Food Chemistry 153 (2014) 60–65
Contents lists available at ScienceDirect
Food Chemistry journal homepage: www.elsevier.com/locate/foodchem
Analytical Methods
Competitive chemiluminescent enzyme immunoassay for vitamin B12 analysis in human milk Daniela Hampel a,b,⇑, Setareh Shahab-Ferdows a, Joseph M. Domek a, Towfida Siddiqua a,b,c, Rubhana Raqib c, Lindsay H. Allen a,b a b c
USDA, ARS Western Human Nutrition Research Center, 430 W. Health Sciences Drive, Davis, CA 95616, USA Department of Nutrition, University of California, One Shields Ave., Davis, CA 95616, USA International Centre for Diarrheal Disease Research, Dhaka, Bangladesh
a r t i c l e
i n f o
Article history: Received 23 March 2013 Received in revised form 27 November 2013 Accepted 7 December 2013 Available online 14 December 2013 Keywords: Cobalamin Vitamin B12 Human milk Haptocorrin Chemiluminescence Radioassay
a b s t r a c t Recent discoveries of matrix interferences by haptocorrin (HC) in human milk and serum show that past analyses of vitamin B12 in samples with high HC content might have been inaccurate (Lildballe et al., 2009; Carmel & Agrawal, 2012). We evaluated two competitive enzyme-binding immunoassays for serum/plasma (IMMULITE and SimulTRAC-SNB) for B12 analysis in human milk. B12-recovery rates (United States Environmental Protection Agency, 2007) were determined to be 78.9 ± 9.1% with IMMULITE and 225 ± 108% (range 116–553%) using SimulTRAC-SNB, most likely due to the presence of excess HC. HC-interferences were not observed with the IMMULITE assay, rendering previously reported mandatory HC-removal (Lildballe et al., 2009) unnecessary. Linearity continued at low B12-concentrations (24–193 pM; r2 > 0.985). Milk B12 concentrations from Bangladeshi women (72–959 pM) were significantly lower than those from California (154–933 pM; p < 0.0001) showing IMMULITE’s robustness against the complex milk matrix and its ability to measure low milk B12 concentrations. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Although vitamin B12 has been analyzed in human milk for the last 40 years, only limited attention has been given to the methods used for analysis. Techniques used for vitamin B12 analysis in human milk include microbiological (Craft, Matthews, & Linnell, 1971; Jadhav, Webb, Vaishnava, & Baker, 1962; Jathar, Kamath, Parikh, Rege, & Satoskar, 1970; Samson & McClelland, 1980; Sandberg, Begley, & Hall, 1981) and competitive protein binding assays (CPBA) developed for serum and plasma samples (Areekul, Quarom, & Doungbarn, 1977; Casterline, Allen, & Ruel, 1997; Donangelo et al., 1989; Keizer, Gibson, & O’Connor, 1995; McPhee, Davidson, Leahy, & Beare, 1988; Patel & Lovelady, 1998; Sneed, Zane, & Thomas, 1981; Specker, Black, Allen, & Morrow, 1990; Thomas, Kawamoto, Sneed, & Eakin, 1979; Thomas et al., 1980; Trugo & Sardinha, 1994; Van Zoeren-Grobben, Schrijver, Van den Berg, & Berger, 1987). The latter have emerged as the method of choice in Abbreviations: HC, haptocorrin; cbi, cobinamide; CPBA, competitive protein binding assay; LOQ, limit of quantitation; RDA, Recommended Dietary Allowance; AI, adequate intake; wk, weeks; mo, months; DTT, dithiothreitol; KCN, potassium cyanide; IF, intrinsic factor; RA, radioassay; RT, room temperature. ⇑ Corresponding author. Tel.: +1 530 752 9519; fax: +1 530 752 5295. E-mail addresses: [email protected], [email protected] (D. Hampel). 0308-8146/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.foodchem.2013.12.033
recent years (Deegan et al., 2012; Honzik et al., 2010; Israel-Ballard et al., 2008; Lildballe, Hardlei, Allen, & Nexo, 2009) even though little has been reported concerning validation and adaptability of this method to the human milk matrix (Areekul et al., 1977; McPhee et al., 1988). The concept of CPBA is based on the release of the vitamin from its binding proteins transcobalamin (TC) and haptocorrin (HC) prior to analysis. Normal serum/plasma TC and HC concentrations are 0.5–1.5nM and 10 nM interfere with the analysis in both serum and breast milk, resulting in falsely high or low B12 concentrations depending on the CPBA used for analysis (Lildballe et al., 2009). Therefore,
D. Hampel et al. / Food Chemistry 153 (2014) 60–65
B12-analyses may have been inaccurate when excess amounts of apo-HC are present as in breast milk, or in serum/plasma samples from patients with myeloproliferative disorders (Ermens, Vlasveld, & Lindemans, 2003; Gimsing, 1998), agreeing with the observation that these types of assays fail to accurately determine low B12 values in such serum samples (Carmel & Agrawal, 2012). However, Lildballe et al. (2009) found that removing apo-HC by cobinamide (cbi)-Sepharose treatment prior to analysis resulted in accurate determination of B12 in milk and serum, and validated the method using CPBA on the Centaur analyzer (Lildballe et al., 2009). Unfortunately, this pre-treatment requires laborious preparation of the cbi-Sepharose and each batch needs to be tested for optimal HC-removal. Moreover, low B12 concentrations in milk from women with poor B12 status challenges the limit of quantitation for automated systems. Only 35% of milk samples from Guatemalan women could be quantified after cbi-pretreatment prior to Centaur analysis as described, because 65% of the samples had B12 concentrations below the limit of quantitation (LOQ) Lildballe et al., 2009; Deegan et al., 2012. This study examined two CPBAs for their suitability to analyze vitamin B12 in human milk, even at low concentrations ( 0.05). Prolonged incubation to 2 and 3 h without changing the cbi-Sepharose showed similar B12 values to the original 1 h incubation (data not shown). Analysis by IMMULITE indicated that the cbi-Sepharose pretreatment is unnecessary. Comparing 20 different samples revealed a good linear agreement between pre-treated and untreated samples indicated by linear regression analysis which yielded a slope of 1.06 and a Pearson’s correlation coefficient of 0.99 (p < 0.0001), and no significant difference (p > 0.05) was observable between the milk reference values and the B12 values obtained with IMMULITE. Therefore, cbi-Sepharose pre-treatment of the samples was not carried out for IMMULITE analyses. Comparing results obtained with RA to the milk reference values (Lildballe et al., 2009) confirmed considerably higher B12 values than the reference when RA was used. Moreover, a significant correlation was observed with total HC in the samples (qRA-total HC: 0.8669, p = 0.0001, Fig. 1). 3.2. IMMULITE/IMMULITE 1000 vs. SimulTRAC-SNB Aliquots of pooled normal plasma (UTAK Laboratories; Inc. Valencia, CA) were analyzed with RA and IMMULITE. Samples analyzed with the RA showed significantly higher B12 concentrations than aliquots analyzed with IMMULITE (405 pg/mL ± 31.4 (n = 22) vs. 379 pg/mL ± 22.4 (n = 23); p = 0.001), and even greater
2500 B12 vs. reference values [pM]
62
Reference
1000 500 0 1
2
3
4
5
6
7 8 Sample
9
10
11
12
13
2000 RA
B12 [pM]
Statistical analysis during method development and sample analysis was carried out using Analysis ToolPack in Microsoft Excel 2007 (Microsoft, Redmond, WA) and SASÒ statistical software 9.3 (SAS Institute, Cary, NC). The Shapiro–Wilk test was used to assess normality. Statistical tests for method development included twotail t-test, paired t-test, one-way ANOVA, and Pearson’s correlation coefficient as indicated in the respective sections below. B12 concentrations obtained from milk samples of Bangladeshi and Californian women could not be normalized; thus they were compared by the Mann–Whitney rank sum test. P-values < 0.05 were considered to be statistically significant.
A
RA
1500
1500
2.8. Statistical analysis
IMMULITE
2000
B
IMMULITE
y ( ) = 0.0066x + 314.52 R² = 0.7515
1000 500 0 y ( ) = 0.0008x - 60.663 R² = 0.047 -500 0
50000
100000 total HC [pM]
150000
200000
Fig. 1. B12 concentration in human milk samples analyzed with reference method (dark grey), IMMULITE (light grey) and RA (white; A) and correlation between total HC and differences of B12 concentration based on analysis method tested (RA, IMMULITE) compared to the reference values (a: Pearson’s correlation coefficient qRA-total HC: 0.8669, p = 0.0001; qIMMULITE-total HC: 0.2168, p > 0.05; B; slopes are significantly different (GLM procedure), p = 0.0023).
63
D. Hampel et al. / Food Chemistry 153 (2014) 60–65
5.6–7.6% over a period of 5 mo (n = 41; Table 2). Centrifuging the samples after the initial heating step to remove minimal precipitation did not produce any significant differences in B12 concentrations (data not shown); therefore, this precipitate was not removed during sample preparation and analysis. The analysis of NIST SRM 1849 infant/adult formula (n = 29) resulted in a mean recovery of 91.6 ± 3.0%, slightly higher than the recovery rates obtained for human milk (Table 2).
1200
y = 0.8695x + 241.88 R² = 0.9903
B12 [pg/mL]
1000 800 600 400 200
3.4. Linearity at low B12 concentrations and limit of detection and quantitation (LOQ)
0 0
100
200
300
400
500
600
700
800
900 1000
Std-Addition [pg/mL]
Analysis of diluted breast milk samples (n = 31, B12 range: 24– 93 pM) showed a mean linear correlation coefficient (r) greater than 0.992 and a mean coefficient of determination (r2) greater than 0.985 (Fig. 3A). To account for the run-to-run variability of the assay, the counts of the milk samples were normalized against the zero calibrator for the respective run. One-way ANOVA analysis showed that the values obtained with different dilution factors were significantly different from each other (p < 0.0001). One-way ANOVA analysis of the diluted calibrator curve revealed that very low B12 concentrations were not significantly different from 0 to 70 pg/mL (Fig. 3B). Values in the 80–170 pg/ mL range were quantifiable but not significantly different in the range of ±20–40 pg/mL.
Fig. 2. Mean vitamin B12 concentrations in human milk in response to B12 standard addition (n = 50 over 10 sets in 3 months, STDEV indicated as bars).
differences with aliquots of pooled human milk (781 pg/mL ± 124, n = 40 vs. 262 pg/mL ± 21.4, n = 50; p < 0.0001) and NIST SRM 1989 infant formula (438 pg/mL ± 136, n = 30 vs. 367 pg/mL ± 12.1, n = 29, p = 0.0074). 3.3. IMMULITE recovery rates and precision Analysis of ten different recovery experiments (n = 50) over three months revealed a linear curve with a mean linear correlation coefficient (r) greater than 0.995 and a coefficient of determination (r2) greater than 0.99 (Fig. 2). The mean recovery rate for B12 overall was 78.9 ± 9.1% with an intraassay precision ranging from 4.6% to 7.4%, depending on the amount of added B12 (Table 1). Calculating recovery rates based on the manufacturer’s method resulted in a B12 recovery of 85.3–94.4%, which is slightly lower than the manufacturer’s stated B12 recovery of 93–115%. The control samples and the calibrator provided by the manufacturer and analyzed with each sample set revealed an interassay variation of
3.5. Application to human milk samples from Bangladesh and California The validated method was applied to existing milk samples from women in Bangladesh (n = 35; 72–959 pM) and California (n = 26, 154–933 pM). Samples obtained from Bangladeshi women were significantly lower in B12 concentration compared to milk B12 from Californian women (p < 0.0001; Fig. 4). Considerably more milk samples from Bangladesh were in the low B12-range (0–220 pM) compared to milk obtained from Californian women.
Table 1 Mean B12 concentration [pg/mL] and mean recovery rates [%] ± StDEV in human milk with and without B12 standard addition in 10 runs over 3 months. STD-Addition [pg/mL]
n
Mean B12 [pg/mL]
STDEV [pg/mL]
Intraassay variation [%]
Mean recov. [%]a
Mean recov. [%]b
0 100 200 300 400 500 600 700 800 900 Total
50 50 50 50 50 49 48 49 50 50 497
262 328 395 482 573 668 814 879 939 991 –
±21.4 ±31.1 ±36.3 ±65.8 ±57.5 ±74.3 ±103 ±108 ±112 ±122 –
5.9 5.7 5.0 6.2 6.0 6.6 7.0 7.4 5.3 4.6 –
– 65.5 66.5 73.3 77.6 81.1 92.0 88.0 84.6 81.0 78.9 ± 9.10
– 90.5 85.5 85.8 86.5 87.6 94.4 91.3 88.4 85.3 88.4 ± 3.14
Recovery rates calculated as described (aR [%] = (Cmeasured – Cendogenous)⁄100/Cadded, bR [%] = (Cobserved/Cexpected)⁄100 according to the manufacturer’s protocol (see section 2.5).
Table 2 Mean B12 concentration [pg/mL] ± StDEV, range, CV and recovery of controls (CON) and calibrators (zero and high) analyzed with every set (n = 41) over 5 months and NIST SRM 1849 infant formula (n = 29).
*
Control
Mean ± StDEV [pg/mL]
Range [pg/mL]
CV [%]
Recovery [%]
CON 4 CON 5 CON 6 Zero calibrator* High calibrator NIST
231 ± 14.7 375 ± 20.9 731 ± 52.1 1.35e07 ± 1.02e06 678 ± 48.0 367 ± 12.1
191–266 313–421 624–878 1.13 e07–1.52 e07 557–782 344–394
6.4 5.6 7.1 7.6 7.1 3.3
99.1 96.2 105.3 n/a n/a 91.6
Mean, StDEV and range expressed in counts.
64
D. Hampel et al. / Food Chemistry 153 (2014) 60–65
1.20
y = -0.0028x + 1.0194 R² = 0.9851
normlaized counts
1.00
A
0.80 0.60 0.40 0.20 0.00 0 1.4E+07
50 100 150 B12 concentration in diluted milk samples [pM] y = 2.1492x3 - 591.84x2 + 19629x + 1E+07 R² = 0.9828
1.3E+07
200
B
counts
1.2E+07 1.1E+07 1.0E+07 9.0E+06 8.0E+06 0
20
40
60
80 100 120 B12 [pg/mL]
140
160
180
Fig. 3. Mean normalized counts responding to B12 concentration in diluted human milk (A; dilution factors: 0.125, 0.25, 0.5, 0.75, 1 (24–193 pM); n P 6 over 9 sets in 3 months; counts normalized against the mean of the zero calibrator (n = 5) for the respective run) and mean counts responding to B12 concentrations (B) of diluted calibrator solution (n = 3 for each concentration; n = 5 for 0 pg/mL B12).
4. Discussion
Frequency [%]
This evaluation of two CPBAs confirmed that the presence of apo-HC can interfere with the analysis. However, when proteins are inactivated and denaturized by heat in the presence of dithiothreitol and potassium cyanide, interferences due to HC or nonspecific protein activities are not observed. Both, the SimulTRAC and Centaur use basic conditions to liberate B12 from HC without heat treatment. While the latter can be used to accurately measure cobalamin after apo-HC removal (Lildballe et al., 2009), this proved not to be true when SimulTRAC-SNB was used for analysis. The strong correlation between the differences in B12 concentrations obtained with RA and reference values and total HC in the samples indicates that HC has a major influence on the assay tested even after apo-HC removal (Fig. 1), suggesting that remaining HC is not efficiently inactivated when samples are treated according to the SimulTRAC protocol. Since the assay is validated to analyze serum and plasma which usually contain significantly lower HC
45% 40% 35% 30% 25% 20% 15% 10% 5% 0%
concentrations, the base treatment might not be sufficient to completely inactivate the excess amounts of HC in breast milk. Results obtained with IMMULITE analysis were comparable to the reference values even without cbi-Sepharose pre-treatment, most likely due to the harsh but efficient heating step. The complexity of the human milk matrix compared to plasma might be responsible for the lower recovery rates of 80%. However, the imprecision is within the manufacturer’s range as described above and no additional pre-treatment is required using the original protocol. Nevertheless, slight modifications to the sample preparation are highly recommended: capping the samples and exact timing of the heating step followed by a 5 min cool down in an ice/water bath produced a considerable improvement in the reproducibility. Linearity of the method could be shown in diluted human milk for B12 concentrations as low as 24 pM. While the diluted milk samples produced distinguishable results, the calibrator curve did not allow differentiation of B12 values below 60 pM. Given the limitations of the calibrator curve analysis (only 3 replicates per point, one time experiment) the results obtained are rather tentative. However, when analyzing B12 at low concentrations, an additional standard curve using breast milk as matrix for the expected low B12 range below the recommended calibration range should be analyzed with the samples to ensure accuracy of the results. The B12 recovery of NIST SRM 1849 formula of 91.6% again demonstrates some interference possible due to the milk based matrix. The relative standard variation of ±3.3% shows high precision of the measurement. Vitamin B12 deficiency and depletion has been identified as a wide-spread problem in populations with a low intake of animal source foods (Allen, 2012). Using IMMULITE, human milk vitamin B12 concentrations can now be compared across these populations. Since maternal B12 deficiency is known to result in neurological and developmental disorders in the exclusively breastfed infant (Dror & Allen, 2008), it is of critical importance to know what concentration of B12 in human milk is protective, whether maternal supplementation during pregnancy and/or lactation is required, and if so, what level and timing of supplementation is appropriate. Also, given that the adequate intake (AI) of the vitamin for infants is based on analysis of very few human milk samples, using uncertain methods and uncertain maternal B12 status (Institute of Medicine, 1998), ‘‘normal’’ values of the vitamin in human milk need to be established. In conclusion, we have described the first application of a commercial CPBA intended for analyzing B12 in serum/plasma to B12 analysis in human milk without laborious pretreatment. The assay was validated for linearity, accuracy, matrix effects, precision and stability. LOQ was determined to be 24 pM demonstrating the IMMULITE assay’s linearity well below the sensitivity of the assay, enabling for the first time quantitative measurement of B12 concentrations below 50 pM in milk from vitamin B12 deficient women. Whether or not analysis of B12 in human milk can be used
Bangladesh
California
150 220 400 500 600 700 800 900 1000
150 220 400 500 600 700 800 900 1000
B12 concentration [pM]
B12 concentration [pM]
Fig. 4. Histogram of milk vitamin B12 concentrations from Bangladeshi (n = 35) and Californian (n = 26) women.
D. Hampel et al. / Food Chemistry 153 (2014) 60–65
as a non-invasive indicator for maternal and/or infant B12 status remains to be determined. Acknowledgments The authors thank Drs. Linda S. Adair and Margaret E. Bentley (University of North Carolina), Bill & Melinda Gates Foundation (OPP53107 and OPP1061055), Nestlé Foundation (FR-112/12/ 745/2(R-1)) and intramural USDA-ARS Project # 5306-51000003-00D. The reference values for B12 and total holo TC in our milk samples were obtained from the study conducted by Lildballe et al. (2009). USDA is an equal opportunity employer and provider. References Lildballe, D. L., Hardlei, T. F., Allen, L. H., & Nexo, E. (2009). High concentrations of haptocorrin interfere with routine measurement of cobalamins in human serum and milk. A problem and its solution. Clinical Chemistry and Laboratory Medicine, 47, 182–187. Jadhav, M., Webb, J. K. G., Vaishnava, S., & Baker, S. J. (1962). Vitamin B12 deficiency in Indian infants. Lancet, 280, 903–907. Jathar, V. S., Kamath, S. A., Parikh, M. N., Rege, D. V., & Satoskar, R. S. (1970). Maternal milk and serum vitamin B12, folic acid, and protein levels in Indian subjects. Archives of Disease in Childhood, 45, 236–241. Craft, I. L., Matthews, D. M., & Linnell, J. C. (1971). Cobalamins in human pregnancy and lactation. Journal of Clinical Pathology, 24, 449–455. Sandberg, D. P., Begley, J. A., & Hall, C. A. (1981). The content, binding, and forms of vitamin B12 in milk. American Journal of Clinical Nutrition, 34, 1717–1724. Samson, R. R., & McClelland, D. B. (1980). Vitamin B12 in human colostrum and milk. Quantitation of the vitamin and its binder and the uptake of bound vitamin B12 by intestinal bacteria. Acta Paediatr Scand, 69, 93–99. Trugo, N. M. F., & Sardinha, F. (1994). Cobalamin and cobalamin-binding capacity in human milk. Nutrition Research, 14, 23–33. Sneed, S. M., Zane, C., & Thomas, M. R. (1981). The effects of ascorbic acid, vitamin B6, vitamin B12, and folic acid supplementation on the breast milk and maternal nutritional status of low socioeconomic lactating women. American Journal of Clinical Nutrition, 34, 1338–1346. Donangelo, C. M., Trugo, N. M., Koury, J. C., Barreto, S. M. I., Freitas, L. A., Feldheim, W., et al. (1989). Iron, zinc, folate and vitamin B12 nutritional status and milk composition of low-income Brazilian mothers. European Journal of Clinical Nutrition, 43, 253–266. Specker, B. L., Black, A., Allen, L. H., & Morrow, F. (1990). Vitamin B-12: Low milk concentrations are related to low serum concentrations in vegetarian women and to methylmalonic aciduria in their infants. American Journal of Clinical Nutrition, 52, 1073–1076. Areekul, S., Quarom, K., & Doungbarn, J. (1977). Determination of vitamin B12 and vitamin B12 binding proteins in human and cow’s milk. Modern Medicine Asia, 13, 17–23. Thomas, M. R., Kawamoto, J., Sneed, S. M., & Eakin, R. (1979). The effects of vitamin C, vitamin B6, and vitamin B12 supplementation on the breast milk and maternal status of well-nourished women. American Journal of Clinical Nutrition, 32, 1679–1685. Thomas, M. R., Sneed, S. M., Wei, C., Nail, P. A., Wilson, M., & Sprinkle, E. (1980). The effects of vitamin C, vitamin B6, vitamin B12, folic acid, riboflavin, and thiamin on the breast milk and maternal status of well-nourished women at 6 months postpartum. American Journal of Clinical Nutrition, 33, 2151–2156. Van Zoeren-Grobben, D., Schrijver, J., Van den Berg, H., & Berger, H. M. (1987). Human milk vitamin content after pasteurisation, storage, or tube feeding. Archives of Disease in Childhood, 62, 161–165.
65
McPhee, A. J., Davidson, G. P., Leahy, M., & Beare, T. (1988). Vitamin B12 deficiency in a breast fed infant. Archives of Disease in Childhood, 63, 921–923. Casterline, J. E., Allen, L. H., & Ruel, M. T. (1997). Vitamin B-12 deficiency is very prevalent in lactating Guatemalan women and their infants at three months postpartum. Journal of Nutrition, 127, 1966–1972. Patel, K. D., & Lovelady, C. A. (1998). Vitamin B12 status of east Indian vegetarian lactating women living in the United States. Nutrition Research, 18, 1839– 1846. Keizer, S. E., Gibson, R. S., & O’Connor, D. L. (1995). Postpartum folic acid supplementation of adolescents: Impact on maternal folate and zinc status and milk composition. American Journal of Clinical Nutrition, 62, 377– 384. Honzik, T., Adamovicova, M., Smolka, V., Magner, M., Hruba, E., & Zeman, J. (2010). Clinical presentation and metabolic consequences in 40 breastfed infants with nutritional vitamin B12 deficiency-What have we learned? European Journal of Paediatric Neurology, 14, 488–495. Israel-Ballard, K. A., Abrams, B. F., Coutsoudis, A., Sibeko, L. N., Cheryk, L. A., & Chantry, C. J. (2008). Vitamin content of breast milk from HIV-1–infected mothers before and after flash-heat treatment. Journal of Acquired Immune Deficiency Syndromes, 48, 444–449. Deegan, K. L., Jones, K. M., Zuleta, C., Ramirez-Zea, M., Lildballe, D. L., Nexo, E., et al. (2012). Breast milk vitamin B-12 concentrations in Guatemalan women are correlated with maternal but not infant vitamin B-12 status at 12 months postpartum. Journal of Nutrition, 142, 112–116. Carmel, R., & Agrawal, Y. P. (2012). Failures of cobalamin assays in pernicious anemia. New England Journal of Medicine, 367, 385–386. Nexø, E., Hansen, M., Poulsen, S. S., & Olsen, P. S. (1985). Characterization and immunohistochemical localization of rat salivary cobalamin-binding protein and comparison with human salivary haptocorrin. Biochimica et Biophysica Acta, 838, 264–269. Siemens (2008). IMMULITE Vitamin B12. Llanberis. UK: Siemens Healthcare Diagnostics Products Ltd. MP. SimulTRAC-SNB. Orangeburg, NY: MP Diagnostics Division. Institute of Medicine (1998). Dietary Reference intakes for thiamin, riboflavin, niacin, vitamin B6, folate, vitamin B12, pantothenic acid, biotin, and choline. Washington, DC: National Academies Press. Allen, L. H. (2012). B vitamins in breast milk: Relative importance of maternal status and intake, and effects on infant status and function. Advances in Nutrition, 3, 362–369. Dror, D. K., & Allen, L. H. (2008). Effect of vitamin B12 deficiency on neurodevelopment in infants: Current knowledge and possible mechanisms. Nutrition Reviews, 66, 250–255. Ermens, A., Vlasveld, L., & Lindemans, J. (2003). Significance of elevated cobalamin (vitamin B12) levels in blood. Clinical Biochemistry, 36, 585–590. Gimsing, P. (1998). Cobalamin metabolism in chronic myelogenous leukemia. Danish Medical Bulletin, 45, 459–479. Lau, K. S., Gottlieb, C., Wassermann, L. R., & Herbert, V. (1965). Measurement of serum vitamin B12 level using radioisotope dilution and coated charcoal. Blood, 26, 202–214. Nexo, E., Christensen, A. L., Petersen, T. E., & Fedosov, S. N. (2000). Measurement of transcobalamin by ELISA. Clinical Chemistry, 46, 1643–1649. Sandberg, D., Begley, J., & Hall, C. (1981). The content, binding, and forms of vitamin B12 in milk. American Journal of Clinical Chemistry, 34, 1717–1724. Van Zoeren-Gobben, D., Schrijver, J., Van den Berg, H., & Berger, H. M. (1987). Human milk vitamin content after pasteurisation, storage, or tube feeding. Archives of Disease in Childhood, 62, 161–165. Wuerges, J., Geremia, S., & Randaccio, L. (2007). Structural study on ligand specificity of human vitamin B12 transporters. Biochemical Journal, 403, 431–439. United States Environmental Protection Agency. Equations utilized in conjunction with SOM01.2. Washington DC, 2007;1–43. Center for Drug Evaluation and Research (CDER). Guidance for industry: bioanalytical method validation, 2001.
Contents lists available at ScienceDirect
Food Chemistry journal homepage: www.elsevier.com/locate/foodchem
Analytical Methods
Competitive chemiluminescent enzyme immunoassay for vitamin B12 analysis in human milk Daniela Hampel a,b,⇑, Setareh Shahab-Ferdows a, Joseph M. Domek a, Towfida Siddiqua a,b,c, Rubhana Raqib c, Lindsay H. Allen a,b a b c
USDA, ARS Western Human Nutrition Research Center, 430 W. Health Sciences Drive, Davis, CA 95616, USA Department of Nutrition, University of California, One Shields Ave., Davis, CA 95616, USA International Centre for Diarrheal Disease Research, Dhaka, Bangladesh
a r t i c l e
i n f o
Article history: Received 23 March 2013 Received in revised form 27 November 2013 Accepted 7 December 2013 Available online 14 December 2013 Keywords: Cobalamin Vitamin B12 Human milk Haptocorrin Chemiluminescence Radioassay
a b s t r a c t Recent discoveries of matrix interferences by haptocorrin (HC) in human milk and serum show that past analyses of vitamin B12 in samples with high HC content might have been inaccurate (Lildballe et al., 2009; Carmel & Agrawal, 2012). We evaluated two competitive enzyme-binding immunoassays for serum/plasma (IMMULITE and SimulTRAC-SNB) for B12 analysis in human milk. B12-recovery rates (United States Environmental Protection Agency, 2007) were determined to be 78.9 ± 9.1% with IMMULITE and 225 ± 108% (range 116–553%) using SimulTRAC-SNB, most likely due to the presence of excess HC. HC-interferences were not observed with the IMMULITE assay, rendering previously reported mandatory HC-removal (Lildballe et al., 2009) unnecessary. Linearity continued at low B12-concentrations (24–193 pM; r2 > 0.985). Milk B12 concentrations from Bangladeshi women (72–959 pM) were significantly lower than those from California (154–933 pM; p < 0.0001) showing IMMULITE’s robustness against the complex milk matrix and its ability to measure low milk B12 concentrations. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Although vitamin B12 has been analyzed in human milk for the last 40 years, only limited attention has been given to the methods used for analysis. Techniques used for vitamin B12 analysis in human milk include microbiological (Craft, Matthews, & Linnell, 1971; Jadhav, Webb, Vaishnava, & Baker, 1962; Jathar, Kamath, Parikh, Rege, & Satoskar, 1970; Samson & McClelland, 1980; Sandberg, Begley, & Hall, 1981) and competitive protein binding assays (CPBA) developed for serum and plasma samples (Areekul, Quarom, & Doungbarn, 1977; Casterline, Allen, & Ruel, 1997; Donangelo et al., 1989; Keizer, Gibson, & O’Connor, 1995; McPhee, Davidson, Leahy, & Beare, 1988; Patel & Lovelady, 1998; Sneed, Zane, & Thomas, 1981; Specker, Black, Allen, & Morrow, 1990; Thomas, Kawamoto, Sneed, & Eakin, 1979; Thomas et al., 1980; Trugo & Sardinha, 1994; Van Zoeren-Grobben, Schrijver, Van den Berg, & Berger, 1987). The latter have emerged as the method of choice in Abbreviations: HC, haptocorrin; cbi, cobinamide; CPBA, competitive protein binding assay; LOQ, limit of quantitation; RDA, Recommended Dietary Allowance; AI, adequate intake; wk, weeks; mo, months; DTT, dithiothreitol; KCN, potassium cyanide; IF, intrinsic factor; RA, radioassay; RT, room temperature. ⇑ Corresponding author. Tel.: +1 530 752 9519; fax: +1 530 752 5295. E-mail addresses: [email protected], [email protected] (D. Hampel). 0308-8146/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.foodchem.2013.12.033
recent years (Deegan et al., 2012; Honzik et al., 2010; Israel-Ballard et al., 2008; Lildballe, Hardlei, Allen, & Nexo, 2009) even though little has been reported concerning validation and adaptability of this method to the human milk matrix (Areekul et al., 1977; McPhee et al., 1988). The concept of CPBA is based on the release of the vitamin from its binding proteins transcobalamin (TC) and haptocorrin (HC) prior to analysis. Normal serum/plasma TC and HC concentrations are 0.5–1.5nM and 10 nM interfere with the analysis in both serum and breast milk, resulting in falsely high or low B12 concentrations depending on the CPBA used for analysis (Lildballe et al., 2009). Therefore,
D. Hampel et al. / Food Chemistry 153 (2014) 60–65
B12-analyses may have been inaccurate when excess amounts of apo-HC are present as in breast milk, or in serum/plasma samples from patients with myeloproliferative disorders (Ermens, Vlasveld, & Lindemans, 2003; Gimsing, 1998), agreeing with the observation that these types of assays fail to accurately determine low B12 values in such serum samples (Carmel & Agrawal, 2012). However, Lildballe et al. (2009) found that removing apo-HC by cobinamide (cbi)-Sepharose treatment prior to analysis resulted in accurate determination of B12 in milk and serum, and validated the method using CPBA on the Centaur analyzer (Lildballe et al., 2009). Unfortunately, this pre-treatment requires laborious preparation of the cbi-Sepharose and each batch needs to be tested for optimal HC-removal. Moreover, low B12 concentrations in milk from women with poor B12 status challenges the limit of quantitation for automated systems. Only 35% of milk samples from Guatemalan women could be quantified after cbi-pretreatment prior to Centaur analysis as described, because 65% of the samples had B12 concentrations below the limit of quantitation (LOQ) Lildballe et al., 2009; Deegan et al., 2012. This study examined two CPBAs for their suitability to analyze vitamin B12 in human milk, even at low concentrations ( 0.05). Prolonged incubation to 2 and 3 h without changing the cbi-Sepharose showed similar B12 values to the original 1 h incubation (data not shown). Analysis by IMMULITE indicated that the cbi-Sepharose pretreatment is unnecessary. Comparing 20 different samples revealed a good linear agreement between pre-treated and untreated samples indicated by linear regression analysis which yielded a slope of 1.06 and a Pearson’s correlation coefficient of 0.99 (p < 0.0001), and no significant difference (p > 0.05) was observable between the milk reference values and the B12 values obtained with IMMULITE. Therefore, cbi-Sepharose pre-treatment of the samples was not carried out for IMMULITE analyses. Comparing results obtained with RA to the milk reference values (Lildballe et al., 2009) confirmed considerably higher B12 values than the reference when RA was used. Moreover, a significant correlation was observed with total HC in the samples (qRA-total HC: 0.8669, p = 0.0001, Fig. 1). 3.2. IMMULITE/IMMULITE 1000 vs. SimulTRAC-SNB Aliquots of pooled normal plasma (UTAK Laboratories; Inc. Valencia, CA) were analyzed with RA and IMMULITE. Samples analyzed with the RA showed significantly higher B12 concentrations than aliquots analyzed with IMMULITE (405 pg/mL ± 31.4 (n = 22) vs. 379 pg/mL ± 22.4 (n = 23); p = 0.001), and even greater
2500 B12 vs. reference values [pM]
62
Reference
1000 500 0 1
2
3
4
5
6
7 8 Sample
9
10
11
12
13
2000 RA
B12 [pM]
Statistical analysis during method development and sample analysis was carried out using Analysis ToolPack in Microsoft Excel 2007 (Microsoft, Redmond, WA) and SASÒ statistical software 9.3 (SAS Institute, Cary, NC). The Shapiro–Wilk test was used to assess normality. Statistical tests for method development included twotail t-test, paired t-test, one-way ANOVA, and Pearson’s correlation coefficient as indicated in the respective sections below. B12 concentrations obtained from milk samples of Bangladeshi and Californian women could not be normalized; thus they were compared by the Mann–Whitney rank sum test. P-values < 0.05 were considered to be statistically significant.
A
RA
1500
1500
2.8. Statistical analysis
IMMULITE
2000
B
IMMULITE
y ( ) = 0.0066x + 314.52 R² = 0.7515
1000 500 0 y ( ) = 0.0008x - 60.663 R² = 0.047 -500 0
50000
100000 total HC [pM]
150000
200000
Fig. 1. B12 concentration in human milk samples analyzed with reference method (dark grey), IMMULITE (light grey) and RA (white; A) and correlation between total HC and differences of B12 concentration based on analysis method tested (RA, IMMULITE) compared to the reference values (a: Pearson’s correlation coefficient qRA-total HC: 0.8669, p = 0.0001; qIMMULITE-total HC: 0.2168, p > 0.05; B; slopes are significantly different (GLM procedure), p = 0.0023).
63
D. Hampel et al. / Food Chemistry 153 (2014) 60–65
5.6–7.6% over a period of 5 mo (n = 41; Table 2). Centrifuging the samples after the initial heating step to remove minimal precipitation did not produce any significant differences in B12 concentrations (data not shown); therefore, this precipitate was not removed during sample preparation and analysis. The analysis of NIST SRM 1849 infant/adult formula (n = 29) resulted in a mean recovery of 91.6 ± 3.0%, slightly higher than the recovery rates obtained for human milk (Table 2).
1200
y = 0.8695x + 241.88 R² = 0.9903
B12 [pg/mL]
1000 800 600 400 200
3.4. Linearity at low B12 concentrations and limit of detection and quantitation (LOQ)
0 0
100
200
300
400
500
600
700
800
900 1000
Std-Addition [pg/mL]
Analysis of diluted breast milk samples (n = 31, B12 range: 24– 93 pM) showed a mean linear correlation coefficient (r) greater than 0.992 and a mean coefficient of determination (r2) greater than 0.985 (Fig. 3A). To account for the run-to-run variability of the assay, the counts of the milk samples were normalized against the zero calibrator for the respective run. One-way ANOVA analysis showed that the values obtained with different dilution factors were significantly different from each other (p < 0.0001). One-way ANOVA analysis of the diluted calibrator curve revealed that very low B12 concentrations were not significantly different from 0 to 70 pg/mL (Fig. 3B). Values in the 80–170 pg/ mL range were quantifiable but not significantly different in the range of ±20–40 pg/mL.
Fig. 2. Mean vitamin B12 concentrations in human milk in response to B12 standard addition (n = 50 over 10 sets in 3 months, STDEV indicated as bars).
differences with aliquots of pooled human milk (781 pg/mL ± 124, n = 40 vs. 262 pg/mL ± 21.4, n = 50; p < 0.0001) and NIST SRM 1989 infant formula (438 pg/mL ± 136, n = 30 vs. 367 pg/mL ± 12.1, n = 29, p = 0.0074). 3.3. IMMULITE recovery rates and precision Analysis of ten different recovery experiments (n = 50) over three months revealed a linear curve with a mean linear correlation coefficient (r) greater than 0.995 and a coefficient of determination (r2) greater than 0.99 (Fig. 2). The mean recovery rate for B12 overall was 78.9 ± 9.1% with an intraassay precision ranging from 4.6% to 7.4%, depending on the amount of added B12 (Table 1). Calculating recovery rates based on the manufacturer’s method resulted in a B12 recovery of 85.3–94.4%, which is slightly lower than the manufacturer’s stated B12 recovery of 93–115%. The control samples and the calibrator provided by the manufacturer and analyzed with each sample set revealed an interassay variation of
3.5. Application to human milk samples from Bangladesh and California The validated method was applied to existing milk samples from women in Bangladesh (n = 35; 72–959 pM) and California (n = 26, 154–933 pM). Samples obtained from Bangladeshi women were significantly lower in B12 concentration compared to milk B12 from Californian women (p < 0.0001; Fig. 4). Considerably more milk samples from Bangladesh were in the low B12-range (0–220 pM) compared to milk obtained from Californian women.
Table 1 Mean B12 concentration [pg/mL] and mean recovery rates [%] ± StDEV in human milk with and without B12 standard addition in 10 runs over 3 months. STD-Addition [pg/mL]
n
Mean B12 [pg/mL]
STDEV [pg/mL]
Intraassay variation [%]
Mean recov. [%]a
Mean recov. [%]b
0 100 200 300 400 500 600 700 800 900 Total
50 50 50 50 50 49 48 49 50 50 497
262 328 395 482 573 668 814 879 939 991 –
±21.4 ±31.1 ±36.3 ±65.8 ±57.5 ±74.3 ±103 ±108 ±112 ±122 –
5.9 5.7 5.0 6.2 6.0 6.6 7.0 7.4 5.3 4.6 –
– 65.5 66.5 73.3 77.6 81.1 92.0 88.0 84.6 81.0 78.9 ± 9.10
– 90.5 85.5 85.8 86.5 87.6 94.4 91.3 88.4 85.3 88.4 ± 3.14
Recovery rates calculated as described (aR [%] = (Cmeasured – Cendogenous)⁄100/Cadded, bR [%] = (Cobserved/Cexpected)⁄100 according to the manufacturer’s protocol (see section 2.5).
Table 2 Mean B12 concentration [pg/mL] ± StDEV, range, CV and recovery of controls (CON) and calibrators (zero and high) analyzed with every set (n = 41) over 5 months and NIST SRM 1849 infant formula (n = 29).
*
Control
Mean ± StDEV [pg/mL]
Range [pg/mL]
CV [%]
Recovery [%]
CON 4 CON 5 CON 6 Zero calibrator* High calibrator NIST
231 ± 14.7 375 ± 20.9 731 ± 52.1 1.35e07 ± 1.02e06 678 ± 48.0 367 ± 12.1
191–266 313–421 624–878 1.13 e07–1.52 e07 557–782 344–394
6.4 5.6 7.1 7.6 7.1 3.3
99.1 96.2 105.3 n/a n/a 91.6
Mean, StDEV and range expressed in counts.
64
D. Hampel et al. / Food Chemistry 153 (2014) 60–65
1.20
y = -0.0028x + 1.0194 R² = 0.9851
normlaized counts
1.00
A
0.80 0.60 0.40 0.20 0.00 0 1.4E+07
50 100 150 B12 concentration in diluted milk samples [pM] y = 2.1492x3 - 591.84x2 + 19629x + 1E+07 R² = 0.9828
1.3E+07
200
B
counts
1.2E+07 1.1E+07 1.0E+07 9.0E+06 8.0E+06 0
20
40
60
80 100 120 B12 [pg/mL]
140
160
180
Fig. 3. Mean normalized counts responding to B12 concentration in diluted human milk (A; dilution factors: 0.125, 0.25, 0.5, 0.75, 1 (24–193 pM); n P 6 over 9 sets in 3 months; counts normalized against the mean of the zero calibrator (n = 5) for the respective run) and mean counts responding to B12 concentrations (B) of diluted calibrator solution (n = 3 for each concentration; n = 5 for 0 pg/mL B12).
4. Discussion
Frequency [%]
This evaluation of two CPBAs confirmed that the presence of apo-HC can interfere with the analysis. However, when proteins are inactivated and denaturized by heat in the presence of dithiothreitol and potassium cyanide, interferences due to HC or nonspecific protein activities are not observed. Both, the SimulTRAC and Centaur use basic conditions to liberate B12 from HC without heat treatment. While the latter can be used to accurately measure cobalamin after apo-HC removal (Lildballe et al., 2009), this proved not to be true when SimulTRAC-SNB was used for analysis. The strong correlation between the differences in B12 concentrations obtained with RA and reference values and total HC in the samples indicates that HC has a major influence on the assay tested even after apo-HC removal (Fig. 1), suggesting that remaining HC is not efficiently inactivated when samples are treated according to the SimulTRAC protocol. Since the assay is validated to analyze serum and plasma which usually contain significantly lower HC
45% 40% 35% 30% 25% 20% 15% 10% 5% 0%
concentrations, the base treatment might not be sufficient to completely inactivate the excess amounts of HC in breast milk. Results obtained with IMMULITE analysis were comparable to the reference values even without cbi-Sepharose pre-treatment, most likely due to the harsh but efficient heating step. The complexity of the human milk matrix compared to plasma might be responsible for the lower recovery rates of 80%. However, the imprecision is within the manufacturer’s range as described above and no additional pre-treatment is required using the original protocol. Nevertheless, slight modifications to the sample preparation are highly recommended: capping the samples and exact timing of the heating step followed by a 5 min cool down in an ice/water bath produced a considerable improvement in the reproducibility. Linearity of the method could be shown in diluted human milk for B12 concentrations as low as 24 pM. While the diluted milk samples produced distinguishable results, the calibrator curve did not allow differentiation of B12 values below 60 pM. Given the limitations of the calibrator curve analysis (only 3 replicates per point, one time experiment) the results obtained are rather tentative. However, when analyzing B12 at low concentrations, an additional standard curve using breast milk as matrix for the expected low B12 range below the recommended calibration range should be analyzed with the samples to ensure accuracy of the results. The B12 recovery of NIST SRM 1849 formula of 91.6% again demonstrates some interference possible due to the milk based matrix. The relative standard variation of ±3.3% shows high precision of the measurement. Vitamin B12 deficiency and depletion has been identified as a wide-spread problem in populations with a low intake of animal source foods (Allen, 2012). Using IMMULITE, human milk vitamin B12 concentrations can now be compared across these populations. Since maternal B12 deficiency is known to result in neurological and developmental disorders in the exclusively breastfed infant (Dror & Allen, 2008), it is of critical importance to know what concentration of B12 in human milk is protective, whether maternal supplementation during pregnancy and/or lactation is required, and if so, what level and timing of supplementation is appropriate. Also, given that the adequate intake (AI) of the vitamin for infants is based on analysis of very few human milk samples, using uncertain methods and uncertain maternal B12 status (Institute of Medicine, 1998), ‘‘normal’’ values of the vitamin in human milk need to be established. In conclusion, we have described the first application of a commercial CPBA intended for analyzing B12 in serum/plasma to B12 analysis in human milk without laborious pretreatment. The assay was validated for linearity, accuracy, matrix effects, precision and stability. LOQ was determined to be 24 pM demonstrating the IMMULITE assay’s linearity well below the sensitivity of the assay, enabling for the first time quantitative measurement of B12 concentrations below 50 pM in milk from vitamin B12 deficient women. Whether or not analysis of B12 in human milk can be used
Bangladesh
California
150 220 400 500 600 700 800 900 1000
150 220 400 500 600 700 800 900 1000
B12 concentration [pM]
B12 concentration [pM]
Fig. 4. Histogram of milk vitamin B12 concentrations from Bangladeshi (n = 35) and Californian (n = 26) women.
D. Hampel et al. / Food Chemistry 153 (2014) 60–65
as a non-invasive indicator for maternal and/or infant B12 status remains to be determined. Acknowledgments The authors thank Drs. Linda S. Adair and Margaret E. Bentley (University of North Carolina), Bill & Melinda Gates Foundation (OPP53107 and OPP1061055), Nestlé Foundation (FR-112/12/ 745/2(R-1)) and intramural USDA-ARS Project # 5306-51000003-00D. The reference values for B12 and total holo TC in our milk samples were obtained from the study conducted by Lildballe et al. (2009). USDA is an equal opportunity employer and provider. References Lildballe, D. L., Hardlei, T. F., Allen, L. H., & Nexo, E. (2009). High concentrations of haptocorrin interfere with routine measurement of cobalamins in human serum and milk. A problem and its solution. Clinical Chemistry and Laboratory Medicine, 47, 182–187. Jadhav, M., Webb, J. K. G., Vaishnava, S., & Baker, S. J. (1962). Vitamin B12 deficiency in Indian infants. Lancet, 280, 903–907. Jathar, V. S., Kamath, S. A., Parikh, M. N., Rege, D. V., & Satoskar, R. S. (1970). Maternal milk and serum vitamin B12, folic acid, and protein levels in Indian subjects. Archives of Disease in Childhood, 45, 236–241. Craft, I. L., Matthews, D. M., & Linnell, J. C. (1971). Cobalamins in human pregnancy and lactation. Journal of Clinical Pathology, 24, 449–455. Sandberg, D. P., Begley, J. A., & Hall, C. A. (1981). The content, binding, and forms of vitamin B12 in milk. American Journal of Clinical Nutrition, 34, 1717–1724. Samson, R. R., & McClelland, D. B. (1980). Vitamin B12 in human colostrum and milk. Quantitation of the vitamin and its binder and the uptake of bound vitamin B12 by intestinal bacteria. Acta Paediatr Scand, 69, 93–99. Trugo, N. M. F., & Sardinha, F. (1994). Cobalamin and cobalamin-binding capacity in human milk. Nutrition Research, 14, 23–33. Sneed, S. M., Zane, C., & Thomas, M. R. (1981). The effects of ascorbic acid, vitamin B6, vitamin B12, and folic acid supplementation on the breast milk and maternal nutritional status of low socioeconomic lactating women. American Journal of Clinical Nutrition, 34, 1338–1346. Donangelo, C. M., Trugo, N. M., Koury, J. C., Barreto, S. M. I., Freitas, L. A., Feldheim, W., et al. (1989). Iron, zinc, folate and vitamin B12 nutritional status and milk composition of low-income Brazilian mothers. European Journal of Clinical Nutrition, 43, 253–266. Specker, B. L., Black, A., Allen, L. H., & Morrow, F. (1990). Vitamin B-12: Low milk concentrations are related to low serum concentrations in vegetarian women and to methylmalonic aciduria in their infants. American Journal of Clinical Nutrition, 52, 1073–1076. Areekul, S., Quarom, K., & Doungbarn, J. (1977). Determination of vitamin B12 and vitamin B12 binding proteins in human and cow’s milk. Modern Medicine Asia, 13, 17–23. Thomas, M. R., Kawamoto, J., Sneed, S. M., & Eakin, R. (1979). The effects of vitamin C, vitamin B6, and vitamin B12 supplementation on the breast milk and maternal status of well-nourished women. American Journal of Clinical Nutrition, 32, 1679–1685. Thomas, M. R., Sneed, S. M., Wei, C., Nail, P. A., Wilson, M., & Sprinkle, E. (1980). The effects of vitamin C, vitamin B6, vitamin B12, folic acid, riboflavin, and thiamin on the breast milk and maternal status of well-nourished women at 6 months postpartum. American Journal of Clinical Nutrition, 33, 2151–2156. Van Zoeren-Grobben, D., Schrijver, J., Van den Berg, H., & Berger, H. M. (1987). Human milk vitamin content after pasteurisation, storage, or tube feeding. Archives of Disease in Childhood, 62, 161–165.
65
McPhee, A. J., Davidson, G. P., Leahy, M., & Beare, T. (1988). Vitamin B12 deficiency in a breast fed infant. Archives of Disease in Childhood, 63, 921–923. Casterline, J. E., Allen, L. H., & Ruel, M. T. (1997). Vitamin B-12 deficiency is very prevalent in lactating Guatemalan women and their infants at three months postpartum. Journal of Nutrition, 127, 1966–1972. Patel, K. D., & Lovelady, C. A. (1998). Vitamin B12 status of east Indian vegetarian lactating women living in the United States. Nutrition Research, 18, 1839– 1846. Keizer, S. E., Gibson, R. S., & O’Connor, D. L. (1995). Postpartum folic acid supplementation of adolescents: Impact on maternal folate and zinc status and milk composition. American Journal of Clinical Nutrition, 62, 377– 384. Honzik, T., Adamovicova, M., Smolka, V., Magner, M., Hruba, E., & Zeman, J. (2010). Clinical presentation and metabolic consequences in 40 breastfed infants with nutritional vitamin B12 deficiency-What have we learned? European Journal of Paediatric Neurology, 14, 488–495. Israel-Ballard, K. A., Abrams, B. F., Coutsoudis, A., Sibeko, L. N., Cheryk, L. A., & Chantry, C. J. (2008). Vitamin content of breast milk from HIV-1–infected mothers before and after flash-heat treatment. Journal of Acquired Immune Deficiency Syndromes, 48, 444–449. Deegan, K. L., Jones, K. M., Zuleta, C., Ramirez-Zea, M., Lildballe, D. L., Nexo, E., et al. (2012). Breast milk vitamin B-12 concentrations in Guatemalan women are correlated with maternal but not infant vitamin B-12 status at 12 months postpartum. Journal of Nutrition, 142, 112–116. Carmel, R., & Agrawal, Y. P. (2012). Failures of cobalamin assays in pernicious anemia. New England Journal of Medicine, 367, 385–386. Nexø, E., Hansen, M., Poulsen, S. S., & Olsen, P. S. (1985). Characterization and immunohistochemical localization of rat salivary cobalamin-binding protein and comparison with human salivary haptocorrin. Biochimica et Biophysica Acta, 838, 264–269. Siemens (2008). IMMULITE Vitamin B12. Llanberis. UK: Siemens Healthcare Diagnostics Products Ltd. MP. SimulTRAC-SNB. Orangeburg, NY: MP Diagnostics Division. Institute of Medicine (1998). Dietary Reference intakes for thiamin, riboflavin, niacin, vitamin B6, folate, vitamin B12, pantothenic acid, biotin, and choline. Washington, DC: National Academies Press. Allen, L. H. (2012). B vitamins in breast milk: Relative importance of maternal status and intake, and effects on infant status and function. Advances in Nutrition, 3, 362–369. Dror, D. K., & Allen, L. H. (2008). Effect of vitamin B12 deficiency on neurodevelopment in infants: Current knowledge and possible mechanisms. Nutrition Reviews, 66, 250–255. Ermens, A., Vlasveld, L., & Lindemans, J. (2003). Significance of elevated cobalamin (vitamin B12) levels in blood. Clinical Biochemistry, 36, 585–590. Gimsing, P. (1998). Cobalamin metabolism in chronic myelogenous leukemia. Danish Medical Bulletin, 45, 459–479. Lau, K. S., Gottlieb, C., Wassermann, L. R., & Herbert, V. (1965). Measurement of serum vitamin B12 level using radioisotope dilution and coated charcoal. Blood, 26, 202–214. Nexo, E., Christensen, A. L., Petersen, T. E., & Fedosov, S. N. (2000). Measurement of transcobalamin by ELISA. Clinical Chemistry, 46, 1643–1649. Sandberg, D., Begley, J., & Hall, C. (1981). The content, binding, and forms of vitamin B12 in milk. American Journal of Clinical Chemistry, 34, 1717–1724. Van Zoeren-Gobben, D., Schrijver, J., Van den Berg, H., & Berger, H. M. (1987). Human milk vitamin content after pasteurisation, storage, or tube feeding. Archives of Disease in Childhood, 62, 161–165. Wuerges, J., Geremia, S., & Randaccio, L. (2007). Structural study on ligand specificity of human vitamin B12 transporters. Biochemical Journal, 403, 431–439. United States Environmental Protection Agency. Equations utilized in conjunction with SOM01.2. Washington DC, 2007;1–43. Center for Drug Evaluation and Research (CDER). Guidance for industry: bioanalytical method validation, 2001.
