Assay of Retinoids in Biological Samples by Reverse-Phase High-Pressure Liquid Chromatography

[CANCER RESEARCH 38. 3327-3332, October 1978]

Anita B. Roberts,1 Margaret D. Nichols, Charles A. Frolik, Dianne L. Newton, and Michael B. Sporn

Lung Cancer Branch, National Cancer Institute, Bethesda, Maryland 20014

ABSTRACT the routine bioassay of the liver storage of retinoids (3), as well as the biochemical investigation of the metabolism of A separation procedure for retinoids based on reverse- various retinoids (7, 10,13, 15). phase high-pressure liquid Chromatography with solvent In the past the only simple and rapid procedures available mixtures of acetonitrile and water is described. The for retinoid determination have been based on either the method may be applied to the screening of synthetic Carr-Price antimony trichloride colorimetrie assay (3) or the retinoids, which have potential for use in the prevention trifluoroacetic acid modification of Dugan ef al. (6), both of of cancer. It is easily adapted to a variety of biological which confound retinol and retinyl esters. A separate esti samples and can be applied to other conventional retinoid mation of retinol and its esters has required lengthy Chro assays in liver and plasma, detecting as little as 1 nmol matographie procedures (10-13, 15, 19). The method pro retinyl esters and less than 0.3 nmol retinol per g tissue. posed here is based on the reverse-phase HPLC2 system The one-step Chromatography results in separation and recently described by Frolik ef al. (9), which can rapidly and simultaneous determination of many of the synthetic reti reproducibly separate and simultaneously quantitate a va noids and all of the natural retinoids, including the retinyl riety of retinoids with little or no degradation. The use of a esters that are separated into their major fatty acid com 2-step elution system results in a clean separation of retinol ponents. The method has been applied to the analysis of from many closely related synthetic derivatives and from retinoid levels in the liver and intestine of vitamin A- other physiological forms of the vitamin such as retinoic deficient hamsters following a p.o. dose (0.5 mg/day for 2 acid, retinal, and retinyl esters. In addition, the retinyl days) of retinyl acetate or of a synthetic vitamin A analog esters are separated into their major fatty acid components, and is predictive of the degree to which various synthetic which by previous methods required 2 extra Chromato retinoids can be converted to retinol and stored in the graphie steps (12). liver as retinyl esters. Because of its speed, excellent With this method it has been shown that 2 synthetic recoveries, and high resolution, the method offers signifi retinoids, N-acetyl retinyl amine (14) and retinal acetylhy- cant advantages over previous, more lengthy procedures. drazone (18) (Chart 1), each undergo conversion to retinol and retinyl esters in the vitamin A-deficient hamster. The INTRODUCTION splitting of these derivatives to a physiological form of the Increasing attention is being drawn to the chemopreven- vitamin could account for the reported vitamin A activity of tive role of retinoids in the development of cancers of the the compounds in the reversal of keratinization in trachÃ©al epithelial tissues. As recently reviewed (22) this family of organ cultures (23). compounds with "vitamin A activity" has been shown to be effective at pharmacological levels in preventing or delaying MATERIALS AND METHODS the development of cancers of the bladder, lung, skin, and Chemicals. Retinol was obtained from Eastman Organic breast of experimental animals, possibly by an enhance Chemicals, Rochester, N. Y.; retinyl palmitate was from ment of its physiological role in the maintenance of normal Sigma Chemical Corporation, St. Louis, Mo.; and axeroph- epithelial cell differentiation. The natural retinoids, how thene was from BASF/AG, Ludwigshafen am Rhein, Ger ever, have several undesirable properties that severely limit many. All other retinoids were a generous gift of Hoffmann- their clinical application. In addition to their intrinsic toxicity (5), their conversion to long-chain retinyl esters can La Roche Inc., Nutley, N. J. Solvents used for HPLC were obtained from Burdick & Jackson Laboratories, Muskegon, result in excessive liver storage and damage to that tissue (21). For circumvention of these properties, synthetic reti Mich., and other reagents used were analytical grade. The water used for HPLC was redistilled over glass. noids are being developed. It was therefore desirable to develop a method that could quickly and accurately deter Dosing of Animals. Weanling male and female Syrian golden hamsters raised on a vitamin A-deficient regimen (4, mine the amounts of retinoids in tissue samples and the degree to which synthetic retinoids might be converted to 9) were used at 30 to 33 days of age. At this time the livers retinyl esters and stored in the liver. Furthermore, the contained no detectable retinol and very small amounts of limited amounts of the trial compounds required that the retinyl esters (average concentration of 1 nmol/g). assay be more sensitive than were previous methods. Other Hamsters were dosed daily with either retinoids or vehicle applications of the assay include the estimation of plasma for 1 or 2 days by stomach intubation of 0.2 ml of a solution retinol and retinyl esters in vitamin A toxicity (17, 21) and of ethanohtrioctanoin (1:3) containing 0.025% butylated hydroxytoluene. 1To whom requests tor reprints should be addressed. Received May 1. 1978; accepted July 18, 1978. ! The abbreviation used is: HPLC, high-pressure liquid Chromatography.

OCTOBER 1978 3327

approximately 10 min following the change of the eluant composition to 98% acetonitrile. The identification of the various peaks was established by chromatography of standard compounds as well as by cochromatography of standard retinoid solutions with the tissue extracts. The integrated peak areas were converted to nmol of retinoid by comparing the integrator number for the peak area at 325 nm of the retinoid of interest with an average retinal acetylhydrazone standard integrator number/nmol of that retinoid, obtained from the chromatography and'peak integration of several Chart 1. Structure of synthetic retinoids. reference samples in the range of 0.1 to 2 /xg. Extraction of Tissues. Animals were killed by CO2 nar cosis 24 hr after receiving the last p.o. dose. Duplicate 1-g RESULTS samples of the liver and a single 1-g sample of the intestine The separation of retinol, retinal acetylhydrazone, N-ace- (the first 15 cm washed free of its contents) were weighed tyl retinyl amine, retinyl acetate, axerophthene (the vitamin and stored for up to 3 weeks in a nitrogen freezer. Storage A hydrocarbon), and retinyl palmitate by reverse-phase for longer periods resulted in a gradual decrease in the HPLC is shown in Chart 2. Depending on which retinoids amount of retinol detected in tissue samples, although the are present in a particular tissue sample, the percentage of esters appeared to be stable for at least 5 weeks. acetonitrile in the initial eluting solvent can be modified to Preparation of the tissue for extraction was by 1 of 2 give separation of the more polar retinoids in the minimal previously described procedures, the method of Ames ef al. time. For example, retinol and retinyl acetate are eluted at (2) or that of Ito ef a/. (13). Ether containing 0.1% butylated 3.3 and 4.6 min, respectively, when the initial solvent is 84% hydroxytoluene was used for extraction. An aliquot of the acetonitrile in water. extract was evaporated to dryness under a stream of nitro With anhydrous ether as the solvent, the sodium sulfate gen, and the residue was redissolved in 100% chloroform, powder procedure of Ames ef al. (2) and the lyophilization followed by the addition of 0.5 volume of methanol to give method of Itoef al. (13) were shown to be equivalent for the a final composition of 2:1. This solution was centrifuged to determination of tissue retinoids, although the lyophiliza clarity and analyzed directly by HPLC. All procedures were tion procedure has distinct advantages when water-misci- carried out in minimal light or under red lights, and sample ble solvents are used for extraction. The relative efficiency solutions were stored in the dark at -20Â°prior to chroma- of ether, ethyl acetate, methanol, or chloroform:methanol tography. For the trifluoroacetic acid determinations, aliquots of the same HPLC solution as above were analyzed by a modification (8) of the procedure of Dugan ef al. (6). For determination of the efficiency of various solvents in .448 - extracting the retinoids, the lyophilization procedure of Ito ef al. (13) was used. Retinyl acetate was added to each extraction as an internal standard. All calculations on the retinol and retinyl esters extracted from these livers were adjusted relative to a constant amount of retinyl acetate to eliminate variance due to evaporative losses. HPLC and Retinoid Determination. HPLC analysis was carried out on a Spectra Physics Model 3500B unit fitted with a Valco sample injection valve. Dual wavelength detec tion was accomplished with a Spectra Physics 8200 UV detector (maximal sensitivity, 0.0025 A unit, full scale) set at 365 nm and a Schoeffel Model 770 variable wavelength detector (maximal sensitivity, 0.01 A unit, full scale) set at 325 nm. The 770 detector was coupled to a Spectra Physics Autolab System I computing integrator to determine peak area. A Spectra Physics 5-fj.m Spherisorb ODS column (5% loading, 3 x 250 mm) was used for all separations. 02468 10 12 14 16 18 20 22 24 26 28 The sample was eluted isocratically at ambient tempera TIME (min) ture with a solvent flow of 1.2 ml/min and a solvent Chart 2. HPLC of retinoid standards. A 50-Â¿ilsample containing 0.65 Â¿ig composition in the range of 70 to 90% acetonitrile in water retinol, 0.69 /jg W-acetyl retinyl amine, 0.77 ^g retinyl acetate, 0.88 (Â¿g axerophthene, and 1.58 ,ig retinyl palmitate in methanol was injected onto until all compounds more polar than the retinyl esters had the reverse-phase column and eluted at a flow of 1.2 ml/min with 79% eluted, followed by a step change to 98% acetonitrile in acetonitrile in water, followed at Point A by 98% acetonitrile in water. water until all esters were eluted from the column (total Chromatography under identical conditions of a mixture of retinol, retinal acetylhydrazone (0.99 ^g), and retinyl acetate was used to establish the time, approximately 20 to 30 min). The esters were very elution time of the acetylhydrazone derivative ( ), which under these strongly retained on the column and were not eluted for conditions is not separated from N-acetyl retinyl amine.

3328 CANCER RESEARCH VOL. 38

(1:1) in extracting retinol and retinyl esters from livers was tively, when dried onto intestinal samples and ground with compared. Differences among solvents were not signifi sodium sulfate. These findings are not inconsistent with the cant. The ease of evaporation of the ether made it the data of Ames ef al. (2), as their reported 99% recovery for solvent of choice for this analysis. However, the effective retinyl acetate was quantitated by the Carr-Price reaction, extraction of certain retinoids may require the use of a which cannot distinguish between retinol and retinyl esters. different solvent. For example, retinoic acid, which is more It follows, however, that the lyophilization procedure, which polar than is retinol, was extracted by ether only 50 to 60% minimizes these effects, is the method of choice for this as well as by methanol. analysis. With the sodium sulfate powder method with ether as the The choice of a solvent for redissolving the sample extracting solvent, the recovery of 261 nmol retinol and 378 following evaporation of an aliquot of the extracting solvent nmol retinyl palmitate was determined. A 100-^1 aliquot of was particularly critical, as shown in Table 1. Both 100% a chloroform solution of each retinoid was either applied methanol and 100% acetonitrile, while adequately dissolv directly onto the liver or added to the ether in the first ing the retinol in the sample, were poor solvent choices for extraction step. Drying of the chloroform solutions onto a the solution of the very nonpolar retinyl esters. Increasing 1-g sample of liver from a control hamster prior to grinding amounts of chloroform in mixture with methanol resulted in the liver with sodium sulfate resulted in 98.5 and 94.5% maximal solubilization of both retinol and retinyl esters. recovery of the retinol and retinyl palmitate, respectively, The limited solubility of chloroform mixtures in acetoni- compared to the recovery of the same amount of standard trile:water solutions, however, necessitated a compromise added to the ether at the time of the first extraction step. between the percentage of acetonitrile in the initial eluant Absolute recoveries in the latter procedure were 109 and (and therefore the degree of resolution desired) and the 111%, respectively. percentage of chloroform in the sample solution. With 90% Unexpectedly, chloroform solutions (100 Â¿ilofa 1-mg/ml acetonitrile as eluant, 10-/xl samples could be injected in solution) of retinyl esters underwent a tissue-dependent 100% chloroform, but if 70% acetonitrile was used (as is hydrolysis to retinol in both the sodium sulfate powder necessary to resolve closely related retinoids) the high method and the lyophilization procedure. Aliquots of a degree of immiscibility of the 100% chloroform sample retinyl acetate solution dried onto the liver prior to grinding solution and the eluant caused a loss of resolution of the with sodium sulfate were hydrolyzed 55 to 92%, whereas an early peaks. The ester peaks, however, were unaffected. aliquot added to the medium prior to homogenization in the Selection of the percentage of acetonitrile in the initial lyophilization method resulted in a lesser extent of hydroly eluant and of the sample solvent composition must there sis (17 to 24%). Further investigation of the hydrolysis of fore depend on the particular array of retinoids present and retinyl esters by the sodium sulfate powder method showed the degree of resolution necessary to quantitate them that aliquots of a retinyl acetate solution that had been accurately. added either to a liver sample that had been boiled in 0.9% A linear relationship was found between the total dose NaCI solution for 15 min or directly to sodium sulfate with administered to the hamster and the liver retinoid concen no liver tissue present were recovered unchanged. Retinyl tration in the dose range tested of 0.2 to 1 mg retinyl acetate palmitate solutions, on the other hand, were hydrolyzed to per day for 1 or 2 days (Chart 3). As can be seen, liver retinol to only a minor degree by either procedure (grinding, storage of retinoids was increased by a multiple dosing 3.1%; lyophilization, 0.5%), and the synthetic retinoid, reti schedule. For example, the liver storage of retinyl esters nal acetylhydrazone, was not hydrolyzed at all by the same was increased approximately 60% by the administration of procedures. Retinyl acetate and retinyl palmitate solutions 0.5 mg retinyl acetate for each of 2 successive days, as were hydrolyzed to retinol approximately 5 and 2%, respec compared to a single 1-mgdose. Whether this phenomenon

Table 1 Effect of solvent composition of the solution of retinol and retinyl esters for HPLC Miscibility6 in parts/100 in acetoni- Relative solubility" trile:water mixtures at: Sample solvent composition100% esters1.0 methanol 100%CH3CN 0.96 0.91 00>60 00 00X>60 CHCI.,:methanol 0.940.97 1.951.99 0033 (1:1) CHCI.,:methanol 9 13 (2:1) 100%CHCI.1Retinol1.0 0.98Retinyl 2.0560%X0013<570%X <580%00 990%X 23 " Results are the average of duplicate determinations on aliquots of an ether extract of a liver containing 60 nmol retinol and 815 nmol retinyl esters perg. An aliquot of the ether extract was taken to dryness under nitrogen and redissolved in 100 /Â¿Iofthe solvent indicated. Sample volumes for HPLC were 10 Â¿il,and the elution was as described in Chart 2. 6 This is an approximate determination done by adding 25-fil portions of the sample solvent to 500 n\ of the acetonitrile:water mixture and noting the point at which turbidity or a distinct 2-phase system was produced.

OCTOBER 1978 3329

lOOOr mitted the collection of samples in the regions identified as Ester 1. Ester 2 + 3, and Ester 4. Each of these samples had a UV absorption spectrum characteristic of retinyl esters (Xmax 325 nm). Furthermore, hydrolysis of each of these fractions yielded retinol as the only 325-nm-absorbing ma-

0.056

0.048

0.040

0.032

O 0.024 OÃ

0.016

0.008

0.4 0.8 1.2 1.6 2.0 2.4 CUMULATIVE DOSE RETINYL ACETATE (mgl O 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 Chart 3. Total retinoid (nmol) per g liver versus cumulative dose of retinyl TIME Iminl acetate. Vitamin A-deficient hamsters (3 in each group) were given daily p.o. doses of either 0.2, 0.5, or 1.0 mg retinyl acetate in 0.2 ml ethanohtrioctanoin (1:3) containing 0.025% butylated hydroxytoluene for either 1 (---- ) or 2 (- ) days prior to sacrifice 24 hr after administration of the last dose. Tissues were extracted with ether by the method of Ames ef a/. (2). The total nmol retinoid is the sum of the values for retinol and all retinyl esters as determined by HPLC (â€¢)orthe value derived from the Irifluoroacetic acid method (A) of Dugan et al. (6) determined on an aliquot of the same sample. results from a decreased ability of the deficient animal to 0.010 store the initial dose of retinyl acetate or from an increase in liver storage with time, as reported by Ames and Harris 0.008 (1), is unknown. The supplementation schedule chosen for synthetic retinoids was a p.o. dose of 0.5 mg/day for 2 days 0006 followed by sacrifice 24 hr after administration of the final dose. O 0004 Since the colorimetrie determination of vitamin A has formed the basis of many assays (1 , 2, 17), it was desirable to compare that method with the present HPLC method. 0002 The amount of sample used for the trifluoroacetic acid determination was on the average 20 times the amount used for the HPLC. As seen in Chart 3, the values were quite O 24 6 8 10 12 14 16 18 20 22 24 26 28 30 consistent with those for the HPLC, averaging about 91% of TIME (mini the HPLC value. Although the trifluoroacetic acid method is Chart 4. HPLC of ether extracts of liver (A) and small intestine (B) of a faster, the HPLC method provides significantly more infor vitamin A-deficient hamster fed 0.5 mg retinal acetylhydrazone per day for 2 mation about the nature of the retinoids involved. days. A 10-jil sample representing '/75th of the ether extract of 1 g of liver or 2/7sth'sof the extract of 1 g of small intestine was injected onto the reverse- The Chromatographie profile of the ether extract of the liver and small intestine of a vitamin A-deficient hamster fed phase column in chlorofomvmethanol (2:1). The sample was eluted with a solvent flow of 1.2 ml/min with 80% acetonitrile in water (1240 psi) for the 2 p.o. doses of 0.5 mg retinal acetylhydrazone per day is first 8 min, followed at Point A by 98% acetonitrile in water (880 psi) until all shown in Chart 4. The parent compound (A,,,ax380 nm) is esters were eluted from the column. The peak at 9 to 14 min (approximately, 0.005 absorbance unit) was caused by a refractive index change resulting easily distinguished from retinol (Xrmix 325 nm) by the from the change to a new solvent at Point A. It is most apparent at the lower relative peak heights of the 365- and the 325-nm detector. absorbance units, full scale, setting (B). The peak immediately preceding Ester 1 in 8 is an artifact present in all intestinal samples. Ester 3 is retinyl In the retinyl ester portion of the chromatogram, only Ester palmitate, and Esters 1, 2, and 4 can be tentatively identified as retinyl 3 was positively identified by cochromatography as retinyl linoleate, oleate, and stÃ©arate,respectively. Dual wavelength detection was palmitate. The other peaks in the retinyl palmitate portion performed with a Spectra Physics 8200 UV detector equipped with a 365 nm filter ( ) and with a Schoeffel Model 770 variable wavelength detector of the chromatogram were determined to be retinyl esters set at 325 nm ( ) to aid further in the identification of retinol (X,,,a,.325 by several criteria. Chromatography of larger samples per nm) and retinal acetyl hydrazone (xm;lx,380 nm).

3330 CANCER RESEARCH VOL. 38

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1978 American Association for Cancer Research. Bioassay of Retinoids by HPLC terial on HPLC. Comparison of the elution order of the ester sents an application of the reverse-phase HPLC system of peaks with known elution patterns from reverse-phase pa Frolik ef al. (9) to the bioassay of tissue retinoids. The per Chromatographie systems (8, 10, 12) permits tentative reverse-phase system was chosen because of the excellent identification of Esters 1, 2, and 4 as retinyl linoleate, recoveries attainable, the reproducibility, and the mild na oleate, and stÃ©arate,respectively. Whereas the parent reti- ture of the Chromatographie process, which minimizes the noid represents over 40% of the total in the intestine, it production of artifacts. The method as described in this represents only about 2% of the total retinoid in the liver, paper is capable of quantitating retinoids at less than 1 possibly implying that hydrolysis occurs prior to absorption. nmol/g tissue, which represents an approximately 20-fold The retinoid concentrations, in nmol/g liver or small increase in sensitivity over previous methods (1). intestine, of hamsters given either retinyl acetate, retinal, or The use of vitamin A-deficient animals is important when the synthetic retinoids, retinal acetylhydrazone or A/-acetyl working within the limits of the sensitivity of this assay, a retinyl aminÃ©,on a dose schedule of 0.5 mg/day for 2 days feature that was also included in the liver storage assay are tabulated in Table 2. The 2 synthetic retinoids are procedure of Ames and Harris (1). The vitamin A-deficient converted to retinol and retinyl esters, although consider hamsters have on the average less than 1 nmol retinyl esters ably less efficiently than is retinyl acetate or retinal. Neither per g liver and therefore make the detection of ester levels retinal nor retinyl acetate was detected in the tissues of greater than 2 nmol/g significant compared to controls. hamsters dosed with those retinoids. This presumably re This permits the conservative use of synthetic retinoids in sults from their rapid conversion to other retinoid species. the assay. In the case of retinyl acetate, hydrolysis during the work-up A major feature of this Chromatographie system is the procedure could account for this, in part. However, the resolution of the long-chain fatty acid esters of retinol. partial hydrolysis in liver extracted by the lyophilization Although absolute identification of the ester peaks other procedure (17 to 24%), coupled with the fact that no retinyl than retinyl palmitate would require cochromatography acetate was detected by either extraction procedure, makes with standard compounds, comparison of the elution pat it unlikely that this artifactual hydrolysis must be taken into tern with published R, values from a reverse-phase paper consideration in the interpretation of the data. Only for the chromatography system (16) allows for tentative identifica synthetic analogs, which seem to have a longer half-life in tion as stated in the legend for Chart 4. Supporting this vivo, is the parent compound still detected 1 day after the tentative identification is the close agreement of the relative last dose. Regardless of the retinoid dosed, the major proportions of the different ester components in the liver retinoid species in the liver are retinyl esters with varying and intestine with that reported by Goodman ef al. (10). smaller amounts of retinol. The slightly different retinyl Most notable is the decrease in the proportion of palmitate ester distribution in the intestine as compared to the liver is and the increase in the proportion of stÃ©aratein the intes tine as compared to the liver. The "critical pair" (20) of the consistent with that reported previously (10). oleate (18:1) and palmitate (16:0) esters could not be sepa DISCUSSION rated completely even under conditions resulting in large separations of this peak from the other ester peaks. HPLC is increasingly becoming the method of choice for A useful refinement of this assay would be the addition of various biochemical assays because of both the high reso an internal reference standard to the extracting solvent to lution attainable and the short elution times as compared to eliminate sample variance due to evaporation and other conventional Chromatographie systems. This work repre losses. The choice of a standard should be limited to

Table 2 Relative retinoid levels in liver and small intestine of vitamin A-deficient hamsters following p.o. administration of 0.5 mg retinoid per day for 2 days esters* esters)S (% of total

RetinoidgivenRetinyl esters44133751.160.78.77.3<0.5d15.55.36.23.45.313.4Â°212.69.39.17.17.67.2370.573.12.077.962.271.7411.512.332.711.724.97.9 acetateRetinalRetinalacetylhydrazoneA/-Acetyl

retinylamineTissueliverliverintestineliverintestineliverintestinenmol/gRetinol34.661.711.511.61.41.4NDtissue"ParentretinoidNDNDND1.47.21.63.0Retinyl

" Each value represents the average of separate determinations on tissues from 3 hamsters. The extraction was by the method of Ames (2) with ether as the extracting solvent. Values below the detection limits of the assay are indicated as ND. The livers and intestines from vehicle-dosed hamsters contained no detectable retinol and an average ester level of 1 nmol/g liver and <0.1 nmol/g intestinal mucosa. 6 Ester 3 is retinyl palmitate, and Esters 1, 2, and 4 can be tentatively identified as retinyl linoleate, oleate, and stÃ©arate,respectively. c Represents the sum of Ester 1 and a larger peak migrating between Ester 1 and Ester 2. d Only retinyl palmitate was detected, and the level was too low to assess accurately.

OCTOBER 1978 3331

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1978 American Association for Cancer Research. A. B. Roberts et al. compounds not found in biological samples and intermedi REFERENCES ate in polarity between retinol and the long-chain retinyl 1. Ames, S. R., and Harris, P. L. Slope-Ratio Liver-Storage Bioassay for esters, so that its recovery would parallel the esters, which Vitamin A. Anal. Chem., 28: 874-878, 1956. are the predominant retinoid species. Two compounds 2. Ames, S. R., Risley, H. A., and Harris, P. L. Simplified Procedure for fulfilling these requirements are retinyl acetate and axeroph- Extraction and Determination of Vitamin A in Liver. Anal. Chem., 26. 1378-1381, 1954. thene, the hydrocarbon analog of retinol. Retinyl acetate 3. Carr, F. H., and Price, E. A. Colour Reactions Attributed to Vitamin A. was used successfully in this capacity, and axerophthene, Biochem. J., 20: 497-501, 1926. 4. Clamon, G. H., Sporn, M. B., Smith, J. M., and Henderson, W. R. Effect which remains unaltered during work-up by either of the of a-Retinyl Acetate on Growth of Hamsters Fed Vitamin A-Deficient described procedures, offers the additional advantage of Diets. J. Nutr., 705. 215-219, 1975. 5. Dingle, J. T., and Lucy, J. A. Vitamin A, Carotenoids and Cell Function. being able to be applied directly to tissue samples. Each of Biol. Rev. Cambridge Phil. Soc., 40: 422-461, 1965. these compounds is easily separated from retinol and most 6. Dugan, R. E., Frigerio. N. A., and Siebert, J. M. Colorimetrie Determi synthetic end group retinoid derivatives (Chart 2). nation of Vitamin A and Its Derivatives with Trifluoracetic Acid. Anal. Chem..36. 114-117,1964. The application of this method to the detection of tissue 7. Fidge, N. H., Shiratori, T., Ganguly, J., and Goodman, D. S. Pathways of retinoids following dosing with synthetic retinoids such as Absorption of Retinal and Retinoic Acid in the Rat. J. Lipid Res., 9. 103- A/-acetyl retinyl amine and retinal acetylhydrazone sheds 109.1968. 8. Frolik, C. A., Tavela, T. E., Peck, G. L., and Sporn, M. B. High-pressure some light on their possible mechanism of action. Both of Liquid Chromatographie Determination of 13-C/s-Retinoic Acid and All- these retinoids have been shown to have vitamin A-like 7>ans-Retinoic Acid in Human Plasma. Anal. Biochem., 86. 743-750, activity in reversing the keratinization of the vitamin A- 1978. 9. Frolik, C. A., Tavela, T. E., and Sporn, M. B. Separation of the Natural deficient hamster trachÃ©alepithelium grown in organ cul Retinoids by High Pressure Liquid Chromatography. J. Lipid Res., 79. ture (23). The relative activity of these 2 retinoids in trachÃ©al 32-37, 1978. 10. Goodman. D. S.. Huang. H. S., and Shiratori. T. Tissue Distribution and organ culture is of the same order of magnitude as their Metabolism of Newly Absorbed Vitamin A in the Rat. J. Lipid Res., 6. relative liver storage in the form of retinyl esters following 390-396. 1965. P.O. dosing. It must therefore be considered that these 2 11. Holasova, M., and Blattna, J. Gel Chromatographie Separation of Re tinol. Retinyl Esters and Other Fat-Soluble Vitamins. J. Chromatography, synthetic retinoids might be converted to a physiological 723. 225-230, 1976. metabolite by cleavage of the end group derivative prior to 12. Huang, H. S.. and Goodman, D. S. Vitamin A and Carotenoids. J. Biol. Chem., 240. 2839-2844, 1965. exerting their biochemical action. The amount of liver 13. Ito. Y. L., Zile. M., Ahrens, H., and Deluca, H. F. Liquid-Gel Partition storage resulting from dosing with these 2 synthetic reti Chromatography of Vitamin A Compounds: Formation of Retinoic Acid noids relative to that from animals dosed with retinyl acetate from Retinyl Acetate in Vivo. J. Lipid Res., 75. 517-524, 1974. 14. Kienzle, R. Die Synthese von AII-TVans-Retinylamm und Anderen Polyen- was about 3 times greater in another experiment in which aminen. Helv. Chim. Acta. 56. 1662-1671, 1973. animals were given a daily dose of 1 mg retinoid for each of 15. Lawrence, C. W., Crain, R. D., Lotspeich, F. J., and Krause, R. F. Absorption. Transport, and Storage of Retinyl-15-"C Palmitate-9. 10-3H 5 days. This may indicate an inductive effect of the enzymes in the Rat. J. Lipid Res., 7. 226-229, 1966. involved in the cleavage of these 2 derivatives. 16. Mahadevan, S., and Ganguly, J. Further Studies on the Absorption of In summary, it is imperative that synthetic retinoids being Vitamin A. Biochem. J., 87: 53-58, 1961. developed for long-term clinical application in the preven 17. Mallia. A. K., Smith. J. E., and Goodman. D. S. Metabolism of Retinol- Binding Protein and Vitamin A during Hypervitaminosis A in the Rat. J. tion of epithelial cancers (24) be screened for their ability to Lipid Res., 76. 180-188. 1975. promote liver storage of retinyl esters, as excessive liver 18. Miki. T., and HarÃ¡.Y. Studies on Ethylenic Compounds. Ill Hydrazone- Type Derivatives of Vitamin A Aldehyde Chem. Pharm. Bull. Tokyo. 70: stores are a potential contributor to toxic effects. The 922-926. 1962. sensitive assay described here would be particularly useful 19. Olson, J. A. The Conversion of Radioactive ^-Carotene into Vitamin A for this and for other general pharmacological studies, by the Rat Intestine in Vivo. J. Biol. Chem.. 236: 349-356, 1961. 20. Roels, 0., and Mahadevan, S. Methods for Separation of the Various since it requires only a small amount of retinoid for testing Members of the Vitamin A Group of Compounds. In: P. GyÃ¶rgyand W. and provides information on all species of retinoids present N. Pearson (eds.). The Vitamins, Ed. 2, Vol. 6, pp. 170-177. New York: Academic Press, 1967. in the tissues. 21. Smith, F. R., and Goodman. D. S. Vitamin A Transport in Human Vitamin AToxicity. NewEngl. J. Med.,294: 805-808, 1976. 22. Sporn, M. B. Retinoids and Carcinogenesis. Nutr. Rev.,35: 65-69, 1977. 23. Sporn, M. B., Dunlop, N. M., Newton, D. L.. and Henderson. W. E. Relationships between Structure and Activity of Retinoids. Nature, 263 ACKNOWLEDGMENTS 110-113, 1976. 24. Sporn, M. B., Dunlop, N. M.. Newton, D. L.. and Smith, J. M. Prevention We thank W. R. Henderson for assistance with dosing the hamsters and of Chemical Carcinogenesis by Vitamin A and Its Synthetic Analogs Margaret B. Mose for preparation of the manuscript. (Retinoids). Fed. Proc.,35: 1332-1338, 1976.

3332 CANCER RESEARCH VOL. 38

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1978 American Association for Cancer Research. Assay of Retinoids in Biological Samples by Reverse-Phase High-Pressure Liquid Chromatography

Anita B. Roberts, Margaret D. Nichols, Charles A. Frolik, et al.

Cancer Res 1978;38:3327-3332.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/38/10/3327

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/38/10/3327. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.