Pharmacokinetic assessment of teratologically effective concentrations of an endogenous retinoic acid metabolite

Multilayer omics analysis reveals a non-classical retinoic acid signaling axis that regulates hematopoietic stem cell identity

Hematopoietic stem cells (HSCs) rely on complex regulatory networks to preserve stemness. Due to the scarcity of HSCs, technical challenges have limited our insights into the interplay between metabolites, transcription, and the epigenome. In this study, we generated low-input metabolomics, transcriptomics, chromatin accessibility, and chromatin immunoprecipitation data, revealing distinct metabolic hubs that are enriched in HSCs and their downstream multipotent progenitors. Mechanistically, we uncover a non-classical retinoic acid (RA) signaling axis that regulates HSC function. We show that HSCs rely on Cyp26b1, an enzyme conventionally considered to limit RA effects in the cell. In contrast to the traditional view, we demonstrate that Cyp26b1 is indispensable for production of the active metabolite 4-oxo-RA. Further, RA receptor beta (Rarb) is required for complete transmission of 4-oxo-RA-mediated signaling to maintain stem cells. Our findings emphasize that a single metabolite controls stem cell fate by instructing epigenetic and transcriptional attributes.

Comparison of the function and expression of CYP26A1 and CYP26B1, the two retinoic acid hydroxylases

All-trans-retinoic acid (atRA) is an important signaling molecule in all chordates. The cytochrome P450 enzymes CYP26 are believed to partially regulate cellular concentrations of atRA via oxidative metabolism and hence affect retinoid homeostasis and signaling. CYP26A1 and CYP26B1 are atRA hydroxylases that catalyze formation of similar metabolites in cell systems. However, they have only 40% sequence similarity suggesting differences between the two enzymes. The aim of this study was to determine whether CYP26A1 and CYP26B1 have similar catalytic activity, form different metabolites from atRA and are expressed in different tissues in adults. The mRNA expression of CYP26A1 and CYP26B1 correlated between human tissues except for human cerebellum in which CYP26B1 was the predominant CYP26 and liver in which CYP26A1 dominated. Quantification of CYP26A1 and CYP26B1 protein in human tissues was in agreement with the mRNA expression and showed correlation between the two isoforms. Qualitatively, recombinant CYP26A1 and CYP26B1 formed the same primary and sequential metabolites from atRA. Quantitatively, CYP26B1 had a lower K(m) (19nM) and V(max) (0.8 pmol/min/pmol) than CYP26A1 (K(m)=50 nM and V(max)=10 pmol/min/pmol) for formation of 4-OH-RA. The major atRA metabolites 4-OH-RA, 18-OH-RA and 4-oxo-RA were all substrates of CYP26A1 and CYP26B1, and CYP26A1 had a 2-10-fold higher catalytic activity towards all substrates tested. This study shows that CYP26A1 and CYP26B1 are qualitatively similar RA hydroxylases with overlapping expression profiles. CYP26A1 has higher catalytic activity than CYP26B1 and seems to be responsible for metabolism of atRA in tissues that function as a barrier for atRA exposure.

Function of retinoid nuclear receptors: lessons from genetic and pharmacological dissections of the retinoic acid signaling pathway during mouse embryogenesis

Retinoic acid (RA) is involved in vertebrate morphogenesis, growth, cellular differentiation, and tissue homeostasis. The use of in vitro systems initially led to the identification of nuclear receptor RXR/RAR heterodimers as possible transducers of the RA signal. To unveil the physiological functions of RARs and RXRs, genetic and pharmacological studies have been performed in the mouse. Together, their results demonstrate that (a) RXR/RAR heterodimers in which RXR is either transcriptionally active or silent are involved in the transduction of the RA signal during prenatal development, (b) specific RXRalpha/RAR heterodimers are required at many distinct stages during early embryogenesis and organogenesis, (c) the physiological role of RA and its receptors cannot be extrapolated from teratogenesis studies using retinoids in excess. Additional cell type-restricted and temporally controlled somatic mutagenesis is required to determine the functions of RARs and RXRs during postnatal life.

Differential teratogenesis of all-trans-retinoic acid administered on gestational day 9.5 to SWV and C57BL/6N mice: Emphasis on limb dysmorphology

BACKGROUND

Mouse strain differences in teratologic response are well documented. However, because retinoids cause similar malformation syndromes across many species, the strain differences may be predicted to be minimal. The goals of this study were to characterize and explain the differences between the C57BL/6N and SWV mouse strains in terms of all-trans-retinoic acid (RA)-induced teratologic effects at the time of gestation that cause postaxial forelimb ectrodactyly.

METHODS

Visceral and skeletal malformations were determined by Wilson's sectioning and double-staining techniques, respectively; developmental staging was performed according to the somite count; and retinoid concentrations were assessed by HPLC.

RESULTS

C57BL/6N mice were more susceptible than SWV mice to induction of embryolethality, cardiovascular defects, and forelimb ectrodactyly, whereas the opposite was true for the induction of ear, thymus, and tail agenesis, and cleft palate, gastroschisis, and anal atresia. As determined by somite counts, 1 strain intercross was developmentally advanced compared to the parental strains and the reciprocal cross. Retinoid susceptibility was equivalent between the reciprocal crosses for some malformations and determined by the maternal genotype for others. Toxicokinetic experiments showed that whole-embryo peak retinoid concentrations did not differ between the strains, but the area under the curve (AUC) for all-trans-RA was 1.3 times higher in C57BL/6N than in SWV embryos.

CONCLUSIONS

The malformation spectrum induced by RA was strain-specific, and the strain sensitivity for forelimb ectrodactyly was consistent with all previously tested teratogenic agents (i.e., C57BL/6N was more sensitive than SWV). The strain differences in teratologic effects were not explained by developmental timing differences or toxicokinetic differences at the whole-embryo level.

Teratogenicity of retinoic acid and its effects on TGF-β2 expression in the developing cerebral cortex of the rat

Vitamin A metabolites are potent teratogens in a wide variety of species, including man. Transforming growth factor betas (TGF-betas) are involved in several mammalian prenatal developmental processes. The aim of this study was to determine the effects of exogenous and excessive all-trans retinoic acid on TGFbeta2 expression in the developing cerebral cortex of the rat. Many of the malformations including exencephaly, exophtalmus, abdominal wall defects, extremity reduction defects observed in this study were dependent on the time of administration of retinoic acid. TGF-beta2 was diversely expressed, as revealed immunohistochemically, in the cerebral cortex and plexus choroideus. The diversity depended on the gestational day and the was affected by the administration of retinoic acid. In the 15-day-old fetus from mothers who had been fed by gavage a single dose of 60 mg/kg body weight of all-trans retinoic acid on the 8th day of gestation, TGF-beta2 immunoreactivity in the brain was decreased. However, by the 18th day of gestation, TGF-beta2 expression increased. The expression of TGF-beta2 in fetuses whose mothers had been given all-trans retinoic acid after the neurulation period (on day 12 of gestation) was generally similar to that in a control group. We conclude that all-trans retinoic acid leads to severe congenital malformations if administered before neurulation whereas if given after neurulation, it is not so teratogenic. Further, retinoic acid has a variable effect on the expression of TGF-beta2.

The role of vitamin A in mammalian reproduction and embryonic development

Since the late 1980s, there has been an explosion of information on the molecular mechanisms and functions of vitamin A. This review focuses on the essential role of vitamin A in female reproduction and embryonic development and the metabolism of vitamin A (retinol) that results in these functions. Evidence strongly supports that in situ-generated all-trans retinoic acid (atRA) is the functional form of vitamin A in female reproduction and embryonic development. This is supported by the ability to reverse most reproductive and developmental blocks found in vitamin A deficiency with atRA, the block in embryonic development that occurs in retinaldehyde dehydrogenase type 2 null mutant mice, and the essential roles of the retinoic acid receptors, at least in embryogenesis. Early studies of embryos from marginally vitamin A-deficient (VAD) pregnant rats revealed a collection of defects called the vitamin A-deficiency syndrome. The manipulation of all-trans retinoic acid (atRA) levels in the diet of VAD female rats undergoing a reproduction cycle has proved to be an important new tool in deciphering the points of atRA function in early embryos and has provided a means to generate large numbers of embryos at later stages of development with the vitamin A-deficiency syndrome. The essentiality of the retinoid receptors in mediating the activity of atRA is exemplified by the many compound null mutant embryos that now recapitulate both the original vitamin A-deficiency syndrome and exhibit a host of new defects, many of which can also be observed in the VAD-atRA-supported rat embryo model and in retinaldehyde dehydrogenase type 2 (RALDH2) mutant mice. A major task for the future is to elucidate the atRA-dependent pathways that are normally operational in vitamin A-sufficient animals and that are perturbed in deficiency, thus leading to the characteristic VAD phenotypes described above.

Co-oxidation of all-trans retinol acetate by human term placental lipoxygenase and soybean lipoxygenase

The oxidation of all-trans retinol acetate (t-RAc) mediated by soybean lipoxygenase (SLO) and affinity purified human term placental lipoxygenase (HTPLO) was investigated. Under optimum assay conditions, SLO, which is used as a model enzyme, exhibited a specific activity of 850 nmol of t-RAc depleted/min/nmol of enzyme. For HTPLO, incubation of 100 microM t-RAc, 2 mM linoleic acid, and 50 micrograms of enzyme protein in Tris buffer at pH 9.0 was essential to observe the optimum rate of t-RAc cooxidation (370 nmol of t-RAc depleted/min/mg protein). The linoleate dependent cooxidation of t-RAc by both SLO and HTPLO was significantly inhibited by micromolar concentrations of nordihydroguaiaretic acid, butylated hydroxytoluene, butylated hydroxyanisole, gossypol, and 5,8,11-eicosatriynoic acid. These results suggest that t-RAc is metabolized via the lipoxygenase pathway. Significant rate of t-RAc oxidation by HTPLO observed in this study suggests that it may constitute an important part of the puzzle worthy of consideration in the understanding of retinoid teratogenicity in humans.

Prenatal ethanol consumption alters the expression of cellular retinol binding protein and retinoic acid receptor mRNA in fetal rat embryo and brain

The mechanism by which prenatal ethanol ingestion causes fetal alcohol syndrome (FAS) is unknown. We hypothesize that ethanol disrupts the normal function of retinoids in embryogenesis and differentiation, resulting in FAS. The present work was designed to determine if prenatal ethanol ingestion affects the expression of cellular retinol binding protein (CRBP) and nuclear retinoic acid receptors (RARs). Paired timed pregnant rats were fed a liquid diet, one group treated with 36% of carbohydrate calories replaced with ethanol. Maternal serum retinol concentrations during pregnancy peaked on the 6th day of pregnancy, but no difference was noted between the ethanol and control group. At the 12th and 20th day of gestation, embryos or fetal brain were removed, and RNA was isolated for Northern hybridization. The abundance of CRBP mRNA was significantly elevated by ethanol consumption. In both the 12-day embryo (relative density of control: 1.00 +/- 0.10; vs. ethanol: 1.87 +/- 0.30, p < 0.05) and 20-day fetal brain (relative density of control: 1.00 +/- 0.09; vs. ethanol: 1.46 +/- 0.09, p < 0.01). In the embryo, ethanol ingestion resulted in a decrease in the level of RAR-beta mRNA (control: 1.00 +/- 0.05; vs. ethanol: 0.71 +/- 0.07, p < 0.01), but had no effect on RAR-alpha or RAR-gamma mRNA. In contrast to the embryo, the expression of both the 3.7- and 2.7-kb RAR-alpha transcripts was significantly greater in day 20 fetal brain of ethanol-treated rats (3.7-kb RAR-alpha control: 1.00 +/- 0.11; vs. ethanol: 1.65 +/- 0.06; p < 0.001; 2.7-kb RAR-alpha control: 1.00 +/- 0.14; vs. ethanol: 1.74 +/- 0.27, p < 0.05), whereas RAR-beta and RAR-gamma expression were not altered. These observations suggest that altered vitamin A function is a potential factor in the embryopathy of prenatal ethanol exposure.

Chromatographic and electrophoretic analysis of biomedically important retinoids

The determination of retinol (vitamin A) and its metabolites, as well as synthetic retinoids, in biological samples is a challenging task due to the sensitivity of these compounds to light, heat and oxygen, high protein binding, separation of geometric isomers and determination of low endogenous levels. Numerous procedures for sample preparation have been published for biological fluids and tissues, consisting of solvent extraction, solid-phase extraction (off-line) and HPLC with column switching (on-line solid-phase extraction). The last-mentioned technique has several advantages, including a high degree of automation, no evaporation of extraction solvents, protection from light and higher sensitivity. Due to the favourable UV characteristics of most retinoids, HPLC with UV detection is most often employed, and photodiode array detection is becoming more and more popular. Fluorescence and electrochemical detection have found only a limited field of application, but the use of LC-MS resulted in a few highly sensitive methods. Reconsideration of GC through the use of better deactivated columns and cold on-column injection and evaluation of new promising separation methods, such as supercritical fluid chromatography and capillary electrophoresis, have shown preliminary encouraging results, but appear to reach the required sensitivity only by coupling to MS. Therefore, HPLC with UV detection is still the method of choice for highly sensitive and selective retinoid determination, as well as for high sample throughput and robustness.

Teratogenesis by retinoic acid analogs positively correlates with elevation of retinoic acid receptor-beta 2 mRNA levels in treated embryos

Retinoic acid (RA) plays an important role during normal embryogenesis, however high doses of RA are teratogenic. Retinoic acid receptor-beta 2 (RAR-beta 2) mRNA and protein levels were previously demonstrated to undergo rapid elevation in susceptible tissues after treatment with teratogenic doses of RA. In this report we compared the effects of a number of retinoids, which represent a wide variety of chemical structures and which differ in their teratogenic potencies, on RAR-beta 2 mRNA levels in mouse embryos 6 hr after treatment. Retinoid treatments which result in a high incidence of limb defects elevated RAR-beta 2 mRNA levels similarly (10-14 fold in the limb buds, 4-8 fold in the head, and 2-4 fold in the remainder of the body). On the other hand, retinoid treatments which cause a low or no incidence of limb defects resulted in minor changes in RAR-beta 2 mRNA levels in each embryonic region. Therefore, a strong positive correlation was found between the elevation of RAR-beta 2 mRNA levels and the retinoids which produce limb defects. This provides further evidence that an elevation of RAR-beta 2 mRNA levels, and subsequently protein levels, is an important event involved in mediating the effects of RA during dysmorphogenesis.

Microinjections of cultured rat conceptuses: studies with 4-oxo-all-trans-retinoic acid, 4-oxo-13-cis-retinoic acid and all-trans-retinoyl-beta-glucuronide

4-Oxo-all-trans-retinoic acid, 4-oxo-13-cis-retinoic acid and all-trans-retinoyl-beta-glucuronide were intraamniotically microinjected in rat embryos on day 10 of gestation and cultured until day 11.5. A comparison of the concentration-effect relationships showed that the dysmorphogenic effects produced by these metabolites were qualitatively similar to those of parent all-trans-retinoic acid. Compared with all-trans-retinoic acid (300 ng/ml), the dysmorphogenic effects were elicited by a 2-fold higher concentration of 4-oxo-all-trans-retinoic acid, an approximately 10-fold higher concentration of 4-oxo-13-cis-retinoic acid and a 16-fold higher concentration of all-trans-retinoyl-beta-glucuronide. A surplus of uridine 5'-diphospho-glucuronic acid, microinjected together with 300 ng/ml all-trans-retinoic acid, decreased the observed embryo-toxicity of all-trans-retinoic acid, suggesting the possibility of glucuronidation in tissues of the conceptus per se. The results of the study provide further support for the hypothesis that 4-oxo-all-trans-retinoic acid and all-trans-retinoic acid are, in contrast to the corresponding cis-isomers and glucuronides, ultimate dysmorphogenic retinoids.

Inhibition of isotretinoin teratogenicity by acetylsalicylic acid pretreatment in mice

Although isotretinoin (ITR) has been suggested to cause malformations via cytopathic effects on embryonic cells, the molecular mechanisms of ITR cytotoxicity in teratogenesis are not clear. Since ITR undergoes metabolism by prostaglandin synthase to a potentially cytotoxic peroxyl free radical, the possible role of prostaglandin synthase metabolism as a modulator of ITR teratogenicity was evaluated. Craniofacial and limb abnormalities were noted in fetuses on day 18.5 of gestation following administration of ITR to pregnant CD-1 mice in a three dose regimen of 100 mg/kg at 4 hr intervals on day 10.5 of gestation (plug day = day 0.5 of gestation). Mice were also treated with acetylsalicylic acid (ASA), an irreversible inhibitor of the cyclooxygenase component of prostaglandin synthase, at doses of 20 and 60 mg/kg body weight 2 hr prior to each ITR dose. ASA pretreatment of mice receiving ITR treatment showed a dose-dependent decrease in the overall incidence of malformations, number of defects per fetus, and the incidence of specific craniofacial and limb defects. Equivalent doses of ASA given to control mice did not cause malformations or alter the incidence of resorptions. These results demonstrate that ASA is able to ameliorate the teratogenic effects of ITR observed in fetal mice near term and indicate that prostaglandin metabolism could play a mechanistic role in ITR teratogenicity.

Temporal distribution of retinoic acid and cellular retinoic acid-binding protein (CRABP) in the fetal hamster

The temporal relationship between the distribution of retinoic acid, a known human and rodent teratogen, and that of cellular retinoic acid-binding protein (CRABP) was investigated from Day 11 to Day 14 of hamster prenatal development. The 11,12-(3)H2 and 15(-14C) forms of all-trans-retinoic acid were used for quantitative distribution studies and autoradiography, respectively, and were evaluated 15 min after a single intravenous injection. Radioactivity was detected in all fetal tissues examined (brain, liver, heart, spinal cord, limb, and skin), and at Day 14, approximately 66% of the total radioactivity was present as parent all-trans-retinoic acid. High concentrations of total radioactivity were observed by autoradiography in the midbrain and hindbrain (mesencephalon, metencephalon, and myelencephalon) and spinal cord, but not in the forebrain. At the earliest time studied, limb buds showed relatively high concentrations of radioactivity. Levels of radioactivity were also high in portions of the developing face, nose, and tongue. Immunohistochemical analyses indicated that the amount of CRABP in Day 14 tissues was the highest in spinal cord followed by limb and skin; heart and liver contained only relatively small amounts of this protein. From Day 11 to Day 14, the amount of CRABP, as measured by high-performance size-exclusion liquid chromatography, in the whole body decreased as gestation progressed. Microscopic immunohistochemical localization of CRABP found the highest concentration in the ventral midbrain and in the ventral and lateral sides of the hindbrain and spinal cord; CRABP was also abundant in tongue, limb, and skin. The distribution of CRABP-positive cells in the central nervous system was similar to the distribution of retinoic acid. The data presented here indicate that fetal CRABP appears to play a role in differential accumulation of retinoic acid in certain structures of the developing hamster. The patterns of tissue retinoid and CRABP distribution observed here are consistent with the patterns of congenital malformations induced by prenatal retinoid exposure.

Elevations in the endogenous levels of the putative morphogen retinoic acid in embryonic mouse limb-buds associated with limb dysmorphogenesis

Retinoic acid, an endogenous metabolite of vitamin A (retinol), possesses striking biological activity akin to a morphogen in developing and regenerating vertebrate limbs. Systemic administration of retinoic acid (RA) to pregnant mammals during the period of limb organogenesis invariably results in dose-dependent dysmorphogenesis. In an attempt to uncover the mode of action of RA in the developing limb bud we analyzed, by HPLC methods, the levels of RA and its metabolic precursor, retinol, in embryonic mouse tissues prior to and following maternal exposure to a teratogenic dose of RA. Detectable levels of both RA and its isomer 13-cis-retinoic acid were found in the limb buds of Day 11 mouse embryos (40 +/- 2 somites). Although retinol was the major retinoid found in ethanolic extracts of either whole embryo or the limb buds, the latter is enriched in RA compared to the whole embryo. This indicated either a higher degree of retinol metabolism or a sequestration of RA in the limb bud compared to the rest of the embryo at this stage of development. A study of the time course of retinoid levels in treated embryos showed that changes occur rapidly, are stable for several hours, and then begin to return to pretreatment levels. After a maternal dose of 10 mg/kg RA, which resulted in a mild degree of limb anomalies, peak RA levels in the limb bud increased 50-fold over the endogenous level; a full 300-fold increase was found after a 100 mg/kg dose which results in 100% incidence of phocomelia. Interestingly, a dose-dependent depression in retinol levels was observed after RA treatment both in maternal plasma as well as the embryo. Studies are in progress to trace the intracellular disposition of both retinol and RA as well as any further active metabolite of RA in the limb buds and other embryonic tissues.