QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Craft, N. E. Articles by West, K. P. Search for Related Content PubMed PubMed Citation Articles by Craft, N. E. Articles by West, K. P., Jr.

---

Article

Retinol Analysis in Dried Blood Spots by HPLC^{1 ,2}

Neal E. Craft^*3, Tom Haitema^*, Lisa K. Brindle^*, Sedigheh Yamini, Jean H. Humphrey and Keith P. West, Jr.

^* Craft Technologies, Inc., Wilson, NC 27893, and Center for Human Nutrition, Johns Hopkins School of Hygiene and Public Health, Baltimore, MD 21205

³To whom correspondence should be addressed.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
There are many advantages to measuring vitamin A in dried blood spots (DBS) from a finger prick as compared to plasma collected by venipuncture. The advantages include easier collection, transport and storage; accessibility to younger and more remote populations; and decreased risk of disease transmission. We describe a method for the extraction of retinol from DBS for analysis by HPLC and initial comparison to plasma retinol. The effects of various buffers, detergents, antioxidants and chelators were evaluated to establish the most effective approach to elute the retinol: retinol binding protein (holo-RBP) complex from the blood collection cards. The process involves ultrasonic agitation to elute holo-RBP into a phosphate buffer containing an antioxidant and metal chelator. The holo-RBP complex was denatured by the addition of ethanol containing additional antioxidants permitting the extraction of free retinol into hexane. Following solvent evaporation, the extract was dissolved in methanol for HPLC analysis. The initial measured retinol levels in freshly collected DBS declined for 6–10 d whether stored at 25, 4 or -20°C, but remained consistent thereafter (homeostatic). By incorporating a "recovery/volume adjustment" factor, measured retinol values in homeostatic DBS were adjusted to the equivalent of plasma retinol. For 17 normal adults, the correlation coefficient was 0.90 between plasma retinol and adjusted DBS retinol in samples that had been stored at -70°C for < 9 mo. The use of this new sample matrix for vitamin A assessment will allow access to previously unavailable populations.

KEY WORDS: • vitamin A • dried blood spot • HPLC • filter paper • nutritional assessment

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
As the twentieth century draws to a close, the ravages of vitamin A deficiency (VAD)⁴ still plague up to 250 million individuals (Underwood and Arthur 1996 ). VAD often results in juvenile xerophthalmia, keratomalacia, Bitot’s spots, night blindness or total blindness while making children and young women more vulnerable to disease (Semba et al. 1994 , Sommer et al. 1984 ). To identify and monitor deficient or marginally deficient populations, an accurate and inexpensive analytical method must be readily available.

Conventional VAD monitoring usually involves analysis of serum retinol derived from venous blood by HPLC. The sampling process is invasive and impractical to carryout in a field situation. Typically, venipuncture is required to obtain the volume of blood necessary (>500 µL of whole blood). Due to fear of needles, disease transmission and/or religious beliefs, this factor alone excludes many from participating. Electricity is necessary for centrifugation and long-term freezer storage. Sample collection would be much easier and less invasive if dried blood spots (DBS) could be used. It does not require needles and the potential for disease transmission is minimal. Sample collection requires only a few drops of blood from a finger- or heel-prick. The blood collection cards are easily labeled, transported and stored. Transportation of samples to an analytical laboratory may not require subzero temperatures, thus reducing costs and inconvenience.

The measurement of vitamin A in DBS was first described by Shi et al. (1995) using high performance capillary electrophoresis (HPCE) with laser-enhanced fluorescence detection. Using a modification of the HPCE method of Ma et al. (1993) , Shi et al. optimized the separation conditions and improved the reliability of the method. More importantly, they developed sample preparation methods for the elution of holo-RBP from DBS, ultimately demonstrating the stability of the complex in dried blood. This observation was an important milestone since the holo-RBP complex was thought to be unstable when exposed to air and iron from the red blood cells. Now vitamin A, like many other biochemical markers, could be measured in blood samples collected from a finger- or heel-prick directly onto collection cards. However, the instrumentation and expertise necessary for the measurement were not practical for population status assessment.

The purpose of the present study was to develop a protocol for the extraction of retinol from DBS for analysis by HPLC. In addition, we examined the stability of retinol in DBS at three temperatures and provide the initial method validation using HPLC to analyze the samples rather than HPCE.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Materials.

Retinyl acetate, ascorbic acid, diethylenetriamine pentaacetic acid (DTPA), sodium phosphate, ammonium acetate, triethylamine, BHT and hydroquinone of reagent grade or better were purchased from Sigma Chemical (St. Louis, MO). Acetonitrile, p-dioxane, ethanol, hexane and methanol of HPLC grade were purchased from Curtin Matheson Scientific (Kennesaw, GA). Tocol, used as an internal standard, was generously donated by Hoffman La-Roche (Basel, Switzerland). Retinol was purchased from US Biochemical (Cleveland, OH). Blood spot cards, identical to those used for neonatal screening, were provided by Schleicher and Schuell (cat. #903, Keene, NH).

Sample collection.

Samples of venous blood were collected in vacuum tubes containing EDTA as an anticoagulant. The 17 healthy U. S. volunteers had provided informed consent to participate in a nutrition study. Sample collection cards were spotted with 50 µL aliquots of whole blood; then the remainder of the blood was centrifuged to collect plasma for comparison. This allowed the comparison of a known volume of blood with plasma samples, critical for the experimental stage of method development. These cards were air-dried overnight in a darkened room, then wrapped individually in tissue and placed in zip-closure plastic bags prior to storage at -70°C.

Chromatographic conditions.

The HPLC consisted of a vacuum solvent degasser, quaternary gradient pump, programmable UV/VIS detector fitted with deuterium and tungsten lamps and an autosampler with refrigerated sample compartment and column oven (Thermo Separation Products, San Jose, CA). The column used in the final HPLC method was a Betasil C8, 3 µm, 4.6 x 150 mm, protected with a Javelin Betasil C8, 3 µm guard column (Keystone Scientific, Bellefonte, PA). The mobile phase consisted of methanol/water (95:5, v/v) at a flow rate of 1.0 mL/min. The detector wavelength was set at 325 nm throughout the run (4.6 min). The injection volume was 20 µL.

The HPLC method of Craft (1996) was used for serum retinol analysis and DBS retinol analysis during the development of the sample extraction. It incorporated a Spherisorb ODS2, 3 µm, 4.6 x 150 mm, protected with a Javelin ODS2, 3 µm guard column (Keystone Scientific). The mobile phase consisted of acetonitrile/dioxane/methanol/triethylamine (83:13:4:0.1, v/v/v/v). Ammonium acetate (150 mmol/L) was present in the methanol component of mobile phase. The flow rate was 1.2 mL/min for DBS analysis, and 1.5 mL/min for serum analysis and column temperature was maintained at 29°C. The detector was programmed at 325 nm until 3.0 min then at 300 nm until the end of the run (6 min). The wavelength change to 300 nm was required to measure tocol as the internal standard. The injection volume was 20 µL.

Methods.

A 0.635 cm (1/4 inch) disk was punched from the center of the DBS on the collection card. A standard hole-punch was used ensuring the same quantity of filter paper was removed from sample to sample. During the development stage, whole DBS consisting of 50 µL aliquots were used, and the entire spot was cut out to ensure a known quantity of blood. The punched spot was placed in a 10 x 75 mm borosilicate test tube, and 1 mL of buffer was added. The final buffer consisted of 150 mmol/L phosphate buffer, pH 7.8, containing 57 mmol/L (10 g/L) ascorbic acid and 2 mmol/L DTPA. The sample in buffer was sonicated 15 min in an ultrasonic bath (Branson Model 1210, Danbury, CT), then 100 µL of internal standard and 900 µL of ethanol containing 100 mmol/L BHT and 50 mmol/L hydroquinone, as antioxidants, were added followed by vortex mixing for 20 s. A 2-mL portion of hexane was added and vortex-mixed for 1 min. The tube was centrifuged at 500 x g for 1 min to separate the phases. The hexane layer was removed and the extraction repeated. The combined hexane layers were evaporated under a nitrogen stream and dissolved in 60 µL of mobile phase by sonicating 10 s and vortex mixing 30 s. The reconstituted sample was placed in a conical insert before HPLC analysis.

Calibration.

Standards of retinol, retinyl acetate and tocol were prepared in ethanol. The stock retinol concentration was calculated based on Beer’s Law using an absorptivity of 1850 E_cm^1%at 325 nm. The spectroscopic concentration was corrected for purity as determined by HPLC. A multiple point, internal standard, calibration curve was generated using the ratio of peak areas. Both plasma and matching blood spots were analyzed for retinol by HPLC. A "recovery/volume adjustment" factor (plasma retinol/DBS retinol) was calculated for a subset of samples. The median of these ratios was used to adjust the concentration of DBS retinol to values equivalent to plasma retinol. We also estimated the plasma volume of a punched sample by analyzing both a 50 µL DBS and 0.635 cm punch from the same individual and correcting for packed red blood cell volume.

Experimental Procedures.

DBS and plasma from 17 healthy subjects were analyzed for retinol by HPLC. Adjusted DBS retinol values were plotted against plasma retinol values to determine the correspondence between the two data sets.

Three different sets of DBS were placed in separate envelopes and maintained at 25, 4, and -20°C. Samples from each set were analyzed in duplicate on d 1, 2, 3, 6, 15, 30 and 80. Specimens for two of the sets were freshly collected at the lab, dried and immediately analyzed to determine time 0 DBS retinol concentrations. The remaining specimen was from the Center for Disease Control’s Neonatal Screening Program (Atlanta, GA) and had been stored in sealed foil pouches at -70°C for ~2 y before this analysis.

Statistical analysis.

Means and SD for the stability data and correlation coefficients comparing the adjusted DBS retinol to the plasma retinol were determined using Microsoft Excel version 6.0 (Microsoft Corporation, Redmond, WA). Confidence limits were determined to ascertain if the regression line for plasma vs. DBS retinol differed from unity. A line having a slope of zero is a horizontal line and is indicative of no change over time. Therefore, DBS retinol values were deemed "homeostatic" when the confidence limits of the change in the slope of DBS retinol concentration vs. time encompassed zero.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
A representative chromatogram of the separation of retinol extracted from DBS is illustrated in Figure 1 .

View larger version (14K): [in this window] [in a new window]   Figure 1. HPLC chromatogram of the separation of retinol extracted from a dried blood spot. Retinol and the internal standard, retinyl acetate, elute at 3.6 and 4.0 min, respectively. HPLC Conditions: Betasil C8, 3 µm, 4.6 x 150 mm column, methanol/water (95:5, v/v), 1.0 mL/min flow rate, 325 nm detection.

View larger version (16K): [in this window] [in a new window]   Figure 2. Effect of temperature on dried blood spot (DBS) retinol measured by HPLC. Representative stability curve of retinol in DBS samples from a single subject illustrating the effect of storage on specimens at 25, 4 and -20°C. Each point represents the means ± SEM of duplicate measurements using the HPLC conditions listed in Figure 1 .

View larger version (18K): [in this window] [in a new window]   Figure 3. Effect of storage time at 4°C on dried blood spot (DBS) retinol measured by HPLC. Illustration of the effect of storage time at 4°C on three DBS specimens. Each point represents the means ± SEM of one sample stored at 4°C for the indicated time measured in duplicate using HPLC conditions listed in Figure 1 . Diamonds represent the "aged" blood spot specimen from the Center for Disease Control (CDC) Neonatal Screening Program; the squares and triangles represent freshly collected blood spot samples from subjects 1 and 2, respectively.

Figure 4 illustrates the comparison of plasma retinol to DBS retinol from 17 healthy U. S. volunteers. The correlation between plasma retinol and DBS retinol for these well-controlled samples was 0.90, and the line of identity was within the confidence intervals of the regression line for plasma retinol vs. DBS retinol.

View larger version (17K): [in this window] [in a new window]   Figure 4. Correlation between plasma retinol and blood spot retinol from 17 healthy adults measured by HPLC. Blood spots and plasma had been collected < 9 mo before analysis and stored at -70°C. The correlation coefficient, r2, is 0.90, and the line of identity is included within the confidence intervals of the regression line for plasma retinol vs. dried blood spot (DBS) retinol.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The relative dose response (RDR) (Loerch et al. 1979 ) and the modified dose response (MRDR) (Tanumihardjo et al. 1990 , 1996 ) have been used to measure body stores of vitamin A. Although these techniques are less susceptible to problems such as hemolysis and inflammation, Russell et al. (1982) pointed out the problem of using the RDR in individuals with protein malnutrition, a common companion of VAD in many developing countries. In addition, both techniques necessitate a time interval (~5 h) from dosing with vitamin A or didehydroretinol until the final blood draw. The RDR requires two blood samples while the MRDR requires only one draw. As in the Futterman method, the collection of whole blood necessitates centrifugation, low-temperature storage and the ability to transport frozen specimens. In 1993 Oliver et al. briefly reported successfully using serum samples spotted on filter paper for the analysis of retinol by HPLC. They not only demonstrated the feasibility of using filter paper-immobilized samples as a matrix for retinol analysis but also that the holo-RBP complex was stable in immobilized serum for at least 5 mo. The authors did not discuss the storage conditions under which the samples were maintained prior to analysis. This method is limited by the rather large 200 µL serum sample and the fact that the use of whole blood was not examined.

When Shi et al. (1995) demonstrated the capability to measure holo-RBP in DBS, the door was opened for much more convenient and far-reaching sample collection. Our newly developed method builds on their demonstration that retinol remains stable in DBS. It has been optimized to efficiently extract and preserve retinol from DBS. As such, it measures free retinol in the same manner as HPLC methods for plasma retinol. Thus many of the limitations of the HPCE method for analysis of holo-RBP in DBS have been overcome, e.g., instrumentation availability, technical expertise and errant sample values.

The set of DBS that was analyzed for retinol concentrations and compared to corresponding plasma values came from ideal collection conditions. The blood collection cards from Johns Hopkins had been spotted and handled in accordance with the accepted guidelines for such specimens (Hannon et al. 1992 ) and maintained at -70°C. Both plasma retinol and DBS retinol were performed at the same location using the same equipment and methodology.

Since no DBS control samples are commercially available with known concentrations of retinol, comparable to National Institute of Standards and Technology (NIST) Standard Reference Material 968: Fat-Soluble Vitamins in Human Serum, (NIST, 1995 ) the accuracy of the DBS method is improved if a subset of plasma samples with matching DBS is available to serve as calibrators. Both plasma and DBS retinol are determined in the matched samples and a "recovery/volume adjustment" factor is determined which permits the DBS retinol concentration to be expressed as plasma retinol. This factor adjusts the DBS retinol concentrations for the volume of plasma in the DBS sample and the extraction efficiency. Due to the novelty of the technique, this is deemed as a necessary precaution since variations in sample collection and populations have not yet been established. Thus far, the median factors from different populations have varied by < 10%. With further validation and the availability of quality control specimens, it may be unnecessary to include any plasma samples with DBS analysis.

Slight variations of this method have been used to measure DBS samples from the USA, Nepal, Bangladesh, Liberia and Guatemala. The HPLC method using the C8 column must be flushed daily with a lipophilic solvent, such as tetrahydrofuran, to elute strongly retained contaminants. DBS samples from CDC’s Neonatal Screening Program stored at -70°C were used as QC samples for larger studies and found to be very stable and reproducible.

Our initial attempt to examine the stability of retinol in DBS at various temperatures revealed that the measured retinol concentration in freshly prepared DBS declined for 6–10 d at all temperatures tested (See Fig. 3 ). At this point the "fresh" samples appear to have reached "homeostasis" and remained constant through d 80. However, samples held at ambient temperature plateaued at a slightly lower concentration. These data indicate an apparent initial degradation or change in extraction efficiency, followed by considerable stability. The measured retinol concentration in "aged" CDC samples remained constant at all temperatures throughout the 80 d.

This is the first report on the development of methodology to measure retinol in DBS using HPLC. In light of the strong correlation between plasma and DBS retinol and the stability observed beyond the homeostatic point, the authors feel it will serve as a useful approach to assess vitamin A status of populations. However, more research is needed to understand the initial decline in the measured concentration of retinol in DBS samples and to determine conditions under which samples should be stored. Many cultures or high-risk subgroups, such as children, who would not participate in venous blood samplings, will become accessible by way of DBS.

The authors wish to appropriately caution those attempting to implement this DBS methodology. The DBS punch contains the equivalent of ~10–12 µL of plasma and less than half is injected for HPLC analysis. It should only be attempted incorporating all the components detailed above using an optimized HPLC system with a highly sensitive detector set at the wavelength maximum of retinol.

	FOOTNOTES

 
¹ Presented in part at Experimental Biology ‘98, April 1998, San Francisco, CA [Dreyfus, M. L., Craft, N. E., Yamini, S., Humphrey, J. H. & West, Jr., K. P. (1998) Vitamin A analysis in dried blood spots by HPLC. FASEB J. 12: A840 (abs. 4668)].

² Supported by grants from Office of Health and Nutrition, USAID, Washington, D. C. and Task Force Sight and Life, Basel, Switzerland.

⁴ Abbreviations used: CDC, Center for Disease Control; DBS, dried blood spot; DTPA, diethylenetriaminepentaacetic acid; HPCE, high-performance capillary electrophoresis; MRDR, modified relative dose response; NIST, National Institute of Standards and Technology; RBP, retinol binding protein; RDR, relative dose response; VAD, vitamin A deficiency.

Manuscript received August 6, 1999. Initial review completed September 15, 1999. Revision accepted December 7, 1999.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Craft N. E. High resolution HPLC method for the simultaneous analysis of carotenoids, retinoids and tocopherols. FASEB J. 1996;10:527abs#3039

2. Futterman S., Swanson D., Kalina R. E. A new rapid fluorometric determination of retinol in serum. Invest. Opthalmol. 1975;14:125-130

3. Hannon W. H., Aziz K. J., Collier F. C., Fisher D. A., Fafara C. E., Knight W. S., Mitchell M. L., Sideman L., Therell , Jr B. L., Wolfson M. Blood collection on filter paper for neonatal screening programs Second edition. 1992:NCCLS Document LA4–A2 vol. 12 no. 13

4. Loerch J. D., Underwood B. A., Lewis K. C. Response of plasma levels of vitamin A to a dose of vitamin A as an indicator of hepatic vitamin A reserves in rats. J. Nutr. 1979;109:778-786

5. Ma Y., Wu Z., Furr H. C., Lammi-Keefe C., Craft N. E. Fast minimicroassay of serum retinol (vitamin A) by capillary zone electrophoresis with laser-excited fluorescence detection. J. Chromatogr. B 1993;616:31-37

6. Marinovic A. C., May W. A., Sowell A. L., Khan L. K., Huff D. L., Bowman B. A. Effect of hemolysis on serum retinol as assessed by direct fluorometry. Am. J. Clin. Nutr. 1997;66:1160-1164[Abstract/Free Full Text]

7. National Institute of Standards and Technology. (1995) Certificate of Analysis: Standard Reference Material 968b, Fat-Soluble Vitamins and Cholesterol in Human Serum. Gaithersburg, MD: NIST Standard Reference Materials Program.

8. Oliver R. W. A., Kafwembe E. M., Mwandu D. Stability of vitamin A circulating complex in spots of dried serum samples absorbed onto filter paper. Clin. Chem. 1993;39:1744-1745[Medline]

9. Russell R. M., Iber F. L., Bustin M., Goldberg N., Miller P., Krasinski S. D. Failure of the relative dose response test (RDR) to predict vitamin A (VA) deficiency in patients with protein calorie malnutrition. Am. J. Clin. Nutr. 1982;35:857

10. Semba R. D., Miooti P. G., Chiphangwi J. D., Saah A. J., Canner J. K., Dallabetta G. A., Hoover D. R. Maternal vitamin A deficiency and mother-to-child transmission of HIV-1. Lancet 1994;343:1593-1597[Medline]

11. Shi H., Ma Y., Humphrey J. H., Craft N. E. Determination of vitamin A in dried human blood spots by high-performance capillary electrophoresis with laser-excited fluorescence detection. J. Chromatogr. B 1995;665:89-96[Medline]

12. Sommer A., Katz J., Tarwotjo I. Increased risk of respiratory disease and diarrhea in children with pre-existing mild vitamin A deficiency. Am. J. Clin. Nutr. 1984;40:1090-1095[Abstract/Free Full Text]

13. Tanumihardjo S. A., Cheng J.-C., Permaesih D., Muherdiyantiningsih , Rustan E., Muhilal , Karyadi D., Olson J. A. Refinement of the modified-relative-dose-response test as a method for assessing vitamin A status in a field setting: Experience with Indonesian children. Am. J. Clin. Nutr. 1996;64:966-971[Abstract/Free Full Text]

14. Tanumihardjo S. A., Furr H. C., Erdman J. W., Jr, Olson J. A. Use of the modified relative dose response (MRDR) assay in rats and its application to humans. Eur. J. Clin. Nutr. 1990;44:219-224[Medline]

15. Underwood B., Arthur P. The Contribution of Vitamin A to Public Health. FASEB J 1996;10:1040-1048[Abstract]

This article has been cited by other articles:

				N. Potischman Biologic and Methodologic Issues for Nutritional Biomarkers J. Nutr., March 1, 2003; 133(3): 875S - 880. [Abstract] [Full Text] [PDF]

				J. G. Erhardt, N. E. Craft, F. Heinrich, and H. K. Biesalski Rapid and Simple Measurement of Retinol in Human Dried Whole Blood Spots J. Nutr., February 1, 2002; 132(2): 318 - 321. [Abstract] [Full Text] [PDF]

				G. L. Coté Noninvasive and Minimally-Invasive Optical Monitoring Technologies J. Nutr., May 1, 2001; 131(5): 1596S - 1604. [Abstract] [Full Text]

				N. E. Craft Innovative Approaches to Vitamin A Assessment J. Nutr., May 1, 2001; 131(5): 1626S - 1630. [Abstract] [Full Text]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Craft, N. E. Articles by West, K. P. Search for Related Content PubMed PubMed Citation Articles by Craft, N. E. Articles by West, K. P., Jr.