|
|
||||||||
1 From the Department of Neurobiology, 2 Brain Research Institute and 3 Jules Stein Eye Institute, School of Medicine, University of California, Los Angeles; and the 4 Department of Ophthalmology and Visual Sciences, The University of Iowa, Center for Macular Degeneration, Iowa City.
| Abstract |
|---|
|
|
|---|
METHODS. Using DNA probes specific for LRAT, a bacterial artificial chromosome (BAC) clone containing the LRAT gene was isolated, subcloned into DNA fragments and relevant subclones characterized by sequencing. Exonintron junctions were determined by comparison with the cDNA sequence previously published. Southern blot analysis was performed on human genomic DNA samples digested with several restriction enzymes. Fluorescence in situ hybridization (FISH) analysis of normal metaphase chromosomes derived from phytohemagglutinin (PHA) stimulated peripheral blood lymphocytes and radiation hybrid mapping were used for localization of the LRAT gene. Single-strand conformation polymorphism analysis (SSCP) was used to screen for potential mutations in patients with age-related macular degeneration, Leber congenital amaurosis, retinitis pigmentosa, and cone-rod dystrophy.
RESULTS. The human LRAT gene is organized into three exons of 219, 541, and 2058 bp and two introns of 103 and 4117 bp. Southern blot analysis of digested genomic DNA revealed a single band, suggesting a single copy of the LRAT gene. The human LRAT gene was localized to chromosome 4q31.2, a locus having no previous association with human eye disease. Additionally, the bovine LRAT homologue sequence was deduced and a general LRAT protein topology is suggested. No polymorphisms that segregated with retinal disease phenotypes were identified in 374 unrelated probands.
CONCLUSIONS. The organization of the LRAT gene, based on cDNA clones derived from the retinal pigment epithelium (RPE) has been determined. Its structure is less complex than other acyltransferases such as lecithin cholesterol acyltransferase (LCAT) and acyl CoA acyltransferase (ACAT). The absence of polymorphisms in the probands examined suggests a very low mutation level in the LRAT gene from the diseases analyzed.
| Introduction |
|---|
|
|
|---|
LRAT activity is fundamental in the continuation of the visual cycle. This important process takes place in the photoreceptor cells, and the RPE and is responsible for the generation of 11-cis-retinaldehyde, which is the chromophore for rhodopsin and the cone photopigments.2 3 The mechanism of action of LRAT suggests that an acyl group, specifically the acyl group at position sn-1 of a membrane phosphotidylcholine, is removed and transferred to all-trans retinol during the formation of all-trans retinyl esters.4 These retinyl esters are not only the storage form of all-trans-retinol (Vitamin A) but are also the substrate for the isomerohydrolase that generates 11-cis-retinol, another important intermediate in the visual cycle.5 6 In addition to RPE, LRAT is expressed in other tissues involved in processing and transport of retinol such as intestine, testis, and liver.7 8 9 Because of the relative instability of LRAT in the detergent solubilized form, its purification has represented a challenge. Nonetheless, by affinity labeling studies using LRAT specific inhibitors, its kinetic properties and substrate specificity have been well-documented.10 11
The LRAT-specific inhibitor [3H] all-trans-retinyl a-bromoacetate was used with partially purified membrane proteins from bovine RPE to detect a product of approximately 25 kDa.12 This observation was confirmed with the use of an additional LRAT-specific inhibitor N-boc-L-biocytinyl-11-aminoundecane chloromethyl ketone.1 Peptide sequences were obtained from this 25-kDa product, and its primary structure was determined by cDNA cloning and sequencing.1 Conceptual translation of the cDNAs demonstrated that in the human RPE, LRAT contains 230 amino acid residues with a calculated mass of 25.3 kDa. This predicted protein mass is in agreement with the LRAT protein detected on Western blot analysis using anti-LRAT polyclonal antibodies.1 Interestingly, comparison of the LRAT sequence with sequences deposited in GenBank database did not show homology with other proteins, particularly with other acyltransferases, including the well-characterized lecithin cholesterol acyltransferase (LCAT) and acyl-CoA: cholesterol acyltransferase (ACAT).13 14 15 16 This observation suggests that LRAT is not a splicing variant of these enzymes and is consequently the product of an independent gene.
To understand the structure of the LRAT gene, we sought to isolate genomic DNA clones containing the gene encoding LRAT to determine its intronexon arrangement and to identify its chromosomal location. We also used this information to search for disease-causing mutations in individuals with various forms of hereditary retinal degenerations.
| Materials and Methods |
|---|
|
|
|---|
Isolation of LRAT Genomic Subclones
DNA was extracted from each of the 95 subclones by using the
Wizard Plus Miniprep DNA purification system (Promega, Madison, WI).
With a 96-well Bio-Dot microfiltration apparatus (Bio-Rad, Hercules,
CA), 200 ng of DNA from each sample was blotted onto Hybond nylon
membranes (Amersham, Piscataway, NJ) and cross-linked using a UV
Stratalinker 1800 apparatus (Stratagene, La Jolla, CA). A subclone
containing LRAT sequences was identified using an EcoRI DNA
fragment labeled with 32P dCTP by Nick
translation. This probe contained the entire LRAT coding sequence and
fragments of the 5' (96 bp) and 3' (254 bp) untranslated regions (UTR).
Hybridization of duplicate filters with the radioactive probe at 60°C
was performed using ExpressHyb hybridization solution according to the
manufacturers instructions (Clontech, Palo Alto, CA). After
hybridization, the filters were rinsed twice at room temperature with a
2x SSC, 0.1% SDS solution for 15 minutes. High stringency washes with
0.1x SSC and 0.1% SDS at 50°C were performed for two rounds of 30
minutes each. Filters were exposed to X-ray film (Fujifilm, Stamford,
CT) using an intensifying screen at -80°C.
Genomic Structure Characterization
DNA sequence analysis of LRAT subclones was performed using the
fluorescence labeled dideoxy nucleotide termination method (Dye
Terminator) in an ABI Model 377 automated DNA sequencer. Exonic
fragments were determined by comparison of the obtained genomic
sequences with the published LRAT cDNA sequence.
Chromosomal Localization
Radiation hybrid analysis was performed with the Genebridge 4
radiation hybrid panel from Research Genetics (Huntsville, AL). A pair
of primers (5'AACATCCTGGTCAATCACCTGGA3' and
5'AAGTGCTCGCAGTTGTTCCACAGC3') corresponding to positions 361 to 383 and
468 to 491 of the published LRAT sequence was used for PCR
amplification of radiation hybrid clones. Final reactions were
electrophoresed on 6% polyacrylamide5% glycerol gels, and after
staining with silver nitrate, the presence of amplified products were
scored. The data were then submitted to the Whitehead Institute Server
(http://www.genome.wi.mit.edu/cgi-bin/contig/rhmapper.pl) for
computational analysis and chromosomal assignment of the gene.
Fluorescence in situ hybridization (FISH) was also used for the
chromosomal localization of the human LRAT gene. This
technique yields fluorescent signals at the same site on both
chromatids of a chromosome. A probe consisting of DNA from BAC clone
417L23 was labeled with digoxigenin dUTP by Nick translation. The
labeled probe was combined with sheared human genomic DNA and
hybridized to normal metaphase chromosomes derived from PHA-stimulated
peripheral blood lymphocytes in a solution containing 50% formamide,
10% dextran sulfate, and 2x SSC. Specific hybridization signals were
detected by incubating the hybridized slides with fluoresceinated
antidigoxigenin antibodies, followed by counterstaining with 4'-6'
diamino-2-phenylindole (DAPI). In a second experiment, a chromosome 4
centromere specific probe was cohybridized with the human LRAT BAC
clone 417L23 for reference.
Southern Blot Analysis
Ten µg of human genomic DNA per assay was digested
with the restriction enzymes EcoRI (E), HindIII
(H), XbaI (X), and SacI (S). Digested DNAs were
electrophoresed in a 1% agarose gel and blotted to Hybond nylon
membranes (Amersham). The filter membrane was hybridized with a 1026 bp
EcoRI LRAT fragment containing 92 bp of 5' UTR, 690 bp of
coding sequence and 254 bp of the 3' UTR. The LRAT probe was
radioactively labeled with 32P-dCTP (Amersham) by
Nick translation (GIBCO/BRL, Grand Island, NY). The hybridization
process, washing solutions, and exposure conditions were the same as
described in the characterization of LRAT genomic subclones section.
Isolation of Bovine RPE cDNAs Encoding LRAT
A fetal bovine RPE cDNA library was constructed in the Uni-ZAP
XR system (Stratagene) by using polyA+ RNA obtained from cultured fetal
bovine RPE cells that were maintained and harvested as described
earlier.17
This cDNA library was screened with a
radiolabeled 32P DNA probe, which included
nucleotides -92 through 1850 of the published LRAT
sequence.1
Using the Sequenase 2.0 System (Amersham),
positive cDNA clones were characterized further by sequencing both
strands with the dideoxy chain-termination method.
Topology
Hydropathy profiles for both bovine and human LRAT amino acid
sequences were performed using the KyteDoolittle
algorithm.18
To predict possible transmembrane domains in
this protein, we also used the TMpred program developed by Hofmann and
Stoffel19
with a minimum of 17 amino acids and maximum of
33 amino acids as range parameters for the formation of a transmembrane
helix. This program is available at
(http://ulrec3.unil.ch/sofware/TMPRED form.html).
Screening for Mutations in Patients with Retinal Degenerations
This study was approved by the Human Subjects Review Committee
at the University of Iowa, and informed consent was obtained from all
study participants following the tenets of the Declaration of Helsinki
for human experimentation. As shown in Table 1
, 374 unrelated probands were screened for the entire coding
region and the intronexon borders of the LRAT gene.
Probands were selected among clinical genotypes for which specific
mutations have not yet been identified. The probands had the following
clinical diagnoses: 38 with Leber congenital amaurosis, 91 with
retinitis pigmentosa, 58 with conerod dystrophy, 93 with age-related
macular degeneration, and 94 normal volunteers over 40 years of age who
had no family or personal history of ocular disease. DNA from all study
individuals was extracted from venous blood using a previously
described protocol.20
Six primers pairs were used to
amplify the entire coding sequence and exonintron boundaries of
the LRAT gene in all 374 individuals. They were:
exon 1 5'CTTATCCGTCTCAATCCCCA3' and 5'GGCTGGGCAAGTTAAGCTC3'; exon
2A 5'ACCTCTCCAAGACGCCCT3' and 5'AGATGCCATAGTGGGTCAGG3';
exon 2B 5'ACCAGCTCTTTCCACCGAG3' and 5'GTAGGCGAAGTCCTCCACTG3'; exon 2C
5'TATTGTCAAAGTGGCCAGCA3' and 5'GAAGTGCTCGCAGTTGTTCC3'; exon 2D
5'AAAAGCTGCTGGGCTTTACC3' and 5'GGGAAGAGAAAAGGTCAGGG3';
and exon 3 5'TCTTCTTGGGTTTAGCCACC3' and
5'TTTACATACAGAATACACACACTGACA3'. All amplimers were analyzed
using single-strand conformation polymorphism analysis (SSCP). All SSCP
gels were scored independently by a minimum of two experienced
investigators. Amplimers showing a bandshift were reamplified and
sequenced bidirectionally using an ABI 377 automated sequencer and
dye-terminator chemistry.
|
| Results |
|---|
|
|
|---|
|
|
|
Comparison of Bovine and Human LRAT Sequences
Two cDNA clones were isolated following screening of a fetal
bovine RPE cDNA library with a human LRAT cDNA probe. These bovine cDNA
clones, designated bLRAT-clone 1 and bLRAT-clone 2 (GenBank database
accession number: AF275344), contained inserts of 2.4 and 2.2 Kb
respectively. We previously determined the first 24 N-terminal amino
acids of bovine LRAT by direct amino acid sequencing
(MKNPMLEAVSLVLEKLLFISYFKF).1
The assembled nucleotide
sequence of the isolated cDNA clones contained the majority of the open
reading frame of bovine LRAT in addition to several hundred bp of 3'
UTR. The protein sequence deduced from the cDNA is in perfect agreement
with the actual N-terminal protein sequence. The first two amino acids
were not encoded by either cDNA clone. However, it is possible to
deduce the missing nucleotides of the bovine sequence from the obtained
N-terminal amino acid sequence as methionine is always encoded by
atg, Lysine by aaa/g and the first nucleotide of
Asparagine is a. (Fig. 3)
.
|
Putative LRAT Topology
Using the TMpred program of Hofmann and Stoffel,19
in
combination with hydropathy plots according to Kyte and
Doolittle,18
two potential transmembrane domains can be
predicted for the LRAT protein (Fig. 4)
. The first putative transmembrane domain was found between residues
Val-9 and Gly-31, and a second domain may be located between Leu-196
and Ile-222. These transmembrane domains were found at the same
position and had the same length in both human and bovine polypeptide
sequences. According to predictions of the TMpred program, both the N
and C terminus of LRAT should be found in the lumen of the Endoplasmic
reticulum (ER), yet the majority of the protein mass is predicted to
reside in the cytosol. This domain includes the four cysteines present
in the LRAT polypeptide at positions 161, 168, 182, and 208 in both
human and cow. Recently, we proposed a role for Cys-161 and Cys-168 in
the catalytic activity of the enzyme, based on results obtained by
site- directed mutagenesis experiments.21
|
| Discussion |
|---|
|
|
|---|
The LRAT gene structure differs considerably from other
acyltransferases such as ACAT and LCAT. The ACAT protein is encoded by
16 exons with seven transmembrane domains distributed along its entire
polypeptide and the LCAT protein with 440 amino acids is encoded by six
different exons with no transmembrane domains.13
22
23
The
enzymatic activities of these proteins are associated with microsomal
membranes. However, in spite of carrying out similar functions, they
have different substrates (e.g., retinol for LRAT, cholesterol for
LCAT, and Acyl CoA for ACAT). Both LCAT and ACAT have been widely
studied and their biochemical properties and mechanism of action are
understood.13
15
24
25
Whereas ACAT and LCAT possess a
similar molecular protein mass (
50 to 60 kDa) and some homology at
the amino acid level, LRAT is a much smaller protein (
25 kDa) and
shares no homology with these acyltransferases.
Furthermore, the predictions of LRAT topology suggest that it is simpler than the other acyltransferases. Only two transmembrane domains are predicted to traverse the ER lipid bilayer, exposing most of the protein mass in the cytosolic compartment where the catalytic site(s) has been proposed to reside.1 21 Interestingly, the LRAT protein does not possess the serine protease/lipase active site domain that is present in LCAT and ACAT. Therefore, it is likely that LRAT has a different mechanism of enzymatic activity. Currently, little is known about the LRAT active site(s) and its catalytic mechanism. We recently determined by site-directed mutagenesis that, of the four cysteines present in LRAT, Cys-161 and Cys-168 are directly involved in the catalytic activity of this enzyme.21
Comparison of the LRAT amino acid sequences reported here and sequences available in the GenBank database reveals a 12-amino acid domain (NCEHFVTYCRYG) that is highly conserved within the human and bovine LRAT sequences, a partial EST mouse sequence (clone IMAGE: 619832), the human TIG3 and rev 107 proteins,26 27 and the rat rev 107 protein.26 A high level of identity between the human and bovine LRAT amino acid sequences was found when these two sequences were aligned (88%). The homology between human and bovine LRAT supports strongly our previous observation of cross-reactivity of a protein in bovine RPE microsomal membranes with the human anti-LRAT antibodies. However, when more phylogenetically distant species such as mouse and rat were used, no reactivity between RPE cells and the anti-LRAT antibody was detected by Western blot analysis (A. Ruiz and D. Bok, unpublished results, February, 2000).
Preliminary results from Western blot analysis using native cells from bovine liver have shown that antibodies directed against LRAT recognize proteins of higher molecular mass than the 25 kDa observed in RPE. Whether these bands represent a larger variant of the LRAT protein or cross-reactivity with distinct, but related proteins, is currently unknown. Northern blot analysis has shown that tissues known to process retinoids express a 5 Kb mRNA transcript that hybridizes to LRAT cDNA probes. Taken together, these data suggest the possibility that LRAT isoforms exist. Clearly, more work at the molecular level is necessary to understand these issues related to tissue specificity.
The localization of the human LRAT gene on chromosome 4 by FISH was also supported by computer analysis performed on BAC sequences deposited in the GenBank database as part of the Human Genome Project. BAC clones AC009567 and AC027377 originating from human chromosome 4 confirmed the exonintron structure reported in this study. In addition, the full sequences of the RPE LRAT introns are also now available.
The data presented here pave the way for a more detailed analysis of the exonic sequences of LRAT in the search for mutations in a gene that could be associated not only with eye disease but disease mechanisms in any tissues where processing of retinol or its derivatives is abnormal. These data will facilitate strategies for the study of LRAT at the molecular level in other species.
| Footnotes |
|---|
Submitted for publication July 20, 2000; revised September 19, 2000; accepted September 22, 2000.
Commercial relationships policy: N.
Corresponding author: Dean Bok, Jules Stein Eye Institute, 100 Stein Plaza, Rm. B-182, UCLA School of Medicine, Los Angeles, CA 90095. bok{at}jsei.ucla.edu
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
A. R. Moise, M. Golczak, Y. Imanishi, and K. Palczewski Topology and Membrane Association of Lecithin: Retinol Acyltransferase J. Biol. Chem., January 19, 2007; 282(3): 2081 - 2090. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. P. Simmons, M. L. Peach, J. R. Friedman, M. M.B. Green, M. C. Nicklaus, and L. M. De Luca Evidence that sequence homologous region in LRAT-like proteins possesses anti-proliferative activity and DNA binding properties: translational implications and mechanism of action Carcinogenesis, April 1, 2006; 27(4): 693 - 707. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Boorjian, S. K. Tickoo, N. P. Mongan, H. Yu, D. Bok, R. R. Rando, D. M. Nanus, D. S. Scherr, and L. J. Gudas Reduced Lecithin: Retinol Acyltransferase Expression Correlates with Increased Pathologic Tumor Stage in Bladder Cancer Clin. Cancer Res., May 15, 2004; 10(10): 3429 - 3437. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. L. Batten, Y. Imanishi, T. Maeda, D. C. Tu, A. R. Moise, D. Bronson, D. Possin, R. N. Van Gelder, W. Baehr, and K. Palczewski Lecithin-retinol Acyltransferase Is Essential for Accumulation of All-trans-Retinyl Esters in the Eye and in the Liver J. Biol. Chem., March 12, 2004; 279(11): 10422 - 10432. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. C. Ross and R. Zolfaghari Regulation of Hepatic Retinol Metabolism: Perspectives from Studies on Vitamin A Status J. Nutr., January 1, 2004; 134(1): 269S - 275. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. C. Ross Retinoid Production and Catabolism: Role of Diet in Regulating Retinol Esterification and Retinoic Acid Oxidation J. Nutr., January 1, 2003; 133(1): 291S - 296. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. P. Simmons, F. Andreola, and L. M. De Luca Human melanomas of fibroblast and epithelial morphology differ widely in their ability to synthesize retinyl esters Carcinogenesis, November 1, 2002; 23(11): 1821 - 1830. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |