sjögren-larsson syndrome: diversity of mutations and polymorphisms in the fatty aldehyde...
TRANSCRIPT
MUTATION UPDATE
Sjogren-Larsson Syndrome: Diversity of Mutationsand Polymorphisms in the Fatty AldehydeDehydrogenase Gene (ALDH3A2)
William B. Rizzon and Gael Carney
Department of Pediatrics, University of Nebraska Medical Center, Omaha, Nebraska
Communicated by Jan P. Kraus
Sjogren-Larsson syndrome (SLS) is an autosomal recessive disorder characterized by ichthyosis, mentalretardation, and spastic diplegia or tetraplegia. The disease is caused by mutations in the ALDH3A2 gene (alsoknown as FALDH and ALDH10) on chromosome 17p11.2 that encodes fatty aldehyde dehydrogenase(FALDH), an enzyme that catalyzes the oxidation of long-chain aldehydes derived from lipid metabolism. InSLS patients, 72 mutations have been identified, with a distribution that is scattered throughout the ALDH3A2gene. Most mutations are private but several common mutations have been detected, which probably reflectfounder effects or recurrent mutational events. Missense mutations comprise the most abundant class (38%)and expression studies indicate that most of these result in a profound reduction in enzyme activity. Deletionsaccount for about 25% of the mutations and range from single nucleotides to entire exons. Twelve splice-sitemutations have been demonstrated to cause aberrant splicing in cultured fibroblasts. To date, more than a dozenintragenic ALDH3A2 polymorphisms consisting of SNPs and one microsatellite marker have beencharacterized, although none of them alter the FALDH protein sequence. The striking mutational diversityin SLS offers a challenge for DNA-based diagnosis, but promises to provide a wealth of information aboutenzyme structure–function correlations. Hum Mutat 26(1), 1–10, 2005. rr 2005 Wiley-Liss, Inc.
KEY WORDS: aldehyde dehydrogenase; ALDH3A2; SLS; ichthyosis; mental retardation; spasticity; fatty alcohol; fattyaldehyde
BACKGROUND
In 1957, Sjogren and Larsson described a clinically distinctsyndrome characterized by congenital ichthyosis, mental retardationand spastic diplegia or tetraplegia [Sjogren and Larsson, 1957]. Mostof the patients were from an isolated region of northern Sweden andtheir genetic lineage could be traced back to a small group ofancestors from the early 1700s. The disease exhibited an autosomalrecessive inheritance. Subsequent patients of non-Swedish ancestrywere identified throughout the world and in all ethnic groups.Today, Sjogren-Larsson syndrome (SLS) is the most widelyrecognized neuroichthyotic disorder [Theile, 1974; Rizzo, 2001].
The symptoms of SLS are fully evident in the first year or two oflife. Most patients are born prematurely [Willemsen et al., 1999a].Dry scaly skin (ichthyosis) is usually apparent at birth and has acharacteristic pruritic nature. Delayed motor development due tospastic diplegia or tetraplegia is usually seen by the end of the firstyear, resulting in impaired ambulation. Mental retardation variesfrom mild to profound. Many patients have retinopathy withdistinctive glistening white dots in a perifoveal distribution andsuffer from photophobia. Exceptional patients may lack one of themajor symptoms of SLS or have a very mild phenotype.
SLS patients have altered lipid metabolism due to deficientactivity of fatty aldehyde dehydrogenase (FALDH; E.C. 1.2.1.48)
[Rizzo et al., 1988; Rizzo and Craft, 1991; Rizzo, 2001], an enzymethat catalyzes the NAD-dependent oxidation of aliphaticaldehydes to fatty acids [Kelson et al., 1997]. FALDH is amember of a large family of aldehyde dehydrogenase (ALDH)enzymes that are subclassified according to amino acid sequence,kinetic properties, and subcellular localization [Perozich et al.,1999; Vasiliou et al., 1999]. FALDH is considered a class 3 ALDHand has been historically known as microsomal ALDH in theliterature [Rizzo et al., 2001]. The enzyme has been purified fromhuman liver [Kelson et al., 1997], rat liver [Nakayasu et al., 1978;Lindahl and Evces, 1984; Mitchell and Petersen, 1989], and rabbitintestine [Ichihara et al., 1986b]. The purified human protein is54 kD and, like other class 3 ALDHs, it is probably catalytically
Received 27 September 2004; accepted revised manuscript12 January 2005.
nCorrespondence to: Dr. Rizzo at Department of Pediatrics, Univer-sity of Nebraska Medical Center, 985456 Nebraska Medical Center,Omaha, NE 68198-5456. E-mail: [email protected]
Grant sponsor: Tobacco Settlement Fund of the State of Nebraska;Grant sponsor: National Institutes of Health; Grant number:AR44552.
DOI10.1002/humu.20181Published online inWiley InterScience (www.interscience.wiley.com).
rr2005 WILEY-LISS, INC.
HUMANMUTATION 26(1),1^10,2005
active only as a homodimer. FALDH has the ability to oxidizealiphatic aldehydes that are 6–24 carbons long, but seems to prefer16–20-carbon substrates [Kelson et al., 1997].
SLS is caused by mutations in the gene that encodes FALDH[De Laurenzi et al., 1996]. This gene has recently been renamedALDH3A2 (MIM# 270200), but was formerly known as FALDHand ALDH10. The human ALDH3A2 gene spans 31 kb andconsists of 11 exons, numbered 1–10 with an additional exon(exon 90) situated between exons 9 and 10 [Chang and Yoshida,1997; Rogers et al., 1997]. Alternate splicing of exon 90 resultsin the production of two transcripts, which encode proteinisoforms that differ at their carboxy-terminal domains [Rogerset al., 1997; Lin et al., 2000]. The most abundant transcript isderived from splicing of exons 1–10 and produces a 485–aminoacid protein. A minor transcript that accounts for less than 10% ofthe total FALDH mRNA is produced by splicing of exon 90
between exons 9 and 10, and gives rise to a variant protein isoform(FALDHv) of 508 amino acids. Expression studies indicate thatboth isoforms of FALDH have enzymatic activity (W.B.R.,unpublished results).
In addition to the two transcripts derived from alternatesplicing, Northern analysis has shown the existence of three sizesof transcripts (2.0, 3.8, and 4 kb) that arise from utilization ofdistinct polyadenylation sites in the 30 untranslated region [Rogerset al., 1997]. The longer transcripts (3.8 and 4 kb) are moreabundant in brain, heart, skeletal muscle, and pancreas, whereasthe liver has an excess of the shorter transcript.
The FALDH promoter lacks a TATA box and has multiple CpG
islands [Chang and Yoshida, 1997; Rogers et al., 1997]. The
transcription start site is at nucleotide –258 in relation to the
translation initiation codon and there is a functional Sp1 binding
site 51 nucleotides further upstream [Rogers et al., 1997]. An
alternate transcription start site may exist at nucleotide –195
[Chang and Yoshida, 1997].Based on Northern analysis of human [Chang and Yoshida,
1997; Rogers et al., 1997] and rodent [Miyauchi et al., 1991; Lin
et al., 2000] species, ALDH3A2 is expressed in a variety of
tissues, particularly liver and intestine. FALDH enzyme activity
correlates with the amount of ALDH3A2 transcript. In mice,
the minor transcript encoding FALDHv generally mirrors the
amount of the major transcript, but it is slightly more abundant
in brain and testes compared to other tissues [Lin et al., 2000].
Ironically, although symptoms of SLS chiefly affect the skin and
nervous system, FALDH enzyme activity is not particularly high in
these two tissues compared to liver, intestine, and kidney.Density gradient centrifugation studies establish that FALDH is
localized to the microsomal fraction in human liver [Kelson et al.,
1997]. The rat FALDH protein is synthesized on free polysomes
and then inserted into the endoplasmic reticulum [Takagi et al.,
1985]. Deletion mutagenesis experiments reveal that the carboxy-
terminal 35 amino acids define a hydrophobic domain that is
necessary for anchoring the enzyme to the microsomal membrane
[Masaki et al., 1994]. In humans and mice, the carboxy-terminal
domains of both protein isoforms derived from alternate splicing
are enriched in hydrophobic amino acid residues, suggesting that
both forms of the protein are membrane-bound.FALDH gene expression is subject to physiologic regulation
in vivo [Demozay et al., 2004]. Gene expression is increased inliver and white adipose tissue of rats treated with insulin, whereasdiabetic mice exhibit a reduced expression of the FALDH genecompared to nondiabetic animals. Moreover, gene expression inrats is increased after feeding compared to fasted animals. In
addition, ALDH3A2 mRNA levels are increased in mouse liver bytreatment with clofibrate, a hypolipidemic drug that interacts withthe peroxisome proliferator-activated receptor [Vasiliou et al.,1996].
MUTATIONSAND POLYMORPHISMS
To date, 72 mutations have been identified in SLS probandsfrom 121 families (Table 1). These include missense and nonsensemutations, deletions and insertions, splicing defects, and severalcomplex mutations with combinations of nucleotide substitutions,deletions, insertions, or duplications. As shown in Figure 1, themutations are distributed throughout the gene and involve allexons, except for exons 90 and 10. A total of 41 of the mutations(56%) are single nucleotide substitutions. Of these, 27 aretransitions and 14 are transversions. Seven nucleotide substitu-tions occur at CpG dinucleotides.
Most mutations appear to be private and 55% of the SLSprobands are homozygous, probably reflecting founder effects orinbreeding. At least one mutant allele was identified in all SLSpatients investigated, but three patients, all compound hetero-zygotes of European descent, possessed one allele with nodetectable mutation in the protein coding region [Tsukamotoet al., 1997; Sillen et al., 1998b; Rizzo et al., 1999; Carney et al.,2004].
It is notable that no benign coding change has been identifiedin the ALDH3A2 gene, perhaps due to ascertainment bias inselecting for SLS patients who had previously been shown to haveFALDH deficiency. Some SLS patients, however, have beeninvestigated by DNA sequencing alone, without prior enzymaticconfirmation of the diagnosis. Except for limited screening ofspecific mutations and polymorphisms, no systematic population-based analysis of the gene has been performed to search for aminoacid changes in normal individuals.
Missense Mutations
Missense mutations account for 38% of the known mutations inALDH3A2 and are scattered throughout the gene (Fig. 1). A totalof 28 mutations have been found to alter 27 amino acid residues,with two mutations (c.551C4G and c.551C4T) changing thesame T184 residue to arginine and methionine, respectively. Atotal of 21 missense mutations have been expressed in hetero-logous systems using Chinese hamster ovary cells [Rizzo et al.,1999; Carney et al., 2004] or baculovirus-infected insect cells [DeLaurenzi et al., 1997]. With one exception (c.798G4C; p.K266N,see below), the mutant enzymes were found to possess little or nocatalytic activity (Table 1).
Missense mutations in ALDH3A2 pinpoint critical amino acidresidues necessary for FALDH catalytic activity. The missensemutations identified to date cover only 7% of the amino acidresidues in the major isoform of FALDH. Although the three-dimensional protein structure of FALDH has not been elucidated,a related rat cytosolic class 3 ALDH has been crystallized and itsstructure solved [Liu et al., 1997]. This rat protein shares 65%amino acid identity with human FALDH, but lacks thehydrophobic carboxy-terminal domain seen in the human protein.Nevertheless, it is highly likely that the basic structure of humanFALDH closely resembles that of the rat protein. Notably, theamino acid residues mutated in SLS patients share 89% identitywith the rat protein, indicating that they tend to be more highlyconserved across species and perhaps less tolerant of change withrespect to enzyme function compared to the protein at large [Rizzo
2 HUMANMUTATION 26(1),1^10,2005
TABLE
1.Mutations
intheALDH3A2Gen
eDetec
tedin
SLS
Type
ofmutation
Intron/exo
nNuc
leotide
chan
gea
Pro
tein
ortran
scriptc
han
ged,e
ALD
H3A2
hap
lotype
fReferen
ce
Misse
nse
Ex1
c.80C4T
p.L27
PWillemse
net
al.[20
01a]
Ex1
c.13
3A4T
p.I45F(9%
residu
alen
zymeac
tivity)
1Rizzo
etal.[19
99]
Ex2
c.19
1T4A
p.V64D
(o1%
residu
alen
zymeac
tivity)
3Rizzo
etal.[19
99]
Ex2
c.31
7T4G
p.L1
06R(o
1%
residu
alen
zymeac
tivity)
3Sillenet
al.[19
98b];Rizzo
etal.[19
99]
Ex2
c.34
1C4T
p.P114L(o
1%
residu
alen
zymeac
tivity)
1Rizzo
etal.[19
99]
Ex2
c.36
2C4T
p.P12
1L(o
1%
residu
alen
zymeac
tivity)
3Rizzo
etal.[19
99]
Ex4
c.55
1C4G
p.T18
4R(o
1%
residu
alen
zymeac
tivity)
2Rizzo
etal.[19
99]
Ex4
c.55
1C4Tb
p.T18
4M
(o1%
residu
alen
zymeac
tivity)
1,2
Rizzo
etal.[19
99];W
illemse
net
al.[20
01a]
Ex4
c.55
4G4C
p.G18
5A(1%
residu
alen
zymeac
tivity)
1Rizzo
etal.[19
99]
Ex4
c.641
G4A
p.C21
4Y
1DeLau
renz
ieta
l.[1996]
Ex4
c.67
8C4G
p.C22
6W
Sillenet
al.[19
98b]
Ex5
c.682
C4Tb
p.R22
8C
(9%
residu
alen
zymeac
tivity)
2,3
Rizzo
etal.[19
99]
Ex5
c.71
0G4A
p.C23
7Y(o
1%
residu
alen
zymeac
tivity)
1Rizzo
etal.[19
99]
Ex5
c.73
3G4Ab
p.D24
5N
(1%
residu
alen
zymeac
tivity)
Sillenet
al.[19
98b];Rizzo
etal.[19
99];
Willemse
net
al.[20
01a]
Ex5
c.79
8G4C
p.K26
6N
(55%
residu
alen
zymeac
tivity);
unstab
letran
script
2Rizzo
etal.[19
99];W
illemse
net
al.[20
01a]
Ex6
c.83
5T4A
p.Y27
9N
(o1%
residu
alen
zymeac
tivity)
3Rizzo
etal.[19
99];K
raus
etal.[20
00]
Ex7
c.94
3C4T
p.P31
5S(o
1%
residu
alen
zymeac
tivity)
1DeLau
renz
ieta
l.[1997
];Sillenet
al.[19
97b];
Rizzo
etal.[19
99];IJlst
etal.[19
99]
Ex7
c.98
4G4C
p.M32
811
Rizzo
etal.[19
99]
Ex7
c.10
94C4Tb
p.S365
L(3%
residu
alen
zymeac
tivity)
1,2
Sillenet
al.[19
98b];Rizzo
etal.[19
99]
Ex8
c.11
39G4A
p.S380
N(o
1%
residu
alen
zymeac
tivity)
3Carne
yet
al.[20
04]
Ex8
c.11
57A4G
p.N386S
Aoki
etal.[20
00]
Ex9
c.12
16G4A
p.G40
6R
2Rizzo
etal.[19
99]
Ex9
c.12
31C4T
p.H41
1Y(o
1%
residu
alen
zymeac
tivity)
4Rizzo
etal.[19
99]
Ex9
c.12
34G4Ab
p.G41
2R
Sillenet
al.[19
98b]
Ex9
c.12
44G4A
p.S41
5N
(o1%
residu
alen
zymeac
tivity)
2Rizzo
etal.[19
99]
Ex9
c.12
56T4C
p.F41
9S(3%
residu
alen
zymeac
tivity)
4Rizzo
etal.[19
99]
Ex9
c.12
68G4Ab
p.R42
3H
(1%
residu
alen
zymeac
tivity)
3Rizzo
etal.[19
99]
Ex9
c.13
39A4G
p.K447
E(1%
residu
alen
zymeac
tivity)
1Rizzo
etal.[19
99]
Deletion
Ex1
c.21
_ 46de
lp.R7fs
Willemse
net
al.[20
01a]
Ex1
c.10
3de
lCp.Q35
fsSillenet
al.[19
98b]
Ex2
c.28
6_ 2
96de
lp.Y96fs
3Rizzo
etal.[19
99]
Ex2
c.33
8de
lAp.Y11
3fs
3Rizzo
etal.[19
99]
Ex2
c.37
4_ 378
del
p.A12
5fs
2Rizzo
etal.[19
99]
Ex3
c.469
delC
p.Q15
7fs
1Rizzo
etal.[19
99]
Ex4
c.52
1delT
p.L1
74fs
1DeLau
renz
ieta
l.[1996]
Ex6
c.80
8de
lGp.G27
0fs
2DeLau
renz
ieta
l.[1996]
Ex6
c.82
1_ 822
delA
Ap.K27
4fs
3Rizzo
etal.[19
99]
Ex6
c.906de
lTp.F30
2fs
Kraus
etal.[20
00]
Ex7
c.968de
lCp.P32
3fs
1Rizzo
etal.[19
99]
Ex7
c.11
00de
lAp.N367
fs3
Rizzo
etal.[19
99]
Ex9
c.12
23de
lGp.G40
8fs
1Carne
yet
al.[20
04]
Ex9
c.12
91_ 129
2de
lAA
p.K43
1fs
2Rizzo
etal.[19
99]
Ex9
c.12
97_ 129
8de
lGA
p.E43
3fs
(o1%
residu
alen
zymeac
tivity)
1Tsuk
amoto
etal.[19
97];Rizzo
etal.[19
97];Sillenet
al.
[1998
b];IJlste
tal.[1999];L
inet
al.(un
publis
hed
)Ex9
c.13
84_ 1387
delG
AAA
p.E462
fsWillemse
net
al.[19
99b];Willemse
net
al.[20
01a]
Ex9
c.Exo
n9de
letion
p.G40
3fs
Sillenet
al.[19
98b];Kraus
etal.[20
00]
HUMANMUTATION 26(1),1^10,2005 3
Insertionan
ddu
plic
ation
Ex2
c.28
9_ 354du
pc
p.L1
22Qins2
21
Carne
yet
al.[20
04]
Ex4
c.487
_ 488insA
p.I163fs
Willemse
net
al.[20
01a]
Ex4
c.62
0_ 6
21insG
p.G20
7fs
3Rizzo
etal.[19
99]
Ex6
c.865
_ 866insT
p.K28
9fs
Sillenet
al.[19
98b]
Ex9
c.12
23_ 122
4insG
p.G40
8fs
4Carne
yet
al.[20
04]
Ex9
c.12
50_ 126
5du
pp.Q42
2fs
1Carne
yet
al.[20
04]
Ex9
c.13
07_ 131
1dup
ACAAA
p.L43
8fs
Tsuk
amoto
etal.[19
97];Rizzo
etal.[19
99]
Com
plex
Int1
-Int2
c.15
3+5_ 3
87-4
08de
lins19
p.S52
_A12
8de
lfsRizzo
etal.[19
99]
Ex5+Ex6
[c.733
G4A;c.901
G4C;
c.906de
lT;c.909T4G]
[p.D
245N
(+)p
.A30
1P(+
)p.F
302fs]
2Sillenet
al.[19
98b]
Ex7
c.94
1_ 943
delC
CCins2
1p.A31
4_P31
5de
lins
GAKSTVGA
1DeLau
renziet
al.[19
96];Tsu
kamoto
etal.[19
97];
Sillenet
al.[19
98b]
Ex8
c.11
08_ 111
6de
linsA
CAG
p.L37
0fs
3Rizzo
etal.[19
99]
Nons
ense
Ex1
c.24
-25CC4TT
p.R9X
1Rizzo
etal.[19
99]
Ex1
c.28
C4T
p.Q10
X3
Rizzo
etal.[19
99]
Ex2
c.23
3G4A
p.W
78X
3Rizzo
etal.[19
99]
Ex4
c.52
9C4Tb
p.R17
7X
2Carne
yet
al.[20
04]
Ex9
c.12
77T4G
p.L42
6X
Kraus
etal.[20
00]
Splic
e-site
Int2
c.38
5+2T4C
Skipsex
on2-
r.154_ 3
85de
land
p.S52
_ 129
delfs
Kraus
etal.[20
00]
Int3
c.47
1+1d
elG
Skipsex
ons
2an
d3-
r.154_ 471
dela
ndp.52
_ 157
del
2Rizzo
etal.[19
99];K
raus
etal.[20
00]
Int3
c.47
1+1G
4C
Skipsex
ons
2an
d3-
r.154_ 471
dela
ndp.52
_ 157
del
Kraus
etal.[20
00]
Int3
c.47
1+2T4G
Skipsex
ons
2an
d3-
r.154_ 471
dela
ndp.52
_ 157
del
3Rizzo
etal.[19
99]
Int3
c.47
2-2
A4G
Skipsinitial3
3bp
inex
on4an
dutilize
sdo
wnstream
cryp
ticac
ceptors
itein
exon
4-
r.472
_ 504
dela
ndp.15
8_ 168de
l
3Rizzo
etal.[19
99]
Int4
c.681
-2A4G
Skipsex
ons
5-
r.681
_ 798
dela
ndp.R22
7fs
3Rizzo
etal.[19
99]
Int4
c.681
-14T4A
Cau
sesutiliza
tionof
ane
wcryp
ticac
ceptor
site
inintron4ad
ding24
bpbe
twee
nex
ons
4an
d5-
r.681
-14u4a,
680
_ 681
ins2
4an
dp.C22
26_R22
7ins8
1Rizzo
etal.[19
99]
Skipsex
on5-
r.681
_ 798
dela
ndp.R22
7fs
Int5
c.79
8+5G4A
Skipsex
on5-
r.681
_ 798
dela
ndp.R22
7fs
Skipsex
ons
4an
d5-
r.472
_ 798
dela
ndp.D15
8_K26
6de
l
3Rizzo
etal.[19
99]
Int5
c.79
8+1de
lGSkipsex
on5-
r.681
_ 798
dela
ndp.R22
7fs
Skipsex
ons
4an
d5-
r.472
_ 798
dela
ndp.D15
8_K26
6de
l
1Rizzo
etal.[19
99]
Int5
c.79
8+1_ 798+6de
lGTTTGT
Skipsex
on5-
r.681
_ 798
dela
ndp.R22
7fs
Skipsex
ons
4an
d5-
r.472
_ 798
dela
ndp.D15
8_K26
6de
l
1Rizzo
etal.[19
99]
Int7
c.11
07+2T4G
Skipsex
on7-
r.941
_ 110
7de
land
p.A31
4fs
Skipsex
ons
6-8
-r.7
99_ 120
7de
land
p.E26
7_G40
3de
lfs
3Carne
yet
al.[20
04]
Minor
amou
nto
fnor
mal
tran
script-
r.=
TABLE
1.(continu
ed)
Typeof
mutation
Intron/exo
nNuc
leotide
chan
gea
Protein
ortrans
criptc
han
ged,e
ALD
H3A2
hap
lotype
fReferen
ce
4 HUMANMUTATION 26(1),1^10,2005
et al., 2001]. In SLS patients, no amino acid substitutions havebeen found in the hydrophobic membrane-binding domain at thecarboxy-terminus of the FALDH protein.
Comparison of the human FALDH protein with the rat class 3ALDH crystal structure provides insight into the effects of aminoacid substitutions on protein structure and catalytic function. Forexample, the c.733G4A (p.D245N) mutation replaces asparticacid 245 with asparagine and results in a protein with 1% ofnormal catalytic activity [Rizzo et al., 1999]. Aspartic acid 245 inhuman FALDH is equivalent to aspartic acid 247 in thehomologous rat class 3 protein [Liu et al., 1997]. In the ratenzyme, aspartic acid 247 is important for maintaining thegeometric structure of the active site by forming key hydrogenbonds with lysine 213 in the catalytic domain and isoleucine 334on the protein chain. Replacement of aspartic acid 247 withasparagine eliminates the hydrogen bonds and results in a mutantenzyme with a profound reduction in the Vmax and a minimalchange in Km [Hempel et al., 2001].
One mutation in ALDH3A2, while ostensibly directing anamino acid replacement, may have pleiotropic effects on geneexpression. The c.798G4C mutation alters the last nucleotide inexon 5 and is predicted to result in a p.K266N replacement in theFALDH protein; however, the mutant protein possesses consider-able (55%) residual catalytic activity when expressed in Chinesehamster cells [Rizzo et al., 1999]. The SLS patient who carried thismutation was a compound heterozygote and had a second allelecontaining the c.943C4T (p.P315S) missense mutation. Uponamplification of the patient’s fibroblast mRNA, only thec.943C4T transcript was detectable, suggesting that the transcriptbearing the c.798G4C mutation was poorly expressed orrapidly degraded, presumably due to aberrant splicing. Thus, it islikely that the c.798G4C mutation is not expressed at the proteinlevel.
The c.943C4T mutation has historical significance because itis the cause of SLS in most Swedish patients originally described bySjogren and Larsson [De Laurenzi et al., 1997; Sillen et al.,1997b]. This mutation is also carried by many northern Europeanpatients, who share an identical ALDH3A2 haplotype with theSwedish patients. These findings are consistent with a foundereffect and prompted speculation that the c.943C4T allele wasspread to the northern European continent by migration of theVikings during the 9th and 10th centuries [De Laurenzi et al.,1997].
Deletions
Approximately one-fourth of the mutations in the ALDH3A2gene are deletions (Table 1). Three deletions are at splice-sitesand are listed as splicing mutations in Table 1. The deletions rangein size from a single nucleotide to an entire exon and aredistributed throughout the gene (Fig. 1). Most deletions haveconsisted of one or two nucleotides only. The larger deletions tendto be more complicated, involving both deletion and insertion ofnucleotides, and are therefore classified as complex mutations inTable 1.
Two deletions in exon 9, c.1291_1292delAA (p.K431fs) andc.1297_1298delGA (p.E433fs), involve the same 9-bp palindromicsequence (AAGAGAGAA). The c.1297_1298delGA deletion isthe most common mutation seen among SLS patients of Europeanheritage, accounting for 26% of the mutant alleles in one series of21 probands [Rizzo et al., 1997].
The largest deletion in the ALDH3A2 gene is 6 kb and resultsin complete loss of exon 9 [Sillen et al., 1998b; Kraus et al., 2000].
Int7
c.11
08-1G4C
Skipsex
on8-
r.110
8_ 120
7de
land
p.L37
0_G40
3fs
1Rizzo
etal.[19
99]
Skipsex
ons
8-9
-r.110
8_ 1443
dela
ndp.L37
0_K48de
laMutations
arenam
edus
ingthecD
NAse
quen
celis
tedin
Gen
Ban
k(acc
essionNM_00
0382
.1)inwhichtheAof
theinitiatorA
TG
codo
nis
design
ated
+1.
b Trans
itionmutationat
CpG
dinu
cleo
tide
.c Thismutationwas
errone
ouslynam
edas
c.29
9_ 3
65du
pbyCarne
yet
al.[20
04].
dAllprotein
chan
geswerede
duce
dfrom
thenu
cleo
tide
chan
ges,wherea
sch
ange
sin
thetran
scripta
risingfrom
splic
e-site
mutations
wereex
perim
entallyde
term
ined
.e N
umbe
rsin
paren
thes
esindica
tetheresidu
alen
zymeca
talyticac
tivity
ofm
utan
tFALDH
protein
expressed
inChines
eham
ster
ovaryce
llsas
ape
rcen
tage
ofthat
seen
withwild-typeFA
LDH
protein.
f Hap
lotypes
wereco
nstructed
using4
intrag
enic
SNPs(see
Table
2)an
darede
¢ne
das
follow
s:hap
lotype1:
[c.153
+1T
;c.47
1+31
C;c.94
0+53
C;c.14
46T];
hap
lotype2:
[c.153
+1C
;c.47
1+31
T;c.94
0+53
G;
c.14
46T];hap
lotype
3:[c.153
+1C
;c.471
+31
T;c.94
0+53
C;c
.1446A];hap
lotype
4:[c.153
+1C
;c.471
+31
T;c.94
0+53
C;c
.1446T].
TABLE
1.(continu
ed)
Typeof
mutation
Intron/exo
nNuc
leotide
chan
gea
Protein
ortrans
criptc
han
ged,e
ALD
H3A2
hap
lotype
fReferen
ce
HUMAN MUTATION 26(1),1^10,2005 5
Kraus et al. [2000] sequenced the deletion breakpoints andidentified an identical 8-bp flanking sequence at the deletionjunctions in intron 8 and intron 9, suggesting a slippage mispairingmechanism for the deletion. The mutation accounted for 5 of 18alleles in the German SLS patients studied.
Insertions and Duplications
Four single-nucleotide insertion mutations have been found inALDH3A2 (Table 1). Three additional duplication mutations,ranging from five nucleotides to 66 nucleotides in length, havebeen identified. Most of these mutations cause frameshifts and arepredicted to result in truncated proteins. The largest duplication,c.289-354dup (p.L122Qins22) is predicted to cause an amino acidsubstitution along with an in-frame insertion of 22 amino acids inthe FALDH protein [Carney et al., 2004].
Complex Mutations
Three complex mutations in ALDH3A2 involve combinationsof deletions and insertions. A common mutation seen in SLSpatients originating from the Mideast is c.941_943delCCCins21(p.A314_P315delinsGAKSTVGA). At least four of these osten-sibly unrelated patients have an identical intragenic haplotype,suggesting that the mutation arose in a shared ancestor [Rizzoet al., 1999].
A consanguineous Turkish patient carried a large deletionspanning more than 2 kb involving intron 1, exon 2, and a portionof intron 2 [Rizzo et al., 1999]. Sequencing across the deletionidentified the precise nucleotide breakpoints (c.153+5_387-408delins19). The downstream breakpoint in intron 2 was foundto include a 19-bp insertion that included a 10-bp duplicationfrom intron 2. A third deletion/insertion mutation(c.1108_1116delinsACAG) (p.L370fs) was identified in an Italianpatient.
An unusual ALDH3A2 allele carried by several SLS patients ofEuropean ancestry consists of four distinct nucleotide changes:c.733G4A; c.901G4C; c.906delT; and c.909T4G [Sillen et al.,1998b; Rizzo et al., 1999; Kraus et al., 2000]. Two of thesenucleotide changes (c.733G4A and c.906delT) are destructive toenzyme function and have also been found as isolated mutationsin different patients.
Splicing Defects
Twelve (16%) of the ALDH3A2 mutations involve splice sites.All have been investigated at the mRNA level and found todisrupt normal splicing [Rizzo et al., 1999; Kraus et al., 2000]. Allof the splicing mutations are predicted to have drastic effects onthe FALDH protein by deleting amino acids, inserting additionalresidues or resulting in a truncated protein.
Splice donor mutations have generally resulted in skipping oneor two upstream exons (Table 1). In contrast, the splice acceptormutations have tended to show more variable splicing errors. Onepatient carrying the c.681–14T4A mutation near the acceptorsplice site at exon 5 was found to utilize a cryptic acceptor sitefurther upstream in intron 4 [Rizzo et al., 1999]. This resulted inthe addition of 24 bp to the transcript and is predicted to cause anin-frame insertion of eight amino acids in the FALDH protein. Inaddition, a second transcript lacked exon 5. The splicing errorsfrom this mutation were not complete, since a very small amountof normal FALDH transcript was also produced. A secondmutation (c.472–2A4G) alters the consensus AG of the splice-acceptor site of exon 4 and mRNA analysis revealed the utilizationof the next available downstream AG as a cryptic acceptor site.This resulted in an in-frame deletion of the first 33 nucleotideswithin exon 4 of the mRNA and a predicted loss of 11 amino acidsfrom the FALDH protein.
SequenceVariations of Uncertain Signi¢cance
Rizzo et al. [1999] reported an SLS patient who carried thec.1268G4A (p.R423H) missense mutation in exon 9 togetherwith a single-nucleotide variation (c.1494G4A) 36 bp down-stream from the stop codon in the 30-untranslated region of exon10. No other nucleotide changes were detected by amplifyingexons from genomic DNA. When the patient’s fibroblast mRNAwas amplified by RT-PCR and sequenced, only the c.1268G4Aallele was seen, suggesting that the transcript bearing thec.1494G4A sequence variation was not expressed or was unstableand rapidly degraded. Expression studies were not performed andother family members were not examined to establish thebiological significance of the c.1494G4A change. It is unclearwhether the c.1494G4A change is a rare polymorphism, since ithas been subsequently detected in one SLS patient carrying the
FIGURE 1. Location of mutations in the ALDH3A2 gene. Exons are indicated with numbered rectangles and are drawn to scale.Introns are indicated by horizontal lines and are not drawn to scale. Only coding regions of exons 1 and 10 are shown, since nocon¢rmed disease mutations have been identi¢ed in 50 and 30 untranslated regions. Nucleotides are numbered according to thecDNA sequence listed in GenBank (accession number NM_000382.1) in which the A of the initiator codon is designated +1.Mutations shown above the gene are missense and nonsense mutations, whereas those below the gene are deletions, insertions,splicing defects, and complex changes. Nonsensemutations are indicated by an asterisk. Splice-sitemutations are underlined.
6 HUMANMUTATION 26(1),1^10,2005
TABLE
2.Polymor
phismsin
theALDH3A2ge
nen
Exo
n/intron
loca
tion
Type
ofpolymor
phism
Nuc
leotide
pos
itiona
(cDNA)
Nuc
leotide
pos
itionb
(gen
omic)
Alle
lefreq
uenc
yReferen
cec
Intron1
SNP
c.15
3+39
C4T
g.41
2762
C4T
C:0
.71
Rizzo
etal.[19
99];refSNPrs46467
93T:
0.29
Intron2
SNP
c.38
6-135
A4T
g.41
601
1A4T
A:0
.588
refSNPrs10
0449
0T:
0.41
2Intron3
SNP
c.47
1-31
T4C
g.41
6262
T4C
T:0.73
Rizzo
etal.[19
99];refSNPrs10
0449
C:0
.27
Intron3
SNP
c.47
2-5
84C4T
g.41
9381
C4T
C:0
.577
refSNPrs962
801d
T:0.42
3Intron3
SNP
c.47
2-308
G4A
g.41
965
7G4A
G:0
.528
refSNPrs20
7233
3d
A:0
.472
Intron3
SNP
c.47
2-211
A4G
g.41
9754A4G
A:0
.275
refSNPrs962
800d
G:0
.725
Intron3
SNP
c.47
2-5
6G4A
g.41
9909G4A
G:0
.572
refSNPrs20
7233
2d
A:0
.428
Intron4
SNP
c.681
-206A4G
g.42
1138
A4G
A:0
.703
refSNPrs10
3489
7d
G:0
.297
Intron6
SNP
c.94
0+53
C4G
g.42
4920
C4G
C:0
.84
Rizzo
etal.[19
99];refSNPrs18
00869
G:0
.16
Intron6
SNP
c.94
1-701
G4A
g.42
6231
G4A
C:0
.442
refSNPrs21
0897
1d
A:0
.558
Intron8
Microsa
tellite
TG
repe
atc.12
07+33
1(TG)15_ 24
g.42
8977
(TG)15_ 24
10alleles;
freq
uenc
iesno
trepor
ted
DeLau
renz
ieta
l.[1997
];Gen
Ban
kZ67
319
(AFMa1
26yd
5)
Intron8
SNP
c.12
07+19
06A4G
g.43
0606A4G
A:0
.442
refSNPrs80
6957
6G:0
.558
Exo
n10
SNP
c.14
46A4T
g.43
9159
A4T
A:0
.51;
0.29
4Chan
gan
dYo
shida[1997
];Sillenet
al.[19
98b];
Rizzo
etal.[19
99];refSNPrs72
16T:
0.49
;0.706
Exo
n10
/30 -U
TR
SNP
c.14
58+16
09A4G
g.440
780A4G
A:0
.340
refSNPrs72
15G:0
.660
Exo
n10
/30 -U
TR
Insertion/deletion
c.14
58+18
76_ 145
8+18
77de
lAC
g.441
047_ 441
048de
lAC
AC:0
.669
refSNPrs16
420
delAC:0
.331
nOnly
those
polymor
phismsarelis
tedforw
hichallele
freq
uenc
ieshav
ebe
ende
rive
dfrom
more
than
100ch
romoso
mes.
acD
NAnu
cleo
tide
numbe
ringsystem
usingGen
Ban
kac
cessionNM_00
0382
.1.
bGen
omic
DNAnu
mberingsystem
derive
dfrom
Gen
Ban
kco
ntigac
cessionNT_03
084
3.c R
eferen
cesareto
pub
lished
literature,
Gen
Ban
kac
cessionnu
mbe
r,an
d/orp
olymor
phismslis
tedin
thedb
SNPda
tabas
eat
www.ncb
i.nlm
.gov
/SNP.
d This
SNPin
thedb
SNPda
tabas
eisreported
usingtheminus
(non-co
ding)D
NAstrand.
Forc
ons
istenc
yin
thetable,these
quen
ceva
riations
arelis
tedforthepositive
strandonly
asreported
inNT_03
0843
.SNP,
singlenu
cleo
tide
polymorp
hism.
HUMANMUTATION 26(1),1^10,2005 7
disease-causing c.1094C4T (p.S365L) mutation (W.B.R., G.C.unpublished observations).
To date, no nucleotide variations have been reported in thepromoter region of ALDH3A2. The mutation detection strategiestaken by most groups, however, have not examined more thanseveral dozen nucleotides upstream of the coding region in exon 1.
Polymorphisms and ALDH3A2 Haplotypes
At least 15 polymorphisms have been identified within theALDH3A2 gene (Table 2). Most of the nucleotide changes areSNPs, but one polymorphism is a microsatellite TG repeat. All ofthe polymorphisms are located in noncoding regions, with theexception of one SNP (c.1446A4T) in exon 10 that alters thethird nucleotide of codon 481 and results in no amino acid change.Allele frequencies of the most characterized SNPs are listed inTable 2. None of the polymorphisms are rare.
Early haplotype studies relied on flanking microsatellite markersto localize the SLS gene and demonstrate genetic relationshipsbetween SLS patients [Pigg et al., 1994; De Laurenzi et al., 1997;Sillen et al., 1997b, 1998a, 1998b]. With the identification ofintragenic polymorphisms, haplotype analysis of the ALDH3A2gene has become more refined. Using haplotypes generated fromfour SNPs, haplotype associations were established in most of themutations reported (see Table 1) [Rizzo et al., 1999; Carney et al.,2004]. Four mutations are each found on multiple haplotypes,suggesting that they arose independently. Three of these mutations(c.551C4T, c.682C4T, and c.733G4A) occur at CpG dinucleo-tides and may represent mutational hot spots. From a practicalpoint of view, intragenic haplotypes may prove useful for diagnosticpurposes and genetic counseling in those rare SLS families inwhich mutation analysis has been unable to identify a mutantallele. For example, the ALDH3A2 allele(s) with no identifiedmutation carried by three European patients mentioned above isassociated with haplotype 1 (see Table 1 for haplotype definition).
BIOLOGICAL RELEVANCE
The study of SLS patients who carry mutations in theALDH3A2 gene has been instrumental in establishing thebiological function of FALDH. This enzyme plays a pivotal rolein metabolism of several lipids, including fatty alcohol, etherglycerolipids, phytanic acid, and leukotriene B4.
In vitro metabolic studies of SLS fibroblasts, together with thedemonstration that SLS patients accumulate hexadecanol andoctadecanol in plasma [Rizzo et al., 1988; Rizzo and Craft, 2000],have provided conclusive evidence that FALDH is responsible foroxidizing long-chain aldehydes derived from metabolism of 16–18-carbon straight-chain fatty alcohols. In this role, FALDH appearsto interact closely with fatty alcohol dehydrogenase and forms anenzyme complex, called fatty alcohol:NAD+ oxidoreductase, thatcatalyzes the complete oxidation of fatty alcohol to fatty acid [Lee,1979; Ichihara et al., 1986a].
In vitro studies of SLS fibroblasts and FALDH-deficientChinese hamster cells also implicate FALDH in the oxidation offatty aldehydes derived from metabolism of ether glycerolipids[Rizzo et al., 2000]. These lipids have an ether-linked alkyl chainattached to the first carbon of a glycerol backbone, and areprominent in certain tissues, such as in myelin, skin and heart.Metabolism of ether glycerolipids proceeds by catalytic hydrolysisof the ether-linked alkyl group released as a fatty aldehyde, whichis subsequently oxidized to fatty acid by FALDH.
There is evidence to suggest that FALDH is involved in twoseparate steps in the metabolism of phytol and phytanic acid. Phytol
is a branched-chain fatty alcohol that is derived from the diet. It ismetabolized by initial oxidation to phytanic acid, a reaction that isimpaired in cultured skin fibroblasts from SLS patients [van denBrink et al., 2004]. Phytanic acid is subsequently metabolized byremoval of the carboxyl group via alpha-oxidation. This processgenerates a fatty aldehyde (pristanal), which is subsequentlyoxidized to pristanic acid. In vitro studies on cultured SLSfibroblasts incubated with radioactive phytanic acid have demon-strated a deficiency of this oxidative step, implicating FALDH inthis second point in the pathway [Verhoeven et al., 1998].
SLS patients accumulate leukotriene B4 (LTB4), a proinflam-matory lipid mediator synthesized from arachidonic acid (a 20-carbon polyunsaturated fatty acid). LTB4 is metabolized by initialomega-hydroxylation forming 20-hydroxy-LTB4, followed byoxidation of the omega-hydroxy group to form a dicarboxylic acidin a reaction that is analogous to straight-chain fatty alcoholoxidation. Metabolism of LTB4 has been shown to be defective inleukocytes of SLS patients [Willemsen et al., 2001c]. Accumula-tion of LTB4 and/or 20-hydroxy-LTB4 appears to be responsible inpart for the pruritic nature of the ichthyosis in SLS, becausetreatment of patients with a selective inhibitor of LTB4 synthesis(Zileuton) reverses its accumulation and ameliorates the itchinessof the skin [Willemsen et al., 2001b]. This drug, however, does notimprove the ichthyosis itself or the neurologic symptoms.
It is hypothesized that the accumulation of fatty alcohol and/orformation of aldehyde adducts in biological membranes contributeto the pathogenesis of SLS in ways that are not yet defined [Rizzo,2001]. Fatty alcohol accumulation in the skin may disrupt theepidermal water barrier resulting in ichthyosis. Owing to theirhighly reactive nature, the fatty aldehyde substrates of FALDH,which cannot be metabolized to fatty acids, have the potential toform covalent Schiff base adducts with other molecules containingfree amino groups. One prominent lipid target for aldehydemodification is phosphatidylethanolamine (PE), forming N-alkyl-PE [James and Zoeller, 1997; Verhoeven et al., 1998]. Thiscovalent modification changes the polar nature of the PE moleculeand may affect membrane structure and function. Aliphaticaldehydes also have the potential to covalently modify lysineresidues in proteins, including membrane-bound enzymes thatmight exhibit altered catalytic properties.
CLINICAL RELEVANCE
Most SLS patients have the major symptoms (ichthyosis, mentalretardation, and spasticity) to more or less varying degrees. Owingto the large proportion of private mutations in SLS and manypedigrees with single cases, the influence of ALDH3A2 genotypeon clinical phenotype is difficult to judge. Furthermore, many ofthe published cases have little or no clinical information. Despitethe limited genotype–phenotype information available so far,however, it would appear that the severity of the clinicalphenotype does not closely correlate with specific mutations orresidual enzyme activity [Willemsen et al., 2001a]. Even amongsiblings who share ALDH3A2 genotypes, considerable clinicalvariation can occur. For example, one atypical SLS family hadthree affected siblings (7, 12, and 14 years of age) with no mentalretardation and quite divergent spasticity and cutaneous disease,including one sibling without ichthyosis [Nigro et al., 1996]. Thesepatients are compound heterozygotes for the commonc.1297_1298delGA mutation and a unique c.984G4C(p.M328I) mutation. In another kindred, a pair of adult siblingswho are compound heterozygotes for c.943C4T and c.551C4Texhibited ichthyosis. The male sibling had late-onset spasticity and
8 HUMANMUTATION 26(1),1^10,2005
was feeble-minded at 36 years of age, whereas his 46-year-old sisterhad normal intellectual development and no spasticity [Willemsenet al., 2001a]. These unusual families highlight the striking clinicalvariation that can occur between SLS siblings and suggest theexistence of other genetic or environmental modifiers.
DIAGNOSTIC RELEVANCE
The ability to diagnose SLS by mutation analysis offers analternative to enzymatic testing. The FALDH enzymatic test is notwidely available and diagnostic centers rely on cultured skinfibroblasts as the source of enzyme, which requires an invasive skinbiopsy and several weeks for cells to grow. In contrast, the DNA-based diagnosis of genetic disease is now routine and themethodology is widely available. Although screening the entireFALDH gene is still time-consuming and expensive, targetedscreening for common mutations in suspected SLS patients fromcertain ethnic or geographic populations is more efficient. Forexample, c.1297-1298delGA and c.943C4T together account for36% of the alleles detected in European SLS probands [Rizzo et al.,1999], whereas the exon 9 deletion and c.906delT are the majormutations causing SLS among Bavarian patients [Kraus et al.,2000]. Similarly, c.682C4T and c.943delCCCins19nt are rela-tively common among patients of Mideastern ancestry [Rizzoet al., 1999]. Rapid screening methods, based on allele-specificPCR or restriction enzyme digestion of PCR amplicons, have beendeveloped to detect these mutations [De Laurenzi et al., 1997;Rizzo et al., 1997, 1999; Kraus et al., 2000].
Prenatal diagnosis of SLS has been accomplished by enzymaticassay of cultured amniocytes or chorionic villi cells [Rizzo et al.,1994] and mutation analysis [Sillen et al., 1997a] (W.B.R.,unpublished results). DNA-based diagnosis requires knowledge ofthe specific mutation, but is faster than enzymatic studies, whichrequire growth of fetal-derived cells.
FUTURE PROSPECTS
A more complete identification of the mutations responsible forSLS will ultimately prove useful for the reliable diagnosis of thisdisease and its prevention through genetic counseling and prenataldiagnosis. At present, the wide mutational diversity in SLS justifiesthe continuing need for a biochemical approach to the diagnosis ofthis disease, except for those populations harboring commonmutations. However, advances in mutation screening technologiesmay lead to DNA-based diagnostic tests replacing enzymaticmethods in most laboratories. In the near term, a more exhaustivedescription of the clinical phenotype of SLS patients, together withtheir mutational genotype, is sorely needed to investigategenotype–phenotype correlations and uncover possible modifiergenes. At the biochemical level, elucidation of the three-dimensional structure of FALDH will be necessary to understandthe precise molecular mechanisms whereby amino acid substitu-tions destroy catalytic function.
Although the molecular basis of SLS has been elucidated atleast in part, the next major goal is to understand the pathogenicmechanisms leading to clinical symptoms and devise methods tointerrupt or modify this process. This will require a more completeunderstanding of the role of FALDH in normal metabolism. Therare nature of the disease, reluctance of patients to provide tissuespecimens, and the lack of an animal model for SLS have hinderedmany studies. Progress has recently been achieved in generating aFALDH-deficient mouse model using gene-targeting methods(Rizzo WB, Carney G, Spieker R, Bridger J, Lin Z, Stribley J,
Salbaum M., unpublished results). The availability of such SLSmice promises to open new avenues for investigating thebiochemical basis for symptoms and developing new therapeuticapproaches to this disease.
REFERENCES
Aoki N, Suzuki H, Ito K, Ito M. 2000. A novel point mutation ofthe FALDH gene in a Japanese family with Sjorgen-Larssonsyndrome. J Invest Dermatol 114:1065–1066.
Carney G, Wei S, Rizzo WB. 2004. Sjogren-Larsson syndrome:seven novel mutations in the fatty aldehyde dehydrogenase geneALDH3A2. Hum Mutat 24:186.
Chang C, Yoshida A. 1997. Human fatty aldehyde dehydrogenasegene (ALDH10): organization and tissue-dependent expression.Genomics 40:80–85.
De Laurenzi V, Rogers GR, Hamrock DJ, Marekov LN, SteinertPM, Compton JG, Markova N, Rizzo WB. 1996. Sjogren-Larssonsyndrome is caused by mutations in the fatty aldehydedehydrogenase gene. Nat Genet 12:52–57.
De Laurenzi V, Rogers GR, Tarcsa E, Carney G, Marekov L, BaleSJ, Compton JG, Markova N, Steinert PM, Rizzo WB. 1997.Sjogren-Larsson syndrome is caused by a common mutation innorthern European and Swedish patients. J Invest Dermatol109:79–83.
Demozay D, Rocchi S, Mas JC, Grillo S, Pirola L, Chavey C, VanObberghen E. 2004. Fatty aldehyde dehydrogenase: potentialrole in oxidative stress protection and regulation of its geneexpression by insulin. J Biol Chem 279:6261–6270.
Hempel J, Kuo I, Perozich J, Wang B-C, Lindahl R, Nicholas H. 2001.Aldehyde dehydrogenase. Maintaining critical active site geometryat motif 8 in the class 3 enzyme. Eur J Biochem 268:722–726.
Ichihara K, Kusunose E, Noda Y, Kusunose M. 1986a. Someproperties of the fatty alcohol oxidation system and reconstitu-tion of microsomal oxidation activity in intestinal mucosa.Biochim Biophys Acta 878:412–418.
Ichihara K, Noda Y, Tanaka C, Kusunose M. 1986b. Purification ofaldehyde dehydrogenase reconstitutively active in fatty alcoholoxidation from rabbit intestinal microsomes. Biochim BiophysActa 878:419–425.
IJlst L, Oostheim W, van Werkhoven M, Willemsen MAAP,Wanders RJA. 1999. Molecular basis of Sjorgen-Larssonsyndrome: frequency of the 1297-1298 del GA and 943C4Tmutation in 29 patients. J Inher Metab Dis 22:319–321.
James PF, Zoeller RA. 1997. Isolation of animal cell mutantsdefective in long-chain fatty aldehyde dehydrogenase. Sensitiv-ity to fatty aldehydes and Schiff’s base modification ofphospholipids: implications for Sjogren-Larsson syndrome. J BiolChem 272:23532–23539.
Kelson TL, Secor M Jr, Rizzo WB. 1997. Human liver fatty aldehydedehydrogenase: microsomal localization, purification, and bio-chemical characterization. Biochim Biophys Acta 1335:99–110.
Kraus C, Braun-Quentin C, Ballhausen WG, Pfeiffer RA. 2000.RNA-based mutation screening in German families withSjogren-Larsson syndrome. Eur J Hum Genet 8:299–306.
Lee T. 1979. Characterization of fatty alcohol:NAD+ oxidor-eductase from rat liver. J Biol Chem 254:2892–2896.
Lin Z, Carney G, Rizzo WB. 2000. Genomic organization,expression, and alternate splicing of the mouse fatty aldehydedehydrogenase gene. Mol Genet Metab 71:496–505.
Lindahl R, Evces S. 1984. Rat liver aldehyde dehydrogenase. I.Isolation and characterization of four high Km normal liverisozymes. J Biol Chem 259:11986–11990.
HUMANMUTATION 26(1),1^10,2005 9
Liu ZJ, Sun YJ, Rose J, Chung YJ, Hsiao CD, Chang WR, Kuo I,Perozich J, Lindahl R, Hempel J, Wang BC. 1997. The firststructure of an aldehyde dehydrogenase reveals novel interactionsbetween NAD and the Rossmann fold. Nat Struct Biol 4:317–326.
Masaki R, Yamamoto A, Tashiro Y. 1994. Microsomal aldehydedehydrogenase is localized to the endoplasmic reticulum via itscarboxyl-terminal 35 amino acids. J Cell Biol 126:1407–1420.
Mitchell DY, Petersen DR. 1989. Oxidation of aldehydic productsof lipid peroxidation by rat liver microsomal aldehyde dehy-drogenase. Arch Biochem Biophys 269:11–17.
Miyauchi K, Masaki R, Taketani S, Yamamoto A, Akayama M,Tashiro Y. 1991. Molecular cloning, sequencing, and expressionof cDNA for rat liver microsomal aldehyde dehydrogenase. J BiolChem 266:19536–19542.
Nakayasu H, Mihara K, Sato R. 1978. Purification and propertiesof a membrane-bound aldehyde dehydrogenase from rat livermicrosomes. Biochem Biophys Res Commun 83:697–703.
Nigro JF, Rizzo WB, Esterly NB. 1996. Redefining the Sjogren-Larsson syndrome: atypical findings in three siblings and implica-tions regarding diagnosis. J Am Acad Dermatol 35:678–684.
Perozich J, Nicholas H, Wang B-C, Lindahl R, Hempel J. 1999.Relationships within the aldehyde dehydrogenase extendedfamily. Protein Sci 8:137–146.
Pigg M, Jagell S, Sillen A, Weissenbach J, Gustavson KH, WadeliusC. 1994. The Sjogren-Larsson syndrome gene is close toD17S805 as determined by linkage analysis and allelic associa-tion. Nat Genet 8:361–364.
Rizzo WB, Dammann AL, Craft DA. 1988. Sjogren-Larssonsyndrome. Impaired fatty alcohol oxidation in cultured fibro-blasts due to deficient fatty alcohol:nicotinamide adeninedinucleotide oxidoreductase activity. J Clin Invest 81:738–744.
Rizzo WB, Craft DA. 1991. Sjogren-Larsson syndrome. Deficientactivity of the fatty aldehyde dehydrogenase component of fattyalcohol:NAD+ oxidoreductase in cultured fibroblasts. J ClinInvest 88:1643–1648.
Rizzo WB, Craft DA, Kelson TL, Bonnefont JP, Saudubray JM,Schulman JD, Black SH, Tabsh K, Dirocco M, Gardner RJ. 1994.Prenatal diagnosis of Sjogren-Larsson syndrome using enzymaticmethods. Prenat Diagn 14:577–581.
Rizzo WB, Carney G, De Laurenzi V. 1997. A common deletionmutation in European patients with Sjogren-Larsson syndrome.Biochem Mol Med 62:178–181.
Rizzo WB, Carney G, Lin Z. 1999. The molecular basis of Sjogren-Larsson syndrome: mutation analysis of the fatty aldehydedehydrogenase gene. Am J Hum Genet 65:1547–1560.
Rizzo WB, Craft DA. 2000. Sjogren-Larsson syndrome: accumula-tion of free fatty alcohols in cultured fibroblasts and plasma.J Lipid Res 41:1077–1081.
Rizzo WB. 2001. Sjogren-Larsson syndrome: fatty aldehydedehydrogenase deficiency. In: Scriver CR, Beckman K, SmallGM, Valle D, editors. The metabolic and molecular bases ofinherited disease. New York: McGraw-Hill. p 2239–2258.
Rizzo WB, Heinz E, Simon M, Craft DA. 2000. Microsomal fattyaldehyde dehydrogenase catalyzes the oxidation of aliphaticaldehyde derived from ether glycerolipid catabolism: implicationsfor Sjogren-Larsson syndrome. Biochim Biophys Acta 1535:1–9.
Rizzo WB, Lin Z, Carney G. 2001. Fatty aldehyde dehydrogenase:genomic structure, expression and mutation analysis in Sjogren-Larsson syndrome. Chem Biol Interact 130-132:297–307.
Rogers GR, Markova NG, De Laurenzi V, Rizzo WB, Compton JG.1997. Genomic organization and expression of the human fattyaldehyde dehydrogenase gene (FALDH). Genomics 39:127–135.
Sillen A, Holmgren G, Wadelius C. 1997a. First prenatal diagnosisby mutation analysis in a family with Sjogren-Larsson syndrome.Prenat Diagn 17:1147–1149.
Sillen A, Jagell S, Wadelius C. 1997b. A missense mutation in theFALDH gene identified in Sjogren-Larsson syndrome patientsoriginating from the northern part of Sweden. Hum Genet100:201–203.
Sillen A, Alderborn A, Pigg M, Jagell S, Wadelius C. 1998a.Detailed genetic and physical mapping in the Sjogren-Larssonsyndrome gene region in 17p11.2. Hereditas 128:245–250.
Sillen A, Anton-Lamprecht I, Braun-Quentin C, Kraus CS, SayliBS, Ayuso C, Jagell S, Kuster W, Wadelius C. 1998b. Spectrum ofmutations and sequence variants in the FALDH gene in patientswith Sjogren-Larsson syndrome. Hum Mutat 12:377–384.
Sjogren T, Larsson T. 1957. Oligophrenia in combination withcongenital ichthyosis and spastic disorders. Acta PsychiatrNeurol Scand 32(Suppl 113):1–113.
Takagi Y, Ito A, Omura T. 1985. Biogenesis of microsomalaldehyde dehydrogenase in rat liver. J Biochem (Tokyo)98:1647–1652.
Theile U. 1974. Sjogren-Larsson syndrome. Oligophrenia–ichthyo-sis–di-tetraplegia. Humangenetik 22:91–118.
Tsukamoto N, Chang C, Yoshida A. 1997. Mutations associated withSjogren-Larsson syndrome. Ann Hum Genet 61(Pt 3):235–242.
van den Brink DM, van Miert JN, Dacremont G, Rontani JF,Jansen GA, Wanders RJ. 2004. Identification of fatty aldehydedehydrogenase in the breakdown of phytol to phytanic acid. MolGenet Metab 82:33–37.
Vasiliou V, Kozak CA, Lindahl R, Nebert DW. 1996. Mousemicrosomal class 3 aldehyde dehydrogenase: AHD3 cDNAsequence, inducibility by dioxin and clofibrate, and geneticmapping. DNA Cell Biol 15:235–245.
Vasiliou V, Bairoch A, Tipton KE, Nebert DW. 1999. Eukaryoticaldehyde dehydrogenase (ALDH) genes: human polymorphisms,and recommended nomenclature based on divergent evolutionand chromosomal mapping. Pharmacogenetics 9:421–434.
Verhoeven NM, Jakobs C, Carney G, Somers MP, Wanders RJ,Rizzo WB. 1998. Involvement of microsomal fatty aldehydedehydrogenase in the alpha-oxidation of phytanic acid. FEBSLett 429:225–228.
Willemsen MA, Rotteveel JJ, van Domburg PH, Gabreels FJ,Mayatepek E, Sengers RC. 1999a. Preterm birth in Sjogren-Larsson syndrome. Neuropediatrics 30:325–327.
Willemsen MAAP, Steijlen PM, de Jong JPN, Rotteveel W, IJlst L,van Werkhoven MA, Wanders RJA. 1999b. A novel 4 bpdeletion mutation in the FALDH gene segregating in a TurkishFamily with Sjorgen-Larsson Syndrome. J Invest Dermatol112:827–828.
Willemsen MA, Ijlst L, Steijlen PM, Rotteveel JJ, de Jong JG, vanDomburg PH, Mayatepek E, Gabreels FJ, Wanders RJ. 2001a.Clinical, biochemical and molecular genetic characteristicsof 19 patients with the Sjogren-Larsson syndrome. Brain124:1426–1437.
Willemsen MA, Lutt MA, Steijlen PM, Cruysberg JR, Van DerGM, Nijhuis-van der Sanden MW, Pasman JW, Mayatepek E,Rotteveel JJ. 2001b. Clinical and biochemical effects of zileutonin patients with the Sjogren-Larsson syndrome. Eur J Pediatr160:711–717.
Willemsen MA, Rotteveel JJ, de Jong JG, Wanders RJ, Ijlst L,Hoffmann GF, Mayatepek E. 2001c. Defective metabolism ofleukotriene B4 in the Sjogren-Larsson syndrome. J Neurol Sci183:61–67.
10 HUMAN MUTATION 26(1),1^10,2005