SRD5A3 Is Required for Converting
Polyprenol to Dolichol and Is Mutated
in a Congenital Glycosylation Disorder
Vincent Cantagrel,1Dirk J. Lefeber,2,3Bobby G. Ng,7Ziqiang Guan,8Jennifer L. Silhavy,1Stephanie L. Bielas,1
Ludwig Lehle,9Hans Hombauer,10Maciej Adamowicz,11Ewa Swiezewska,13Arjan P. De Brouwer,4Peter Blu ¨mel,14
Jolanta Sykut-Cegielska,12Scott Houliston,7Dominika Swistun,1Bassam R. Ali,15William B. Dobyns,17
Dusica Babovic-Vuksanovic,18Hans van Bokhoven,4,5Ron A. Wevers,2Christian R.H. Raetz,8Hudson H. Freeze,7
E´va Morava,6Lihadh Al-Gazali,15,16,* and Joseph G. Gleeson1,*
1Neurogenetics Laboratory, Institute for Genomic Medicine, Howard Hughes Medical Institute,
Department of Neurosciences and Pediatrics, University of California, San Diego, La Jolla, CA 92093, USA
2Department of Laboratory Medicine, Institute for Genetic and Metabolic Disease
3Department of Neurology
4Department of Human Genetics, Nijmegen Centre for Molecular Life Sciences
5Department of Cognitive Neuroscience, Donders Institute for Brain, Cognition, and Behaviour
6Department of Paediatrics, Institute for Genetic and Metabolic Disease
Radboud University Nijmegen Medical Centre, 6500 HB Nijmegen, The Netherlands
7Genetic Disease Program, Sanford Children’s Health Research Center, Sanford-Burnham Medical Research Institute,
La Jolla, CA 92037, USA
8Department of Biochemistry, Duke University Medical Center, Durham, NC 27710, USA
9Universita ¨t Regensburg, Lehrstuhl fu ¨r Zellbiologie und Pflanzenbiochemie, D-93053 Regensburg, Germany
10Ludwig Institute for Cancer Research, Department of Medicine, Department of Cellular and Molecular Medicine and Cancer Center,
University of California, San Diego, School of Medicine, La Jolla, CA 92093, USA
11Department of Biochemistry and Experimental Medicine
12Department of Metabolic Diseases, Endocrinology, and Diabetology
The Children’s Memorial Health Institute, 04-730 Warsaw, Poland
13Institute of Biochemistry and Biophysics, Polish Academy of Sciences, 02-106 Warsaw, Poland
14Preyer’sches Kinderspital, 1100 Vienna, Austria
15Department of Pathology
16Department of Pediatrics
United Arab Emirates University, School of Medicine and Health Sciences, 17666 Al Ain, United Arab Emirates
17Department of Human Genetics, Neurology, and Pediatrics, University of Chicago, Chicago, IL 60637, USA
18Departments of Medical Genetics, Pediatric Neurology, Laboratory Genetics, Pediatric Endocrinology, and Dermatology,
Mayo Clinic, Rochester, MN 55905, USA
*Correspondence: firstname.lastname@example.org (L.A.-G.), email@example.com (J.G.G.)
N-linked glycosylation is the most frequent modifica-
tion of secreted and membrane-bound proteins in
eukaryotic cells, disruption of which is the basis of
the congenital disorders of glycosylation (CDGs).
We describeanew type of CDG caused bymutations
in the steroid 5a-reductase type 3 (SRD5A3) gene.
and cerebellar defects. We found that SRD5A3 is
necessary for the reduction of the alpha-isoprene
unit of polyprenols to form dolichols, required for
synthesis of dolichol-linked monosaccharides, and
the oligosaccharide precursor used for N-glycosyla-
tion. The presence of residual dolichol in cells
depleted for this enzyme suggests the existence of
an unexpected alternative pathway for dolichol de
novo biosynthesis. Our results thus suggest that
SRD5A3 is likely to be the long-sought polyprenol
reductase and reveal the genetic basis of one of the
earliest steps in protein N-linked glycosylation.
N-glycosylation occurs on certain asparagine residues present
on nascent polypeptides in all eukaryotic cells. The glycan struc-
tures resulting from this process show an incredible variability
depending on the protein, cell type, and species. This essential
posttranslational modification occurs on most secreted and
plasma membrane proteins and is involved in protein folding
and trafficking with implications for cell-cell and cell-matrix
interactions and intracellular signaling (Freeze, 2006; Helenius
and Aebi, 2001). The process of N-linked protein glycosylation
is localized in the endoplasmic reticulum (ER) and the Golgi
Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc. 203
compartment. Three separate phases can be distinguished:
First, an oligosaccharide precursor, a block of 14 monosaccha-
rides (Glc3Man9GlcNAc2), is assembled on the lipid carrier
dolichol-phosphate (Dol-P) in the ER membrane. Second, this
glycan is transferred cotranslationally or posttranslationally to
dedicated asparagine residues of nascent glycoproteins (Ruiz-
Canada et al., 2009). In this reaction, oligosaccharyltransferase
(OST) recognizes the acceptor sequence NX[S/T] (where X can
be any amino acid except proline) on nascent polypeptides
and catalyzes the transfer of the glycan precursor en bloc from
its lipid carrier to the protein (Chavan and Lennarz, 2006). Third,
the N-linked glycan is further modified by a series of trimming
and elongation reactions beginning in the ER and ending in the
late Golgi compartment.
The early steps of this pathway are present not only in eukary-
otic cells but also in archae and bacteria, all relying on a lipid to
build an oligosaccharide precursor (Jones et al., 2009). This
carrier lipid, a polyisoprenoid, is assembled from a variable
number of isoprene units, linked head to tail. The length of the
carrier polyisoprenoid varies across evolution: bacteria posses
a single predominant polyprenol, usually undecaprenol (com-
posed of 11 isoprene units), but in eukaryotic cells these lipids
typically occur as mixtures of different lengths, depending
upon the species. In mammalian cells, dolichols are predomi-
nantly 18–21 isoprene units in length.
A requirement in eukaryotic organisms is the reduction of
the precursor polyprenol to dolichol on the terminal isoprene
unit (alpha) (Swiezewska and Danikiewicz, 2005), followed by
phosphorylation to generate Dol-P. The identification of Dol-P
as glycosyl carrier lipids in glycosylation was described 40 years
ago (Behrens and Leloir, 1970), but the role of the free lipid,
broadly distributed in mammalian cells (Rip et al., 1985), and
some of its biosynthetic steps remain elusive. Using prenol
labeling studies, a pathway for dolichol biosynthesis was
proposed (Sagami et al., 1993); however, several enzymes
were still missing, including a polyprenol reductase. In this
article, the term ‘‘polyprenol’’ will be restricted to alpha-unsatu-
rated compounds, despite its more general meaning, to distin-
guish themfromdolichol, asit iscommonly donein theliterature.
Several glycosylation-defective cell lines, generated in vitro,
(Acosta-Serrano et al., 2004; Rosenwald et al., 1993) and sug-
gested that polyprenol reductionwas the rate-limiting stepin do-
lichol synthesis, with major consequences on N-glycosylation.
disorders of glycosylation (CDGs), a growing class of hereditary
disorders (Freeze, 2006; Gru ¨newald and Matthijs, 2000; Haeup-
tle and Hennet, 2009; Jaeken and Matthijs, 2007). Defects in the
maturation and transfer of the glycan precursor, located in the
ER, have been grouped in the past as CDG type I, and disorders
affecting the subsequent N-glycan processing steps grouped as
CDG type II. A recently proposed alternate nomenclature uses
only the gene name together with a CDG suffix (Jaeken et al.,
2009). These diseases show wide symptomatology and severity.
The main features are psychomotor retardation, cerebellar
ataxia, seizures, retinopathy, liver fibrosis, coagulopathies,
failure to thrive, and dysmorphic features including abnormal
fat distribution and ophthalmological anomalies (Eklund and
Freeze, 2006). Even though a multisystem phenotype is often
observed, several cases have been reported with primary
neurological involvement including cerebellar ataxia (Vermeer
et al., 2007), suggesting that cerebellar disease maybe a sensi-
tive measure of defective N-glycosylation.
In this study, we identify SRD5A3 orthologs as necessary for
and promoting the reduction of polyprenol to dolichol in human,
mouse, and yeast and describe a new syndrome of CDG type I
in seven families caused by a defect in this newly identified
Loss of Function Mutations of the SRD5A3 Gene Cause
a Multisystemic Syndrome with Eye Malformations,
Cerebellar Vermis Hypoplasia, and Psychomotor Delay
We identified a large consanguineous Emirati family of Baluchi
(Southern Iran) origin (CVH-385, Figure 1A) (Al-Gazali et al.,
2008). All affected children displayed ocular colobomas, ich-
thyosis, heart defect, developmental delay, and brain malforma-
tions including cerebellar vermis hypoplasia. We performed a
genome-wide linkage analysis and mapped the disease locus
on chromosome 4q12 with a multipoint logarithm (base 10) of
odds (LOD) score of 4.2 (Figure 1B). This mapping defined an
interval of 53.8–57.4 MB (between the markers rs751266 and
rs899631) encompassing 42 genes (based on NCBI genome
browser, build 36 version 3) (Figure 1C), which were screened
with a systematic mutational analysis of candidates with bidirec-
tional sequencing. Analysis of the steroid 5a-reductase 3
function, identified a molecular rearrangement with a homozy-
gous 3 bp deletion and a 10 bp insertion resulting in a predicted
stop codon at amino acid 96 (Figure 1D). To exclude the possi-
bility that this change represented a common polymorphism,
we tested 96 DNAs (192 chromosomes) from geographically
matched controls but identified no carriers. Although CHIME
syndrome (colobomas of the eye, heart defects, ichthyosiform
dermatosis, mental retardation, and ear defects or epilepsy)
was the closest related disease without a known molecular
tested negative for SRD5A3 mutations. However, another family
of Baluchi origin (MR3) also living in the Emirates with a compa-
rable phenotype (Table 1), displayed the same molecular
rearrangement in SRD5A3 (despite denying known relationship
with CVH-385 family), suggesting the existence of a common
founder mutation. Due to the phenotypic similarity with CDG,
we tested the N-glycosylation status of transferrin using mass
spectrometry (O’Brien et al., 2007), a reliable screening test
for type I CDG patients. Transferrin is a serum protein with two
N-glycosylation sites, fully occupied in control individuals.
CVH-385 and MR3 patients showed a very clear defect with,
respectively, about 45% and 25% of monoglycosylated trans-
ferrin, suggesting that the SRD5A3 mutation leads to a type I
CDG (Figure 2A and Figure S2A available online). We also found
a defect in extra-cellular secretion of N-glycosylated DNase I in
index patients’ fibroblasts (Figure S2B), a sensitive measure of
Vleugels et al., 2009).
204 Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc.
Consequently, we screened 38 patients with CDG type I-x
(CDG type I negative for known gene mutations), and we identi-
fied five other independent homozygous or compound heterozy-
gous mutations and one genomic rearrangement (Figure 1F,
Table 1, and Figure S1). Among the mutations identified were
a 2 bp deletion and four single-base substitutions resulting in
truncation of the gene encompassing within exon 5, in the 30part
of SRD5A3 open reading frame (ORF). Further expression
analysis in patients’ fibroblasts showed partial nonsense-medi-
ated messenger RNA (mRNA) decay of SRD5A3 transcript in
some patients (Figure 1G).
Based on the phenotype of 11 children from seven families,
the most striking features observed were the presence of con-
genital eye malformations with variable degree of visual loss,
nystagmus, muscle hypotonia, motor delay, mental retardation,
and facial dysmorphism. Microcytic anemia, elevated levels
of liver enzyme activities, coagulation abnormalities, and de-
creased antithrombin III levels were detected in nine evaluated
cases. Most children presented with ocular coloboma or hypo-
plasia of the optic disc (unique features in the CDGs) (Morava
et al., 2009), with striking cerebellar atrophy or vermis malforma-
malformations were sporadically present. Midline malformations
and endocrine anomalies were only present in the index patients
(Al-Gazali et al., 2008) (Table 1). The relative uniformity in the
biochemical and clinical phenotype associated with frequent
early truncating mutations
SRD5A3 as the genetic mechanism.
suggests loss-of-function of
Patients with Mutation in SRD5A3 Show an Early Defect
in Lipid-Linked Oligosaccharide Synthesis
The absence of whole glycan chains on proteins indicated that
the metabolic block occurred early in the N-glycosylation path-
oligosaccharide (LLO), to recipient proteins. Epitope tagged
SRD5A3 localized predominantly to the ER (Figure 2B), where
LLO synthesis occurrs (Aebi and Hennet, 2001).
An abnormal composition of the glycan precursor impairs
its transfer to acceptor proteins. Accordingly, we investigated
the size of the LLO glycans using HPLC after [2-3H]-mannose
metabolic labeling with fibroblasts from index patients CVH-
385-IV-11 and CVH-385-IV-13. Since no major structural abnor-
malities in LLO were detected (Figure S2E), we also determined
theamountofradiolabeled LLO.Wedetected aseverereduction
in the amount of newly synthesized LLO in four of the five
patients tested compared to three control cell lines (Figure 2C),
suggesting that the N-glycosylation block occurs prior the
glycan transfer step.
The reduced levels of LLO could be explained by a limited
availability of Dol-P. To test this hypothesis, we used an
in vitro assay to assess the production of Dol-PP-GlcNAc1and
Dol-PP-GlcNAc2, the first two reactions of LLO synthesis. With
fibroblast homogenates used as a source of enzyme and UDP-
[14C]GlcNAc as glycosyl donor, all SRD5A3 deficient patient
samples showed a reduced synthesis of Dol-PP-GlcNAc1/2
without addition of exogenous Dol-P, compared to controls
(Figure 2D). However, when exogenous Dol-P was added to
the incubation mixture, formation of Dol-PP-GlcNAc1/2 was
Fibroblasts from patients with other known CDG-I defects
(CDG-Ik or CDG-Io) behaved comparable to controls, showing
no evidence of Dol-P mediated rescue (Figure 2D). Elongation
of Dol-PP-GlcNAc2 to Dol-PP-GlcNAc2-Man5 was unremark-
able. Similarly, OST activity was normal (data not shown).
Altogether, the rescue of the enzymatic GlcNAc transferases
deficiencies by exogenous Dol-P indicates that the amount of
Dol-P islimiting in the patients’ fibroblasts and suggests a defect
in Dol or Dol-P biosynthesis.
SRD5A3 Is the Human Ortholog
of the Yeast DFG10/YIL049W Gene
A yeast mutant for the DFG10 gene, called dfg10-100, was
previously isolated by a genetic screen for mutant strains defec-
tive for filamentous growth (dfg), using insertional mutagenesis
(Mo ¨sch and Fink, 1997). The product of this gene shows 25%
amino acid identity and 43% similarity with the human SRD5A3
protein (Blastp, NCBI). To determine whether DFG10 is the yeast
ortholog of SRD5A3 (Figure 3A), we first asked whether the
dfg10-100 mutant displays a lack of N-glycan modifications.
Carboxypeptidase Y (CPY) is a secreted enzyme with a mature
form that contains four N-glycan sites, all of which are occupied
under optimal growth conditions (Hasilik and Tanner, 1978) and
all of which can be removed with PNGase F treatment. In
contrast with the WT strain (L5366), the dfg10-100 mutants
(diploid and homozygous at the DFG10 locus) produced hypo-
glycosylated CPY, with the detection of tri-, di- and monoglyco-
sylated CPY (Figure 3B). Because the dfg10-100 mutant is a
result of a transposon insertion into the DFG10 promoter, it
was possible that still some protein was expressed, thus
accounting for the nonlethal phenotype. This possibility was
excluded by engineering a deletion of the whole DFG10 ORF,
which produced an identical growth delay and CPY phenotype
(Figure S3). The identification of a similar biochemical defect in
yeast and human suggests a conserved function for SRD5A3
In human, five partially homologous genes compose the
steroid 5a-reductase family, including the well-characterized
SRD5A1, SRD5A2 involved in testosterone reduction (Russell
and Wilson, 1994), encoding for proteins targeted in treatments
against prostatic hyperplasia and male pattern hair loss (Aggar-
wal et al., 2010), and also SRD5A2L2, GPSN2, SRD5A3 that are
less characterized. Bioinformatics comparison showed that the
DFG10 sequence shares most identity with SRD5A3 (BLASTP,
E value = 2e-13) whereas SRD5A1, SRD5A2, SRD5A2L2 and
GPSN2 show E values of, respectively 6e-04, 3e-04, 3e-04
and 5e-05 (Figure 3A). To test for functional conservations we
expressed each mammalian ORF under the control of a strong
constitutive yeast promoter (Alber and Kawasaki, 1982) in the
dfg10-100 mutant. The mutant transformed with yeast DFG10
showed a full correction of the CPY underglycosylation. Further-
more, SRD5A3 was the only homolog able to rescue the pheno-
type (Figure 3C). This experiment shows that SRD5A3 is the
diverged human ortholog of the yeast DFG10 gene and suggests
a specific role for SRD5A3 in protein glycosylation compared
with other family members.
Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc. 205
1 2 3 4 5 6 7 8 9 10 11 12 13
ThrThr GluGlu Ser LeuSer Leu Phe Phe LeuLeu GlyGlyAla ProAla Pro
ThrThrStop 3 bp deletion/10 bp insertionStop 3 bp deletion/10 bp insertion
1 2 3 4 5
9101112 13 14 15 1617 18 19 202122X
Figure 1. Identification of Mutations in the SRD5A3 Gene in Patients with Multisystemic Syndrome Including Cerebellar Hypoplasia
(A) Pedigree of family CVH-385 showing several levels of consanguinity with cousin marriages. The two branches each produced two affected offspring
represented by filled symbols in generation IV.
(B) Whole-genome analysis of linkage results with chromosomal position (x axis) and multipoint LOD score (y axis) showing a peak LOD score of 4.2 on
chromosome 4 (arrowhead).
(C) Expanded view of the candidate interval on chromosome 4q12, containing 42 candidate genes including SRD5A3 (red), spanning 25.5 kb of genomic DNA
with 5 exons. A mutation in exon 2 was identified in family CVH-385.
206 Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc.
Phylogenetic analysis of proteins with a steroid 5a-reductase
domain from multiple species indicates that this steroid reductase
family can be separated in three main groups consisting of (1) the
SRD5A1-SRD5A2 group, (2) the SRD5A3 group containing
ing the idea that different classes of lipids can be substrates for
(D) DNA sequence of exon 2 of SRD5A3 from a control individual, an obligate carrier, and an affected family member from CVH-385. The mutation consists of a
3 bp deletion associated with a 10 bp insertion, resulting in a frame shift and premature termination at amino acid 96 of 318, within the second of six
(E) Brain MRI midline sagittal view showing cerebellar vermis hypoplasia (red arrowhead) in SRD5A3 mutated patients.
(F) Topology model of SRD5A3 withmutations indicated and six transmembrane domains. Mutations were scattered throughout theORF, all leading topredicted
protein termination before the steroid reductase domain (in red).
all patients CVH-385-IV-13, CVH-385-IV-11, 07-0153compared to control, suggesting nonsense-mediated mRNA decay. Noexpressionwasdetectedinpatient
AK0295, as a result of a homozygous genomic rearrangement. RT?, no reverse transcriptase; water, no cDNA.
See also Figure S1.
Table 1. Clinical Phenotype Associated with SRD5A3 Mutations
Ethnic backgroundBaluchi BaluchiPolishTurkishPolish TurkishIranian
Muscle hypotonia/motor retardation+/?
Hypoplasia or coloboma/
Other eye malformation
Dry skin/atopic dermatitis+
Failure to thrive +
Microcytic anemia ++
Elevated liver enzymes+ NA+++
Abnormal coagulation studiesNANA+++++
3/protein C and S levels
Mutation cDNAc.286_288 delins
c.292_293 del c.320 G/Ahet c.424
Mutation at protein levelp.Gln96 delinsXp.Gln96 delinsXp.Leu98
NA, data not available.
Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc. 207
Control 1 Control 1Control 2 Control 2Control 3 CVH358-IV-11 CVH358-IV-13Control 3 CVH358-IV-11 CVH358-IV-1307-0153 07-015308-048708-0487 AK0295 AK0295
Intensity , cps
72000 74000 76000 78000 80000
72000 74000 76000 78000 80000
72000 74000 76000 78000 80000
Example of CDG type II
72000 74000 76000 78000 80000
Intensity , cps
Intensity , cps
Intensity , cps
% Dol-PP- GlcNAc relative to control
% Dol-PP-GlcNAc relative to control
Ratio of activity -/+ Dol-P
endogenous Dol-Pexogenous Dol-P
% incorporation compared to control
Figure 2. SRD5A3 Mutated Patients Have a Congenital Disorder of Glycosylation Type I Caused by a Defect in LLO Synthesis, Rescued
In Vitro with Exogenous Dolichol Phosphate
(A) Mass spectra of transferrin, normally N-glycosylated on two sites, Asn-432 and Asn-630 (control). Transferrin containing a single N-glycan in SRD5A3 patient
(depicted by lack of certain sugar moieties), is shown for comparison.
(B) Intracellular localization of SRD5A3 containing a N terminus DsRed tag (center panel) in COS7 cells costained with antibody against ER-specific marker,
calreticulin, ERGIC-specific marker ERGIC53, and Golgi-specific marker Giantin. DsRed-SRD5A3 colocalized with most of the ER whereas Giantin staining
did not colocalize. The scale bar represents 10mm.
(C) Incorporation of [3H]-mannose into LLO after labeling of human fibroblasts. The results indicate severely reduced levels of LLO in four out of five patient
samples. Error bars represent the mean ± standard deviation from three experiments.
(D) Rescue of LLO precursor levels with exogenous Dol-P. GlcNAc transferase activity in fibroblasts was measured. Microsomal fractions from fibroblasts were
incubated with radioactive GlcNAc, and then Dol-PP-GlcNAc1/2formation was analyzed by TLC. Extracts from patients’ fibroblasts produced a reduced amount
of Dol-PP-GlcNAc1/2. However, the addition of exogenous Dol-P rescued this defect.
See also Figure S2.
208 Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc.
Srd5a3 Mutation Is Lethal in Mouse and Results in an
Activation of the Unfolded Protein Response Pathway
We found that mice homozygous for a LacZ gene trap (Gt)
insertion in intron 3 of Srd5a3 were recovered at embryonic
stages, up to embryonic day 12.5 (E12.5) but not beyond (Fig-
ures 4A and 4B). At E10.5, Srd5a3Gt(betaGeo)703Lex/Gt(betaGeo)703Lex
embryos (abbreviated Srd5a3
undergo axial rotation observed at E8.5 in WT littermates. Anal-
ysis with b-gal colorimetric staining in asymptomatic heterozy-
gous carriers showed strong expression in the yolk sac, eyes,
heart, and neural tube (Figures S4A–S4D). In keeping with this,
homozygous mutants frequently presented dilated hearts
(Figure 4A) and open neural tubes, which is consistent with the
broad phenotypes observed in patients.
To identify the misregulated pathways underlying these devel-
opmental defects,wecarried out expressionmicroarray analysis
ical differences appeared (Figure 4C). Whole-transcriptome
analysis revealed that among the 50 most upregulated tran-
response (UPR) or are activated in this pathway (Tables S1 and
S2). An activation of the UPR pathway in E8.5 Srd5a3Gt/Gt
embryos was confirmed by real-time RT-PCR and with an E9.5
mouse embryonic fibroblast cell line treated with tunicamycin
Gt/Gt) were smaller and failed to
wt dfg10-100 wt dfg10-100
Figure 3. N-Glycosylation Phenotype of
dfg10-100 Yeast Mutant and Rescue with
(A) Phylogenic tree representation of the yeast
protein DFG10 and human proteins presenting a
steroid 5a-reductase domain, with branch support
value indicated in red.
(B) N-glycosylation status of the yeast protein CPY
in yeast WT and dfg10-100 mutant strain, mutated
by transposon insertion. Two colonies from each
strain were tested. CPY is posttranslationally
modified by the addition of four glycan chains. In
the dfg10-100 mutants, a protein lacking one,
two, or three glycan chains is detected. Protein
extracts were treated with PNGase F to remove
(C) Glycosylation status of CPY in dfg10-100
mutant transformed with each steroid 5a-reduc-
tase domain containing gene from human. Only
SRD5A3 showed rescue effect. Positions of
mature CPY (mCPY) and the different glycoforms
(?1, ?2, ?3, ?4) are indicated. Vector indicates
empty vector control.
See also Figure S3.
used as a positive control (activates the
UPR pathway by blocking N-glycosyla-
BiP, is upregulated at the transcript and
protein levels in Srd5A3Gt/Gtembryos,
with a particularly high expression in neu-
roepithelial cells (Figure 4E). Srd5a3
expression was not detected in mutant
embryos; however, in cells inhibited
for N-glycosylation by tunicamycin treat-
ment, its expression increased significantly (Figure 4D). We
also confirmed UPR activation by determining enrichment of
gene ontology (GO) categories by all the genes significantly
misregulated in Srd5A3Gt/Gtembryos compared with littermate
controls. UPR was the biological process most significantly
enriched for the genes upregulated (Table S3), whereas genes
involved in general cellular metabolic processes and specific
embryonic developmental program like regionalization were
the most significantly downregulated (Table S3). These observa-
tions suggest that Srd5a3 is required for ER protein folding,
a primary role of N-glycan during development.
DFG10 and SRD5A3 Are Necessary for Conversion
of Polyprenol to Dolichol in Yeast, Mouse, and Human
During the denovo synthesisof dolichol in eukaryotes, the farne-
syl pyrophosphate (FPP), a product of the mevalonate pathway,
is elongated by its successive condensation with isopentenyl
pyrophosphate (IPP), catalyzed by a cis-isopentenyltransferase
named dehydrodolichyl diphosphate synthase (DHDD) (Fig-
ure 5A). According to the current model, when the chain reaches
target length, the pyrophosphate and phosphate groups are
removed, although the phosphatases are not yet identified
(Kato et al., 1980; Wolf et al., 1991). The alpha-isoprene unit of
polyprenol is subsequently reduced by an NADPH-dependent
Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc. 209
22.214.171.124 1.01.1 1.1 1.21.2 1.31.3 1.4 1.4
E10.5 Srd5a3 Gt/Gt
E10.5 Srd5a3 +/+
E10.5 Srd5a3 Gt/Gt
Log2 ratio of expression
n=21 n=27 n=12 n=18 n=7 n=32
E9.5E9.5 E9.5E10.5 E11.5 E13.5 E10.5 E11.5 E13.5E10.5 E11.5 E13.5P0P0P0
MEFs + tunicamicyn
E8.5 W T embryo
E8.5 Mutant embryo
E8.5 Srd5a3 Gt/+ E8.5 Srd5a3 Gt/Gt
Percent recovered based on genotype
Age (Embryonic day)
Normalized transcript level
E9.5 Srd5a3 +/+
Figure 4. Characterization of Homozygous Srd5a3Gt/GtGene Trap Mouse Embryos
(A) Phenotype at E10.5 shows failure to rotate, ventral body wall defect (arrow), and dilated heart (arrowhead). The scale bar represents 1 mm.
(B) Graphic representation of the genotype obtained from the progeny of heterozygous mating, with lethality appearing between E11.5 and E13.5.
(C) Genes overexpressed in Srd5a3Gt/Gtat E8.5 detected with 44k mouse genome oligo microarray (1 is the log 2 of a 2-fold expression increase; error bars
representthemean±standarddeviationfromfourindependentexperiments).Amongtenofthemost upregulated genes,five(inblue)areinvolvedintheunfolded
protein response pathway (UPR). Morphology of heterozygous and homozygous Srd5a3 mutant embryos at E8.5, before embryo axial rotation. The scale bar
represents 500 mm.
(D) Real-time RT-PCR confirming activation of the UPR pathway (Errors bars are means ± standard deviations, asterisks indicate p < 0.05, n = 3).
210 Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc.
microsomal reductase (Sagami et al., 1993), but the enzyme
involved has not been identified. Finally, a dolichol kinase
(hDK), mutated in CDG-Im, transfers phosphate from CTP to
dolichol (Allen et al., 1978; Kranz et al., 2007). The unidentified
polyprenol reductase enzyme made SRD5A3 a likely candidate
for this function.
in SRD5A3 or DFG10 mutated cells and to unequivocally identify
the last step of dolichol synthesis, we used liquid chromatog-
raphy-mass spectrometry (LC-MS) (Garrett et al., 2007) to
analyze polyprenols in WT and dfg10-100 yeast strains, E11.5
WT and Srd5a3Gt/Gtmouse embryos, and fibroblasts and leuko-
cytes from controls and patients. Polyprenol was not detected in
any samples of WT origin as reported (Swiezewska and Danikie-
wicz, 2005) but was easily detected in the yeast and mouse
mutants, in the same molar range as the dolichol naturally
present in control samples (Figures 5B and 5C), suggesting a
block in the polyprenol reduction step. In reference to an internal
standard, by correcting polyprenol isotopic contribution, we de-
tected a 70% decrease in dolichol in the dfg10-100 mutant
right end). Given these striking results in both mouse and yeast,
we were surprised to find no clear change in prenol profiles in
patient fibroblasts or leukocytes (data not shown). Because
one possible explanation might be that the normal exogenous
fetal calf serum used in tissue culture supplied dolichol to over-
come this metabolic block, we instead analyzed directly patient
fresh plasma. We found an increased level of polyprenoids in
patients’ samples versus controls and in other CDG-I patients
with a significant increased of polyprenols-18,19,20/dolichols-
18,19,20 ratios (Figure 5D), indicating a defect in polyprenols
metabolism in all organisms tested.
SRD5A3 Promotes the Reduction of Polyprenol
We next tested whether SRD5A3 was capable of reducing poly-
prenol to dolichol. We assessed polyisoprenoid levels in yeast
transformed with vectors expressing the human and the yeast
enzymes (Figures 6A–6E), cultured in minimal media and
harvested during the log phase. The important accumulation of
polyprenol detected in the dfg10-100 strain (Figure 6B) is effi-
ciently and specifically corrected in yeast transformed with the
SRD5A3 gene (Figure 6D) compared to other human steroid
5a-reductases (Figure S5), whereas a mutant SRD5A3 (H296G)
encoding an enzyme predicted to be inactive (Wigley et al.,
1994) did not show any reduction of the accumulated polyprenol
(Figure 6E). To evaluate whether SRD5A3 was able to facilitate
polyprenol reduction, we used lysates of transfected HEK293T
cells overexpressing tagged SRD5A3. Exogenous polyprenol-
18 was added in a buffer containing different detergents and
mixed with lysates containing NADPH (Figures 6F–6H0). Some
exogenous polyprenol-18 was elongated to polyprenol-19,
indicating a good incorporation of this lipid in the protein-lipid
complexes from the lysate. The most efficient in vitro reduction
was obtained with 0.1% Triton X-100; the reduction efficiency
by the lysate of transfected cells with an empty vector was
comparable with the previously described assay (Sagami et al.,
1993). In control HEK293T cells, we found ?28% exogenous
polyprenol reduced to dolichol, whereas in lysates overexpress-
Figure S6E). These results suggest that transfected SRD5A3
promotes efficient reduction of polyprenol.
Biological Activity of SRD5A3
SRD5A3 sequence predicts a steroid 5a-reductase domain, and
some enzymes with this domain are able to reduce a variety of
steroid hormones with a delta4,5,3-oxo structure (Russell and
Wilson, 1994). Mutation of the SRD5A2 gene in human causes
male pseudohermaphroditism as a result of an enzymatic block
of testosterone to dihydrotestosterone conversion (Andersson
et al., 1991), and mutation of the Srd5a1 gene in mice affects
androgen metabolism (Mahendroo and Russell, 1999). Interest-
ingly, a previous study suggested that cell extract containing
overexpressed SRD5A3 was able to reduce testosterone to
dihydrotestosterone (Uemura et al., 2008), albeit at a slow rate.
However, both our biochemical and clinical investigations in
the patients with SRD5A3 mutations indicate that the nature of
the substrate of the SRD5A3 enzyme is not related to the
steroids. Our patients displayed no abnormal sexual abnormali-
ties that would have suggested a primary defect of steroid
metabolism. Moreover, karyotype analysis excluded the possi-
bility of sex reversal in all (data not shown). These observations
be a different lipid. Polyprenols share a common origin with
cholesterol because they are also built from isoprene units.
Another enzyme with a predicted steroid 5a-reductase
domain, Tsc13/GPSN2, has been shown to be an enoyl
(Kohlwein et al., 2001). This study also illustrates that the pre-
dicted steroid 5a-reductase domain is involved in the reduction
of a nonsteroid lipid and suggests that the full spectrum of lipid
Current Models for Dolichol Biosynthesis
Several mechanisms have been proposed for the last steps of
dolichol biosynthesis. One postulated an initial dephosphoryla-
tion of polyprenol diphosphate followed by reduction to Dol-P,
then dephosphorylation of Dol-P to produce dolichol (Chojnacki
and Dallner, 1988). However, several studies demonstrated the
tion of Dol-P (Heller et al.,1992; Rossignol etal., 1983). Asecond
proposal suggested that the final condensation reaction of
Pol-PP uses isopentenol instead of isopentenyl-PP. In this reac-
tion, Pol-PP is directly transformed to a one isoprene unit longer
(E) Immunofluorescence staining showing expression of BiP protein (red), a marker of the UPR pathway activation, in the neuroepithelium of the forebrain vesicle
(arrowheads). The scale bar represents 50 mm.
See also Figure S4.
Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc. 211
E11.5 Srd5a3 +/+ embryos
E11.5 Srd5a3 Gt/Gt embryos
Transfer of the glycan part on
Ratios of plasma polypreoids
Figure 5. Analysis of Polyprenols and Dolichols, in Yeast, Mouse, and Human with LC-MS
the building block for all polyprenoids. IPP molecules are added sequentially in trans-configuration on dimethylallyl pyrophosphate (DMAPP) via the farnesyl pyro-
bothphosphate residues are released by unidentifiedphosphatases. The alpha-isoprene unit of the polyprenol is subsequentlyreduced by an NADPH-dependent
microsomal reductase. For this step, the corresponding aldehydes have also been suggested as intermediates (Sagami et al., 1996). Finally the dolichol-specific
kinase (SEC59/DK) transfers a phosphate from CTP to dolichol. Dol-P is used to build the lipid linked oligosaccharide (LLO). Once the oligosaccharide structure
is transferred to specific asparagine residues, Dol-P is released on the luminal leaflet of the ER and is dephosphorylated by a pyrophosphatase (CWH8).
212 Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc.
dolichol, thus circumventing the dephosphorylation steps
(Ekstro ¨m et al., 1987). The third proposal is the most widely
accepted (Figure 5A) based on the finding of high concentrations
of polyprenol during the initial phase of dolichol biosynthesis
(Ekstro ¨m et al., 1984) and the detection of a basal polyprenol
reductase activity, in vitro (Sagami et al., 1993). However, the
reductase postulated in this reaction had not been identified,
and thus these models could not be directly evaluated. Our
results suggest that SRD5A3 is the polyprenol reductase, which
is consistent with the last model, confirming that the reduction of
polyprenol is the major pathway for dolichol biosynthesis.
Residual Dolichol in SRD5A3 Mutants
DFG10 mutants, suggesting the existence of another de novo
biosynthetic pathway for dolichol production. The presence of
whichwas reportedto benegligible inrat (Keller etal.,1982),and
the nature of the mutations in human, mouse, and yeast sug-
gests that these organisms have null mutations for this gene.
These observations indicate the existence of an alternative
pathway for de novo synthesis of dolichol in eukaryotic cells.
Disruption of LLO biosynthesis due to mutation in Dpagt1 in
mouse results in embryonic lethality at E5 (Marek et al., 1999),
a more severe phenotype than that observed in Srd5a3 mutant
mouse embryos, consistent with an alternative pathway. One
candidate is the TSC13 gene, the only other gene in S. cerevisiae
encoding a steroid 5a-reductase domain (Pfam database). We
tested whether the tsc13 mutant had abnormal CPY glycosyla-
tion and whether the dfg10/tsc13 double mutant showed further
increase of the polyprenol/dolichol ratio, but found no effect of
either (Figure S3), suggesting that the alternative pathway for
dolichol synthesis is independent of these genes. Interestingly,
among the pathways activated in embryonic mouse mutants
was the mevalonate pathway, including the isoprenoid biosyn-
thetic enzymes (Tables S1 and S2). This could suggest a positive
feedback mechanism, which might help organisms overcome
a partial block of these pathways.
Phenotypic Spectrum Resulting from Disruption
of Dolichol Metabolism
Tissues affected in patients with SRD5A3 mutations, such as
nervous system, ocular structures, skin, or coagulation factors,
reflect sensitivity for alteration in N-glycosylation. Such congen-
ital defects and the detection of a restricted expression pattern
of Srd5a3 in mouse embryo suggest a spatial-temporal require-
ment during development. N-glycan number and branching
regulate surface glycoprotein levels, affecting cell proliferation
and differentiation (Lau et al., 2007). N-glycosylation may help
regulate specific developmental pathways yet to be discovered.
A comparable multisystem disorder has been recently mapped
to the same locus, suggesting that these patients have the
same genetic defect (Kahrizi et al., 2009).
Although we find defects in the N-glycosylation pathway,
dolichol is also required for the synthesis of O-mannose-linked
synthesis, and some of the pathology may derive from these
defects, not explored here. Furthermore, little is known about
the glycosylation-independent functions of dolichol, considered
as a general membrane component in mammalian cells (Rip
et al., 1985).
The pathogenesis and phenotypic specificity of CDGs
deserves further investigations. However, our results point to
an unsuspected role for a steroid reductase-like enzyme in the
pathogenesis of one type of CDG, presumably mediated by
a requirement for dolichol synthesis.
All patients were enrolled according to approved human subjects protocol at
respective institutions. DNA was extracted from peripheral blood leukocytes
by salt extraction, genotyped with the Illumina Linkage IVb mapping panel
(Murray et al., 2004), and analyzed with easyLINKAGE-Plus software (Hoff-
mann and Lindner, 2005). Parameters were set to autosomal recessive with
full penetrance and disease allele frequency of 0.001. Genomic regions with
LOD scores over 2 were considered as candidate intervals. Linkage simula-
tions were performed with Allegro 1.2c under the same parameters, with
5000 markers at average 0.64 cM intervals, codominant allele frequencies,
and parametric calculations (Hoffmann and Lindner, 2005).
Mutation and CDG Screening
We performed direct bidirectional sequencing of the five coding exons and
splice junction sites of SRD5A3 via BigDye Terminator cycle sequencing
(Applied Biosystems). We screened 31 patients with CDG-Ix and seven
patients from a cohort with CDG-Ix and either strong clinical overlap such as
severe congenital eye malformation and/or indications for a dolichol-phos-
phate biosynthesis defect. Clinical description of patients 08-0486, 08-0487,
and 07-0419 was previously reported, corresponding respectively to patients
3, 5, and 7 (Morava et al., 2008) and 25, 26, and 27 (Morava et al., 2009). CDG
transferrin (O’Brien et al.,2007) or with transferrin isoelectric focusing (de Jong
et al., 1994).
GlcNAc-Transferase Assays in Fibroblasts
Skin fibroblasts from the patients and controls were cultured in Dulbecco’s
modified Eagle’s medium 10% fetal calf serum. Microsomal membranes
were prepared as described (Thiel et al., 2002) and suspended in 20 mM
Tris-HCl (pH 7.1), 10 mM MgCl2, and 1 mM dithiothreitol (DTT).
For measurement of the transfer of GlcNAc from UDP-GlcNAc to endogenous
lipid acceptor in microsomal membranes, the reaction contained 50 mM
Tris-HCl (pH 7.5), 0.1 mCi UDP-[14C]GlcNAc (specific activity 262 mCi/ mmol),
triple arrows highlight natural isotopic distribution, each offset by one atomic mass unit (amu). Polyprenol was detected only in mutant (red), by identification of
a spectrum that partially overlapped with dolichol. Quantification of the dolichol content of yeast mutant indicates 70% reduction (after subtraction of polyprenol
isotopic contribution; see the Experimental Procedures; white bar, WT; gray bar, mutant).
(C) Scans of Dolichol-18,19 in WT and E11.5 Srd5a3Gt/Gtembryos showing the accumulation of corresponding polyprenols in the mutant embryos.
statistical analysis. Color bars represent individual measurement for five patients with SRD5A3 mutation. Error bar was not generated for pol-20/dol-20 ratio in the
group CDGI-a,c as polyprenol levels were undetectable in three out of four patients.
Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc. 213
Polyprenol-18 + HEK293T with GFP
Polyprenol-18 + HEK293T with GFP-SRD5A3
s t n
, y t i s
e t n I
dfg10-100 mutant + wt SRD5A3
dfg10-100 mutant strain + vector
wt strain + vector
dfg10-100 mutant strain + wt DFG10 gene
dfg10-100 mutant + SRD5A3 H296G
Figure 6. In Vivo and In Vitro Polyprenol Reduction Promoting Activity of SRD5A3
(A–E) LC-MS analysis of lipid extracts from yeast cultured in minimal media.
(A) Only dolichol is detected in WT yeast strains transformed with pYX212 empty vector.
(B) In dfg10-100 strain transformed with pYX212 empty vector, accumulation of polyprenol relative to dolichol is evident. An additional compound (arrows) was
tentatively identified as polyprenal, previously suggested as an intermediate in yeast during in vitro dolichol biosynthesis (Sagami et al., 1996).
(C) In the dfg10-100 strain transformed with the WT DFG10 gene, no polyprenol accumulation was detected.
214 Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc.
15 mM MgCl2, 0.8 mM DTT, 26% glycerol, and 150 mg protein in a final volume
of 60 ml. After 15 min at 24?C, the reaction was stopped with chloroform/meth-
anol (3/2, by volume) and processed by phase partitioning (Sharma et al.,
1982). Radioactive glycolipids were separated on silica gel 60 plates (Merck)
developed in chloroform/methanol/water (65/25/4, by volume). Radioactivity
was detected and quantified by phosphorimager (Molecular Dynamics).
For determination of the transfer to exogenous Dol-P, the reaction contained
3.6 mg Dol-P, 2.4 mM diheptanoyl-phosphatidylcholine, 38 mM Tris-HCl
(pH 7.5), 0.1 mCi UDP-[14C]GlcNAc (specific activity 262 mCi/ mmol), 11 mM
MgCl2, 0.7 mM DTT, 25% glycerol, and 150 mg protein in a final volume of
60 ml. Incubation and processing of the reaction was as in assay I.
Construction of DsRed-SRD5A3 and GFP-SRD5A3 Expression
The ORF of SRD5A3 was amplified from a human fetal brain complementary
DNA (cDNA) library and cloned in phrGFP II-N (Stratagene) and pDsRed2-
C1 (Clontech) vectors. Site-directed mutagenesis was performed with
QuickChange II site-directed mutagenesis kit (Stratagene) according to the
Analysis of Srd5a3Gt/GtEmbryos, Microarray Experiments,
and Quantitative PCRs
Frozen embryos (129/SvEvBrd 3 C57BL6/J mix) carrying a gene trap insertion
in one allele of Srd5a3 were obtained from Lexicon and transferred in pseudo-
pregnant female (Renard and Babinet, 1984). Genotyping was performed
by PCR with yolk sac-extracted DNA. Whole-transcriptome analysis was
performed with the 44k Agilent genome oligo microarray kit with four embryos
Srd5a3Gt/Gtand four littermate controls from two different litters. For these
experiments, E8.5 embryos, which had not started the turning process and
didnot yet show morphological defect, werechosen. Real-timePCR reactions
were performed in the LightCycler 480 system (Roche) with the SYBR Green I
Master Kit (Roche). Seven of the eight genes found misregulated with the
microarray experiment and tested by real-time PCR were found to be compa-
rably and significantly misregulated. Animals were used in compliance with
approved institutional policies.
Mass-Spectrometry Analysis of Yeast and Mouse Samples
Lipid extraction (Bligh and Dyer, 1959) and LC-MS analysis was performed as
described (Garrett et al., 2007). For quantitative measurements, nor-dolichol
(Avanti Polar Lipids) (Garrett et al., 2007) was added to the sample before lipid
extraction. For a detailed description, please see the Extended Experimental
Analysis of Plasma Polyprenoids
Plasma (500 ml) samples from controls (n = 10), all with normal transferrin
isofocusing profile), CDG-I patients with known defect (n = 4), and SRD5A3
patients (n = 5) were subjected to saponification (Yasugi and Oshima, 1994)
by addition of 5 M KOH in water (500 ml) and MeOH (1500 ml) for 1 hr at
100?C under nitrogen atmosphere. Lipids were extracted with hexane (2 3
1.5 ml), the organic phase was washed with 1.5 ml water, and dried. Samples
were dissolved in hexane-MeOH (100 ml, 1:2 v/v), and 5 ml was injected on a
50 3 2 mm monolythic column (C18, Onyx) coupled to a Quattro LC-ESI
tandem mass spectrometer (MicroMass). Polyprenoids were eluted with a
MeOH/Isopropanol gradient containing 1% 50 mM LiI. MRM transitions
[Dol-n]Li+ / 162+ and [Pren-n]Li+ / [Pren-n-H2O]Li+ were used to calculate
response areas/ml plasma of respectively dolichols and polyprenols
(n, number of isoprene units) (D’Alexandri et al., 2006).
Polyprenol Reduction Assay
The assay was performed as a modification of the procedure described previ-
ously (Sagami et al., 1993). The reaction mixture consisted of 50 mM Tris-HCl
and 4 mg/ml polyprenol C90, previously dissolved in ethanol. After 15 min
sonication in a bath apparatus, the reaction was started with the addition of
5 mM NADPH and 700 mg crude cell-extract proteins, to a final volume of
250 ml. Reactions were incubated for 12 hr at 37?C. Samples were lipid
extracted (Bligh and Dyer, 1959) after mixing with an internal standard of
nor-dolichol and analyzed by LC-MS.
Phylogenetic tree representation was done with phylogeny (Dereeper et al.,
2008). (http://www.phylogeny.fr/version2_cgi/index.cgi). Topology prediction
was performed with TMHMM, a program for predicting membrane-spanning
segments based on hidden Markov model (http://www.cbs.dtu.dk/services/
tase domain (PF02544) (http://pfam.sanger.ac.uk/family/PF02544/).
Microarray data are available in the Array Express Archive database with
accession number E-MEXP-2713 (http://www.ebi.ac.uk/microarray-as/ae/).
Supplemental Information includes Extended Experimental Procedures,
six figures, and three tables and can be found with this article online at
We thank G.R. Fink, A. Jansen, F. Karst, K. Gable, T.M. Dunn, and R. Kolodner
for providing yeast strains and helpful advice. We are grateful to J.H. Lin for
valuable discussion. We thank D. Matern and the Mayo Clinic for providing
complete results of patients’ transferrin analysis and also C. Sault for addi-
tional clinical results from family CVH-385. We thank the University of Califor-
nia, San Francisco, microscopy core (P30 NS047101 and DK80506) and the
Biomedical Genomics Core for help in imaging and microarray data analysis.
A. de Rooij and K. Huyben are gratefully acknowledged for technical assis-
tance. H.H.F. is a Sanford Research Professor. He and B.N. are supported
by the Rocket Fund, R01 DK55615, and the Sanford Children’s Health
Research Center. Financial support from Euroglycanet (LSHM-CT2005-
512131) to R.W. and Metakids and the Netherlands Brain Foundation to D.L.
are kindly acknowledged. L.L. was supported by grants from the Deutsche
Forschungsgemeinschaft and the Ko ¨rber-Stiftung. The mass-spectrometry
facility in the Department of Biochemistry of the Duke University Medical
Center and Z.G. are supported by the LIPID MAPS Large Scale Collaborative
Grant GM-069338 from the National Institutes of Health. V.C. is supported by
a fellowship from Fondation pour la Recherche Me ´dicale, and J.G.G. is an
(D) Transformation of the dfg10-100 strain with the human SRD5A3 gene corrects polyprenol accumulation, although low levels are still detected.
(E) Transformation of the dfg10-100 strain with the human SRD5A3 enzymatically null H296G mutation fails to correct the polyprenol accumulation.
(F–H) LC-MS analysis of a lipid extract from an in vitro experiment performed with the exogenous substrate, polyprenol-18, in which cell lysates were used as the
source of enzyme.
(F and F0) Polyprenol-18 spectrum after incubation in the reaction buffer without cell lysate. No polyprenol-19 or any forms of dolichol are evident.
(G and G0) After incubation with HEK293T cell lysate transfected with GFP, part of polyprenol-18 is elongated in polyprenol-19 and 28% is reduced to the
(H and H0) In the presence of lysate from cell overexpressing GFP-SRD5A3, 67% of the initial polyprenol is reduced to dolichol (arrows). Both cell lysates show
similar elongation of polyprenols.
See also Figures S5 and S6.
Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc. 215
Investigator of the Howard Hughes Medical Institute. L.A.-G. ascertained and
phenotyped family CVH-385 and MR3 and proposedthe collaboration to iden-
tify the defective gene. W.B.D. reviewed brain imaging and suggested CDG
defects, and H.F. suggested SRD5A3 as the polyprenol reductase. D.S.,
A.P.D.B., and H.v.B. coordinated homozygosity mapping. J.L.S. identified
the SRD5A3 mutation, and S.L.B. performed linkage analysis and initiated
generation of the gene trap mouse line. E.M. performed CDG Ix phenotype
assessment on the remaining patients. D.J.L. performed biochemical diag-
nosis, functional studies, and dolichol analysis. R.A.W. coordinated biochem-
ical analysis and homozygosity mapping. L.L. suggested a defect in dolichol
biosynthesis and performeddolicholrescue experiments. E.S. helped with do-
lichol standards. D.B.-V., M.A., P.B., and J.S.C. contributed patients. Mass-
spectrometry analysis was performed by Z.G. under the guidance of
C.R.H.R. B.G.N. performed lipid extraction and patient LLO analysis. S.H.
and B.R.A. provided technical assistance, discussions, and control samples.
H.H. provided assistance with standard yeast techniques. V.C. performed all
other experiments. J.G.G. directed the project. V.C. and J.G.G. wrote the
manuscript, with help from the other authors.
Received: January 30, 2010
Revised: March 26, 2010
Accepted: May 6, 2010
Published online: July 15, 2010
Acosta-Serrano, A., O’Rear, J., Quellhorst, G., Lee, S.H., Hwa, K.Y., Krag,
S.S., and Englund, P.T.(2004). Defects intheN-linked oligosaccharide biosyn-
thetic pathway in a Trypanosoma brucei glycosylation mutant. Eukaryot. Cell
Aebi, M., and Hennet, T. (2001). Congenital disorders of glycosylation: genetic
model systems lead the way. Trends Cell Biol. 11, 136–141.
Aggarwal, S., Thareja, S., Verma, A., Bhardwaj, T.R., and Kumar, M. (2010).
An overview on 5alpha-reductase inhibitors. Steroids 75, 109–153.
Al-Gazali, L., Hertecant, J., Algawi, K., El Teraifi, H., and Dattani, M. (2008).
A new autosomal recessive syndrome of ocular colobomas, ichthyosis, brain
malformations and endocrine abnormalities in an inbred Emirati family. Am.
J. Med. Genet. A. 146, 813–819.
Alber, T., and Kawasaki, G. (1982). Nucleotide sequence of the triose
phosphate isomerase gene of Saccharomycescerevisiae.J.Mol. Appl. Genet.
Allen, C.M., Jr., Kalin, J.R., Sack, J., and Verizzo, D. (1978). CTP-dependent
dolichol phosphorylation by mammalian cell homogenates. Biochemistry 17,
of steroid 5 alpha-reductase 2 gene in male pseudohermaphroditism. Nature
Behrens, N.H., and Leloir, L.F. (1970). Dolichol monophosphate glucose:
an intermediate in glucose transfer in liver. Proc. Natl. Acad. Sci. USA 66,
Bligh, E.G., and Dyer, W.J. (1959). A rapid method of total lipid extraction and
purification. Can. J. Biochem. Physiol. 37, 911–917.
Chavan, M., and Lennarz, W. (2006). The molecular basis of coupling of
translocation and N-glycosylation. Trends Biochem. Sci. 31, 17–20.
Chojnacki, T., and Dallner, G. (1988). The biological role of dolichol. Biochem.
J. 251, 1–9.
D’Alexandri, F.L., Gozzo, F.C., Eberlin, M.N., and Katzin, A.M. (2006).
via Li+ cationization. Anal. Biochem. 355, 189–200.
de Jong, G., van Noort, W.L., and van Eijk, H.G. (1994). Optimized separation
and quantitation of serum and cerebrospinal fluid transferrin subfractions
defined by differences in iron saturation or glycan composition. Adv. Exp.
Med. Biol. 356, 51–59.
Dereeper, A., Guignon, V., Blanc, G., Audic, S., Buffet, S., Chevenet, F.,
Dufayard, J.F., Guindon, S., Lefort, V., Lescot, M., et al. (2008). Phylogeny.fr:
robust phylogenetic analysis for the non-specialist. Nucleic Acids Res.
36(Web Server issue), W465–W469.
Eklund, E.A., and Freeze, H.H. (2006). The congenital disorders of glycosyla-
tion: a multifaceted group of syndromes. NeuroRx 3, 254–263.
Ekstro ¨m, T.J., Chojnacki, T., and Dallner, G. (1984). Metabolic labeling of
dolichol and dolichyl phosphate in isolate hepatocytes. J. Biol. Chem. 259,
Ekstro ¨m, T.J., Chojnacki, T., and Dallner, G. (1987). The alpha-saturation and
terminal events in dolichol biosynthesis. J. Biol. Chem. 262, 4090–4097.
Freeze, H.H. (2006). Genetic defects in the human glycome. Nat. Rev. Genet.
Garrett, T.A., Guan, Z., and Raetz, C.R. (2007). Analysis of ubiquinones,
dolichols, and dolichol diphosphate-oligosaccharides by liquid chromatog-
raphy-electrospray ionization-mass spectrometry. Methods Enzymol. 432,
Gru ¨newald, S., and Matthijs, G. (2000). Congenital disorders of glycosylation
(CDG): a rapidly expanding group of neurometabolic disorders. Neuropediat-
rics 31, 57–59.
Haeuptle, M.A., and Hennet, T. (2009). Congenital disorders of glycosylation:
an update on defects affecting the biosynthesis of dolichol-linked oligosac-
charides. Hum. Mutat. 30, 1628–1641.
and perturbation of its biosynthesis. Eur. J. Biochem. 91, 567–575.
Helenius, A., and Aebi, M. (2001). Intracellular functions of N-linked glycans.
Science 291, 2364–2369.
Heller, L., Orlean, P., and Adair, W.L., Jr. (1992). Saccharomyces cerevisiae
sec59 cells are deficient in dolichol kinase activity. Proc. Natl. Acad. Sci.
USA 89, 7013–7016.
Hoffmann, K., and Lindner, T.H. (2005). easyLINKAGE-Plus—automated
linkage analyses using large-scale SNP data. Bioinformatics 21, 3565–3567.
Jaeken, J., and Matthijs, G. (2007). Congenital disorders of glycosylation:
a rapidly expanding disease family. Annu. Rev. Genomics Hum. Genet. 8,
Jaeken, J., Hennet, T., Matthijs, G., and Freeze, H.H. (2009). CDG nomencla-
ture: time for a change!. Biochim. Biophys. Acta 1792, 825–826.
Jones, M.B., Rosenberg, J.N., Betenbaugh, M.J., and Krag, S.S. (2009).
Structure and synthesis of polyisoprenoids used in N-glycosylation across
the three domains of life. Biochim. Biophys. Acta 1790, 485–494.
Kahrizi, K., Najmabadi, H., Kariminejad, R., Jamali, P., Malekpour, M.,
Garshasbi, M., Ropers, H.H., Kuss, A.W., and Tzschach, A. (2009). An auto-
somal recessive syndrome of severe mental retardation, cataract, coloboma
and kyphosis maps to the pericentromeric region of chromosome 4. Eur. J.
Hm. Genet. 17, 125–128.
Kato, S., Tsuji, M., Nakanishi, Y., and Suzuki, S. (1980). Enzymatic
dephosphorylation of dolichyl pyrophosphate—the bacitracin-sensitive,
rate-limiting step for dolichyl mannosyl phosphate synthesis in rat liver
microsomes. Biochem. Biophys. Res. Commun. 95, 770–776.
Keller, R.K., Jehle, E., and Adair, W.L., Jr. (1982). The origin of dolichol in the
liver of the rat. Determination of the dietary contribution. J. Biol. Chem. 257,
Kohlwein, S.D., Eder, S., Oh, C.S., Martin, C.E., Gable, K., Bacikova, D., and
Dunn, T. (2001). Tsc13p is required for fatty acid elongation and localizes
to a novel structure at the nuclear-vacuolar interface in Saccharomyces
cerevisiae. Mol. Cell. Biol. 21, 109–125.
Kranz, C., Jungeblut, C., Denecke, J., Erlekotte, A., Sohlbach, C., Debus, V.,
Kehl, H.G., Harms, E., Reith, A., Reichel, S., et al. (2007). A defect in dolichol
phosphate biosynthesis causes a new inherited disorder with death in early
infancy. Am. J. Hum. Genet. 80, 433–440.
Lau, K.S., Partridge, E.A., Grigorian, A., Silvescu, C.I., Reinhold, V.N.,
Demetriou, M., and Dennis, J.W. (2007). Complex N-glycan number and
216 Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc.
degree of branching cooperate to regulate cell proliferation and differentiation.
Cell 129, 123–134.
Mahendroo, M.S., and Russell, D.W. (1999). Male and female isoenzymes of
steroid 5alpha-reductase. Rev. Reprod. 4, 179–183.
Marek, K.W., Vijay, I.K., and Marth, J.D. (1999). A recessive deletion in the
GlcNAc-1-phosphotransferase gene results in peri-implantation embryonic
lethality. Glycobiology 9, 1263–1271.
Morava, E., Wosik, H., Ka ´rteszi, J., Guillard, M., Adamowicz, M., Sykut-
Cegielska, J., Hadzsiev, K., Wevers, R.A., and Lefeber, D.J. (2008). Congenital
disorder of glycosylation type Ix: review of clinical spectrum and diagnostic
steps. J. Inherit. Metab. Dis. 31, 450–456.
Morava, E., Wosik, H.N., Sykut-Cegielska, J., Adamowicz, M., Guillard, M.,
Wevers, R.A., Lefeber, D.J., and Cruysberg, J.R. (2009). Ophthalmological
abnormalities in children with congenital disorders of glycosylation type I.
Br. J. Ophthalmol. 93, 350–354.
Mo ¨sch, H.U., and Fink, G.R. (1997). Dissection of filamentous growth by trans-
poson mutagenesis in Saccharomyces cerevisiae. Genetics 145, 671–684.
Murray, S.S., Oliphant, A., Shen, R., McBride, C., Steeke, R.J., Shannon, S.G.,
Rubano, T., Kermani, B.G., Fan, J.B., Chee, M.S., and Hansen, M.S. (2004).
A highly informative SNP linkage panel for human genetic studies. Nat.
Methods 1, 113–117.
Nishikawa, A., and Mizuno, S. (2001). The efficiency of N-linked glycosylation
origin. Biochem. J. 355, 245–248.
O’Brien, J.F., Lacey, J.M., and Bergen, H.R., 3rd. (2007). Detection of hypo-N-
glycosylation using mass spectrometry of transferrin. Curr. Protoc. Hum.
Genet., Chapter 17, Unit 17.4.
Renard, J.P., and Babinet, C. (1984). High survival of mouse embryos after
rapid freezing and thawing inside plastic straws with 1-2 propanediol as
cryoprotectant. J. Exp. Zool. 230, 443–448.
Rip, J.W., Rupar, C.A., Ravi, K., and Carroll, K.K. (1985). Distribution, metab-
olism and function of dolichol and polyprenols. Prog. Lipid Res. 24, 269–309.
Rosenwald, A.G.,Stanley,P.,McLachlan,K.R., and Krag, S.S.(1993). Mutants
indolichol synthesis: conversion of polyprenoltodolichol appears tobe arate-
limiting step in dolichol synthesis. Glycobiology 3, 481–488.
Rossignol, D.P., Scher, M., Waechter, C.J., and Lennarz, W.J. (1983).
Metabolic interconversion of dolichol and dolichyl phosphate during develop-
ment of the sea urchin embryo. J. Biol. Chem. 258, 9122–9127.
Ruiz-Canada, C., Kelleher, D.J., and Gilmore, R. (2009). Cotranslational and
posttranslational N-glycosylation of polypeptides by distinct mammalian
OST isoforms. Cell 136, 272–283.
Russell, D.W., and Wilson, J.D. (1994). Steroid 5 alpha-reductase: two genes/
two enzymes. Annu. Rev. Biochem. 63, 25–61.
Sagami, H., Kurisaki, A., and Ogura, K. (1993). Formation of dolichol from
dehydrodolichol is catalyzed by NADPH-dependent reductase localized in
microsomes of rat liver. J. Biol. Chem. 268, 10109–10113.
Sagami, H., Igarashi, Y., Tateyama, S., Ogura, K., Roos, J., and Lennarz, W.J.
(1996). Enzymatic formation of dehydrodolichal and dolichal, new products
related to yeast dolichol biosynthesis. J. Biol. Chem. 271, 9560–9566.
Sharma, C.B., Lehle, L., and Tanner, W. (1982). Solubilization and character-
ization of the initial enzymes of the dolichol pathway from yeast. Eur. J.
Biochem. 126, 319–325.
Sidbury, R., and Paller, A.S. (2001). What syndrome is this? CHIME syndrome.
Pediatr. Dermatol. 18, 252–254.
Swiezewska, E., and Danikiewicz, W. (2005). Polyisoprenoids: structure,
biosynthesis and function. Prog. Lipid Res. 44, 235–258.
Thiel, C., Schwarz, M., Hasilik, M., Grieben, U., Hanefeld, F., Lehle, L.,
von Figura, K., and Ko ¨rner, C. (2002). Deficiency of dolichyl-P-Man:Man7Glc-
NAc2-PP-dolichyl mannosyltransferase causes congenital disorder of glyco-
sylation type Ig. Biochem. J. 367, 195–201.
Uemura, M., Tamura, K., Chung, S., Honma, S., Okuyama, A., Nakamura, Y.,
overexpressed in hormone-refractory prostate cancer. Cancer Sci. 99, 81–86.
Vermeer, S., Kremer, H.P., Leijten, Q.H., Scheffer, H., Matthijs, G., Wevers,
R.A., Knoers, N.A., Morava, E., and Lefeber, D.J. (2007). Cerebellar ataxia
and congenital disorder of glycosylation Ia (CDG-Ia) with normal routine
CDG screening. J. Neurol. 254, 1356–1358.
Vleugels, W., Haeuptle, M.A., Ng, B.G., Michalski,J.C., Battini, R., Dionisi-Vici,
C., Ludman, M.D., Jaeken, J., Foulquier, F., Freeze, H.H., et al. (2009). RFT1
deficiency in three novel CDG patients. Hum. Mutat. 30, 1428–1434.
Wigley, W.C., Prihoda, J.S., Mowszowicz, I., Mendonca, B.B., New, M.I.,
Wilson, J.D., and Russell, D.W. (1994). Natural mutagenesis study of the
human steroid 5 alpha-reductase 2 isozyme. Biochemistry 33, 1265–1270.
Wolf, M.J., Rush, J.S., and Waechter, C.J. (1991). Golgi-enriched membrane
fractions from rat brain and liver contain long-chain polyisoprenyl pyrophos-
phate phosphatase activity. Glycobiology 1, 405–410.
Yasugi, E., and Oshima, M. (1994). Sequential microanalyses of free dolichol,
dolichyl fatty acid ester and dolichyl phosphate levels in human serum.
Biochim. Biophys. Acta 1211, 107–113.
Cell 142, 203–217, July 23, 2010 ª2010 Elsevier Inc. 217