Mutation Research, 263 (1991) 63-67
63
© 1991 Elsevier Science Publishers B.V. 0165-7992/91/$03.50 ADONIS 016579929100046A MUTLET 0491
Effect of residual splenic function and folate levels on the frequency of micronucleated red blood cells in splenectomized humans Dina M. Schreinemachers and Richard B. Everson Epidemiology Branch, Human Studies Division, Health Effects Research Laboratory, U.S. EnvironmentalProtection Agency, Research Triangle Park, NC 27711 (U.S.A.)
(Received 5 October 1990) (Revision received2 January 1991) (Accepted 10 January 1991)
Keywords: Micronuclei; Folate; Splenectomy
Summary Frequencies o f micronucleated erythrocytes in the peripheral blood o f splenectomized individuals can be used as an index o f genetic damage to erythrocyte precursor cells in the bone marrow. This is in contrast to non-splenectomized humans, whose micronucleated erythrocytes are removed by the spleen. M a n y subjects whose spleen has been removed surgically have residual spleen tissue and consequent residual spleen function (RSF), which can be measured by the percentage of 'pitted' peripheral red blood cells. In this study evidence o f RSF was associated with decreased frequencies o f micronucleated erythrocytes. Analysis o f data limited to subjects with minimal spleen function suggested an inverse association between the incidence of micronucleated erythrocytes and serum folate levels that was not apparent in the absence of stringent control for RSF.
The research described in this article has been reviewed by the Health Effects Research Laboratory, U.S. Environmental Protection Agency and has been approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the Agency. Collection of micronuclei and folate data used in this study was supported by Public Health Service contract ES-25018 from the National Institute of Environmental Health Sciences, National Institutes of Health, Department of Health and Human Services. Analysis of the micronuclei was supported by the Western Regional Research Center, Department of Agriculture.. Correspondence: Dr. Richard B. Everson, Epidemiology Branch/HSD, HERL, U.S. EPA, Mail Drop 55A, Research Triangle Park, NC 27711 (U.S.A.).
Micronuclei (extranuclear chromosomes or c h r o m o s o m a l fragments) in circulating erythrocytes are formed in nucleated erythrocyte precursors. In mice micronucleated erythrocytes remain in the peripheral blood (Schlegel and MacGregor, 1983). Consequently levels of micronuclei in peripheral blood o f mice can be used to determine c h r o m o s o m a l damage and abnormal cytokinesis occurring during erythropoiesis. In rats and humans, however, micronuclei in RBC are rapidly removed by the spleen, such that micronuclei in circulating erythrocytes can be used to measure cytogenetic damage only in the absence of
64
splenic function (Schlegel and MacGregor, 1982, 1984; Schlegel at al., 1986). The advantages of the micronuclei assay in peripheral RBC, which is widely used in experimental animal systems, include the ability to score RBCs rapidly, requirements for only a small blood sample, and elimination of the need for cell culture procedures which both simplifies the procedure and minimizes possible artifacts from in vitro manipulation. Previous studies have shown that the human micronucleus assay is sensitive to genotoxic exposures; while the background of micronucleated erythrocytes in splenectomized humans is about 2 per 1000, chemotherapy causes this to increase as much as 10-fold (Schlegel et al., 1986). The two red cell populations which can be scored for micronuclei are newly-formed RNA-containing erythrocytes (reticulocytes) and fully matured RNA-negative erythrocytes with approximate lifespans of 2 and 120 days, respectively (Wintrobe et al., 1981). Because of these lifespans the micronucleated RNA-positive cells can be used to measure acute effects, while the micronucleated RNA-negative cells measure damage occurring over several months (Everson et al., 1988). Individuals whose spleen has been removed may have recurrent spleen function (splenosis) due to residual spleen tissue caused by trauma or rupture during surgery. The level of residual spleen function can be assessed by the frequency of 'pitted' RBC, which are erythrocytes with surface indentations. The percentage 'pitted' RBC increases with decreasing residual spleen function. People with normal spleen function have low frequencies of 'pitted' RBC, typically less than 1°70 (Pearson et al., 1978). Since residual spleen function would remove micronucleated RBC, the amount of splenosis has to be accounted for when using this approach to assess genetic damage. In this analysis measurements of spleen function were examined closely. After eliminating subjects with evidence of spleen function, an effect of folate was sought to extend our observations from a previous case study on the association between folate and RBC
micronuclei (Everson et al., 1988), as well as other reports (Reidy et al., 1983; Chen et al., 1989) of the effects of folate deficiency on chromosomal breakage.
Subjects and methods Procedures for enrolling subjects and methods for obtaining and scoring samples have been described previously (Schlegel et al., 1986; Everson et al., 1988). Briefly, the 31 subjects in this study were the control subjects described by Schlegel et al. (1986). They were enrolled by newspaper advertisement and had undergone splenectomy because of traumatic rupture of the spleen but were without other known disorders or malignant disease. The individual with borderline clinical folate deficiency (reported in 1988 by Everson et al.) was deleted from this study. Frequencies of micronucleated RBC were determined by scoring approximately 2000 RNA-positive and 10000 RNA-negative erythrocytes in fixed blood smears stained with acridine orange. Cells were classified as RNApositive or RNA-negative based on the presence or absence of orange fluorescent cytoplasm (Schlegel et al., 1986; Everson et al., 1988). Serum folate was determined with a radio-assay by Smith Kline BioSciences Laboratory using Immo Phase (Corning Medical, Medfield, MA). They established that the normal range for serum folate for all ages was between 2.4 and 17.7 ng/ml. Values for 2 subjects reported to be over 20 ng/ml, which was the upper limit of the assay used, were set to 20 ng/ml for the purpose of the statistical analysis. The frequency of 'pitted' erythrocytes was scored to determine residual spleen function according to a method described by Pearson et al. (1978) and was performed by his laboratory at Yale University School of Medicine. According to this method splenectomized subjects with less than 12°70 'pitted' erythrocytes still have residual spleen function, while those with 120/0 or more 'pitted' RBC do not. In the statistical analysis the procedure for general linear models from the SAS Institute Inc., Cary, NC was used.
65
Results
The incidence of micronucleated RNA-positive and RNA-negative erythrocytes increased with increasing levels of percentage 'pitted' RBC" Kendali's tau correlation coefficient for the percentage 'pitted' RBC and frequencies of micronucleated RNA-negative and RNA-positive erythrocytes are respectively 0.52 (P-value=0.0001) and 0.28 (Pvalue = 0.03). The mean frequencies and standard error of micronucleated RNA-negative and RNApositive erythrocytes for subjects with 'pitted' RBC levels of 0-11.9070, 12-23.9070, 24070 and over, are respectively 0.36 + 0.21, 1.95 _+0.54, 2.61 + 0.31 for RNA-negative red cells, and 2.85+0.68, 3.25 _+0.42, 4.45 + 0.67 for RNA-positive red cells. Both the mean frequencies of micronucleated RNA-positive and RNA-negative erythrocytes for subjects with ___24070 'pitted' RBC are higher than the frequencies for subjects with 12-23.9070 'pitted' RBC, indicating that residual spleen function (RSF) may still be affecting frequencies of micronucleated erythrocytes in some subjects even though their frequency of 'pitted' RBC is over 1207o (Fig. 1). The interval of 12°70 was selected because Pearson et al. (1978) previously used this level to
separate subjects with RSF from those without. The fact that frequencies of micronucleated cells among RNA-positive erythrocytes are in general higher than among RNA-negative erythrocytes might be explained by the difference in lifespan of these 2 erythrocyte populations, which gives RSF the chance to filter out micronucleated RNAnegative erythrocytes more thoroughly than RNApositive erythrocytes. Using 12070 'pitted' RBC as the cutoff for residual spleen function and excluding 3 subjects whose folate data were unreliable due to hemolysis
8 u~
I
A.
+ ,,< Z6 ,,,t,-
"03 ee
g = 2
4
6
S
1
12
14
1
1
20
16
18
20
Serum Folate (ng/mi)
8
B.
0
~5
+ ,< Z6 rr
m ne
13
=
~2
..yy
...-'"
i
~,
.........
:E
|, :S
..~ 1
0 - 11.9
I
/
12 - 23.9
> =24
% Pitted RBC
Fig. 1. Association between frequencies (mean ± standard error) of micronucleated RNA-negative (m) and RNA-positive (e) erythrocytes and percentage 'pitted' RBC, grouped 0-11.9~o, 12-23.9°1o, ~24¢'/o. The numbers of subjects are respectively 11, 10, and 9.
0 2
4
6
8 Serum
1
1
1
Folate (ng/ml)
Fig. 2. Association between frequencies of micronucleated RNA-positive erythrocytes and serum folate in subjects with 'pitted' RBC _>1201o (A), and in subjects with 'pitted' RBC _>24O/o (B) (i.e. with increasingly stringent criteria for the absence of discernible spleen function).
66 o f their serum sample, and one subject for whom only 1030 RNA-positive erythrocytes could be scored and an exceptionally high frequency (40070) of 'pitted' RBC was found (suggesting possible technical problems with the assay), no association was found between serum folate and frequencies of micronucleated RNA-positive erythrocytes for the remaining 15 subjects (Fig. 2A, R2=0.04 for a linear model). After using in addition a more stringent cutoff of 2407o 'pitted' RBC, however, an inverse relation between serum folate and frequencies of micronucleated RNA-positive red cells became apparent for the 7 remaining subjects (Fig. 2 B , R E = 0.83, P-value = 0.004). While these data suggest a folate effect, the association was not statistically significant if data for subjects with 'pitted' RBC at 12070 and above were analyzed using modeling procedures to control for the effect o f RSF. A linear model was used with folate and percentage 'pitted' RBC as the independent variables, and frequencies of micronucleated RNA-positive erythrocytes as the dependent variable, which yielded an R 2 of 0.07. No association between serum folate and frequencies o f micronucleated RNA-negative erythrocytes was seen for subjects with 'pitted' RBC at or above either 12070 or 24070 (the R 2 for both was 0.02). Discussion
Previous studies established the use of 'pitted' peripheral erythrocytes to measure residual spleen function in splenectomized subjects (Pearson et al., 1978). The spleen typically is completely removed surgically if the reason for splenectomy is hematological, but is often incompletely removed when splenectomy is performed because of trauma or rupture. In Pearson's study, 18 of the 40 patients had their spleen removed for hematological reasons and 22 for trauma. The percentage 'pitted' RBC for all 18 non-trauma subjects in that study was over 12070. Of the 22 trauma patients, 9 subjects had 'pitted' RBC in the same range as the non-trauma group, and 13 had 6o10 or less 'pitted' RBC. That study established cutoffs of 'pitted' RBC below 1°70 as normal spleen function, 1-8070
as splenosis, and > 12°70 as asplenic. Splenosis was confirmed in the 1-8°70 'pitted' RBC group with a sulfur colloid scan. Serum folate measurements reflect recent levels of folate availability and red cell folate indicates the average folate level over several months. Likewise increased frequencies of micronucleated RNA-positive erythrocytes indicate effects o f recent exposures and increased frequencies of micronucleated RNA-negative erythrocytes measure accumulated exposure effects. Since only serum folate was available for the subjects in this study, we focused our investigation on the possible association between serum folate and frequencies o f micronucleated RNA-positive red cells. Use of increasingly more stringent criteria for exclusion of subjects with RSF in our data suggests an association between levels of folate and chromosomal breakage or aneuploidy in vivo among individuals with clinically normal folate levels. We were not able to find this association in models including subjects with 'pitted' RBC at 1207o and above. Possible reasons include statistical noise caused by interaction between the subject's RSF level and the short lifespan o f RNA-positive erythrocytes (which does not allow the frequencies of micronucleated RNA-positive red cells to reach a steady state) and the small number of subjects involved. Lack of association between serum folate and frequencies o f micronucleated RNA-negative cells for subjects with 'pitted' RBC at 24070 or above may be due in part to the fact that serum folate may not reflect folate availability during the development of older erythrocytes (as would red cell folate levels) or to the small number of subjects and consequent lack o f power. Folate plays a crucial role in DNA synthesis. Low levels of folate are associated with impaired cell division (Herbert and Colman, 1979; Eto and Krumdieck, 1986) and increased levels of chromosomal aberrations in vitro (Reidy et al., 1983) and in vivo (Chen et al., 1989). The findings here extend our previous observation that folate replacement decreased frequencies of micronucleated erythrocytes in a timeseries study of an individual with borderline folate deficiency.
67
Population surveys have demonstrated that a substantial portion of certain sizeable subsets of the U.S. population have low levels of folate (Committee on Dietary Allowances, 1980; Senti and Pilch, 1984; Bailey et al., 1979) suggesting that if the association between folate and micronucleated erythrocytes is confirmed, folate at preclinical levels of depletion may be an important determinant for in vivo chromosomal damage in humans, and may be an important contributor to chromosomal breakage or aneuploidy related to disordered cytokinesis on a population basis. The results in this study were obtained in a very small number of highly selected subjects, and clearly need to be studied in a larger number of subjects. Also the effect of RBC folate levels on frequencies of micronucleated RNA-negative erythrocytes should be studied, as well as the effect of vitamin B12 in serum, since vitamin B12 and folate deficiency may have similar effects on micronuclei formation in red cells.
Acknowledgements We thank Dr. James MacGregor of SRI International for helpful comments and editorial suggestions; Ms. Genie Corton and Ms. Lyle Lansdell of Survey Research Associates for obtaining the laboratory specimens; both Dr. MacGregor and Ms. Carol Wehr of the USDA Western Region Research Center (currently at Bruce Ames Laboratory at the University of California at Berkeley) for scoring the micronucleated erythrocytes; Dr. Howard A. Pearson and Mr. Edward Sullivan, Department of Pediatrics, Yale University, for scoring the 'pitted' RBC; Dr. Woodrow Setzer of HERL, U.S. EPA, and Dr. David Shore of SRAT for statistical advice.
References Bailey, L.B., P.A. Wagner, G.J. Christakis et al. (1979)Folacin and iron status and hematological findings in predominantly black elderly persons from urban low-income households,
Am. J. Clin. Nutr., 32, 2346-2353. Chen, A.T.L., J.A. Reidy, J.L. Annest, T.K. Welty and H. Zhou (1989) Increased chromosome fragility as a consequence of blood folate levels, smoking status, and coffee consumption, Environ. Mol. Mutagen., 13, 319-324. Committee on Dietary Allowances, Food and Nutrition Board, Division of Biological Sciences, Assembly of Life Sciences, National Research Council, Recommended dietary allowances, National Academy of Sciences, Washington, DC, 1980, pp. 106-113. Eto, I., and C.L. Krumdieck (1986) Role of vitamin B12 and folate deficiencies in carcinogenesis, Advan. Exp. Med. Biol., 206, 313-330. Everson, R.B., C.M. Wehr, G.L. Erexson and J.T. MacGregor (1988) Association of marginal folate depletion with increased human chromosomal damage in vivo: Demonstration by analysis of micronucleated erythrocytes, J. Natl. Cancer Inst., 80, 525-529. Herbert, V., and N. Colman (1979) Hematological aspects of folate deficiency, in: M.I. Botez and E.H. Reynolds (Eds.), Folic Acid in Neurology, Psychiatry and Internal Medicine, Raven, New York, pp. 63-74. Pearson, H.A., D. Johnston, K.A. Smith and R.J. Touloukian (1978) The born-again spleen: Return of splenic function after splenectomy for trauma, N. Engl. J. Med., 298, 1389-1392. Reidy, J.A., X. Zhou and A.T.L. Chen (1983) Folic acid and chromosome breakage, I. Implications for genotoxicity studies, Mutation Res., 122, 217-221. Schlegel, R., and J.T. MacGregor (1982) The persistence of micronuclei in peripheral blood erythrocytes: Detection of chronic chromosome breakage in mice, Mutation Res., 104, 367-369. Schlegel, R., and J.T. MacGregor (1983) A rapid screen for cumulative chromosomal damage in mice: accumulation of circulating micronucleated erythrocytes, Mutation Res., 113, 481-487. Schlegel, R., and J.T. MacGregor (1984) The persistence of micronucleated erythrocytes in the peripheral circulation of normal and splenectomized Fischer 344 rats: Implications for cytogenetic screening, Mutation Res., 127, 169-174. Schlegel, R., J.T. MacGregor and R.B. Everson (1986) Assessment of cytogenetic damage by quantitation of micronuclei in human peripheral blood erythrocytes, Cancer Res., 46, 3717-3721. Senti, F.R., and S.M. Pilch (1984) Assessment of the folate nutritional status of the U.S. population based on data collected in the second National Health and Nutrition Examination Survey, 1976-1980, FDA, Washington, DC. Wintrobe, M.M., G.R. Lee, D.R. Boggs et al. (1981) Clinical Hematology, Lea and Febiger, Philadelphia, pp. 108-135. Communicated by J.W. Allen
63
© 1991 Elsevier Science Publishers B.V. 0165-7992/91/$03.50 ADONIS 016579929100046A MUTLET 0491
Effect of residual splenic function and folate levels on the frequency of micronucleated red blood cells in splenectomized humans Dina M. Schreinemachers and Richard B. Everson Epidemiology Branch, Human Studies Division, Health Effects Research Laboratory, U.S. EnvironmentalProtection Agency, Research Triangle Park, NC 27711 (U.S.A.)
(Received 5 October 1990) (Revision received2 January 1991) (Accepted 10 January 1991)
Keywords: Micronuclei; Folate; Splenectomy
Summary Frequencies o f micronucleated erythrocytes in the peripheral blood o f splenectomized individuals can be used as an index o f genetic damage to erythrocyte precursor cells in the bone marrow. This is in contrast to non-splenectomized humans, whose micronucleated erythrocytes are removed by the spleen. M a n y subjects whose spleen has been removed surgically have residual spleen tissue and consequent residual spleen function (RSF), which can be measured by the percentage of 'pitted' peripheral red blood cells. In this study evidence o f RSF was associated with decreased frequencies o f micronucleated erythrocytes. Analysis o f data limited to subjects with minimal spleen function suggested an inverse association between the incidence of micronucleated erythrocytes and serum folate levels that was not apparent in the absence of stringent control for RSF.
The research described in this article has been reviewed by the Health Effects Research Laboratory, U.S. Environmental Protection Agency and has been approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the Agency. Collection of micronuclei and folate data used in this study was supported by Public Health Service contract ES-25018 from the National Institute of Environmental Health Sciences, National Institutes of Health, Department of Health and Human Services. Analysis of the micronuclei was supported by the Western Regional Research Center, Department of Agriculture.. Correspondence: Dr. Richard B. Everson, Epidemiology Branch/HSD, HERL, U.S. EPA, Mail Drop 55A, Research Triangle Park, NC 27711 (U.S.A.).
Micronuclei (extranuclear chromosomes or c h r o m o s o m a l fragments) in circulating erythrocytes are formed in nucleated erythrocyte precursors. In mice micronucleated erythrocytes remain in the peripheral blood (Schlegel and MacGregor, 1983). Consequently levels of micronuclei in peripheral blood o f mice can be used to determine c h r o m o s o m a l damage and abnormal cytokinesis occurring during erythropoiesis. In rats and humans, however, micronuclei in RBC are rapidly removed by the spleen, such that micronuclei in circulating erythrocytes can be used to measure cytogenetic damage only in the absence of
64
splenic function (Schlegel and MacGregor, 1982, 1984; Schlegel at al., 1986). The advantages of the micronuclei assay in peripheral RBC, which is widely used in experimental animal systems, include the ability to score RBCs rapidly, requirements for only a small blood sample, and elimination of the need for cell culture procedures which both simplifies the procedure and minimizes possible artifacts from in vitro manipulation. Previous studies have shown that the human micronucleus assay is sensitive to genotoxic exposures; while the background of micronucleated erythrocytes in splenectomized humans is about 2 per 1000, chemotherapy causes this to increase as much as 10-fold (Schlegel et al., 1986). The two red cell populations which can be scored for micronuclei are newly-formed RNA-containing erythrocytes (reticulocytes) and fully matured RNA-negative erythrocytes with approximate lifespans of 2 and 120 days, respectively (Wintrobe et al., 1981). Because of these lifespans the micronucleated RNA-positive cells can be used to measure acute effects, while the micronucleated RNA-negative cells measure damage occurring over several months (Everson et al., 1988). Individuals whose spleen has been removed may have recurrent spleen function (splenosis) due to residual spleen tissue caused by trauma or rupture during surgery. The level of residual spleen function can be assessed by the frequency of 'pitted' RBC, which are erythrocytes with surface indentations. The percentage 'pitted' RBC increases with decreasing residual spleen function. People with normal spleen function have low frequencies of 'pitted' RBC, typically less than 1°70 (Pearson et al., 1978). Since residual spleen function would remove micronucleated RBC, the amount of splenosis has to be accounted for when using this approach to assess genetic damage. In this analysis measurements of spleen function were examined closely. After eliminating subjects with evidence of spleen function, an effect of folate was sought to extend our observations from a previous case study on the association between folate and RBC
micronuclei (Everson et al., 1988), as well as other reports (Reidy et al., 1983; Chen et al., 1989) of the effects of folate deficiency on chromosomal breakage.
Subjects and methods Procedures for enrolling subjects and methods for obtaining and scoring samples have been described previously (Schlegel et al., 1986; Everson et al., 1988). Briefly, the 31 subjects in this study were the control subjects described by Schlegel et al. (1986). They were enrolled by newspaper advertisement and had undergone splenectomy because of traumatic rupture of the spleen but were without other known disorders or malignant disease. The individual with borderline clinical folate deficiency (reported in 1988 by Everson et al.) was deleted from this study. Frequencies of micronucleated RBC were determined by scoring approximately 2000 RNA-positive and 10000 RNA-negative erythrocytes in fixed blood smears stained with acridine orange. Cells were classified as RNApositive or RNA-negative based on the presence or absence of orange fluorescent cytoplasm (Schlegel et al., 1986; Everson et al., 1988). Serum folate was determined with a radio-assay by Smith Kline BioSciences Laboratory using Immo Phase (Corning Medical, Medfield, MA). They established that the normal range for serum folate for all ages was between 2.4 and 17.7 ng/ml. Values for 2 subjects reported to be over 20 ng/ml, which was the upper limit of the assay used, were set to 20 ng/ml for the purpose of the statistical analysis. The frequency of 'pitted' erythrocytes was scored to determine residual spleen function according to a method described by Pearson et al. (1978) and was performed by his laboratory at Yale University School of Medicine. According to this method splenectomized subjects with less than 12°70 'pitted' erythrocytes still have residual spleen function, while those with 120/0 or more 'pitted' RBC do not. In the statistical analysis the procedure for general linear models from the SAS Institute Inc., Cary, NC was used.
65
Results
The incidence of micronucleated RNA-positive and RNA-negative erythrocytes increased with increasing levels of percentage 'pitted' RBC" Kendali's tau correlation coefficient for the percentage 'pitted' RBC and frequencies of micronucleated RNA-negative and RNA-positive erythrocytes are respectively 0.52 (P-value=0.0001) and 0.28 (Pvalue = 0.03). The mean frequencies and standard error of micronucleated RNA-negative and RNApositive erythrocytes for subjects with 'pitted' RBC levels of 0-11.9070, 12-23.9070, 24070 and over, are respectively 0.36 + 0.21, 1.95 _+0.54, 2.61 + 0.31 for RNA-negative red cells, and 2.85+0.68, 3.25 _+0.42, 4.45 + 0.67 for RNA-positive red cells. Both the mean frequencies of micronucleated RNA-positive and RNA-negative erythrocytes for subjects with ___24070 'pitted' RBC are higher than the frequencies for subjects with 12-23.9070 'pitted' RBC, indicating that residual spleen function (RSF) may still be affecting frequencies of micronucleated erythrocytes in some subjects even though their frequency of 'pitted' RBC is over 1207o (Fig. 1). The interval of 12°70 was selected because Pearson et al. (1978) previously used this level to
separate subjects with RSF from those without. The fact that frequencies of micronucleated cells among RNA-positive erythrocytes are in general higher than among RNA-negative erythrocytes might be explained by the difference in lifespan of these 2 erythrocyte populations, which gives RSF the chance to filter out micronucleated RNAnegative erythrocytes more thoroughly than RNApositive erythrocytes. Using 12070 'pitted' RBC as the cutoff for residual spleen function and excluding 3 subjects whose folate data were unreliable due to hemolysis
8 u~
I
A.
+ ,,< Z6 ,,,t,-
"03 ee
g = 2
4
6
S
1
12
14
1
1
20
16
18
20
Serum Folate (ng/mi)
8
B.
0
~5
+ ,< Z6 rr
m ne
13
=
~2
..yy
...-'"
i
~,
.........
:E
|, :S
..~ 1
0 - 11.9
I
/
12 - 23.9
> =24
% Pitted RBC
Fig. 1. Association between frequencies (mean ± standard error) of micronucleated RNA-negative (m) and RNA-positive (e) erythrocytes and percentage 'pitted' RBC, grouped 0-11.9~o, 12-23.9°1o, ~24¢'/o. The numbers of subjects are respectively 11, 10, and 9.
0 2
4
6
8 Serum
1
1
1
Folate (ng/ml)
Fig. 2. Association between frequencies of micronucleated RNA-positive erythrocytes and serum folate in subjects with 'pitted' RBC _>1201o (A), and in subjects with 'pitted' RBC _>24O/o (B) (i.e. with increasingly stringent criteria for the absence of discernible spleen function).
66 o f their serum sample, and one subject for whom only 1030 RNA-positive erythrocytes could be scored and an exceptionally high frequency (40070) of 'pitted' RBC was found (suggesting possible technical problems with the assay), no association was found between serum folate and frequencies of micronucleated RNA-positive erythrocytes for the remaining 15 subjects (Fig. 2A, R2=0.04 for a linear model). After using in addition a more stringent cutoff of 2407o 'pitted' RBC, however, an inverse relation between serum folate and frequencies of micronucleated RNA-positive red cells became apparent for the 7 remaining subjects (Fig. 2 B , R E = 0.83, P-value = 0.004). While these data suggest a folate effect, the association was not statistically significant if data for subjects with 'pitted' RBC at 12070 and above were analyzed using modeling procedures to control for the effect o f RSF. A linear model was used with folate and percentage 'pitted' RBC as the independent variables, and frequencies of micronucleated RNA-positive erythrocytes as the dependent variable, which yielded an R 2 of 0.07. No association between serum folate and frequencies o f micronucleated RNA-negative erythrocytes was seen for subjects with 'pitted' RBC at or above either 12070 or 24070 (the R 2 for both was 0.02). Discussion
Previous studies established the use of 'pitted' peripheral erythrocytes to measure residual spleen function in splenectomized subjects (Pearson et al., 1978). The spleen typically is completely removed surgically if the reason for splenectomy is hematological, but is often incompletely removed when splenectomy is performed because of trauma or rupture. In Pearson's study, 18 of the 40 patients had their spleen removed for hematological reasons and 22 for trauma. The percentage 'pitted' RBC for all 18 non-trauma subjects in that study was over 12070. Of the 22 trauma patients, 9 subjects had 'pitted' RBC in the same range as the non-trauma group, and 13 had 6o10 or less 'pitted' RBC. That study established cutoffs of 'pitted' RBC below 1°70 as normal spleen function, 1-8070
as splenosis, and > 12°70 as asplenic. Splenosis was confirmed in the 1-8°70 'pitted' RBC group with a sulfur colloid scan. Serum folate measurements reflect recent levels of folate availability and red cell folate indicates the average folate level over several months. Likewise increased frequencies of micronucleated RNA-positive erythrocytes indicate effects o f recent exposures and increased frequencies of micronucleated RNA-negative erythrocytes measure accumulated exposure effects. Since only serum folate was available for the subjects in this study, we focused our investigation on the possible association between serum folate and frequencies o f micronucleated RNA-positive red cells. Use of increasingly more stringent criteria for exclusion of subjects with RSF in our data suggests an association between levels of folate and chromosomal breakage or aneuploidy in vivo among individuals with clinically normal folate levels. We were not able to find this association in models including subjects with 'pitted' RBC at 1207o and above. Possible reasons include statistical noise caused by interaction between the subject's RSF level and the short lifespan o f RNA-positive erythrocytes (which does not allow the frequencies of micronucleated RNA-positive red cells to reach a steady state) and the small number of subjects involved. Lack of association between serum folate and frequencies o f micronucleated RNA-negative cells for subjects with 'pitted' RBC at 24070 or above may be due in part to the fact that serum folate may not reflect folate availability during the development of older erythrocytes (as would red cell folate levels) or to the small number of subjects and consequent lack o f power. Folate plays a crucial role in DNA synthesis. Low levels of folate are associated with impaired cell division (Herbert and Colman, 1979; Eto and Krumdieck, 1986) and increased levels of chromosomal aberrations in vitro (Reidy et al., 1983) and in vivo (Chen et al., 1989). The findings here extend our previous observation that folate replacement decreased frequencies of micronucleated erythrocytes in a timeseries study of an individual with borderline folate deficiency.
67
Population surveys have demonstrated that a substantial portion of certain sizeable subsets of the U.S. population have low levels of folate (Committee on Dietary Allowances, 1980; Senti and Pilch, 1984; Bailey et al., 1979) suggesting that if the association between folate and micronucleated erythrocytes is confirmed, folate at preclinical levels of depletion may be an important determinant for in vivo chromosomal damage in humans, and may be an important contributor to chromosomal breakage or aneuploidy related to disordered cytokinesis on a population basis. The results in this study were obtained in a very small number of highly selected subjects, and clearly need to be studied in a larger number of subjects. Also the effect of RBC folate levels on frequencies of micronucleated RNA-negative erythrocytes should be studied, as well as the effect of vitamin B12 in serum, since vitamin B12 and folate deficiency may have similar effects on micronuclei formation in red cells.
Acknowledgements We thank Dr. James MacGregor of SRI International for helpful comments and editorial suggestions; Ms. Genie Corton and Ms. Lyle Lansdell of Survey Research Associates for obtaining the laboratory specimens; both Dr. MacGregor and Ms. Carol Wehr of the USDA Western Region Research Center (currently at Bruce Ames Laboratory at the University of California at Berkeley) for scoring the micronucleated erythrocytes; Dr. Howard A. Pearson and Mr. Edward Sullivan, Department of Pediatrics, Yale University, for scoring the 'pitted' RBC; Dr. Woodrow Setzer of HERL, U.S. EPA, and Dr. David Shore of SRAT for statistical advice.
References Bailey, L.B., P.A. Wagner, G.J. Christakis et al. (1979)Folacin and iron status and hematological findings in predominantly black elderly persons from urban low-income households,
Am. J. Clin. Nutr., 32, 2346-2353. Chen, A.T.L., J.A. Reidy, J.L. Annest, T.K. Welty and H. Zhou (1989) Increased chromosome fragility as a consequence of blood folate levels, smoking status, and coffee consumption, Environ. Mol. Mutagen., 13, 319-324. Committee on Dietary Allowances, Food and Nutrition Board, Division of Biological Sciences, Assembly of Life Sciences, National Research Council, Recommended dietary allowances, National Academy of Sciences, Washington, DC, 1980, pp. 106-113. Eto, I., and C.L. Krumdieck (1986) Role of vitamin B12 and folate deficiencies in carcinogenesis, Advan. Exp. Med. Biol., 206, 313-330. Everson, R.B., C.M. Wehr, G.L. Erexson and J.T. MacGregor (1988) Association of marginal folate depletion with increased human chromosomal damage in vivo: Demonstration by analysis of micronucleated erythrocytes, J. Natl. Cancer Inst., 80, 525-529. Herbert, V., and N. Colman (1979) Hematological aspects of folate deficiency, in: M.I. Botez and E.H. Reynolds (Eds.), Folic Acid in Neurology, Psychiatry and Internal Medicine, Raven, New York, pp. 63-74. Pearson, H.A., D. Johnston, K.A. Smith and R.J. Touloukian (1978) The born-again spleen: Return of splenic function after splenectomy for trauma, N. Engl. J. Med., 298, 1389-1392. Reidy, J.A., X. Zhou and A.T.L. Chen (1983) Folic acid and chromosome breakage, I. Implications for genotoxicity studies, Mutation Res., 122, 217-221. Schlegel, R., and J.T. MacGregor (1982) The persistence of micronuclei in peripheral blood erythrocytes: Detection of chronic chromosome breakage in mice, Mutation Res., 104, 367-369. Schlegel, R., and J.T. MacGregor (1983) A rapid screen for cumulative chromosomal damage in mice: accumulation of circulating micronucleated erythrocytes, Mutation Res., 113, 481-487. Schlegel, R., and J.T. MacGregor (1984) The persistence of micronucleated erythrocytes in the peripheral circulation of normal and splenectomized Fischer 344 rats: Implications for cytogenetic screening, Mutation Res., 127, 169-174. Schlegel, R., J.T. MacGregor and R.B. Everson (1986) Assessment of cytogenetic damage by quantitation of micronuclei in human peripheral blood erythrocytes, Cancer Res., 46, 3717-3721. Senti, F.R., and S.M. Pilch (1984) Assessment of the folate nutritional status of the U.S. population based on data collected in the second National Health and Nutrition Examination Survey, 1976-1980, FDA, Washington, DC. Wintrobe, M.M., G.R. Lee, D.R. Boggs et al. (1981) Clinical Hematology, Lea and Febiger, Philadelphia, pp. 108-135. Communicated by J.W. Allen
