- Email: [email protected]
Patterns of CYP26 expression in human prenatal cephalic and hepatic tissues indicate an important role during early brain development
Developmental Brain Research 120 Ž2000. 7–16 www.elsevier.comrlocaterbres
Research report
Patterns of CYP26 expression in human prenatal cephalic and hepatic tissues indicate an important role during early brain development Marina E. Trofimova-Griffin, Monica R. Brzezinski, Mont R. Juchau
)
Department of Pharmacology, UniÕersity of Washington, Box 357280, School of Medicine, Seattle, WA 98195, USA Accepted 23 November 1999
Abstract CYP26 Ž P450RAI. catalyzes catabolic retinoic acid ŽRA. hydroxylation and thereby appears to play a critical role in retinoid signaling pathways during development. In this study, a quantitative competitive reverse transcriptase–polymerase chain reaction ŽRT-PCR. assay was developed for evaluation of CYP26 message levels in human prenatal tissues. Statistical analyses of transcription levels in 12 prenatal human brains and six prenatal human livers demonstrated good sensitivity and reproducibility. Quantitative profiles of CYP26 gene expression in early Žgestational days 57–110. prenatal cephalic and hepatic tissues and comparisons with adult counterparts are reported for the first time. Prenatal cephalic tissues at days 57–67 exhibited values of 1950" 420 ŽCYP26 moleculesr10 6 GAPDH molecules. whereas prenatal cephalic tissues at days 105–110 exhibited values of 22 300 " 4450 ŽCYP26 moleculesr10 6 GAPDH molecules., indicating a sharp developmental increase Žapproximately 11-fold.. Levels in human adult cephalic tissues were slightly less than the prenatal cephalic levels measured during the earliest stages of gestation and were approximately 3-fold lower than those measured in adult human hepatic tissues. Levels in human prenatal hepatic tissues at days 63–110 gestation were less than 800 ŽCYP26 moleculesr10 6 GAPDH molecules. and did not exhibit developmental increases. Considered together, the data have strong implications for the importance of CYP26 in early development of the human brain. q 2000 Elsevier Science B.V. All rights reserved. Keywords: CYP26; P450RAI; Human prenatal; Brain development; Retinoid
1. Introduction Retinoic acids ŽRA. are a group of active metabolites of vitamin A which play highly important roles in regulating cell growth, differentiation and morphogenesis w3,4x. There is increasing evidence indicating their involvement in fundamental aspects of early development of the vertebrate central nervous system ŽCNS. such as: rostrocaudal specification of the forebrain, midbrain, hindbrain and spinal cord; migration of neural crest cells into specific areas of the embryo and stimulation of axonal outgrowth. Exposure of vertebrate embryos to excess retinoids produces both neurobehavioral and morphologic developmental defects in the CNS and depends on the stage in which this exposure occurs. Both excess and deficient retinoids in very early developmental stages of Xenopus and mouse embryos can inhibit the morphogenesis of forebrain and head structures.
) Corresponding author. [email protected]
Fax:
q1-206-685-3822;
e-m ail:
RA treatments after gastrulation have much less effect on the specification of the forebrain, midbrain and eyes, although brain differentiation may still be highly abnormal w5,28,29x. It has long been known that RA can stimulate neurite outgrowth in neuroblastoma cells and embryonal carcinoma cells w7,27x providing evidence for an important role for retinoids in neurogenesis. Overexpression of retinoid receptors in the Xenopus zygote also resulted in the development of ectopic primary neurons internally in the embryo w26x. Information about different aspects of retinoid signal transduction such as identification, localization and developmental expression of retinoid-related receptors, binding proteins and enzymes catalyzing retinoid metabolism is very important for understanding retinoid function during development. Recently, a breakthrough was achieved by the cloning from regenerating zebrafish fin Žby means of differential display. of a novel RA-inducible P450 cytochrome, termed P450RAI, with specificity for RA metabolism w36x. This new RA hydroxylase was defined as CYP26 and became the first representative of a novel P450 family. During the
0165-3806r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 3 8 0 6 Ž 9 9 . 0 0 1 8 5 - 6
8
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
past three years, human w35x, mouse w9,24x and Xenopus w14x homologs of CYP26 have been cloned. CYP26 is now characterized as an RA-inducible hydroxylase that specifically catalyzes hydroxylation of all-trans RA, but does not recognize the 13-cis and 9-cis isomers as substrates w18,31x. Major metabolic products of the enzymatic reaction were 4-oxo RA and 4-hydroxy RA w1,35,36x. In the adult mouse, CYP26 appears to be expressed in liver and brain; interestingly, it was induced severalfold by RA in the liver but not in the brain w24x. CYP26 message was detected in most human adult tissues but hybridization signals varied in intensity with highest levels of transcription in liver, heart, pituitary gland, adrenal gland, placenta and regions of the brain w33x. Chromosomal localization of the human CYP26 gene was determined, assigning it to chromosome 10q23– q24 w34x. Several disease loci have been described in the same region of chromosome 10, including split hand–split foot w12x and infantile onset spinocerebellar atrophy w32x. Increasing evidence implicates CYP26 importantly in regulatory functions for RA signaling during vertebrate development. Both its substrate Žall-trans RA. and a major product of the reaction Ž4-oxo RA. can bind with high affinity to and activate retinoid receptors, which also have been identified during different stages of vertebrate embryogenesis w6x. The expression of CYP26 has been investigated in zebrafish embryos and appears to be higher during early gastrulation Žwhen zebrafish are highly sensitive to exogenous retinoid exposure. than at later stages w36x. A separate group of scientists was able to detect CYP26 transcripts with reverse transcriptase–polymerase chain reaction ŽRT-PCR. as early as embryonic day 8.5 in mouse embryos w24x. A third group of investigators demonstrated that the CYP26 gene is not expressed uniformly prenatally. Rather, it is expressed in a stage- and region-specific manner during mouse development with major expression domains in the posterior neural plate and neural crest cells for cranial ganglia w9x. Comparisons of the temporal and spatial CYP26 expression patterns of mouse and Xenopus homologues during embryogenesis indicated that most CYP26 gene transcription domains are conserved between these two species, although there are also some clear differences w25x. Recently, it was proposed that CYP26 is involved in patterning of the dorso-ventral axis of the mouse retina by forming a narrow horizontal boundary between the dorsal and ventral retinaldehyde dehydrogenases, thereby creating two regions with very high ventral and moderately high dorsal RA levels. This organization of the embryonic vertebrate retina into dorsal and ventral territories divided by a horizontal boundary has parallels to the division of the Drosophila eye disc into dorsal, equatorial and ventral zones, indicating similar topographical patterns in eye morphogenesis. Interestingly, the high expression of CYP26 mRNA during mouse embryogenesis declines rapidly postnatally and appears to leave no significant imprint on the morphology and function of the adult retina w19x.
Data on the role of retinoids during early human development are much more limited. Thus far, it is known that either deficit or excess exposure of human embryos to retinoids often leads to reduced survival rates, behavioral alterations, spina bifida and other malformations of the CNS, most commonly in the hindbrain w30x. CYP26 message was detected in human fetal brain Ž20–25 weeks. w33x but there appear to be no published data on the presence and expression levels of this RA-metabolizing cytochrome P450 during earlier periods of human development. The purpose of this study, therefore was to ascertain the extent to which CYP26 would be expressed in human cephalic and hepatic tissues between the late embryonic period the early fetal period Ždays 57–110.. We also wished to evaluate levels and patterns of CYP26 gene transcription during this period of human brain development.
2. Materials and methods 2.1. Human tissues Tissues from normal human embryos and fetuses were snap-frozen in liquid nitrogen and were provided by the Birth Defects Research Laboratory of the University of Washington ŽDepartment of Pediatrics., Seattle, WA. Handling of these tissues was in accordance with the guidelines of the Human Subjects Review Committee at the same institution. Immediately after surgical procedures Ždilatation and curettage., tissues were snap-frozen and delivered within 3–4 h to the laboratory, where they were stored under liquid nitrogen until analyzed. Gestational ages ranged between 57 and 110 days as estimated from measurements of foot lengths. 2.2. Isolation of total RNA from human prenatal tissues Liquid nitrogen-preserved human prenatal tissues were weighed and immediately processed for total RNA isolation using an RNeasy Midi Kit ŽQiagen, Valencia, CA.. Total RNA samples were dissolved in DEPC-treated water, aliquotted and stored at y808C. The concentrations of RNA were estimated by spectrophotometry using UV absorbance at 260 nm. The integrity of total RNA was determined by comparisons of the 18S and 28S ribosomal RNA after denaturing agarose gel electrophoresis and ethidium bromide staining. Total RNA from all human prenatal tissues was isolated shortly prior to quantitative competitive RT-PCR to ensure uniformity. 2.3. Total RNA from human adult tissues Human adult whole brain total RNA ŽClontech, Palo Alto, CA. was analyzed from a pool of two normal tissue specimens from 43- and 47-year-old Caucasian males. Human adult liver total RNA ŽClontech. was analyzed
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
from a pool of two normal tissue specimens — from a 15-year-old Caucasian male and a 35-year-old Caucasian female. 2.4. Oligonucleotides for quantitatiÕe competitiÕe RT-PCR and generation of competitiÕe standards Primers were synthesized and purified by Gibco BRL ŽGrand Island, NY.. Optimal sequences were selected in conservative regions of CYP26 cDNA w33x and were chosen from published sequences for glyceraldehyde-3-phosphate dehydrogenase ŽGAPDH. w8x. A CYP26 forward primer ŽCYP26FP. and a CYP26 reverse primer ŽCYP26RP. were used to amplify both native and competitive standard templates, resulting in PCR product lengths of 369 and 300 bp, respectively. PCR products from competitive standard templates were sufficiently shorter than native templates Ž369r300s 1.23. to allow good separation with agarose gel electrophoresis. CYP26FP and CYP26RP sequences were 5X-TTTGGAGGACACGAAACCAC-3X and 5X-CAGCATGAATCGGTCAGGAT-3X , respectively. Forward ŽCYP26FPlinker1. and reverse ŽCYP26RPlinker2. oligonucleotides used for the generation of CYP26 competitive standard were 5X-TTTGGAGGACACGAAACCACCCTCCGGAACTATGG-3X and 5X-CAGCATGAATCGGTCAGGATTGTAAAGCTGGAGCCAGG-3X , respectively. These primers were comprised of sequences from CYP26 primers on their 5X ends and linker parts that would anneal to CYP2E1 sequences ŽGenEMBL Accession N D10014. on their 3X ends. As a result, a CYP26 heterologous competitive standard was produced ŽFig. 1A.. A GAPDH forward primer ŽGAPDHFP. and a GAPDH reverse primer ŽGAPDHRP. were used to amplify both native and competitive standard templates, resulting in PCR product lengths of 785 and 628 bp, respectively. PCR products from competitive standard templates were sufficiently shorter than native templates Ž785r628s 1.25. to allow good separation with agarose gel electrophoresis. GAPDHFP and GAPDHRP sequences were 5X-GGTCGGAGTCAACGGATTTG-3X and 5X-TCCGACGCCTGCTTCACCAC-3X , respectively. Forward ŽGAPDHFPlinker3.
9
and reverse ŽGAPDHRPlinker4. oligonucleotides used for the generation of GAPDH competitive standard were 5XGGTCGGAGTCAACGGATTTGCCTGGAAGCCACAGAAA-3X and 5X-TCCGACGCCTGCTTCACCACCATTCAGGAAGTGTTCTGGC-3X , respectively. These primers were composed of sequences from GAPDH primers on their 5X ends and linker parts that would anneal to CYP2E1 sequences on their 3X ends. As a result, a GAPDH heterologous competitive standard was produced ŽFig. 1B.. 2.5. CompetitiÕe standard generation Forward and reverse linker oligonucleotides were used for PCR amplifications with full length human CYP2E1 plasmid DNA as a template ŽFig. 1.. The PCR mixture for CYP26 competitive standard synthesis contained 20 ng of template DNA, 200 mM of each dNTP ŽBoehringer Mannheim, Indianapolis, IN . , 10 m M of each CYP26FPlinker1 and CYP26RPlinker2 primers, 2.5 units of Taq polymerase ŽBoehringer Mannheim. in a 50-ml reaction volume. The PCR mixture for GAPDH competitive standard synthesis was the same except for primers; 10 mM of each GAPDHFPlinker3 and GAPDHRPlinker4 were used. Amplifications were performed with a DNA Thermal cycler ŽPerkin-Elmer Cetus, Norfolk, CT.. The first PCR cycle was 948C for 3 min followed by annealing and elongation at 728C for 7 min. The next 24 cycles were performed as follows: 948C for 40 s Ždenaturation., 588C for 40 s Žannealing., and 728C for 1.5 min Želongation.. Post-PCR conditions included 728C for 7 min, then 48C until the collection of samples. PCR products of expected size Ž300 bp for CYP26 and 628 bp for GAPDH. were purified from 1.2% agarose gels using a GeneClean 2 Kit ŽBio 101, Vista, CA. and were cloned using a TA Cloning Kit ŽInvitrogen, San Diego, CA.. Plasmid DNAs were purified using a Plasmid Mini Purification Kit ŽQiagen.. cDNA inserts were sequenced at the Biochemistry DNA Sequencing Facility ŽUniversity of Washington. to ensure proper structure of both competitive standards. Concentrations of the purified plasmids were determined by spectrophotometry. The competitive standards were aliquotted and stored at y808C until used for quantitative RT-PCR analyses. 2.6. RT
Fig. 1. Schematic view of heterologous competitive standards generation. ŽA. CYP26 structure. ŽB. GAPDH structure.
Two micrograms of total RNA from human prenatal or adult tissues were mixed with 20 units RNAsin ŽBoehringer Mannheim., 50 mM Oligo dŽT.16 ŽPerkin-Elmer, Branchburg, NJ. and DEPC-treated water in a total volume of 9 ml then incubated at 908C for 5 min, 48C for 5 min, and room temperature for 20 min. 10 mM DTT, 500 mM of each dNTP, 1 = First Strand buffer ŽGibco BRL. and 200 units of Super Script Reverse Transcriptase ŽGibco
10
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
BRL. were added to each tube Žtotal volume 20 ml. and were incubated at 428C for 60 min. Reverse transcriptase was inhibited by heating at 958C for 5 min and samples were either used immediately for PCR amplifications or were stored at y208C until further analysis. 2.7. QuantitatiÕe competitiÕe PCR Each sample for CYP26 PCR amplification contained 1 ml of cDNA mixture after RT, 200 mM of each dNTP, 1 = PCR buffer ŽBoehringer Mannheim., 1.5 mM MgCl 2 , 10 mM concentrations of each primer ŽCYP26FP and CYP26RP. and 2.5 units of Taq polymerase in a 50-ml reaction volume. Each sample for GAPDH PCR amplification was the same except for primers — 10 mM concentrations of each primer for GAPDHFP and GAPDHRP were used. Master mixes were prepared for each set of PCR reactions. Negative controls were included for each PCR experiment, when instead of RNA, an equivalent volume of DEPC-treated water was added to the RT reaction mixture and was processed with the other samples, following the same RT reaction protocol. One microliter of this sample was used as a negative control for each set of PCR amplifications. Controls for genomic DNA contamination of total RNA samples also were performed. Two micrograms of total RNA were subjected to the RT reaction but without the addition of reverse transcriptase. One microliter of this mixture was amplified by PCR with the other samples and then analyzed. All PCR amplifications were performed with a DNA Thermal cycler ŽPerkin-Elmer Cetus.. The first PCR cycle was 948C for 3 min followed by annealing and elongation at 728C for 7 min. The next 24 cycles for GAPDH amplifications or 34 cycles for CYP26 amplifications were performed as follows: 948C for 40 s Ždenaturation., 588C for 40 s Žannealing., and 728C for 1.5 min Želongation.. Post-PCR conditions included 728C for 7 min then 48C until analysis of samples. 2.8. Analysis of PCR products CYP26 products Ž369 and 300 bp. and GAPDH products Ž785 and 628 bp. obtained after quantitative competitive PCR were electrophoretically separated on 2% or 1.2% high resolution agarose ŽSigma, St. Louis, MO. gels, respectively. Gels were stained with ethidium bromide, visualized in UV light and photographed on positiver negative Polaroid film, type 665 ŽPolaroid, Cambridge, MA.. The negatives were scanned using Scan Jet II cxrT ŽHewlett-Packard, USA. and DeskScan II version 2.0 software ŽHewlett-Packard.. NIH Image version 1.58 software ŽNational Institutes of Health, USA. was used to perform quantitative analyses of band volumes. Digital images of the agarose gels were used to quantify CYP26 and GAPDH levels of expression in densitometric units ŽDU..
3. Results Specific PCR primers were designed and tested in RT-PCR reactions to ensure that products generated by amplifications were single bands with predicted sizes of 369 or 785 bp for CYP26 or GAPDH, respectively. Contamination of the PCR products is potentially problematic for interpretation and quantitation of the results, therefore, a negative, template-minus control was included for each set of PCR assays. Absence of the bands confirmed that there was no product contamination after PCR amplifications. Controls for genomic DNA contamination of total RNA samples also were performed. After amplification, these samples yielded no bands with the expected sizes, indicating that there was no significant DNA contamination of the RNA samples tested in this study. Errors introduced by differences in the integrity of individual RNA samples and in the efficiency of the RT reaction were sharply reduced by using a ‘‘housekeeping’’ gene, against which the other target gene was compared w2,10,37x. For our studies of CYP26 expression, GAPDH was chosen from among other ‘‘housekeeping’’ genes because it is not inducible by agents known to affect P450 concentrations w13x. Heterologous competitive standards were synthesized using CYP2E1 plasmid DNA as a template and either a pair of CYP26 linker primers ŽFig. 1A. or GAPDH linker primers ŽFig. 1B. during 25 cycles of PCR. Standards were designed to be approximately 20% shorter in their sizes than native products. This allowed good resolution on the agarose gel. PCR products were cloned into a plasmid and sequenced to verify their structure Žparticularly in primerbinding regions. to ensure their identity with native sequences. The next step was to determine an optimal number of PCR cycles for both CYP26 and GAPDH amplifications. This number was evaluated based on two criteria: first that sufficient amount of PCR product was produced for facile ultraviolet visualization on agarose gel, and secondly, that amplifications were performed only in the beginning of the plateau effect. Optimal numbers of cycles for CYP26 and GAPDH amplifications were determined as 35 ŽFig. 2A. and 25 cycles ŽFig. 2B., respectively, and were used for all further quantitative competitive PCR experiments. In order to standardize conditions for comparisons of message levels for CYP26 and GAPDH at different gestational stages and in adult tissues, we determined an amount of the competitive standards that would have to be used in all quantitative PCR experiments. A constant amount of 0.1 mg of reverse-transcribed total RNA from 59-day human fetal brain was co-amplified with serial dilutions of competitive standards during 35 PCR cycles for CYP26 or during 25 cycles for GAPDH. Negative photographs of the gels were analyzed by densitometry in order to quantify each band ŽFig. 3.. Due to the smaller size of the PCR products from competitive standards as compared with
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
Fig. 2. Relationship between product amplification and number of PCR cycles. Total RNA from human prenatal brain Ž59 days gestation. was reverse-transcribed and used as a template for different numbers of PCR amplifications. PCR products were electrophoretically separated, photographed and band volumes were quantitated in DU. ŽA. CYP26 PCR product Ž369 bp., amplifications were performed with primers CYP26FP and CYP26RP. ŽB. GAPDH PCR product Ž785 bp., amplifications were performed with primers GAPDHFP and GAPDHRP as described in Section 2.
wild type cDNA in the agarose gels, the decrease in the incorporation of ethidium bromide must be taken into account w20x. The corrections were made for comparisons of relative band sizes — the competitive standard for CYP26 was multiplied by 369r300s 1.23 when compared to native CYP26 PCR product and the competitive standard for GAPDH was multiplied by 785r628s 1.25 when compared to native GAPDH PCR product. Intensity of the native band divided by the size-adjusted intensity of standard band was plotted as a function of the number of competitive standard molecules per 0.1 mg of total RNA in double logarithmic coordinates. A straight line was fitted to these points and the amount of standard Žwhere the log 10 ratio of the intensities was equal to zero. was interpolated. The amounts of competitive standards when intensities of
Fig. 3. Co-amplification of increasing quantities of competitive standards with a constant amount of reverse-transcribed total RNA from human prenatal brain Ž59 days.. PCR bands were electrophoretically separated and analyzed. The log 10 intensity of native bands measured in DU divided by size-adjusted intensity of standard bands measured in DU Žnativerstandard. were plotted as a function of log 10 amount of standard Žmolecules of standardr0.1 mg of total RNA.. The resulting regression was solved for x when y s 0 to estimate quantity of competitive standard molecules per 0.1 mg of total RNA. ŽA. Co-amplification of native CYP26 with competitive standard during 35 cycles of PCR, primers CYP26FP and CYP26RP were used. ŽB. Co-amplification of native GAPDH with competitive standard during 25 cycles of PCR, primers GAPDHFP and GAPDHRP were used.
11
native and standard bands are equal were estimated as 2.4 = 10 3 Žmolecules of CYP26 standardr0.1 mg of total RNA. and 1.8 = 10 6 Žmolecules of GAPDH standardr0.1 mg of total RNA. and were used for all further quantitative RT-PCR experiments. CYP26 expression levels were evaluated in 12 human prenatal cephalic tissues Ždays 57–110. and six human prenatal hepatic tissues Ždays 63–110. by quantitative competitive RT-PCR. Comparisons also were made with human adult brain and liver, each of which was a pooled sample from two different donors. All tested tissues were normal Ži.e., no detected pathology.. GAPDH levels of expression as ‘‘housekeeping’’ gene also were determined in all human tissues for standardization purposes. An example of the quantitative analyses for CYP26 and GAPDH message levels in a 63-day human prenatal brain is shown in Fig. 4. Constant amounts of the competitive standards determined earlier were co-amplified with a series of dilutions of reverse-transcribed total RNA during 35 PCR cycles for CYP26 mRNA ŽFig. 4A. and during 25 PCR cycles for GAPDH mRNA ŽFig. 4C.. PCR products were separated electrophoretically and quantified by den-
Fig. 4. Competitive RT-PCR analyses for quantitation of CYP26 and GAPDH expression levels in human fetal brain Ž63 days.. ŽA. Ethidium bromide-stained PCR products were separated on 2% agarose gel after constant amounts of the CYP26 competitive standard Ž2.4=10 3 moleculesrreaction. were co-amplified with four decreasing amounts of reverse-transcribed total RNA Žlane 1: 0.12 mg; lane 2: 0.07 mg; lane 3: 0.04 mg; lane 4: 0.02 mg. during 35 cycles of PCR. ŽB. Bands from the gel were analyzed and the intensities of CYP26 native bands measured in DU divided by the respective size-adjusted intensities of standard bands measured in DU Žnativerstandard. were plotted as function of the amount of total RNA Žmg.. ŽC. Ethidium bromide-stained PCR products were separated on 1.2% agarose gel after constant amounts of the GAPDH competitive standard Ž1.8=10 6 moleculesrreaction. were co-amplified with four decreasing amounts of reverse-transcribed total RNA Žlane 1: 0.14 mg; lane 2: 0.09 mg; lane 3: 0.05 mg; lane 4: 0.03 mg. during 25 cycles of PCR. ŽD. Bands from the gel were analyzed and the intensities of GAPDH native bands measured in DU divided by the respective size-adjusted intensities of standard bands measured in DU Žnativerstandard. were plotted as function of the amount of total RNA Žmg..
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
12
Table 1 CYP26 and GAPDH message levels in human prenatal and adult cephalic tissues Gestational stage Ždays.
No. CYP26 moleculesr 1 mg of total RNAa
No. GAPDH moleculesr 1 mg of total RNAa
No. CYP26 moleculesr 10 6 GAPDH molecules
57 59 63 67 79 80 91 98 100 101 105 110 Adult brain
140 = 10 3 " 28.3 = 10 3 Ž n s 2. 108 = 10 3 " 25.5 = 10 3 Ž n s 3. 42 = 10 3 " 2.8 = 10 3 Ž n s 2. 67.5 = 10 3 " 3.5 = 10 3 Ž n s 2. 109 = 10 3 " 9.2 = 10 3 Ž n s 3. 92.5 = 10 3 " 4.9 = 10 3 Ž n s 2. 460 = 10 3 " 28.3 = 10 3 Ž n s 2. 413 = 10 3 " 102 = 10 3 Ž n s 3. 845 = 10 3 " 63.6 = 10 3 Ž n s 2. 460 = 10 3 " 28.3 = 10 3 Ž n s 2. 715 = 10 3 " 120 = 10 3 Ž n s 2. 990 = 10 3 " 68.8 = 10 3 Ž n s 4. 55.3 = 10 3 " 5 = 10 3 Ž n s 3.
55.5 = 10 6 " 6.4 = 10 6 Ž n s 2. 55 = 10 6 " 7.1 = 10 6 Ž n s 2. 28 = 10 6 " 11.3 = 10 6 Ž n s 2. 38 = 10 6 " 2.8 = 10 6 Ž n s 2. 25 = 10 6 " 3.6 = 10 6 Ž n s 3. 22 = 10 6 " 1.4 = 10 6 Ž n s 2. 33 = 10 6 " 4.2 = 10 6 Ž n s 2. 24 = 10 6 " 8.5 = 10 6 Ž n s 2. 40.5 = 10 6 " 6.4 = 10 6 Ž n s 2. 40.5 = 10 6 " 6.4 = 10 6 Ž n s 2. 37.5 = 10 6 " 10.6 = 10 6 Ž n s 2. 39 = 10 6 " 8.5 = 10 6 Ž n s 2. 36 = 10 6 " 6 = 10 6 Ž n s 3.
2500 2000 1500 1800 4400 4200 14 000 17 000 20 900 11 400 19 100 25 400 1500
a
Values are given as averages" S.D. with n values in parentheses.
sitometry. Ratios of target bands to size-adjusted standard bands were plotted against quantities of reverse-transcribed total RNA used in PCR reactions. The point at which this ratio is equal to 1 is the point where the amount of target gene is equal to a known amount of the competitor Ž2.4 = 10 3 molecules and 1.8 = 10 6 molecules for CYP26 and GAPDH competitors, respectively.. The amount of total RNA necessary to reach the equivalence point was interpolated from the same graph and estimated as 0.055 mg of total RNA for CYP26 mRNA ŽFig. 4B. and 0.09 mg of total RNA for GAPDH mRNA ŽFig. 4D.. The numbers of CYP26 and GAPDH molecules per 1 mg of total RNA were determined by the following equations: 2.4 = 10 3 Žmolecules of CYP26 competitive standard.r0.055 Žmg of total RNA. s 44 = 10 3 ŽCYP26 moleculesr1 mg of total RNA. and 1.8 = 10 6 Žmolecules of GAPDH competitive standard.r0.09 Žmg of total RNA. s 20 = 10 6 ŽGAPDH moleculesr1 mg of total RNA., respectively. In order to obtain an accurate measure of the equivalence point it was necessary to make a tight dilution series of reverse-transcribed total RNA used in the PCR reaction because the closer the ratio of target gene to competitive standard to equivalence point, the more linear the curve, and, therefore, the more precise the determination.
In this study, we demonstrated that GAPDH may be used as a ‘‘housekeeping’’ gene not only in adult tissues but also during human development. We found that GAPDH expression levels in human prenatal cephalic tissues exhibited some relatively small variations Ž2.5-fold., which were probably due to differences among individuals andror some variability in the total RNA preparations and RT-PCR testing. The average value for human prenatal cephalic tissues Ž36.5 = 10 6 GAPDH moleculesr1 mg of total RNA. was about the same as the average value for human adult whole brain Ž36 = 10 6 GAPDH moleculesr1 mg of total RNA. ŽTable 1.. GAPDH expression levels in human prenatal hepatic tissues also exhibited some variations Ž; 4-fold. among the different tissues with an average value of 23 = 10 6 ŽGAPDH moleculesr1 mg of total RNA., which is similar to 14 = 10 6 ŽGAPDH moleculesr1 mg of total RNA. for human adult liver ŽTable 2.. We did not observe any gestational stage-dependent changes in GAPDH expression in either cephalic or hepatic tissues. This allowed us to use GAPDH as a ‘‘housekeeping’’ gene for comparisons of CYP26 expression levels between human prenatal and adult cephalic and hepatic tissues. Average values for CYP26 and GAPDH message levels in each individual tissue were determined in at least two indepen-
Table 2 CYP26 and GAPDH message levels in human prenatal and adult hepatic tissues Gestational stage Ždays.
No. CYP26 moleculesr 1 mg of total RNAa
No. GAPDH moleculesr 1 mg of total RNAa
No. CYP26 moleculesr 10 6 GAPDH molecules
63 67 98 101 105 110 Adult liver
- 10 4 Ž n s 2. - 10 4 Ž n s 2. - 10 4 Ž n s 2. - 10 4 Ž n s 2. - 10 4 Ž n s 2. - 10 4 Ž n s 2. 65 = 10 3 " 23.5 = 10 3 Ž n s 4.
29 = 10 6 " 4.2 = 10 6 Ž n s 2. 25.5 = 10 6 " 14.8 = 10 6 Ž n s 2. 12 = 10 6 " 1.4 = 10 6 Ž n s 2. 48 = 10 6 " 4.2 = 10 6 Ž n s 2. 14.5 = 10 6 " 7.8 = 10 6 Ž n s 2. 15.5 = 10 6 " 0.7 = 10 6 Ž n s 2. 14 = 10 6 " 5.3 = 10 6 Ž n s 3.
- 350 - 400 - 800 - 200 - 700 - 650 4600
a
Values are given as averages" S.D. with n values in parentheses.
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
Fig. 5. Profiles of CYP26 expression levels in human cephalic and hepatic tissues.
dent RT-PCR experiments and summarized data are presented in Tables 1 and 2. In order to reduce errors introduced by differences in the integrity of individual RNA samples and in the efficiency of the RT reaction, we also expressed CYP26 levels as the numbers of CYP26 molecules per 10 6 GAPDH molecules. Our results revealed that CYP26 message levels appear to sharply increase Žapproximately 11-fold. during this early period of human prenatal brain development. The least abundance of CYP26 mRNA Žaverages 1950 CYP26 moleculesr10 6 GAPDH molecules. was determined in human fetal brain Ž57–67 days. and was comparable with the message level determined in the human adult brain Ž1500 CYP26 moleculesr10 6 GAPDH molecules.. The highest level of CYP26 expression Žaverages 22 300 CYP26 moleculesr10 6 GAPDH molecules. was found in human fetal brain Ž105–110 days.. CYP26 expression in human prenatal hepatic tissues was too low to quantify Ž- 800 CYP26 moleculesr10 6 GAPDH molecules. with the competitive standard concentration used in this study but was estimated as 4600 ŽCYP26 moleculesr10 6 GAPDH molecules. in human adult liver. Interestingly, during the same period of development Ž63–110 days., CYP26 message levels were much higher in human prenatal cephalic tissues than in human prenatal hepatic tissues but human adult livers exhibited approximately three times more CYP26 mRNA than human adult whole brain. Patterns of CYP26 expression in human cephalic and hepatic tissues are presented in Fig. 5.
4. Discussion The goal of this research was to investigate the presence and dynamics of CYP26 transcription in human cephalic and hepatic tissues during the late embryonic and early fetal periods of development Ždays 57–110.. In preliminary experiments, we were unable to detect CYP26 message in either prenatal tissue with Northern blot analyses using total RNA. In contrast, relatively strong bands were detected in human prenatal brain using RT-PCR. In order to evaluate CYP26 mRNA levels in human fetal tissues, a
13
quantitative competitive RT-PCR assay was developed. Competitive PCR is a highly sensitive method for obtaining quantitative data on specific nucleic acid sequences and relies on the use of standards to compensate for the high number of factors affecting the yield of PCR products. The internal standard is amplified with the same primers Žand thus presumably with the same efficiency. as the endogenous target sequence, and is distinguished from the product derived from the native sequence by either size or restriction site. The standards fall into two categories: homologous and heterologous competitor fragments. One potential problem with homologous competitor fragments is the formation of heteroduplexes between the standard and target sequences during the amplification process, an artifact that could interfere with quantitation. In contrast, heterologous competitor fragments differ from the target except for the primer-binding regions, which are identical, thus, heteroduplex formation cannot occur. Heterologous competitive standards were designed and produced for both CYP26 and GAPDH. Theoretically, the amount of product doubles during each cycle of PCR, but in actuality, beyond a certain number of cycles the efficiency of amplification decreases with increasing cycle number, resulting in a plateau effect. The number of PCR cycles at which the plateau effect occurs varies greatly with the particular DNA sequence being amplified and, therefore, should be individually and empirically determined for each target sequence. Optimal numbers of cycles for CYP26 and GAPDH amplifications were determined as 35 and 25 cycles, respectively, and were used for all further quantitative competitive PCR experiments. It was particularly important to establish an exact relationship between product amplification and the number of PCR cycles because it is commonly accepted that the nature of competitive PCR makes it possible to obtain useful data after the reaction has reached the plateau phase. However, this might not always be true when product is measured well after the plateau phase, especially when the heterologous competitor has been used. We determined amounts of the competitive standards that would have to be used in all quantitative RT-PCR experiments for target and ‘‘housekeeping’’ genes. In order to estimate levels of transcription, competitive standards were co-amplified with a dilution series of reverse-transcribed total RNA Žfrom each particular tissue. during the same number of PCR cycles determined earlier. The levels of expression were estimated from the equivalence point where the amount of native gene was equal to a known amount of the competitor. In the present study, CYP26 expression in human fetal cephalic and hepatic tissues was investigated from days 57 to 110. For proper and consistent evaluation of the data, it is important to adequately identify gestational stage. In this study, we based determinations of the fetal age on one of the commonly used prenatal measurements, fetal foot length w23x. The distinction between the embryonic and the
14
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
fetal period at 8 post-ovulatory weeks in humans is based on multiple internal and external criteria and the fact that 90% of the more than 4500 named structures of the adult body have appeared by that time w22x. The beginning of the fetal period is commonly marked by complete closure of the palate. In humans, the fetal period extends from approximately 58–60 days of gestation until parturition at approximately 267 days of gestation. For comparison, in mice the fetal period extends from approximately day 15 of gestation until parturition at day 19 w16x. Our results revealed that CYP26 expression rapidly increases Žmore than 10-fold. in the human prenatal brain between gestational days 57 and 110. Interestingly, these data are consistent with prior morphological observations of considerable human fetal brain enlargement, especially in the cerebral hemisphere during the same period of development. The period near 90 post-ovulatory days was suggested as a critical time in human fetal development based on mathematical and hormonal criteria w23x. At the end of the embryonic period, the human brain is far more advanced morphologically than is generally appreciated; to such an extent that functional considerations are compelling. The arrangement of the rhombencephalic nuclei and tracts at this period is very similar to that observed in the newborn. Nevertheless, the fetal period is characterized by rapid growth and differentiation, including a series of complex subdivisions, flexures, and cellular differentiation that is not complete until well into the first few years of neonatal life. The most noticeable external changes in the human fetal brain are Ž1. the union of the cerebellar halves and the definition of the vermis; Ž2. the increasing concealment of the diencephalon and mesencephalon; Ž3. further approach of the frontal and temporal poles around the insula; Ž4. the appearance of sulci on the hemispheric surface Žat about the middle of prenatal life.; and Ž5. the increasing eclipse or disappearance of the cervical, pontine, and mesencephalic flexures. Internal changes also occur; one of most noticeable is the caudally directed growth of the corpus callosum w21,23x. Moreover, CYP26 mRNA levels in human prenatal cephalic tissue Žday 110. are approximately 17-fold higher than in adult whole brain, also suggesting a highly important role for this cytochrome P450 during early human brain development. Recent data on CYP26 mRNA distribution in human adult brain indicated that this enzyme was present in all regions of the brain tested, with higher expression levels in the occipital lobe, putamen, substantia nigra, temporal lobe and thalamus w33x. Furthermore, we demonstrated that CYP26 transcription levels in human fetal hepatic tissues were practically undetectable and much lower than in fetal cephalic tissues at all ages examined Ždays 63–110.. Remarkably, CYP26 message levels in human adult livers were approximately 3-fold higher than in human adult whole brain. These data are consistent with the fact that the adult liver is an organ in which the majority of the body’s retinoids are stored w3x. Because RA shows strong physiologic ac-
tions, it seems important to maintain tissue levels of retinoids within tightly regulated limits through coordinated biosynthesis and inactivation. It has been demonstrated that CYP26 is expressed differentially during early Xenopus development. During embryogenesis, CYP26 transcripts were detected at all stages analyzed; the RT-PCR analysis suggested a slight increase in transcript levels during gastrulationrneurulation. Whole-mount in situ hybridization revealed that, during gastrulation, CYP26 is not expressed in the area of the prospective neural plate that develops into hindbrain, whereas the prospective anterior neural and the underlying mesodermal prechordal plate are primary sites for CYP26 expression. In Xenopus embryos, ectopic CYP26 expression can rescue developmental defects in head formation induced by application of exogenous RA, suggesting that the enzymatic modifications introduced inhibit RA signaling activities in vivo. The authors proposed that CYP26mediated modification of RA modulates its signaling activity and helps to establish boundaries of differentially responsive and non-responsive territories w14x. In that study, which investigated the endogenous distribution of RA by high pressure liquid chromatography in chick embryos, sharp onroff boundaries rather than smooth gradients across a wide range of embryonic stages were emphasized. Some tissues, such as heart, had very little RA while others, such as the neural tube, had very high levels. The total variation between these two was 29-fold. The developing brain did not appear to generate RA, but the spinal part of the neural tube generated it at very high levels. It would seem that there must be a sharp onroff boundary in the region of the hindbrainrspinal cord junction; the mesenchyme surrounding the hindbrain generated RA whereas the hindbrain itself did not w17x. By contrast, results from a different group of investigators supported the idea that developing hindbrain in Xenopus embryo may be patterned by a retinoid gradient. The sequentially more posterior hindbrain patterning genes were induced effectively by sequentially higher RA concentrations w11x. Iulianella et al. w15x demonstrated that CYP26 expression was abundant in the retinoid-poor region of the mouse caudal embryo. It was rapidly induced by RA treatment in vivo suggesting that it plays a critical role in retinoid signaling. The authors concluded that CYP26 does not necessarily abolish normal retinoid signaling, rather, it may serve to limit the effects of RA to specific areas andror specific times in the mouse embryo. In summary, the quantitation of CYP26 transcription levels in human fetal cephalic and hepatic tissues are potentially highly important findings. Taken together, the results indicate a significant role for this RA-catabolizing cytochrome P450 during the late embryonic and early fetal period of human brain development. However, a tremendous amount of research remains to be accomplished in this area. It will be very interesting to determine the specific localization of CYP26 message and protein in
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
regions of human brain during various stages of embryogenesis and fetogenesis and to investigate its possible role as a modulator of congenital CNS malformations and neurobehavioral defects.
Acknowledgements The authors thank Birth Defects Laboratory and expert technical assistance of J. Pascoe-Mason and M. Eisenhauer. We also acknowledge S. Lee for technical assistance. This research was supported by NIH Grant ES-04041 and NIH, National Research Service Awards T32 E507032 from the National Institute of Environmental Health Sciences.
References w1x S.S. Abu-Abed, B.R. Beckett, H. Chiba, J.V. Chithalen, G. Jones, D. Metzger, P. Chambon, M. Petkovich, Mouse P450RAI ŽCYP26. expression and retinoic acid–inducible retinoic acid metabolism in F9 cells are regulated by retinoic acid receptor g and retinoid X receptor a , J. Biol. Chem. 273 Ž1998. 2409–2415. w2x M.J. Apostolakos, W.H.T. Schuermann, M.W. Frampton, M.J. Utell, J.C. Willey, Measurement of gene expression by multiplex competitive polymerase chain reaction, Anal. Biochem. 213 Ž1993. 277–284. w3x W.S. Blaner, J.A. Olson, Retinol and retinoic acid metabolism, in: M.S. Sporn, A.B. Roberts, D.S. Goodman ŽEds.., The Retinoids: Biology, Chemistry, and Medicine, Raven Press, New York, 1994, pp. 229–255. w4x L.M. De Luca, N. Darwiche, G. Celli, K. Kosa, C. Jones, S. Ross, L.C. Chen, Vitamin A in epithelial differentiation and skin carcinogenesis, Nutr. Rev. 52 Ž1994. 45–52. w5x A.J. Durston, J.P.M. Timmermans, W.J. Hage, H.F.J. Hendriks, N.J. de Vries, M. Heideveld, P.D. Nieuwkoop, Retinoic acid causes an antero-posterior transformation in the developing central nervous system, Nature 340 Ž1989. 140–144. w6x A.J. Durston, J. van der Wees, W.W.M. Pijnappel, S.F. Godsave, Retinoids and related signals in early development of the vertebrate central nervous system, Curr. Top. Dev. Biol. 40 Ž1998. 111–175. w7x M.K.S. Edwards, M.W. McBurney, The concentration of retinoic acid determines the differentiated cell types formed by a teratocarcinoma cell line, Dev. Biol. 98 Ž1983. 187–191. w8x M.J. Fasco, C. Treanor, L.S. Kaminsky, Cytochrome P450 mRNA induction: quantitation by RNA-polymerase chain reaction using capillary electrophoresis, Methods Enzymol. 272 Ž1996. 401–413. w9x H. Fujii, T. Sato, S. Kaneko, O. Gotoh, Y. Fujii-Kuriyama, K. Osawa, S. Kato, H. Hamada, Metabolic inactivation of retinoic acid by a novel P450 differentially expressed in developing mouse embryos, EMBO J. 16 Ž1997. 4163–4173. w10x W.C. Gause, J. Adamovicz, Use of PCR to quantitate relative differences in gene expression, in: C.W. Dieffenbach, G.S. Dveksler ŽEds.., PCR Primer: A Laboratory Manual, Cold Spring Harbor Laboratory Press, New York, 1995, pp. 293–311. w11x S.F. Godsave, C.H. Koster, A. Getahun, M. Mathu, M. Hooiveld, J. van der Wees, J. Hendriks, A.J. Durston, Graded retinoid responses in the developing hindbrain, Dev. Dyn. 213 Ž1998. 39–49. w12x F. Gurrieri, P. Prinos, D. Tackels, M.W. Kilpatrick, J. Allanson, M. Genuardi, A. Vuckov, L. Nanni, E. Sangiorgi, G. Garofalo, M.E. Nunes, G. Neri, C. Schwartz, P. Tsipouras, A split hand–split foot ŽSHFM3. gene is located at 10q24–25, Am. J. Med. Genet. 62 Ž1996. 427–436.
15
w13x B. Hogan, R. Beddington, F. Costantini, E. Lacy ŽEds.., Manipulating the Mouse Embryo, A Laboratory Manual, Cold Spring Harbor Laboratory Press, New York, 1994. w14x T. Hollemann, Y. Chen, H. Grunz, T. Pieler, Regionalized metabolic activity establishes boundaries of retinoic acid signaling, EMBO J. 17 Ž1998. 7361–7372. w15x A. Iulianella, B. Beckett, M. Petkovich, D. Lohnes, A molecular basis for retinoic acid-induced axial truncation, Dev. Biol. 205 Ž1999. 33–48. w16x M.R. Juchau, H. Boutelet-Bochan, Y. Huang, Cytochrome-P450-dependent biotransformation of xenobiotics in human and rodent embryonic tissues, Drug Metab. Rev. 30 Ž1998. 541–568. w17x M. Maden, E. Sonneveld, P.T. van der Saag, E. Gale, The distribution of endogenous retinoic acid in the chick embryo: implications for developmental mechanisms, Development 125 Ž1998. 4133– 4144. w18x Y. Marikar, Z. Wang, E.A. Duell, M. Petkovich, J.J. Voorhees, G.J. Fisher, Retinoic acid receptors regulate expression of retinoic acid 4-hydroxylase that specifically inactivates all-trans retinoic acid in human keratinocyte HaCaT cells, J. Invest. Dermatol. 111 Ž1998. 434–439. w19x P. McCaffery, E. Wagner, J. O’Neil, M. Petkovich, U.C. Drager, Dorsal and ventral retinal territories defined by retinoic acid synthesis, break-down and nuclear receptor expression, Mech. Dev. 82 Ž1999. 119–130. w20x S. Menzo, P. Bannarelli, M. Giacca, A. Manzin, P.E. Varaldo, M. Clementi, Absolute quantitation of viremia in human immunodeficiency virus infection by competitive reverse transcription and polymerase chain reaction, J. Clin. Microbiol. 30 Ž1992. 1752–1757. w21x K.L. Moore, The Developing Human, W.B. Saunders, Philadelphia, 1988. w22x R. O’Rahilly, F. Muller, Developmental Stages in Human Embryos, Carnegie Institution of Washington, 1987. w23x R. O’Rahilly, F. Muller, The Embryonic Human Brain: An Atlas of Developmental Stages, Wiley-Liss, New York, 1994. w24x W.J. Ray, G. Bain, M. Yao, D.I. Gottlieb, CYP26, a novel mammalian cytochrome P450, is induced by retinoic acid and defines a new family, J. Biol. Chem. 272 Ž1997. 18702–18708. w25x K. Roos, E. Sonneveld, B. Compaan, D. ten Berge, A.J. Durston, P.T. van der Saag, Expression of retinoic acid 4-hydroxylase ŽCYP26. during mouse and Xenopus laeÕis embryogenesis, Mech. Dev. 82 Ž1999. 205–211. w26x C.R. Sharpe, K. Goldstone, Retinoid receptors promote primary neurogenesis in Xenopus, Development 124 Ž1997. 515–523. w27x N. Sidell, Retinoic acid-induced growth inhibition and morphologic differentiation of human neuroblastoma cells in vitro, J. Natl. Cancer Inst. 68 Ž1982. 589–593. w28x A. Simeone, V. Avantaggiato, M.C. Moroni, F. Mavilio, C. Arra, F. Cotelli, V. Nigro, D. Acampora, Retinoic acid induces stage-specific antero-posterior transformation of rostral central nervous system, Mech. Dev. 51 Ž1995. 83–98. w29x H.L. Sive, B.W. Draper, R.M. Harland, H. Weintraub, Identification of a retinoic acid-sensitive period during primary axis formation in Xenopus laeÕis, Genes Dev. 4 Ž1990. 932–942. w30x W.J. Slikker, L.W. Chang, Handbook of Developmental Neurotoxicology, Academic Press, San Diego, 1998. w31x E. Sonneveld, C.E. van den Brink, B.-J.M. van der Leede, R.-K.A.M. Schulkes, M. Petkovich, B. van der Burg, P.T. van der Saag, Human retinoic acid ŽRA. 4-hydroxylase ŽCYP26. is highly specific for all-trans-RA and can be induced through RA receptors in human breast and colon carcinoma cells, Cell Growth Differ. 9 Ž1998. 629–637. w32x A. Suomalainen, J. Kaukonen, P. Amati, R. Timonen, M. Haltia, J. Weissenbach, M. Zeviani, H. Somer, L. Peltonen, An autosomal locus predisposing to deletions of mitochondrial DNA, Nat. Genet. 9 Ž1995. 146–151. w33x M.E. Trofimova-Griffin, M.R. Juchau, Expression of cytochrome
16
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
P450RAI ŽCYP26. in human fetal hepatic and cephalic tissues, Biochem. Biophys. Res. Commun. 252 Ž1998. 487–491. w34x J.A. White, B. Beckett, S.W. Scherer, J.-A. Herbrick, M. Petkovich, P450RAI ŽCYP26A1. maps to human chromosome 10q23–q24 and mouse chromosome 19C2–3, Genomics 48 Ž1998. 270–272. w35x J.A. White, B. Beckett-Jones, Y.-D. Guo, F.J. Dilworth, J. Bonasoro, G. Jones, M. Petkovich, cDNA cloning of human retinoic acid-metabolizing enzyme Žh P450RAI. identifies a novel family of cytochromes P450 ŽCYP26., J. Biol. Chem. 272 Ž1997. 18538– 18541. w36x J.A. White, Y.-D. Guo, K. Baetz, B. Beckett-Jones, J. Bonasoro,
K.E. Hsu, F.J. Dilworth, G. Jones, M. Petkovich, Identification of the retinoic acid-inducible all-trans-retinoic acid 4-hydroxylase, J. Biol. Chem. 271 Ž1996. 29922–29927. w37x J.C. Willey, E.L. Coy, M.W. Frampton, A. Torres, M.J. Apostolakos, G. Hoehn, W.H. Schuermann, W.G. Thilly, D.E. Olson, J.R. Hammersley, C.L. Crespi, M.J. Utell, Quantitative RT-PCR measurement of cytochromes P450 1A1, 1B1, and 2B7, microsomal epoxide hydrolase, and NADPH oxidoreductase expression in lung cells of smokers and nonsmokers, Am. J. Respir. Cell Mol. Biol. 17 Ž1997. 114–124.
Research report
Patterns of CYP26 expression in human prenatal cephalic and hepatic tissues indicate an important role during early brain development Marina E. Trofimova-Griffin, Monica R. Brzezinski, Mont R. Juchau
)
Department of Pharmacology, UniÕersity of Washington, Box 357280, School of Medicine, Seattle, WA 98195, USA Accepted 23 November 1999
Abstract CYP26 Ž P450RAI. catalyzes catabolic retinoic acid ŽRA. hydroxylation and thereby appears to play a critical role in retinoid signaling pathways during development. In this study, a quantitative competitive reverse transcriptase–polymerase chain reaction ŽRT-PCR. assay was developed for evaluation of CYP26 message levels in human prenatal tissues. Statistical analyses of transcription levels in 12 prenatal human brains and six prenatal human livers demonstrated good sensitivity and reproducibility. Quantitative profiles of CYP26 gene expression in early Žgestational days 57–110. prenatal cephalic and hepatic tissues and comparisons with adult counterparts are reported for the first time. Prenatal cephalic tissues at days 57–67 exhibited values of 1950" 420 ŽCYP26 moleculesr10 6 GAPDH molecules. whereas prenatal cephalic tissues at days 105–110 exhibited values of 22 300 " 4450 ŽCYP26 moleculesr10 6 GAPDH molecules., indicating a sharp developmental increase Žapproximately 11-fold.. Levels in human adult cephalic tissues were slightly less than the prenatal cephalic levels measured during the earliest stages of gestation and were approximately 3-fold lower than those measured in adult human hepatic tissues. Levels in human prenatal hepatic tissues at days 63–110 gestation were less than 800 ŽCYP26 moleculesr10 6 GAPDH molecules. and did not exhibit developmental increases. Considered together, the data have strong implications for the importance of CYP26 in early development of the human brain. q 2000 Elsevier Science B.V. All rights reserved. Keywords: CYP26; P450RAI; Human prenatal; Brain development; Retinoid
1. Introduction Retinoic acids ŽRA. are a group of active metabolites of vitamin A which play highly important roles in regulating cell growth, differentiation and morphogenesis w3,4x. There is increasing evidence indicating their involvement in fundamental aspects of early development of the vertebrate central nervous system ŽCNS. such as: rostrocaudal specification of the forebrain, midbrain, hindbrain and spinal cord; migration of neural crest cells into specific areas of the embryo and stimulation of axonal outgrowth. Exposure of vertebrate embryos to excess retinoids produces both neurobehavioral and morphologic developmental defects in the CNS and depends on the stage in which this exposure occurs. Both excess and deficient retinoids in very early developmental stages of Xenopus and mouse embryos can inhibit the morphogenesis of forebrain and head structures.
) Corresponding author. [email protected]
Fax:
q1-206-685-3822;
e-m ail:
RA treatments after gastrulation have much less effect on the specification of the forebrain, midbrain and eyes, although brain differentiation may still be highly abnormal w5,28,29x. It has long been known that RA can stimulate neurite outgrowth in neuroblastoma cells and embryonal carcinoma cells w7,27x providing evidence for an important role for retinoids in neurogenesis. Overexpression of retinoid receptors in the Xenopus zygote also resulted in the development of ectopic primary neurons internally in the embryo w26x. Information about different aspects of retinoid signal transduction such as identification, localization and developmental expression of retinoid-related receptors, binding proteins and enzymes catalyzing retinoid metabolism is very important for understanding retinoid function during development. Recently, a breakthrough was achieved by the cloning from regenerating zebrafish fin Žby means of differential display. of a novel RA-inducible P450 cytochrome, termed P450RAI, with specificity for RA metabolism w36x. This new RA hydroxylase was defined as CYP26 and became the first representative of a novel P450 family. During the
0165-3806r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 3 8 0 6 Ž 9 9 . 0 0 1 8 5 - 6
8
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
past three years, human w35x, mouse w9,24x and Xenopus w14x homologs of CYP26 have been cloned. CYP26 is now characterized as an RA-inducible hydroxylase that specifically catalyzes hydroxylation of all-trans RA, but does not recognize the 13-cis and 9-cis isomers as substrates w18,31x. Major metabolic products of the enzymatic reaction were 4-oxo RA and 4-hydroxy RA w1,35,36x. In the adult mouse, CYP26 appears to be expressed in liver and brain; interestingly, it was induced severalfold by RA in the liver but not in the brain w24x. CYP26 message was detected in most human adult tissues but hybridization signals varied in intensity with highest levels of transcription in liver, heart, pituitary gland, adrenal gland, placenta and regions of the brain w33x. Chromosomal localization of the human CYP26 gene was determined, assigning it to chromosome 10q23– q24 w34x. Several disease loci have been described in the same region of chromosome 10, including split hand–split foot w12x and infantile onset spinocerebellar atrophy w32x. Increasing evidence implicates CYP26 importantly in regulatory functions for RA signaling during vertebrate development. Both its substrate Žall-trans RA. and a major product of the reaction Ž4-oxo RA. can bind with high affinity to and activate retinoid receptors, which also have been identified during different stages of vertebrate embryogenesis w6x. The expression of CYP26 has been investigated in zebrafish embryos and appears to be higher during early gastrulation Žwhen zebrafish are highly sensitive to exogenous retinoid exposure. than at later stages w36x. A separate group of scientists was able to detect CYP26 transcripts with reverse transcriptase–polymerase chain reaction ŽRT-PCR. as early as embryonic day 8.5 in mouse embryos w24x. A third group of investigators demonstrated that the CYP26 gene is not expressed uniformly prenatally. Rather, it is expressed in a stage- and region-specific manner during mouse development with major expression domains in the posterior neural plate and neural crest cells for cranial ganglia w9x. Comparisons of the temporal and spatial CYP26 expression patterns of mouse and Xenopus homologues during embryogenesis indicated that most CYP26 gene transcription domains are conserved between these two species, although there are also some clear differences w25x. Recently, it was proposed that CYP26 is involved in patterning of the dorso-ventral axis of the mouse retina by forming a narrow horizontal boundary between the dorsal and ventral retinaldehyde dehydrogenases, thereby creating two regions with very high ventral and moderately high dorsal RA levels. This organization of the embryonic vertebrate retina into dorsal and ventral territories divided by a horizontal boundary has parallels to the division of the Drosophila eye disc into dorsal, equatorial and ventral zones, indicating similar topographical patterns in eye morphogenesis. Interestingly, the high expression of CYP26 mRNA during mouse embryogenesis declines rapidly postnatally and appears to leave no significant imprint on the morphology and function of the adult retina w19x.
Data on the role of retinoids during early human development are much more limited. Thus far, it is known that either deficit or excess exposure of human embryos to retinoids often leads to reduced survival rates, behavioral alterations, spina bifida and other malformations of the CNS, most commonly in the hindbrain w30x. CYP26 message was detected in human fetal brain Ž20–25 weeks. w33x but there appear to be no published data on the presence and expression levels of this RA-metabolizing cytochrome P450 during earlier periods of human development. The purpose of this study, therefore was to ascertain the extent to which CYP26 would be expressed in human cephalic and hepatic tissues between the late embryonic period the early fetal period Ždays 57–110.. We also wished to evaluate levels and patterns of CYP26 gene transcription during this period of human brain development.
2. Materials and methods 2.1. Human tissues Tissues from normal human embryos and fetuses were snap-frozen in liquid nitrogen and were provided by the Birth Defects Research Laboratory of the University of Washington ŽDepartment of Pediatrics., Seattle, WA. Handling of these tissues was in accordance with the guidelines of the Human Subjects Review Committee at the same institution. Immediately after surgical procedures Ždilatation and curettage., tissues were snap-frozen and delivered within 3–4 h to the laboratory, where they were stored under liquid nitrogen until analyzed. Gestational ages ranged between 57 and 110 days as estimated from measurements of foot lengths. 2.2. Isolation of total RNA from human prenatal tissues Liquid nitrogen-preserved human prenatal tissues were weighed and immediately processed for total RNA isolation using an RNeasy Midi Kit ŽQiagen, Valencia, CA.. Total RNA samples were dissolved in DEPC-treated water, aliquotted and stored at y808C. The concentrations of RNA were estimated by spectrophotometry using UV absorbance at 260 nm. The integrity of total RNA was determined by comparisons of the 18S and 28S ribosomal RNA after denaturing agarose gel electrophoresis and ethidium bromide staining. Total RNA from all human prenatal tissues was isolated shortly prior to quantitative competitive RT-PCR to ensure uniformity. 2.3. Total RNA from human adult tissues Human adult whole brain total RNA ŽClontech, Palo Alto, CA. was analyzed from a pool of two normal tissue specimens from 43- and 47-year-old Caucasian males. Human adult liver total RNA ŽClontech. was analyzed
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
from a pool of two normal tissue specimens — from a 15-year-old Caucasian male and a 35-year-old Caucasian female. 2.4. Oligonucleotides for quantitatiÕe competitiÕe RT-PCR and generation of competitiÕe standards Primers were synthesized and purified by Gibco BRL ŽGrand Island, NY.. Optimal sequences were selected in conservative regions of CYP26 cDNA w33x and were chosen from published sequences for glyceraldehyde-3-phosphate dehydrogenase ŽGAPDH. w8x. A CYP26 forward primer ŽCYP26FP. and a CYP26 reverse primer ŽCYP26RP. were used to amplify both native and competitive standard templates, resulting in PCR product lengths of 369 and 300 bp, respectively. PCR products from competitive standard templates were sufficiently shorter than native templates Ž369r300s 1.23. to allow good separation with agarose gel electrophoresis. CYP26FP and CYP26RP sequences were 5X-TTTGGAGGACACGAAACCAC-3X and 5X-CAGCATGAATCGGTCAGGAT-3X , respectively. Forward ŽCYP26FPlinker1. and reverse ŽCYP26RPlinker2. oligonucleotides used for the generation of CYP26 competitive standard were 5X-TTTGGAGGACACGAAACCACCCTCCGGAACTATGG-3X and 5X-CAGCATGAATCGGTCAGGATTGTAAAGCTGGAGCCAGG-3X , respectively. These primers were comprised of sequences from CYP26 primers on their 5X ends and linker parts that would anneal to CYP2E1 sequences ŽGenEMBL Accession N D10014. on their 3X ends. As a result, a CYP26 heterologous competitive standard was produced ŽFig. 1A.. A GAPDH forward primer ŽGAPDHFP. and a GAPDH reverse primer ŽGAPDHRP. were used to amplify both native and competitive standard templates, resulting in PCR product lengths of 785 and 628 bp, respectively. PCR products from competitive standard templates were sufficiently shorter than native templates Ž785r628s 1.25. to allow good separation with agarose gel electrophoresis. GAPDHFP and GAPDHRP sequences were 5X-GGTCGGAGTCAACGGATTTG-3X and 5X-TCCGACGCCTGCTTCACCAC-3X , respectively. Forward ŽGAPDHFPlinker3.
9
and reverse ŽGAPDHRPlinker4. oligonucleotides used for the generation of GAPDH competitive standard were 5XGGTCGGAGTCAACGGATTTGCCTGGAAGCCACAGAAA-3X and 5X-TCCGACGCCTGCTTCACCACCATTCAGGAAGTGTTCTGGC-3X , respectively. These primers were composed of sequences from GAPDH primers on their 5X ends and linker parts that would anneal to CYP2E1 sequences on their 3X ends. As a result, a GAPDH heterologous competitive standard was produced ŽFig. 1B.. 2.5. CompetitiÕe standard generation Forward and reverse linker oligonucleotides were used for PCR amplifications with full length human CYP2E1 plasmid DNA as a template ŽFig. 1.. The PCR mixture for CYP26 competitive standard synthesis contained 20 ng of template DNA, 200 mM of each dNTP ŽBoehringer Mannheim, Indianapolis, IN . , 10 m M of each CYP26FPlinker1 and CYP26RPlinker2 primers, 2.5 units of Taq polymerase ŽBoehringer Mannheim. in a 50-ml reaction volume. The PCR mixture for GAPDH competitive standard synthesis was the same except for primers; 10 mM of each GAPDHFPlinker3 and GAPDHRPlinker4 were used. Amplifications were performed with a DNA Thermal cycler ŽPerkin-Elmer Cetus, Norfolk, CT.. The first PCR cycle was 948C for 3 min followed by annealing and elongation at 728C for 7 min. The next 24 cycles were performed as follows: 948C for 40 s Ždenaturation., 588C for 40 s Žannealing., and 728C for 1.5 min Želongation.. Post-PCR conditions included 728C for 7 min, then 48C until the collection of samples. PCR products of expected size Ž300 bp for CYP26 and 628 bp for GAPDH. were purified from 1.2% agarose gels using a GeneClean 2 Kit ŽBio 101, Vista, CA. and were cloned using a TA Cloning Kit ŽInvitrogen, San Diego, CA.. Plasmid DNAs were purified using a Plasmid Mini Purification Kit ŽQiagen.. cDNA inserts were sequenced at the Biochemistry DNA Sequencing Facility ŽUniversity of Washington. to ensure proper structure of both competitive standards. Concentrations of the purified plasmids were determined by spectrophotometry. The competitive standards were aliquotted and stored at y808C until used for quantitative RT-PCR analyses. 2.6. RT
Fig. 1. Schematic view of heterologous competitive standards generation. ŽA. CYP26 structure. ŽB. GAPDH structure.
Two micrograms of total RNA from human prenatal or adult tissues were mixed with 20 units RNAsin ŽBoehringer Mannheim., 50 mM Oligo dŽT.16 ŽPerkin-Elmer, Branchburg, NJ. and DEPC-treated water in a total volume of 9 ml then incubated at 908C for 5 min, 48C for 5 min, and room temperature for 20 min. 10 mM DTT, 500 mM of each dNTP, 1 = First Strand buffer ŽGibco BRL. and 200 units of Super Script Reverse Transcriptase ŽGibco
10
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
BRL. were added to each tube Žtotal volume 20 ml. and were incubated at 428C for 60 min. Reverse transcriptase was inhibited by heating at 958C for 5 min and samples were either used immediately for PCR amplifications or were stored at y208C until further analysis. 2.7. QuantitatiÕe competitiÕe PCR Each sample for CYP26 PCR amplification contained 1 ml of cDNA mixture after RT, 200 mM of each dNTP, 1 = PCR buffer ŽBoehringer Mannheim., 1.5 mM MgCl 2 , 10 mM concentrations of each primer ŽCYP26FP and CYP26RP. and 2.5 units of Taq polymerase in a 50-ml reaction volume. Each sample for GAPDH PCR amplification was the same except for primers — 10 mM concentrations of each primer for GAPDHFP and GAPDHRP were used. Master mixes were prepared for each set of PCR reactions. Negative controls were included for each PCR experiment, when instead of RNA, an equivalent volume of DEPC-treated water was added to the RT reaction mixture and was processed with the other samples, following the same RT reaction protocol. One microliter of this sample was used as a negative control for each set of PCR amplifications. Controls for genomic DNA contamination of total RNA samples also were performed. Two micrograms of total RNA were subjected to the RT reaction but without the addition of reverse transcriptase. One microliter of this mixture was amplified by PCR with the other samples and then analyzed. All PCR amplifications were performed with a DNA Thermal cycler ŽPerkin-Elmer Cetus.. The first PCR cycle was 948C for 3 min followed by annealing and elongation at 728C for 7 min. The next 24 cycles for GAPDH amplifications or 34 cycles for CYP26 amplifications were performed as follows: 948C for 40 s Ždenaturation., 588C for 40 s Žannealing., and 728C for 1.5 min Želongation.. Post-PCR conditions included 728C for 7 min then 48C until analysis of samples. 2.8. Analysis of PCR products CYP26 products Ž369 and 300 bp. and GAPDH products Ž785 and 628 bp. obtained after quantitative competitive PCR were electrophoretically separated on 2% or 1.2% high resolution agarose ŽSigma, St. Louis, MO. gels, respectively. Gels were stained with ethidium bromide, visualized in UV light and photographed on positiver negative Polaroid film, type 665 ŽPolaroid, Cambridge, MA.. The negatives were scanned using Scan Jet II cxrT ŽHewlett-Packard, USA. and DeskScan II version 2.0 software ŽHewlett-Packard.. NIH Image version 1.58 software ŽNational Institutes of Health, USA. was used to perform quantitative analyses of band volumes. Digital images of the agarose gels were used to quantify CYP26 and GAPDH levels of expression in densitometric units ŽDU..
3. Results Specific PCR primers were designed and tested in RT-PCR reactions to ensure that products generated by amplifications were single bands with predicted sizes of 369 or 785 bp for CYP26 or GAPDH, respectively. Contamination of the PCR products is potentially problematic for interpretation and quantitation of the results, therefore, a negative, template-minus control was included for each set of PCR assays. Absence of the bands confirmed that there was no product contamination after PCR amplifications. Controls for genomic DNA contamination of total RNA samples also were performed. After amplification, these samples yielded no bands with the expected sizes, indicating that there was no significant DNA contamination of the RNA samples tested in this study. Errors introduced by differences in the integrity of individual RNA samples and in the efficiency of the RT reaction were sharply reduced by using a ‘‘housekeeping’’ gene, against which the other target gene was compared w2,10,37x. For our studies of CYP26 expression, GAPDH was chosen from among other ‘‘housekeeping’’ genes because it is not inducible by agents known to affect P450 concentrations w13x. Heterologous competitive standards were synthesized using CYP2E1 plasmid DNA as a template and either a pair of CYP26 linker primers ŽFig. 1A. or GAPDH linker primers ŽFig. 1B. during 25 cycles of PCR. Standards were designed to be approximately 20% shorter in their sizes than native products. This allowed good resolution on the agarose gel. PCR products were cloned into a plasmid and sequenced to verify their structure Žparticularly in primerbinding regions. to ensure their identity with native sequences. The next step was to determine an optimal number of PCR cycles for both CYP26 and GAPDH amplifications. This number was evaluated based on two criteria: first that sufficient amount of PCR product was produced for facile ultraviolet visualization on agarose gel, and secondly, that amplifications were performed only in the beginning of the plateau effect. Optimal numbers of cycles for CYP26 and GAPDH amplifications were determined as 35 ŽFig. 2A. and 25 cycles ŽFig. 2B., respectively, and were used for all further quantitative competitive PCR experiments. In order to standardize conditions for comparisons of message levels for CYP26 and GAPDH at different gestational stages and in adult tissues, we determined an amount of the competitive standards that would have to be used in all quantitative PCR experiments. A constant amount of 0.1 mg of reverse-transcribed total RNA from 59-day human fetal brain was co-amplified with serial dilutions of competitive standards during 35 PCR cycles for CYP26 or during 25 cycles for GAPDH. Negative photographs of the gels were analyzed by densitometry in order to quantify each band ŽFig. 3.. Due to the smaller size of the PCR products from competitive standards as compared with
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
Fig. 2. Relationship between product amplification and number of PCR cycles. Total RNA from human prenatal brain Ž59 days gestation. was reverse-transcribed and used as a template for different numbers of PCR amplifications. PCR products were electrophoretically separated, photographed and band volumes were quantitated in DU. ŽA. CYP26 PCR product Ž369 bp., amplifications were performed with primers CYP26FP and CYP26RP. ŽB. GAPDH PCR product Ž785 bp., amplifications were performed with primers GAPDHFP and GAPDHRP as described in Section 2.
wild type cDNA in the agarose gels, the decrease in the incorporation of ethidium bromide must be taken into account w20x. The corrections were made for comparisons of relative band sizes — the competitive standard for CYP26 was multiplied by 369r300s 1.23 when compared to native CYP26 PCR product and the competitive standard for GAPDH was multiplied by 785r628s 1.25 when compared to native GAPDH PCR product. Intensity of the native band divided by the size-adjusted intensity of standard band was plotted as a function of the number of competitive standard molecules per 0.1 mg of total RNA in double logarithmic coordinates. A straight line was fitted to these points and the amount of standard Žwhere the log 10 ratio of the intensities was equal to zero. was interpolated. The amounts of competitive standards when intensities of
Fig. 3. Co-amplification of increasing quantities of competitive standards with a constant amount of reverse-transcribed total RNA from human prenatal brain Ž59 days.. PCR bands were electrophoretically separated and analyzed. The log 10 intensity of native bands measured in DU divided by size-adjusted intensity of standard bands measured in DU Žnativerstandard. were plotted as a function of log 10 amount of standard Žmolecules of standardr0.1 mg of total RNA.. The resulting regression was solved for x when y s 0 to estimate quantity of competitive standard molecules per 0.1 mg of total RNA. ŽA. Co-amplification of native CYP26 with competitive standard during 35 cycles of PCR, primers CYP26FP and CYP26RP were used. ŽB. Co-amplification of native GAPDH with competitive standard during 25 cycles of PCR, primers GAPDHFP and GAPDHRP were used.
11
native and standard bands are equal were estimated as 2.4 = 10 3 Žmolecules of CYP26 standardr0.1 mg of total RNA. and 1.8 = 10 6 Žmolecules of GAPDH standardr0.1 mg of total RNA. and were used for all further quantitative RT-PCR experiments. CYP26 expression levels were evaluated in 12 human prenatal cephalic tissues Ždays 57–110. and six human prenatal hepatic tissues Ždays 63–110. by quantitative competitive RT-PCR. Comparisons also were made with human adult brain and liver, each of which was a pooled sample from two different donors. All tested tissues were normal Ži.e., no detected pathology.. GAPDH levels of expression as ‘‘housekeeping’’ gene also were determined in all human tissues for standardization purposes. An example of the quantitative analyses for CYP26 and GAPDH message levels in a 63-day human prenatal brain is shown in Fig. 4. Constant amounts of the competitive standards determined earlier were co-amplified with a series of dilutions of reverse-transcribed total RNA during 35 PCR cycles for CYP26 mRNA ŽFig. 4A. and during 25 PCR cycles for GAPDH mRNA ŽFig. 4C.. PCR products were separated electrophoretically and quantified by den-
Fig. 4. Competitive RT-PCR analyses for quantitation of CYP26 and GAPDH expression levels in human fetal brain Ž63 days.. ŽA. Ethidium bromide-stained PCR products were separated on 2% agarose gel after constant amounts of the CYP26 competitive standard Ž2.4=10 3 moleculesrreaction. were co-amplified with four decreasing amounts of reverse-transcribed total RNA Žlane 1: 0.12 mg; lane 2: 0.07 mg; lane 3: 0.04 mg; lane 4: 0.02 mg. during 35 cycles of PCR. ŽB. Bands from the gel were analyzed and the intensities of CYP26 native bands measured in DU divided by the respective size-adjusted intensities of standard bands measured in DU Žnativerstandard. were plotted as function of the amount of total RNA Žmg.. ŽC. Ethidium bromide-stained PCR products were separated on 1.2% agarose gel after constant amounts of the GAPDH competitive standard Ž1.8=10 6 moleculesrreaction. were co-amplified with four decreasing amounts of reverse-transcribed total RNA Žlane 1: 0.14 mg; lane 2: 0.09 mg; lane 3: 0.05 mg; lane 4: 0.03 mg. during 25 cycles of PCR. ŽD. Bands from the gel were analyzed and the intensities of GAPDH native bands measured in DU divided by the respective size-adjusted intensities of standard bands measured in DU Žnativerstandard. were plotted as function of the amount of total RNA Žmg..
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
12
Table 1 CYP26 and GAPDH message levels in human prenatal and adult cephalic tissues Gestational stage Ždays.
No. CYP26 moleculesr 1 mg of total RNAa
No. GAPDH moleculesr 1 mg of total RNAa
No. CYP26 moleculesr 10 6 GAPDH molecules
57 59 63 67 79 80 91 98 100 101 105 110 Adult brain
140 = 10 3 " 28.3 = 10 3 Ž n s 2. 108 = 10 3 " 25.5 = 10 3 Ž n s 3. 42 = 10 3 " 2.8 = 10 3 Ž n s 2. 67.5 = 10 3 " 3.5 = 10 3 Ž n s 2. 109 = 10 3 " 9.2 = 10 3 Ž n s 3. 92.5 = 10 3 " 4.9 = 10 3 Ž n s 2. 460 = 10 3 " 28.3 = 10 3 Ž n s 2. 413 = 10 3 " 102 = 10 3 Ž n s 3. 845 = 10 3 " 63.6 = 10 3 Ž n s 2. 460 = 10 3 " 28.3 = 10 3 Ž n s 2. 715 = 10 3 " 120 = 10 3 Ž n s 2. 990 = 10 3 " 68.8 = 10 3 Ž n s 4. 55.3 = 10 3 " 5 = 10 3 Ž n s 3.
55.5 = 10 6 " 6.4 = 10 6 Ž n s 2. 55 = 10 6 " 7.1 = 10 6 Ž n s 2. 28 = 10 6 " 11.3 = 10 6 Ž n s 2. 38 = 10 6 " 2.8 = 10 6 Ž n s 2. 25 = 10 6 " 3.6 = 10 6 Ž n s 3. 22 = 10 6 " 1.4 = 10 6 Ž n s 2. 33 = 10 6 " 4.2 = 10 6 Ž n s 2. 24 = 10 6 " 8.5 = 10 6 Ž n s 2. 40.5 = 10 6 " 6.4 = 10 6 Ž n s 2. 40.5 = 10 6 " 6.4 = 10 6 Ž n s 2. 37.5 = 10 6 " 10.6 = 10 6 Ž n s 2. 39 = 10 6 " 8.5 = 10 6 Ž n s 2. 36 = 10 6 " 6 = 10 6 Ž n s 3.
2500 2000 1500 1800 4400 4200 14 000 17 000 20 900 11 400 19 100 25 400 1500
a
Values are given as averages" S.D. with n values in parentheses.
sitometry. Ratios of target bands to size-adjusted standard bands were plotted against quantities of reverse-transcribed total RNA used in PCR reactions. The point at which this ratio is equal to 1 is the point where the amount of target gene is equal to a known amount of the competitor Ž2.4 = 10 3 molecules and 1.8 = 10 6 molecules for CYP26 and GAPDH competitors, respectively.. The amount of total RNA necessary to reach the equivalence point was interpolated from the same graph and estimated as 0.055 mg of total RNA for CYP26 mRNA ŽFig. 4B. and 0.09 mg of total RNA for GAPDH mRNA ŽFig. 4D.. The numbers of CYP26 and GAPDH molecules per 1 mg of total RNA were determined by the following equations: 2.4 = 10 3 Žmolecules of CYP26 competitive standard.r0.055 Žmg of total RNA. s 44 = 10 3 ŽCYP26 moleculesr1 mg of total RNA. and 1.8 = 10 6 Žmolecules of GAPDH competitive standard.r0.09 Žmg of total RNA. s 20 = 10 6 ŽGAPDH moleculesr1 mg of total RNA., respectively. In order to obtain an accurate measure of the equivalence point it was necessary to make a tight dilution series of reverse-transcribed total RNA used in the PCR reaction because the closer the ratio of target gene to competitive standard to equivalence point, the more linear the curve, and, therefore, the more precise the determination.
In this study, we demonstrated that GAPDH may be used as a ‘‘housekeeping’’ gene not only in adult tissues but also during human development. We found that GAPDH expression levels in human prenatal cephalic tissues exhibited some relatively small variations Ž2.5-fold., which were probably due to differences among individuals andror some variability in the total RNA preparations and RT-PCR testing. The average value for human prenatal cephalic tissues Ž36.5 = 10 6 GAPDH moleculesr1 mg of total RNA. was about the same as the average value for human adult whole brain Ž36 = 10 6 GAPDH moleculesr1 mg of total RNA. ŽTable 1.. GAPDH expression levels in human prenatal hepatic tissues also exhibited some variations Ž; 4-fold. among the different tissues with an average value of 23 = 10 6 ŽGAPDH moleculesr1 mg of total RNA., which is similar to 14 = 10 6 ŽGAPDH moleculesr1 mg of total RNA. for human adult liver ŽTable 2.. We did not observe any gestational stage-dependent changes in GAPDH expression in either cephalic or hepatic tissues. This allowed us to use GAPDH as a ‘‘housekeeping’’ gene for comparisons of CYP26 expression levels between human prenatal and adult cephalic and hepatic tissues. Average values for CYP26 and GAPDH message levels in each individual tissue were determined in at least two indepen-
Table 2 CYP26 and GAPDH message levels in human prenatal and adult hepatic tissues Gestational stage Ždays.
No. CYP26 moleculesr 1 mg of total RNAa
No. GAPDH moleculesr 1 mg of total RNAa
No. CYP26 moleculesr 10 6 GAPDH molecules
63 67 98 101 105 110 Adult liver
- 10 4 Ž n s 2. - 10 4 Ž n s 2. - 10 4 Ž n s 2. - 10 4 Ž n s 2. - 10 4 Ž n s 2. - 10 4 Ž n s 2. 65 = 10 3 " 23.5 = 10 3 Ž n s 4.
29 = 10 6 " 4.2 = 10 6 Ž n s 2. 25.5 = 10 6 " 14.8 = 10 6 Ž n s 2. 12 = 10 6 " 1.4 = 10 6 Ž n s 2. 48 = 10 6 " 4.2 = 10 6 Ž n s 2. 14.5 = 10 6 " 7.8 = 10 6 Ž n s 2. 15.5 = 10 6 " 0.7 = 10 6 Ž n s 2. 14 = 10 6 " 5.3 = 10 6 Ž n s 3.
- 350 - 400 - 800 - 200 - 700 - 650 4600
a
Values are given as averages" S.D. with n values in parentheses.
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
Fig. 5. Profiles of CYP26 expression levels in human cephalic and hepatic tissues.
dent RT-PCR experiments and summarized data are presented in Tables 1 and 2. In order to reduce errors introduced by differences in the integrity of individual RNA samples and in the efficiency of the RT reaction, we also expressed CYP26 levels as the numbers of CYP26 molecules per 10 6 GAPDH molecules. Our results revealed that CYP26 message levels appear to sharply increase Žapproximately 11-fold. during this early period of human prenatal brain development. The least abundance of CYP26 mRNA Žaverages 1950 CYP26 moleculesr10 6 GAPDH molecules. was determined in human fetal brain Ž57–67 days. and was comparable with the message level determined in the human adult brain Ž1500 CYP26 moleculesr10 6 GAPDH molecules.. The highest level of CYP26 expression Žaverages 22 300 CYP26 moleculesr10 6 GAPDH molecules. was found in human fetal brain Ž105–110 days.. CYP26 expression in human prenatal hepatic tissues was too low to quantify Ž- 800 CYP26 moleculesr10 6 GAPDH molecules. with the competitive standard concentration used in this study but was estimated as 4600 ŽCYP26 moleculesr10 6 GAPDH molecules. in human adult liver. Interestingly, during the same period of development Ž63–110 days., CYP26 message levels were much higher in human prenatal cephalic tissues than in human prenatal hepatic tissues but human adult livers exhibited approximately three times more CYP26 mRNA than human adult whole brain. Patterns of CYP26 expression in human cephalic and hepatic tissues are presented in Fig. 5.
4. Discussion The goal of this research was to investigate the presence and dynamics of CYP26 transcription in human cephalic and hepatic tissues during the late embryonic and early fetal periods of development Ždays 57–110.. In preliminary experiments, we were unable to detect CYP26 message in either prenatal tissue with Northern blot analyses using total RNA. In contrast, relatively strong bands were detected in human prenatal brain using RT-PCR. In order to evaluate CYP26 mRNA levels in human fetal tissues, a
13
quantitative competitive RT-PCR assay was developed. Competitive PCR is a highly sensitive method for obtaining quantitative data on specific nucleic acid sequences and relies on the use of standards to compensate for the high number of factors affecting the yield of PCR products. The internal standard is amplified with the same primers Žand thus presumably with the same efficiency. as the endogenous target sequence, and is distinguished from the product derived from the native sequence by either size or restriction site. The standards fall into two categories: homologous and heterologous competitor fragments. One potential problem with homologous competitor fragments is the formation of heteroduplexes between the standard and target sequences during the amplification process, an artifact that could interfere with quantitation. In contrast, heterologous competitor fragments differ from the target except for the primer-binding regions, which are identical, thus, heteroduplex formation cannot occur. Heterologous competitive standards were designed and produced for both CYP26 and GAPDH. Theoretically, the amount of product doubles during each cycle of PCR, but in actuality, beyond a certain number of cycles the efficiency of amplification decreases with increasing cycle number, resulting in a plateau effect. The number of PCR cycles at which the plateau effect occurs varies greatly with the particular DNA sequence being amplified and, therefore, should be individually and empirically determined for each target sequence. Optimal numbers of cycles for CYP26 and GAPDH amplifications were determined as 35 and 25 cycles, respectively, and were used for all further quantitative competitive PCR experiments. It was particularly important to establish an exact relationship between product amplification and the number of PCR cycles because it is commonly accepted that the nature of competitive PCR makes it possible to obtain useful data after the reaction has reached the plateau phase. However, this might not always be true when product is measured well after the plateau phase, especially when the heterologous competitor has been used. We determined amounts of the competitive standards that would have to be used in all quantitative RT-PCR experiments for target and ‘‘housekeeping’’ genes. In order to estimate levels of transcription, competitive standards were co-amplified with a dilution series of reverse-transcribed total RNA Žfrom each particular tissue. during the same number of PCR cycles determined earlier. The levels of expression were estimated from the equivalence point where the amount of native gene was equal to a known amount of the competitor. In the present study, CYP26 expression in human fetal cephalic and hepatic tissues was investigated from days 57 to 110. For proper and consistent evaluation of the data, it is important to adequately identify gestational stage. In this study, we based determinations of the fetal age on one of the commonly used prenatal measurements, fetal foot length w23x. The distinction between the embryonic and the
14
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
fetal period at 8 post-ovulatory weeks in humans is based on multiple internal and external criteria and the fact that 90% of the more than 4500 named structures of the adult body have appeared by that time w22x. The beginning of the fetal period is commonly marked by complete closure of the palate. In humans, the fetal period extends from approximately 58–60 days of gestation until parturition at approximately 267 days of gestation. For comparison, in mice the fetal period extends from approximately day 15 of gestation until parturition at day 19 w16x. Our results revealed that CYP26 expression rapidly increases Žmore than 10-fold. in the human prenatal brain between gestational days 57 and 110. Interestingly, these data are consistent with prior morphological observations of considerable human fetal brain enlargement, especially in the cerebral hemisphere during the same period of development. The period near 90 post-ovulatory days was suggested as a critical time in human fetal development based on mathematical and hormonal criteria w23x. At the end of the embryonic period, the human brain is far more advanced morphologically than is generally appreciated; to such an extent that functional considerations are compelling. The arrangement of the rhombencephalic nuclei and tracts at this period is very similar to that observed in the newborn. Nevertheless, the fetal period is characterized by rapid growth and differentiation, including a series of complex subdivisions, flexures, and cellular differentiation that is not complete until well into the first few years of neonatal life. The most noticeable external changes in the human fetal brain are Ž1. the union of the cerebellar halves and the definition of the vermis; Ž2. the increasing concealment of the diencephalon and mesencephalon; Ž3. further approach of the frontal and temporal poles around the insula; Ž4. the appearance of sulci on the hemispheric surface Žat about the middle of prenatal life.; and Ž5. the increasing eclipse or disappearance of the cervical, pontine, and mesencephalic flexures. Internal changes also occur; one of most noticeable is the caudally directed growth of the corpus callosum w21,23x. Moreover, CYP26 mRNA levels in human prenatal cephalic tissue Žday 110. are approximately 17-fold higher than in adult whole brain, also suggesting a highly important role for this cytochrome P450 during early human brain development. Recent data on CYP26 mRNA distribution in human adult brain indicated that this enzyme was present in all regions of the brain tested, with higher expression levels in the occipital lobe, putamen, substantia nigra, temporal lobe and thalamus w33x. Furthermore, we demonstrated that CYP26 transcription levels in human fetal hepatic tissues were practically undetectable and much lower than in fetal cephalic tissues at all ages examined Ždays 63–110.. Remarkably, CYP26 message levels in human adult livers were approximately 3-fold higher than in human adult whole brain. These data are consistent with the fact that the adult liver is an organ in which the majority of the body’s retinoids are stored w3x. Because RA shows strong physiologic ac-
tions, it seems important to maintain tissue levels of retinoids within tightly regulated limits through coordinated biosynthesis and inactivation. It has been demonstrated that CYP26 is expressed differentially during early Xenopus development. During embryogenesis, CYP26 transcripts were detected at all stages analyzed; the RT-PCR analysis suggested a slight increase in transcript levels during gastrulationrneurulation. Whole-mount in situ hybridization revealed that, during gastrulation, CYP26 is not expressed in the area of the prospective neural plate that develops into hindbrain, whereas the prospective anterior neural and the underlying mesodermal prechordal plate are primary sites for CYP26 expression. In Xenopus embryos, ectopic CYP26 expression can rescue developmental defects in head formation induced by application of exogenous RA, suggesting that the enzymatic modifications introduced inhibit RA signaling activities in vivo. The authors proposed that CYP26mediated modification of RA modulates its signaling activity and helps to establish boundaries of differentially responsive and non-responsive territories w14x. In that study, which investigated the endogenous distribution of RA by high pressure liquid chromatography in chick embryos, sharp onroff boundaries rather than smooth gradients across a wide range of embryonic stages were emphasized. Some tissues, such as heart, had very little RA while others, such as the neural tube, had very high levels. The total variation between these two was 29-fold. The developing brain did not appear to generate RA, but the spinal part of the neural tube generated it at very high levels. It would seem that there must be a sharp onroff boundary in the region of the hindbrainrspinal cord junction; the mesenchyme surrounding the hindbrain generated RA whereas the hindbrain itself did not w17x. By contrast, results from a different group of investigators supported the idea that developing hindbrain in Xenopus embryo may be patterned by a retinoid gradient. The sequentially more posterior hindbrain patterning genes were induced effectively by sequentially higher RA concentrations w11x. Iulianella et al. w15x demonstrated that CYP26 expression was abundant in the retinoid-poor region of the mouse caudal embryo. It was rapidly induced by RA treatment in vivo suggesting that it plays a critical role in retinoid signaling. The authors concluded that CYP26 does not necessarily abolish normal retinoid signaling, rather, it may serve to limit the effects of RA to specific areas andror specific times in the mouse embryo. In summary, the quantitation of CYP26 transcription levels in human fetal cephalic and hepatic tissues are potentially highly important findings. Taken together, the results indicate a significant role for this RA-catabolizing cytochrome P450 during the late embryonic and early fetal period of human brain development. However, a tremendous amount of research remains to be accomplished in this area. It will be very interesting to determine the specific localization of CYP26 message and protein in
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
regions of human brain during various stages of embryogenesis and fetogenesis and to investigate its possible role as a modulator of congenital CNS malformations and neurobehavioral defects.
Acknowledgements The authors thank Birth Defects Laboratory and expert technical assistance of J. Pascoe-Mason and M. Eisenhauer. We also acknowledge S. Lee for technical assistance. This research was supported by NIH Grant ES-04041 and NIH, National Research Service Awards T32 E507032 from the National Institute of Environmental Health Sciences.
References w1x S.S. Abu-Abed, B.R. Beckett, H. Chiba, J.V. Chithalen, G. Jones, D. Metzger, P. Chambon, M. Petkovich, Mouse P450RAI ŽCYP26. expression and retinoic acid–inducible retinoic acid metabolism in F9 cells are regulated by retinoic acid receptor g and retinoid X receptor a , J. Biol. Chem. 273 Ž1998. 2409–2415. w2x M.J. Apostolakos, W.H.T. Schuermann, M.W. Frampton, M.J. Utell, J.C. Willey, Measurement of gene expression by multiplex competitive polymerase chain reaction, Anal. Biochem. 213 Ž1993. 277–284. w3x W.S. Blaner, J.A. Olson, Retinol and retinoic acid metabolism, in: M.S. Sporn, A.B. Roberts, D.S. Goodman ŽEds.., The Retinoids: Biology, Chemistry, and Medicine, Raven Press, New York, 1994, pp. 229–255. w4x L.M. De Luca, N. Darwiche, G. Celli, K. Kosa, C. Jones, S. Ross, L.C. Chen, Vitamin A in epithelial differentiation and skin carcinogenesis, Nutr. Rev. 52 Ž1994. 45–52. w5x A.J. Durston, J.P.M. Timmermans, W.J. Hage, H.F.J. Hendriks, N.J. de Vries, M. Heideveld, P.D. Nieuwkoop, Retinoic acid causes an antero-posterior transformation in the developing central nervous system, Nature 340 Ž1989. 140–144. w6x A.J. Durston, J. van der Wees, W.W.M. Pijnappel, S.F. Godsave, Retinoids and related signals in early development of the vertebrate central nervous system, Curr. Top. Dev. Biol. 40 Ž1998. 111–175. w7x M.K.S. Edwards, M.W. McBurney, The concentration of retinoic acid determines the differentiated cell types formed by a teratocarcinoma cell line, Dev. Biol. 98 Ž1983. 187–191. w8x M.J. Fasco, C. Treanor, L.S. Kaminsky, Cytochrome P450 mRNA induction: quantitation by RNA-polymerase chain reaction using capillary electrophoresis, Methods Enzymol. 272 Ž1996. 401–413. w9x H. Fujii, T. Sato, S. Kaneko, O. Gotoh, Y. Fujii-Kuriyama, K. Osawa, S. Kato, H. Hamada, Metabolic inactivation of retinoic acid by a novel P450 differentially expressed in developing mouse embryos, EMBO J. 16 Ž1997. 4163–4173. w10x W.C. Gause, J. Adamovicz, Use of PCR to quantitate relative differences in gene expression, in: C.W. Dieffenbach, G.S. Dveksler ŽEds.., PCR Primer: A Laboratory Manual, Cold Spring Harbor Laboratory Press, New York, 1995, pp. 293–311. w11x S.F. Godsave, C.H. Koster, A. Getahun, M. Mathu, M. Hooiveld, J. van der Wees, J. Hendriks, A.J. Durston, Graded retinoid responses in the developing hindbrain, Dev. Dyn. 213 Ž1998. 39–49. w12x F. Gurrieri, P. Prinos, D. Tackels, M.W. Kilpatrick, J. Allanson, M. Genuardi, A. Vuckov, L. Nanni, E. Sangiorgi, G. Garofalo, M.E. Nunes, G. Neri, C. Schwartz, P. Tsipouras, A split hand–split foot ŽSHFM3. gene is located at 10q24–25, Am. J. Med. Genet. 62 Ž1996. 427–436.
15
w13x B. Hogan, R. Beddington, F. Costantini, E. Lacy ŽEds.., Manipulating the Mouse Embryo, A Laboratory Manual, Cold Spring Harbor Laboratory Press, New York, 1994. w14x T. Hollemann, Y. Chen, H. Grunz, T. Pieler, Regionalized metabolic activity establishes boundaries of retinoic acid signaling, EMBO J. 17 Ž1998. 7361–7372. w15x A. Iulianella, B. Beckett, M. Petkovich, D. Lohnes, A molecular basis for retinoic acid-induced axial truncation, Dev. Biol. 205 Ž1999. 33–48. w16x M.R. Juchau, H. Boutelet-Bochan, Y. Huang, Cytochrome-P450-dependent biotransformation of xenobiotics in human and rodent embryonic tissues, Drug Metab. Rev. 30 Ž1998. 541–568. w17x M. Maden, E. Sonneveld, P.T. van der Saag, E. Gale, The distribution of endogenous retinoic acid in the chick embryo: implications for developmental mechanisms, Development 125 Ž1998. 4133– 4144. w18x Y. Marikar, Z. Wang, E.A. Duell, M. Petkovich, J.J. Voorhees, G.J. Fisher, Retinoic acid receptors regulate expression of retinoic acid 4-hydroxylase that specifically inactivates all-trans retinoic acid in human keratinocyte HaCaT cells, J. Invest. Dermatol. 111 Ž1998. 434–439. w19x P. McCaffery, E. Wagner, J. O’Neil, M. Petkovich, U.C. Drager, Dorsal and ventral retinal territories defined by retinoic acid synthesis, break-down and nuclear receptor expression, Mech. Dev. 82 Ž1999. 119–130. w20x S. Menzo, P. Bannarelli, M. Giacca, A. Manzin, P.E. Varaldo, M. Clementi, Absolute quantitation of viremia in human immunodeficiency virus infection by competitive reverse transcription and polymerase chain reaction, J. Clin. Microbiol. 30 Ž1992. 1752–1757. w21x K.L. Moore, The Developing Human, W.B. Saunders, Philadelphia, 1988. w22x R. O’Rahilly, F. Muller, Developmental Stages in Human Embryos, Carnegie Institution of Washington, 1987. w23x R. O’Rahilly, F. Muller, The Embryonic Human Brain: An Atlas of Developmental Stages, Wiley-Liss, New York, 1994. w24x W.J. Ray, G. Bain, M. Yao, D.I. Gottlieb, CYP26, a novel mammalian cytochrome P450, is induced by retinoic acid and defines a new family, J. Biol. Chem. 272 Ž1997. 18702–18708. w25x K. Roos, E. Sonneveld, B. Compaan, D. ten Berge, A.J. Durston, P.T. van der Saag, Expression of retinoic acid 4-hydroxylase ŽCYP26. during mouse and Xenopus laeÕis embryogenesis, Mech. Dev. 82 Ž1999. 205–211. w26x C.R. Sharpe, K. Goldstone, Retinoid receptors promote primary neurogenesis in Xenopus, Development 124 Ž1997. 515–523. w27x N. Sidell, Retinoic acid-induced growth inhibition and morphologic differentiation of human neuroblastoma cells in vitro, J. Natl. Cancer Inst. 68 Ž1982. 589–593. w28x A. Simeone, V. Avantaggiato, M.C. Moroni, F. Mavilio, C. Arra, F. Cotelli, V. Nigro, D. Acampora, Retinoic acid induces stage-specific antero-posterior transformation of rostral central nervous system, Mech. Dev. 51 Ž1995. 83–98. w29x H.L. Sive, B.W. Draper, R.M. Harland, H. Weintraub, Identification of a retinoic acid-sensitive period during primary axis formation in Xenopus laeÕis, Genes Dev. 4 Ž1990. 932–942. w30x W.J. Slikker, L.W. Chang, Handbook of Developmental Neurotoxicology, Academic Press, San Diego, 1998. w31x E. Sonneveld, C.E. van den Brink, B.-J.M. van der Leede, R.-K.A.M. Schulkes, M. Petkovich, B. van der Burg, P.T. van der Saag, Human retinoic acid ŽRA. 4-hydroxylase ŽCYP26. is highly specific for all-trans-RA and can be induced through RA receptors in human breast and colon carcinoma cells, Cell Growth Differ. 9 Ž1998. 629–637. w32x A. Suomalainen, J. Kaukonen, P. Amati, R. Timonen, M. Haltia, J. Weissenbach, M. Zeviani, H. Somer, L. Peltonen, An autosomal locus predisposing to deletions of mitochondrial DNA, Nat. Genet. 9 Ž1995. 146–151. w33x M.E. Trofimova-Griffin, M.R. Juchau, Expression of cytochrome
16
M.E. TrofimoÕa-Griffin et al.r DeÕelopmental Brain Research 120 (2000) 7–16
P450RAI ŽCYP26. in human fetal hepatic and cephalic tissues, Biochem. Biophys. Res. Commun. 252 Ž1998. 487–491. w34x J.A. White, B. Beckett, S.W. Scherer, J.-A. Herbrick, M. Petkovich, P450RAI ŽCYP26A1. maps to human chromosome 10q23–q24 and mouse chromosome 19C2–3, Genomics 48 Ž1998. 270–272. w35x J.A. White, B. Beckett-Jones, Y.-D. Guo, F.J. Dilworth, J. Bonasoro, G. Jones, M. Petkovich, cDNA cloning of human retinoic acid-metabolizing enzyme Žh P450RAI. identifies a novel family of cytochromes P450 ŽCYP26., J. Biol. Chem. 272 Ž1997. 18538– 18541. w36x J.A. White, Y.-D. Guo, K. Baetz, B. Beckett-Jones, J. Bonasoro,
K.E. Hsu, F.J. Dilworth, G. Jones, M. Petkovich, Identification of the retinoic acid-inducible all-trans-retinoic acid 4-hydroxylase, J. Biol. Chem. 271 Ž1996. 29922–29927. w37x J.C. Willey, E.L. Coy, M.W. Frampton, A. Torres, M.J. Apostolakos, G. Hoehn, W.H. Schuermann, W.G. Thilly, D.E. Olson, J.R. Hammersley, C.L. Crespi, M.J. Utell, Quantitative RT-PCR measurement of cytochromes P450 1A1, 1B1, and 2B7, microsomal epoxide hydrolase, and NADPH oxidoreductase expression in lung cells of smokers and nonsmokers, Am. J. Respir. Cell Mol. Biol. 17 Ž1997. 114–124.