ 3, ~ ~ ~q (O ~ 41) ~ a~ 3~n 3 fi 3 ~ oo oo ._ ~ ~ ._ a, C) ·~ ~ ~ C) ·- ~ · - 3 m . . - 00 ^00 _ - ·t z Z _ _ mo~ ~ 0 ~ o cn c, a a 2 2 ~c, ._ ._ c, .g .= 8 ~._ c ~ ^ 00 · ~ _ d _ '- ~ .t ·t Z Z 04 y a ~ ~4 0\ ~ - , ', 00 ~ · , c ~ 3 \0 o =- ~ 0 ~ 3 g .~ c. t_ ._ u, _ 8 .8 ·g ~ a, '= . .~m ~ a~ ~ 00 Oi · ~.t °s cn z ~ o g c~o o 0D 3 (Q ~ _ _ C) C) ~0, 0 ~C) ._ ._ .g C) m m . · _ · _ - 2 <,, e>4 ca ·m .g ~q 3
105 N ox en, Go - ~I,, ~ A ~C. ·uq ~ ~4 ~4 of a a, 04 to ~ ~ ~ 3 ·_. CD ~ ~ 2 8 ~ . ~ ~ U. .g CQ CD · _ 3 oo CO · _ .~ al al ~4 ho is to 04 E ~ A ~04 g 04 ~ _4 Do g ~ ~ Go _ C. ~ 3 . - C' ·a ~ a) an .~ a~ ~7 ~4 m t- 00 at ~ ._ · U) o P4 o ~: 3 ~ ~n ;^ - E ., ~ E ~ E a, a.g ~ .1 g ~ ~ ~ U -m cr' _. . .t ;> ~ . ~·~ a, t4 t4 Z ~ oo o o o .~ ~n ~ ~ oo ox . ·~ ~e ~ .a ·= ~ a ~ ;~^ ~ O ·~ t1) '= ~ ·_ ~ ~ ~ ~ ~ ~ C · _ . ~ o 04 8 C~ o a, ~ 3 c) 3 8 ~ ·= ~ ~ m ~ a~ C) C) .d .g C) a~ :e 0 ·e . ~ ~ 11 · ~ ~ V . _ C~
106 Health Effects of Ingested Fluoride bone-marrow sister-chromatic-exchange study in which NaF at 1, 10, 50, or 75 mg/L of drinking water was administered to male Chinese hamsters for 21 weeks. No significant increases in sister chromatic exchanges were observed, even though fluoride concentrations in bone and plasma increased with NaF treatment. Micronuclei contain one (usually) or several chromosomes or chromo- some fragments that are not included in the nuclei formed at mitotic cell division. They persist in one of the daughter cells for only one or several subsequent divisions. Thus, micronuclei represent loss of genetic material from a body cell (as opposed to reproductive cells: sperm and eggs), which eventually leads to the death of that cell. Because micronu- clei are scored at the long interphase stage of the cell cycle, they serve as a rapid screen for agents that interfere with normal mitotic chromo- some and cell division. Micronuclei were not induced in the bone marrow of mice injected once intraperitoneally with NaF at 30 mg/kg or Na2FPO3 at 80 mg/kg (Hayashi et al., 1988) or injected twice with SnF2 at up to 39.5 mg/kg (Gocke et al., 1981~. When Li and co-workers (1987c) administered NaF by gavage to male and female B6C3F~ mice in doses up to the maximum tolerated dose (MTD) (80 mg/kg for males and ~15 mg/kg for females), the mice did not show increased micronucleated polychromatic erythrocytes in their bone marrow. Pati and Bhunya (1987) reported increases in micronuclei in bone marrow polychromatic erythrocytes of Swiss mice that received two intraperitoneal injections of NaF at 10-40 mg/kg. No micronuclei were induced in the bone marrow of male AP rats that received NaF at 500 or 1,000 mg/kg by gavage (Albanese, 1987~. Dunipace et al. (1989) did not observe micronuclei in polychrom- atic erythrocytes of B6C3F~ mice that were given NaF in drinking water at concentrations of I-75 mg/L for 21 weeks. Germ Cents In the Mohamed and Chandler (1982) study in which BALB/c mice were given drinking water containing fluoride at 0, I, 5, 10, 50, 100, or 200 mg/1~ for 3-6 weeks, cytological tests showed that NaF at i-200 mg/L induced chromosomal aberrations in spermatocytes in a dose- dependent manner. However, Martin et al. (1979) did not observe
Genotoxicity of Fluoride 107 chromosomal aberrations in mitotic or meiotic cells of testes of mice that were administered fluoride at I-100 mg/L of drinking water for 6 weeks or maintained for several generations on 50 mg/L of drinking water. Inhalation of HE at 0. ~ mg/m3, 6 hours per day, 6 days per week for 2-4 weeks did not induce aberrations in meiotic chromosomes of the testes of white mice (Voroshilin et al., 1975~. Dominant lethal mutations were not observed in male white mice that inhaled HF at concentrations of ~ mg/m3 for 2-4 weeks (Voroshilin et al., 1975~. A study on sperm-head morphology in Swiss mice exposed intraperitoneally to NaF at 10-40 mg/kg for 5 days and then sampled 35 clays later reported a large dose- dependent increase in abnormal sperm (Pati ant} Bhunya, 1987~. How- ever, no morphological abnormalities were observed in sperm from mice that drank water containing fluoride at concentrations of 75 g/L for 21 weeks (Dunipace et al., 1989) or that were given fluoride at up to 70 mg/kg by gavage for 5 days (Li et al., 19871~. No DNA single-strand breaks were observed in testicular cells of rats that were aciministerec] NaF by gavage at up to 84 mg/kg per (lay for 5 days (Skare et al., 1986b). Genotoxicity of fluoride in germ cells of mammals cannot be evaluated because of insufficient data. Human Studies There are no published studies in the literature on the genetic or cytogenetic effects of fluoride in humans. PROPOSED MECHANISMS OF GENOTOXICITY The mechanism of genotoxicity of fluoride is not known. A number of possible mechanisms have been postulated to explain the observer! genotoxicity of fluoride (NTP, 1990; PHS, 1991~. Fluoride is not a typical mutagen. It cannot intercalate (invade and insert) into the DNA molecule and therefore cannot form DNA abducts, which can result in mutation. Furthermore, the results of many studies conducted to cleter- mine the genotoxicity of fluoride often conflict, making it difficult to explain the mechanisms of genotoxicity of fluoride. Nonetheless, clues to how fluoride affects genetic integrity can be found in the interaction
108 Health Effects of Ingested Fluoride of fluoride with cellular components. Speculations on mechanisms of genetic toxicity have been based on the observed reactions of fluoride in solution with divalent cations or nucleotides or the physiological and biochemical responses of cells treated with fluoride. NaF inhibits protein and DNA synthesis in cultured mammalian cells (Holland, 1979a). The inhibition of DNA synthesis might be a secondary effect of the inhibition of protein synthesis or a direct effect of He inhibition of DNA poly- merase or other DNA synthesis-associated enzymes. Fluoride can react with divalent cations in He cell to affect enzyme activities that are neces- sary for DNA or RNA synthesis or chromosomal metabolism or mainten- ance; it might react directly with DNA as part of a complex; or it can disrupt other cellular processes, such as cell differentiation or energy metabolism (Hellung-Larsen and Klenow, 1969; Harper et al., 1974; Holland, 1979a,b; Srivastava et al., 1981; Tmai et al., 19X3; Edwards and Parry, 19X6~. A hypothesis of secondary effects on DNA or chrom- osomes is attractive because there is no apparent mechanism by which many of the genotoxic effects observed can be induced by direct interac- tion of fluoride with DNA. SUMMARY The in vitro data indicate that (~) the genotoxicity of fluoride is limited primarily to doses much higher than those to which humans are exposed, (2) even at high doses, genotoxic effects are not always observed, and (3) the preponderance of the genotoxic effects that have been reporter! are of the types that probably are of no or negligible genetic significance. In vivo tests in rodents for genotoxicity of fluoride provide mixed results that cannot be resolved readily because of differences in protocols and insufficient detail in some reports to allow a thorough analysis.