Localized Folate and Vitamin B-12 Deficiency in Squamous Cell Lung Cancer Is Associated With Global DNA Hypomethylation

Abstract: We measured the concentrations of folate and vitamin B-12 in paired tissue samples of squamous cell cancer (SCC) and adjacent grossly normal-appearing uninvolved bronchial mucosa (from which SCC developed and also "at risk" of developing SCC) of the lung in 12 subjects to determine the involvement of these vitamins in 1) lung carcinogenesis and 2) global DNA methylation. The folate concentrations were significantly lower in SCCs than in uninvolved tissues (p = 0.03). The vitamin B-12 concentrations were also significantly lower in SCCs than in uninvolved tissues (p = 0.02). The radiolabeled methyl incorporation (inversely related to the degree of in vivo DNA methylation) was significantly higher in SCCs than in uninvolved tissues (p < 0.0001). The correlation between folate and radiolabeled methyl incorporation was inverse and statistically significant in SCCs (p = 0.03). The correlation between vitamin B-12 and radiolabeled methyl incorporation also was inverse and statistically significant in SCCs (p = 0.009). The relationship between tissue vitamin B-12 and DNA methylation was minimal in uninvolved tissues. The relationship between folate and DNA methylation, however, was inverse in uninvolved tissues. In the multiple regression models that included both vitamins, only folate was inversely associated with radiolabeled methyl incorporation in uninvolved and cancerous tissues. These results suggested that folate might be the limiting vitamin for proper DNA methylation in SCC as well as in tissues at risk of developing SCC. Several possible mechanisms of folate deficiency, including inactivation of the vitamin by exposure to carcinogens of cigarette smoke and underexpression or absence of folate receptor in SCCs and associated premalignant lesions, are discussed in light of these findings.

Introduction
Krumdieck ( 1) previously hypothesized that repeated exposure of the respiratory epithelium to cigarette smoke might result in localized deficiency of folate and vitamin B-12-derived coenzymes, making the cells more susceptible to the carcinogens present in tobacco smoke. This hypothesis was based on studies demonstrating that cyanide, nitrous oxide, and organic nitrite intake associated with cigarette smoking adversely affects folate and vitamin B-12 nutritional status ( 2-10). These observations and other mechanistic considerations suggest that it is biologically plausible to expect a low folate and vitamin B-12 status in smokers compared with nonsmokers. In support of this, others and we have shown lower circulating folate ( 11-14) and vitamin B-12 ( 4, 5) concentrations in smokers than in nonsmokers. Additionally, plasma folate concentrations were even lower in smokers with potentially premalignant bronchial squamous metaplasia than in smokers without metaplasia ( 15). Supplementation with folic acid and vitamin B-12 was shown to reduce the severity of smoking-induced bronchial metaplasia in humans ( 16, 17). None of these studies, however, measured folate/vitamin B-12 concentrations directly in the tissues affected by the putative localized deficiency. The existence of localized deficiencies of these vitamins in affected tissues, therefore, was supported almost entirely by indirect evidence.

To obtain more direct estimates, Heimburger and coworkers ( 15) measured the levels of folate in bronchial mucosal biopsies of nine smokers and two nonsmokers and reported no differences in folate levels between smokers and nonsmokers. Although the number of samples analyzed was too small to draw valid conclusions, they noted higher folate levels per milligram of tissue in smaller biopsies than in larger biopsies. This assumption was verified by analyzing the folate content of microtome slices of dog trachea cut parallel to the surface of the epithelium in 100-mum increments. As documented ( 15), the most superficial 100 mum indeed had a markedly higher folate content, indicating the need to pay careful attention to mucosal depth or to use an alternative denominator to express folate concentrations in tissues. We approached this limitation by assaying folate and vitamin B-12 concentrations in oral mucosal scrapings and by expressing the concentrations of these vitamins per milligram of protein and documented significantly lower buccal mucosal folate and vitamin B-12 concentrations in smokers than in nonsmokers ( 14). Because we did not correlate the cellular levels of these vitamins with markers of increased malignant transformation in the affected cells, the results have limited clinical significance. The present study addresses these limitations by assessing global DNA hypomethylation, which is a marker of increased risk for malignancy in relation to folate and vitamin B-12 concentrations in tissues. We chose to express vitamin concentrations per microgram of DNA instead of per milligram of protein, since DNA methylation is expressed per microgram of DNA as well. However, because milligrams of protein is a commonly used denominator to express cellular vitamin concentrations, we present vitamin results per milligram of protein and also express the correlation between vitamin results using both denominators so that our results can be compared with other studies, which may use protein as the denominator.

We now report that folate and vitamin B-12 concentrations are lower in squamous cell cancer (SCC) of the lung than in adjacent uninvolved tissues of the same subject and that the concentrations of the two vitamins are related to global DNA methylation. In contrast to earlier studies, we compared samples of solid tumor and adjacent normal-appearing tissue from the same patients, circumventing the possible individual variation from person to person in DNA methylation and folate/vitamin B-12 status of a given type of tissue.

Materials and Methods
Samples of Tissues
Twelve subjects with an age range of 54-84 years who developed SCC of the lung were identified from a National Institutes of Health-funded Early Detection NetWork database through the tissue procurement facility of the University of Alabama at Birmingham. One subject reported active smoking status and one reported nonsmoking status at the time of surgery. The rest of the subjects had been long-term, heavy smokers who quit smoking at various times before surgery. Four subjects were women, and eight subjects were men. Eleven subjects were white, and one subject was an African-American. Frozen tissue samples of SCC (n = 12) and uninvolved (grossly normal-appearing) tissues ( 2-3samples/subject, for a total of 27 uninvolved tissue samples) from the same subjects were available.

The average tumor size of lung cancers was 4.8 +/- 0.7 cm. The status of tumor differentiation was moderate to poor for all lung SCCs. Only two cases of lung SCCs had regional lymph node metastasis, and distant metastasis was not documented for any of these cases.

Sample Preparation and Laboratory Methods
The tissue samples were processed individually without knowledge of the pathological diagnosis. After the wet weight of the tissues was obtained, the frozen tissues were pulverized and dissolved in 400 mul of phosphate-buffered saline (PBS). Total protein concentration in PBS cell homogenate was measured using a protein assay kit (Bio-Rad Laboratories, Richmond, CA). As described below, 100-mul aliquots of PBS cell homogenate each were used for the extraction of folate, vitamin B-12, and DNA.

Folate and vitamin B-12: Methods developed for the extraction and determination of folate and vitamin B-12 in tissues and their reproducibility have been published previously ( 14, 18). The 96-well plate adaptation of the Lactobacillus casei microbiological assay was used to measure the total folate concentrations in the appropriately treated samples of lung tissues ( 19, 20). The Immunophase vitamin B-12 57Co radioassay kit ( 14, 18) measured the total vitamin B-12 concentrations in appropriately treated lung tissues.

DNA extraction: DNA was extracted from lung tissues with a DNA extraction kit (Puregene, Gentra Systems). Briefly, a 100-mul aliquot of cell suspension was mixed with 600 mul of cell lysis buffer, and 1.5 gl of proteinase K (20 mg/ml) were added. The samples were incubated at 55 degrees C until the solutions were clear (~2 h). Ribonuclease (3 mul, 4 mg/ml) was added to each sample, and the samples were incubated at 37 degrees C for 45 minutes. A protein precipitation solution (200 mul; Puregene) was added, and samples were vortexed for 20 seconds and placed on ice for 5 minutes. The samples were then centrifuged at 13,000 g for 15 minutes. The resulting supernatant containing genomic DNA was poured into another tube containing 600 mul of isopropanol, and 1 mul of glycogen (20 mg/ml) was added. The tubes were gently inverted 50 times to precipitate DNA and then centrifuged at 13,000 g for 20 minutes to pellet the DNA. The pellets were washed with 70% ethanol, air-dried, and hydrated overnight in 30 mul of DNA hydration solution (Puregene). The concentration of DNA was determined by measuring the absorbance at 260 nm (A260) of a diluted sample. DNA concentration was calculated using the following formula: 1.0 A260 unit of DNA is equivalent to 50 mug/ml. The concentration of DNA in the original cell homogenate was calculated and used to express folate (ng/mug of DNA), vitamin B-12 (pg/mug of DNA), and global methylation [counts per minute (cpm)/mug of DNA]. The amount of DNA per milligram of cancerous as well as per milligram of uninvolved tissues was calculated using the wet weight of the tissues.

Global DNA methylation analysis: The assay for global DNA methylation has been described by Kim and co-workers ( 21). Briefly, 2 mug of lung tissue genomic DNA were incubated with 5 muCi of S-[methyl-3H]adenosylmethionine ( 3-10Ci/mmol; New England Nuclear), 4 U of Sss I methylase (New England Biolabs, Beverly, MA), and 20 mul of methylation buffer [50 mM NaCl, 10 mM tris(hydroxymethyl)aminomethane. HCl, pH 7.9, 10 mM MgCl2, 1 mM dithiothreitol] in a total reaction volume of 50 mul for three hours at 37 degrees C. The incubation mixtures were washed onto disks of Whatman DE-81 paper with use of a vacuum filtration apparatus and then soaked in 50 ml of sodium phosphate dibasic for 45 minutes. The disks were then dried at 95 degrees C for 30 minutes, and the resulting radioactivity of the DNA retained on the disks was measured by scintillation counting. Controls (which we find typically to be <1% of the sample values) included incubation mixtures lacking DNA, as well as mixtures lacking Sss I methylase. This assay quantitates in vitro transfer of radiolabeled methyl groups from S-adenosylmethionine (SAM) to sites in DNA that were not methylated in vivo. Endogenous DNA methylation status and exogenous radiolabeled methyl 3H incorporation are thus inversely related.

Folate concentrations were expressed as nanograms of folate per milligram of protein and per microgram of DNA. Vitamin B-12 concentrations were expressed as picograms of vitamin B-12 per milligram of protein and per microgram of DNA. The radiolabeled methyl incorporation was expressed as counts per minute per microgram of DNA.

Statistical Analysis
Wilcoxon rank sum test was employed to determine the statistical significance of the differences in micrograms of DNA per milligram of tissue, vitamin concentrations, and radiolabeled methyl incorporation between paired samples of cancerous and uninvolved tissues. Nonparametric (Spearman) correlation analyses were employed to determine the association between folate and vitamin B-12 expressed per milligram of protein and per microgram of DNA. The association between folate and vitamin B-12 concentrations expressed per microgram of DNA and radiolabeled methyl incorporation expressed per microgram of DNA was also evaluated using the Spearman correlation analysis.

Multiple regression analysis was employed to assess simultaneously the effects of folate and vitamin B-12 on DNA methylation. Alternative models were used, in which the dependent variable was the DNA methylation index or its logarithm. The independent variables were evaluated in the scale of their measurements (i.e., as continuous variables, folate in ng/mug DNA and B-12 in pg/mug DNA) or were converted to ordinal scales corresponding to the quartiles of their distribution. The latter specification of the predictors is less sensitive to extreme values and provides a robust assessment of the association with DNA methylation. To facilitate interpretation of the results of multiple regression models, the effect of vitamin levels on methylation was expressed as a percent increase in DNA methylation: Delta% = 100 * [exp(beta) - 1], where beta is the regression coefficient estimate. In the models presented, the Delta% and its 95% confidence interval (CI) indicate the likely range of variation in DNA methylation corresponding to the difference between adjacent quartiles of the distribution of vitamin levels. All regression models were fitted separately with the data from cancerous tissues and with the data from uninvolved tissues, because the ranges of vitamin and DNA methylation levels were very different in the two series of specimens.

Results
The difference in micrograms of DNA per milligram of tissue between cancerous and uninvolved tissues was not statistically significant. The Spearman correlation between the vitamin concentrations expressed per milligram of protein and per microgram of DNA was statistically significant for folate (r = 0.71,p = 0.0001) and vitamin B-12 (r = 0.77,p = 0.0001). The folate concentrations were significantly lower in cancerous tissues (0.05 +/- 0.03 ng/mug DNA) than in uninvolved tissues (0.12 +/- 0.02 ng/mug DNA, p = 0.03; Figure 1A). Folate concentrations expressed per milligram of protein also were significantly lower in cancerous tissues (1.7 +/- 0.24 ng/mg protein) than in uninvolved tissues (2.5 +/0.49 ng/mg protein, p = 0.04). The vitamin B-12 concentrations were also significantly lower in cancerous tissues (3.98 +/- 1.3 pg/mug DNA) than in uninvolved tissues (8.83 +/- 1.3 pg/+/-g DNA, p = 0.02; Figure 1B). Similarly, vitamin B-12 concentrations expressed per milligram of protein also were significantly lower in cancerous tissues (108.9 +/- 34 ng/mg protein) than in uninvolved tissues (208 +/- 37 ng/mg protein, p = 0.03). The radiolabeled methyl incorporation (inversely related to the degree of in vivo DNA methylation) was significantly higher in cancerous tissues ( 35,409+/- 3,819 cpm/mug DNA) than in uninvolved tissues ( 10,508+/- 1,480 cpm/mug DNA, p < 0.0001; Figure 2). As shown in Figure 3, the Spearman correlation between tissue folate and vitamin B-12 concentrations and the radiolabeled methyl incorporation were inverse and statistically significant in cancerous tissues. The relationship between folate and DNA methylation was inverse (r = -0.30) but statistically nonsignificant in uninvolved tissues (Figure 4A). There was no relationship between tissue vitamin B-12 and DNA methylation in uninvolved tissues (r = -0.03, p = 0.89; Figure 4B).

All multiple regression models evaluating DNA methylation levels as a function of folate and vitamin B-12 yielded similar results. The models specifying the dependent variable as the logarithmically transformed DNA methylation level and the independent variables as ordinal scales indicating the quartiles of the distribution of folate and vitamin B-12 were chosen for presentation in Tables 1 and 2. In the multiple regression models that included folate and vitamin B-12 quartile scores, only folate was inversely associated with radiolabeled methyl incorporation, approaching statistical significance in uninvolved and cancerous tissues. In SCC tissue, the effect of a change in folate levels to the next higher quartile of the distribution was associated with a 16% decrease in DNA hypomethylation (95% CI = -31% to +3%; Table 1). In normal tissue, the effect of a change in folate levels to the next-higher quartile of the distribution was associated with a 22% decrease in DNA hypomethylation (95% CI = -41% to +3%; Table 2). Although the range of methylation and folate levels were different in SCC and normal tissue, the estimated effect of increasing folate levels was very similar. Unfortunately, given the small number of specimens available for this study, the effect of folate was of borderline statistical significance, and the confidence intervals of the folate effects in either type of tissue are relatively wide.

Discussion
Neoplastic cells may simultaneously harbor widespread (global) genomic hypomethylation, regional areas of hypermethylation, and increased DNA methyltransferase activity. Although the precise roles of these components are unclear, each component of this "methylation imbalance" may fundamentally contribute to tumor progression. The most likely mechanisms through which global DNA hypomethylation may induce neoplastic transformation are genomic instability ( 22, 23), abnormal chromosomal structures ( 24), and potential activation of oncogenes ( 25-29). Several studies have evaluated global DNA methylation in primary human malignancies. Gama-Sosa and co-workers ( 30), using a high-performance liquid chromatography technique, reported a 6% overall reduction in genomic 5-methylcytosine content when comparing a large number of primary malignancies of various types with "normal" tissues. A quantitative estimate of overall 5-methylcytosine content by densitometric comparison of ethidium bromide-stained, electrophoresed Hpa II and Hha I digests showed no significant difference between normal and neoplastic tissues of the colon ( 31). Feinberg and colleagues ( 32), using a more-sensitive quantitative measurement of 5-methylcytosine content by high-performance liquid chromatography, demonstrated that colon neoplasms are significantly hypomethylated compared with the adjacent normal tissues from the same subject. The level of 5-methylcytosine in DNA in malignant tissue was shown to be one-half that of normal lung, the difference being statistically significant ( 33). Recent studies that determined the extent of in vivo DNA methylation by analyzing DNA methyl-accepting capacity in the presence of S-[methyl-3H]adenosylmethionine and Sss I methylase have demonstrated incremental changes in DNA methylation with different stages of cancer. These changes were shown to occur early in gastric carcinoma ( 34) and late in cervical carcinoma ( 35) and have been unreported in other cancers. A recent study showed that DNA hypomethylation in breast carcinoma correlated with prognostic factors and tumor progression ( 36).

There are several reports of the influence of methyl availability on DNA methylation. SAM is the body's primary donor of methyl groups for methylation of DNA. Folate and vitamin B-12 are essential cofactors in the synthesis of SAM. However, despite the fact that folate and vitamin B-12 are essential cofactors in the synthesis of SAM and the metabolism of these two vitamins is intimately related, studies have assessed only the critical role of folate in DNA methylation. Several studies have shown that experimental folate deficiency significantly reduces SAM levels ( 37, 38). Studies that examined the effect of folate deficiency on global DNA methylation with animal models, however, have produced conflicting results: hypomethylation in DNA from liver ( 39, 40) and brain ( 41) of rats but no effect on hepatic and colonic DNA of rats ( 42). In addition, Kim and associates ( 43) reported hypomethylation within the p53 gene with no effects on global DNA methylation. Pilot data, performed in a prospective, placebo-controlled trial in 11 patients with colonic adenomas, however, showed an 80% improvement in DNA hypomethylation in colonic adenomas after six months of supplementation with folic acid ( 44).

There are few reports of folate status in tissues from human cancers. Meenan and others ( 45) reported significantly lower folate concentrations in colonic carcinoma than in adjacent normal tissues. Kim and associates ( 46) reported that folate concentrations in the normal rectosigmoid mucosa were significantly lower in persons with adenomatous polyps than in those with hyperplastic polyps. None of these studies, however, showed any direct association between cellular folate concentrations and DNA methylation. We recently showed that cervical DNA hypomethylation was correlated with cervical tissue and serum folate levels, although these associations were weak ( 47).

An estimated 85% of lung cancers are attributable to cigarette smoking ( 48, 49), especially small cell, squamous cell, and adenocarcinomas ( 50, 51). Poor dietary habits of smokers ( 52), along with associated alcohol consumption, could increase the risk for folate deficiency ( 53) in smokers. Furthermore, direct adverse effects of components of cigarette smoke on folate and vitamin B-12 could lead to local deficiencies of these two vitamins in the lungs of smokers. However, only a few studies have evaluated the effects of folate or vitamin B-12 in modifying the risk of lung cancer. The relationship between diet and the risk of lung cancer was evaluated among the participants of the New York State Cohort ( 54). The results of this study showed a protective effect of folate and vitamin C intake for the development of lung cancer. The relationship observed for folate was stronger for heavy smokers and seemed to be limited to squamous cell carcinomas. A cross-sectional study of postoperative nonsmall cell lung cancer patients that examined the possible effects of vitamins on disease-free survival showed that patients with higher circulating folate concentrations were more likely to be long-term survivors ( 55). Antitumor effects and longer survival by administration of vitamin B-12 have also been shown in animal models and cancer cell lines ( 56). None of these studies, however, measured vitamin concentrations in lung tissues, where the putative deficiency should be most pronounced.

To our knowledge, this is the first report documenting the concentrations of folate and vitamin B-12 in human lung tissues of SCC and adjacent uninvolved tissues. Folate and vitamin B-12 concentrations were significantly lower in SCC of the lung than in adjacent uninvolved bronchial mucosa. We expressed lung folate and vitamin B-12 concentrations per milligram of protein, as well as per microgram of DNA, and showed significant correlations between vitamin values expressed with both denominators. It is noteworthy that the differences in both vitamins between SCC and adjacent uninvolved tissues were consistently significant irrespective of the denominator used to express these vitamins. However, using DNA as a denominator may artificially increase the difference in cellular vitamin concentrations between cancer and normal cells if cancer cells have more DNA than normal cells. It is likely that we do not see a significant difference in cellular DNA levels between cancerous and uninvolved tissue, because the uninvolved tissues were obtained adjacent to the SCC and cannot be considered normal. Also, the amount of DNA in a cancer cell may vary depending on the proportion of diploid, tetraploid, or aneuploid cells present in the tumor, which could be related to the stage of the disease.

The decreased folate and vitamin B-12 concentrations in SCC of the lung may not be entirely explained by the increased requirement for these vitamins within rapidly proliferating tissues for the following reasons. We did not observe a similar decrease in these vitamins in SCCs of the oral cavity (compared with matched uninvolved tissues, n = 15 pairs, data not shown), another tissue exhibiting rapid proliferation, similar to lung SCCs. Also, some vitamins have been reported to be increased, rather than decreased, in cancerous tissues compared with normal tissues, for example, higher ascorbic acid concentrations in neoplastic tissues of the breast than in matched nonneoplastic tissues ( 57) and higher concentrations of alpha- and gamma-tocopherol in cancerous than in noncancerous tissues of the cervix ( 58).

Several mechanisms may account for the lower levels of folate and vitamin B-12 in SCC tissues observed in this study. Both vitamins are likely to be deficient in smoking-associated SCC tissues because of their biological inactivation by exposure to carcinogens present in cigarette smoke ( 2-10). In addition, the degree of expression of folate receptor may play an important role in determining the concentration of especially folate in these tissues. Human folate receptor, which is responsible for the internalization of folate from the extracellular medium, was shown to be overexpressed in several cancers ( 59), including adenocarcinoma of the lung, but was shown to be underexpressed or absent in SCCs of the lung ( 60). Franklin and co-workers ( 61) also demonstrated that the underexpression of folate receptor occurs early in the neoplastic process of SCC of the lung. Because the blocking of the synthesis of folate receptor was shown to inhibit the cellular uptake of folate ( 62), our observations of lower levels of folate in SCCs may be partly explained by a loss of folate receptor activity in these tissues. The extent of expression of folate receptor has not been reported for SCCs of other sites, including SCCs of the oral cavity. Although SCCs of the lung and oral cavity share some similar features, we expect that the effect of cigarette smoking in inactivating folates could be more pronounced in the lung than in the oral cavity, since lung tissues are more directly exposed to cigarette smoke than are oral tissues. It is, therefore, possible that a combined effect of local inactivation of folate by exposure to cigarette smoke and a loss of folate receptor activity in SCCs of the lung leads to lower levels of folate in SCCs of the lung, but not SCCs of the oral cavity.

Another mechanism that could explain the decreased folate levels in SCC of the lung may be decreased activity of folate-metabolizing enzymes. A decreased activity of methylenetetrahydrofolate reductase (MTHFR) in a given tissue would be expected to result in lower levels of SAM availability for methylation of DNA. Decreased MTHFR activity would be expected to lead to accumulation and eventual leakage of 5-methyltetrahydrofolate from cells. Although a polymorphism in MTHFR has been inversely associated with risk of colon cancer ( 63), the same polymorphism increased the risk of endometrial cancer ( 64). In addition, loss of heterozygosity at the MTHFR locus, which was associated with a decrease in MTHFR activity, was observed in ovarian carcinoma ( 65). Further studies are warranted to evaluate whether loss of heterozygosity at the MTHFR locus could account for altered folate status in SCC of the lung.

To our knowledge, this is also the first report documenting the concentrations of folate and vitamin B-12 in human lung tissues and their relation to genome-wide DNA methylation. DNA methylation changes in premalignant lung lesions and invasive cancers have focused mostly on regional hypermethylation ( 66). In the present study we measured global DNA methylation in cancerous and uninvolved tissues from the same subjects. Such analyses can give insights that may be missed in studies limited to specific genes, in which the state of methylation may not mirror most of the methylation changes in the genome. Also, because we attempt to correlate methylation status with methyl availability through SAM, evaluation of global methylation is more appropriate.

The correlation results showed that the availability of cellular folate and vitamin B-12 is important for DNA methylation in cancerous tissues. The weak but inverse association between folate and DNA methylation and no association between vitamin B-12 and DNA methylation in uninvolved tissues may be indicative of marginal inadequacy of only folate for DNA methylation in these tissues. The uninvolved tissues analyzed in our study are grossly normal appearing but may possibly contain preneoplastic lesions and, hence, be at higher risk of developing SCCs. The loss of folate receptor activity in preneoplastic lesions, as suggested by Franklin and co-workers ( 61), could have contributed to a marginal folate deficiency in these tissues. The multiple regression models that included both vitamins also suggested that folate may be the limiting vitamin for DNA methylation in cancerous tissues as well as in uninvolved tissues that are at high risk for the development of cancers. This observation may indicate that exposure to risk factors (e.g., cigarette smoking for the development of SCC) leading to preneoplastic lesions in the exposed tissues of the lung might lower the content of folate, but not vitamin B-12 concentrations, to a degree that interferes with DNA methylation.

In conclusion, SCC tissues show localized deficiencies of folate and vitamin B-12 compared with adjacent uninvolved lung tissues. Although the significant inverse correlation between folate and vitamin B-12 concentrations and global DNA methylation in SCC tissues indicates the importance of both vitamins for DNA methylation, folate may be the limiting vitamin for DNA methylation in cancerous tissues and at-risk tissues. Several possible mechanisms of folate deficiency, including inactivation of the vitamin by exposure to cigarette smoke and underexpression or absence of folate receptor in SCCs and associated premalignant lesions, are discussed in light of these findings.

Acknowledgments and Notes
The authors greatly appreciate the valuable comments and suggestions of Dr. Carlos Krumdieck in interpretation of data and in preparation of the manuscript. This work was supported by National Cancer Institute Grant K07-CA-70160. Address correspondence to Chandrika J. Piyathilake, Div. of Nutritional Biochemistry and Molecular Biology, Dept. of Nutrition Sciences, University of Alabama at Birmingham, Birmingham, AL 35294.

Submitted 3 December 1999; accepted in final form 23 February 2000.

Table 1. Estimated Variation in Radiolabeled Methyl Incorporation According to Folate and Vitamin B-12 Concentrations in Lung Tissues of SCC[a]

Variable Estimate, Delta% P Value

Folate[b] (ng/mug DNA) -16 0.095
Vitamin B-12[c] (pg/mug DNA) -2 0.842
a: Estimated in a regression model with dependent variable expressed as radiolabeled methyl incorporation of cpm/mug DNA. SCC, squamous cell cancer.

b: Quartiles of tissue folate.

c: Quartiles of tissue vitamin B-12.

Table 2. Estimated Variation in Radiolabeled Methyl Incorporation According to Folate and Vitamin B-12 Concentrations in Uninvolved Lung Tissues[a]

Variable Estimate, Delta% P Value

Folate[b] (ng/mug DNA) -22 0.084
Vitamin B-12[c] (pg/mug DNA) -5 0.626
a: Estimated in a regression model with dependent variable expressed as radiolabeled methyl incorporation of cpm/mug DNA.

b: Quartiles of tissue folate.

c: Quartiles of tissue vitamin B-12.

GRAPHS: Figure 1. Folate (A) and vitamin B-12 (B) concentrations of normal and cancerous tissues of the lung. Folate and vitamin B-12 concentrations were significantly higher in normal than in cancerous tissues (p = 0.03 and 0.02, respectively, by Wilcoxon rank sum test).

GRAPH: Figure 2. Radiolabeled methyl incorporation in cancerous and uninvolved tissues of lung. Radiolabeled methyl incorporation was significantly higher in cancerous than in uninvolved tissues (p

GRAPHS: Figure 3. Correlation between radiolabeled methyl incorporation and lung folate (A) and vitamin B-12 (B) concentrations of cancerous tissues of lung. *, Endogenous DNA methylation status and exogenous radiolabeled methyl incorporation are inversely related. **, Spearman correlation coefficient.

GRAPHS: Figure 4. Correlation between radiolabeled methyl incorporation and lung folate (A) and vitamin B-12 (B) concentrations of uninvolved lung. See Figure 3 legend for explanation of symbols.

References
1. Krumdieck, CL: Folic acid. In Present Knowledge in Nutrition, 6th ed, ML Brown (ed). Washington, DC: Nutr Found, 1990, pp 179-188.

2. Stedman, RL: The chemical composition of tobacco and tobacco smoke. Chem Rev 68, 153-207, 1968.

3. Francis, KT, Thompson, RW, and Krumdieck, CL: Reaction of tetrahydrofolic acid with cyanate from urea solutions. Formation of an inactive folate derivative. Am J Clin Nutr 30, 2028-2032, 1977.

4. Dastur, DK, Quadros, EV, Wadia, NH, Desai, MM, and Bharucha, EP: Effect of vegetarianism and smoking on vitamin B-12, thiocyanate, and folate levels in the blood of normal subjects. Br Med J 3, 260-263, 1972.

5. Linnell, JC, Smith, ADM, Smith, CL, Wilson, J, and Matthews, DM: Effects of smoking on metabolism and excretion of vitamin B-12. Br Med J 2, 215-216, 1986.

6. Falk, HL: Chemical agents in cigarette smoke. In Handbook of Physiology. Reaction to Environmental Agents, DHK Lee (ed). Bethesda, MD: Am Physiol Soc, 1977, pp 199-211.

7. Kondo, H, Osborne, ML, Kolhouse, JF, Binder, MJ, Podell, ER, et al.: Nitrous oxide has multiple deleterious effects on cobalamin metabolism and causes decreases in activities of both mammalian cobalamin-dependent enzymes in rats. J Clin Invest 67, 1270-1283, 1981.

8. Frasca, V, Riazzi, BS, and Matthews, RG: In vitro inactivation of methionine synthase by nitrous oxide. J Biol Chem 261, 15823-15826, 1986.

9. Fujii, K, Nagasaki, T, Vitols, KS, and Huennekens, FM: Polyglutamation as a factor in the trapping of 5-methyl-tetrahydrofolate by cobalamin-deficient L1210 cells. In Folyl and Antifolyl Polyglutamates, ID Goldman, BA Chabner, and JR Bertino (eds). New York: Plenum, 1983, pp 375-397.

10. Khaled, MA, Watkins, CL, and Krumdieck, CL: Inactivation of B-12 and folate coenzymes by butyl nitrite as observed by NMR: implications on one-carbon transfer mechanism. Biochem Biophys Res Commun 135, 201-207, 1986.

11. Witter, FR, Blake, DA, Baumgadner, R, Mellits, ED, and Niebyl, JR: Folate, carotene, and smoking. Am d Obstet Gynecol 144, 857, 1982.

12. Nakazawa, Y, Chiba, K, Imatoh, N, Kotorii, T, Sakamoto, T, et al.: Serum folic acid levels and antipyrine clearance rates in smokers and non-smokers. Drug Alcohol Depend 11, 201-207, 1983.

13. Chen, ATL, Reidy, JA, Annest, JL, Welty, TK, and Zhou, HG: Increased chromosome fragility as a consequence of blood folate levels. Smoking status and coffee consumption. Environ Mol Mutagen 13, 319-324, 1989.

14. Piyathilake, CJ, Macaluso, M, Hine, RJ, Richards, WE, and Krumdieck, CL: Local and systemic effects of cigarette smoking on folate and vitamin B-12. Am J Clin Nutr 60, 559-566, 1994.

15. Heimburger, DC, Krumdieck, CL, Alexander, CB, Dill, SR, and Bailey, WC: Localized folic acid deficiency and bronchial metaplasia in smokers: hypothesis and preliminary report. Nutr Int 3, 54-60, 1987.

16. Heimburger, DC, Alexander, CB, Birch, R, Butterworth, CE, Jr, Bailey, WC, et al.: Improvement in bronchial squamous metaplasia in smokers treated with folate and vitamin B-12. JAMA 259, 1525-1530, 1988.

17. Saito, M, Kato, H, Tsuchida, T, and Konaka, C: Chemoprevention effects on bronchial squamous metaplasia by folate and vitamin B-12 in heavy smokers. Chest 106, 496-499, 1994.

18. Piyathilake, CP, Hine, RJ, Dasnayake, AP, Richard, WE, Freeberg, LE, et al.: Effect of smoking on folate levels in buccal mucosal cells. Int d Cancer 52, 566-569, 1992.

19. Tamura, T: Microbiological assay of folates. In Folic Acid Metabolism in Health and Disease. Contemporary Issues in Clinical Nutrition, MF Picciano, ELR Stokstad, and JF Gregory (eds). New York: Wiley-Liss, 1990, pp 121-137.

20. Scott, JM, Ghanta, V, and Herbert, V: Trouble-free microbiologic serum and red cell folate assays. Am J Med Technol 40, 125-133, 1974.

21. Kim, Y-I, Giuliano, A, Hatch, KD, Schneider, A, Nour, MA, et al.: Global DNA hypomethylation increases progressively in cervical dysplasia and carcinoma. Cancer 74, 893-899, 1994.

22. Laird, PW, and Jaenisch, R: The role of DNA methylation in cancer genetics and epigenetics. Annu Rev Genet 30, 441-464, 1996.

23. Chen, RZ, Pettersson, U, Beard, C, Jackson-Grushby, L, and Jaenisch, R: DNA hypomethylation leads to elevated mutation rates. Nature 395, 89-92, 1998.

24. Lewis, J, and Bird, A: DNA methylation and chromatin structure. FEBS Lett 285, 155-159, 1991.

25. Feinberg, AP, and Vogelstein, B: Hypomethylation of ras oncogenes in primary human cancers. Biochem Biophys Res Commun 111, 47-54, 1983.

26. Feinberg, AP, and Vogelstein, B: Hypomethylation distinguishes genes of some human cancers from their normal counterparts. Nature 301, 89-92, 1983.

27. Baylin, SB, Herman, JG, Graft, JR, Vertino, PM, and Issa, JP: Alterations in DNA methylation: a fundamental aspect of neoplasia. Adv Cancer Res 72, 141-196, 1998.

28. Razin, A, and Cedar, H: DNA methylation and gene expression. Microbiol Rev 55, 451-458, 1991.

29. Weissbach, A, Ward, C, and Bolden, A: Eukaryotic DNA methylation and gene expression. Curr Top Cell Regul 30, 1-21, 1989.

30. Gama-Sosa, MA, Slagel, VA, Trewyn, RW, Oxenhandler, R, Kuo, KC, et al.: The 5-methylcytosine content of DNA from human tumors. Nucleic Acids Res 11, 6883-6894, 1983.

31. Goelz, SE, and Vogelstein, B: Hypomethylation of DNA from benign and malignant human colon neoplasms. Science 228, 187-190, 1985.

32. Feinberg, AP, Gehrke, CW, Kuo, KC, and Ehrlich, M: Reduced genomic 5-methylcytosine content in human colonic neoplasia. Cancer Res 48, 1159-1161, 1988.

33. Kliasheva, RI: DNA methylation in human lung tumors. Voprosy Ono kologii 36, 186-189, 1990.

34. Cravo, M, Pinto, R, Fidalgo, P, Chaves, P, Gloria, L, et al.: Global DNA hypomethylation occurs in the early stages of intestinal-type gastric carcinoma. Gut 39, 434-438, 1996.

35. Kim, Y-I, Giuliano, A, Hatch, KD, Schneider, A, Nour, MA, et al.: Global DNA hypomethylation increases progressively in cervical dysplasia and carcinoma. Cancer 74, 893-899, 1994.

36. Soares, J, Pinto, AE, Cunha, CV, Andre, S, Barao, I, et al.: Global DNA hypomethylation in breast caminoma--correlation with prognostic factors and tumor progression. Cancer 85, 112-118, 1999.

37. Miller, JW, Nadeau, MR, Smith, J, Smith, D, and Selhub, J: Folate-deo ficiency-induced homocysteinaemia in rats: disruption of S-adenosylmethionine's co-ordinate regulation of homocysteine metabolism. Biochem J 298, 415-419, 1994.

38. Balaghi, M, Home, DW, and Wagner, C: Hepatic one-carbon metabolism in early folate deficiency in rats. Biochem J 291, 145-149, 1993.

39. Balaghi, M, and Wagner, C: DNA methylation in folate deficiency: use of CpG methylase. Biochem Biophys Res Commun 193, 1184-1190, 1993.

40. Cravo, M, Mason, J, and Salomon, R: Folate deficiency in rats causes hypomethylation of liver DNA (abstr). FASEB J 4, A914, 1991.

41. Alonso-aperte, E, and Varela-Moreiras, G: Brain folates and DNA methylation in rats fed a choline-deficient diet or treated with low doses of methotrexate. Int J Vitam Nutr Res 66, 232-236, 1996.

42. Kim, Y-I, Christman, JK, Fleet, JC, Cravo, ML, Salomon, RN, et al.: Moderate folate deficiency does not cause global hypomethylation of hepatic and colonic DNA or c-myc-specific hypomethylation of colonic DNA in rats. Am d Clin Nutr 61, 1083-1090, 1995.

43. Kim, Y-I, Pogribny, IP, Basnakian, AG, Miller, JW, Selhub, J, et al.: Folate deficiency in rats induces DNA strand breaks and hypomethylation within the p53 tumor suppressor gene. Am J Clin Nutr 65, 46-52, 1995.

44. Cravo, M, Fidalgo, P, Pereira, AD, Gouveia-Oliveira, A, Chaves, P, et al.: DNA methylation as an intermediate biomarker in colorectal cancer: modulation by folic acid supplementation. Eur d Cancer Prey 3, 473-479, 1994.

45. Meenan, J, O'Hallinan, E, Scott, J, and Weir, DG: Epithelial cell folate depletion occurs in neoplastic but not adjacent normal colon mucosa. Gastroenterology 112, 1163-1168, 1997.

46. Kim, Y-I, Fawaz, K, Knox, T, Lee, YM, Norton, R, et al.: Colonic mucosal concentrations of folate correlate well with blood measurements of folate status in persons with colorectal polyps. Am J Clin Nutr 68, 866-872, 1998.

47. Short, B, Giuliano, A, Piyathilake, CJ, Nour, M, and Hatch, K: Hypomethylation: an early event in cervical carcinogenesis. Cancer Epidemiol Biomarkers Prev 7, 901-906, 1998.

48. Miller, BA, Ries, LAG, Hankey, BF, Kosary, CL, Harras, A, et al. (eds): SEER Cancer Statistics Review. 1973-1990. Bethesda, MD: Dept of Health and Human Services, 1993. (NIH Publ 93-2789)

49. Shopland, DR, Eyre, HJ, and Pechacek, TF: Smoking-attributable cancer mortality in 1991: Is lung cancer now the leading cause of death among smokers in the United States? JNCI 83, 1142-1148, 1991.

50. Ihde, DC, and Minna, JD: Non small cell lung cancer. 1. Biology, diagnosis and staging. Curr Problems Cancer 15, 65-147, 1991.

51. Thun, MJ, Lally, CA, Flannery, JT, Calle, EE, Flanders, WD, et al.: Cigarette smoking and changes in the histopathology of lung cancer. JNCI 89, 1580-1586, 1997.

52. Subar, AF, Harlan, LC, and Mattson, ME: Food and nutrient intake differences between smokers and non smokers in the US. Am J Public Health 80, 1323-1329, 1990.

53. Halsted, CH: Alcohol and folate interactions: clinical implications. In Folate in Health and Disease, LB Bailey (ed). New York: Dekker, 1995, pp 313-327.

54. Bandera, EV, Freudenheim, JL, Marshall, JR, Zielezny, M, Priore, RL, et al.: Diet and alcohol consumption and lung cancer risk in the New York State Cohort (United States). Cancer Causes Control 8, 828-840, 1997.

55. Jatoi, A, Daly, BD, Kramer, G, and Mason, JB: A cross-sectional study of vitamin intake in postoperative non-small cell lung cancer patients. J Surg Oncol 68, 231-236, 1998.

56. Shimizu, N, Hamazoe, R, Kanayama, H, Maeta, M, and Koga, S: Experimental study of antitumor effect of methyl-B-12. Oncology 44, 169-173, 1987.

57. Langemann, H, Torhorst J, Kabiersch, A, Krenger, W, and Honegger, CG: Quantitative determination of water- and lipid-soluble antioxidants in neoplastic and nonneoplastic human breast tissue. Int J Cancer 43, 1169-1173, 1989.

58. Peng, YM, Peng, YS, Childers, JM, Hatch, KD, Roe, DJ, et al.: Concentrations of carotenoids, tocopherols, and retinol in paired plasma and cervical tissue of patients with cervical cancer, precancer, and noncancerous diseases. Cancer Epidemiol Biomarkers Prev 7, 347-350, 1998.

59. Ross, JF, Chaudhuri, PK, and Ratnam, M: Differential regulation of folate receptor isoforms in normal and malignant tissues in vivo and in established cell lines. Physiologic and clinical implications. Cancer 73, 2432-2443, 1994.

60. Franklin, WA, Waintrub, M, Edwards, D, Christensen, K, Prendegrast, P, et al.: New anti-lung-cancer antibody cluster 12 reacts with human folate receptors present on adenocarcinoma. Int J Cancer 8, 89-95, 1994.

61. Franklin, WA, Folkvord, JM, Varella-Garcia, M, Kennedy, T, Proudfoot, S, et al.: Expansion of bronchial epithelial cell populations by in vitro culture of explants from dysplastic and histologically normal sites. Am J Respir Cell Mol Biol 15, 297-304, 1996.

62. Wolf, G: Inhibition of cellular uptake of folate by blocking synthesis of the membrane folate receptor. Nutr Rev 56, 86-87, 1998.

63. Ma, J, Stampfer, MJ, Giovannucci, E, Artigas, C, Hunter, DJ, et al.: Methylenetetrahydrofolate reductase polymorphism, dietary interactions, and risk of colorectal cancer. Cancer Res 57, 1098-1102, 1997.

64. Esteller, M, Garcia, A, Martinez-Palones, JM, Xercavins, J, and Reventos, J: Germ line polymorphisms in cytochrome P-450 1A1 (C4887 CYP1A1) and methylenctetrahydrofolate reductase (MTHFR) genes and endometrial cancer susceptibility. Carcinogenesis 18, 2307-2311, 1997.

65. Viel, A, Dall'Agnese, L, Simone, F, Canzonieri, V, Capozzi, E, et al.: Loss of heterozygosity at the 5,10-methylenetetrahydrofolate reductase locus in human ovarian carcinomas. Br J Cancer 75, 1105-1110, 1997.

66. Kashiwabara, K, Oyama, T, Sano, T, Fukuda, T, and Nakajima, T: Correlation between methylation status of the p16/CDKN2 gene and the expression of p 16 and Rb proteins in primary non-small cell lung cancers. Int J Cancer 79, 215-220, 1998.

~~~~~~~~

By Chandrika J. Piyathilake; Gary L. Johanning; Maurizio Macaluso; Martin Whiteside; Denise K. Oelschlager; Douglas C. Heimburger and William E. Grizzle

C. J. Piyathilake, G. L. Johanning, M. Whiteside, and D. C. Heimburger are affiliated with the Department of Nutrition Sciences, M. Macaluso with the Department of Epidemiology and International Health, and D. K. Oelschlager and W. E. Grizzle with the Department of Pathology, University of Alabama at Birmingham, Birmingham, AL 35294.

Share this with your friends