(Received for publication, January 10, 1997, and in revised form, May 28, 1997)
From the Department of Pediatrics, University of
California San Francisco, San Francisco, California 94143 and the
§ Department of Biological Chemistry, University of
California Los Angeles, Los Angeles, California 90024
X-linked ichthyosis is the result of steroid sulfatase (STS) deficiency. While most affected individuals have extensive deletions of the STS gene, point mutations have been reported in three patients (1). In this study, we identify an additional three point mutations and characterize the effects of all six mutations on STS activity and expression. All six are unique single base pair substitutions. The mutations are located in a 105-amino acid region of the C-terminal half of the polypeptide. Five of the six mutations involve the substitutions of Pro or Arg for Trp372, Arg for His444, Tyr for Cys446, or Leu for Cys341. The other mutation is in a splice junction and results in a frameshift causing premature termination of the polypeptide at residue 427. All the affected residues are conserved to some degree within the sulfatase family. The six mutations were reproduced in normal STS cDNA and transiently expressed in STS-deficient cells. All six mutant vectors direct the expression of STS protein that lacks enzymatic activity. The mutant polypeptides show a shift in mobility on SDS-PAGE and resistance to proteinase K digestion when translated in the presence of dog pancreas microsomes, indicating glycosylation and normal translocation.
Steroid sulfatase (STS;1
steryl-sulfatase or steryl-sulfate sulfohydrolase, EC 3.1.6.2) is a
membrane-bound microsomal enzyme, ubiquitously expressed in mammalian
tissues, which hydrolyzes various 3-hydroxysteroid sulfates (2).
X-linked ichthyosis (XLI) is caused by steroid sulfatase deficiency,
one of the most prevalent human inborn errors of metabolism. The
estimated incidence of this disease is between 1/6000 and 1/2000 males
and occurs with a similar frequency in all ethnic groups studied to
date (3). Clinically, the absence of STS activity during fetal life results in diminished fetal-maternal estrogen production. This has
sometimes been associated with delayed progression of parturition (3).
Postnatally patients have abnormal levels of sulfated steroids in
tissues and body fluids, ichthyosis, and corneal opacities (2).
The STS gene has been cloned and characterized (4). It contains 10 exons spread over 146 kilobase pairs of the short arm of the X chromosome. The encoded polypeptide has a molecular mass of 62 kDa and contains four predicted possible N-linked glycosylation sites (4, 5). Only two of these sites are used (5). Although the majority of STS-deficient patients have complete deletions of the STS gene (4, 6-8), 10% show normal genomic DNA and RNA hybridization patterns when probed with STS cDNA (9). Previous immunoprecipitation and immunoblotting experiments failed to identify STS cross-reacting material in the cells of these patients (1, 9). The mutations in these patients are of interest, since they may help define regions of the STS polypeptide that are important for structure or function.
In this report we describe the identification of point mutations in three unrelated patients with XLI. All three resulted in amino acid sequence alterations in the mutant STS proteins. The three mutations as well as an additional three mutations identified by Basler et al. (1) were each produced in an STS cDNA expression vector using site-directed mutagenesis or subcloning. The mutant STS constructs were used to study the effects of the six different mutations on STS activity and processing in mammalian cells and in a rabbit reticulocyte lysate system.
Cell lines from four patients with XLI were obtained from a variety of clinical sources and geographical locations, originally for diagnostic purposes. Z, T, and the deletion cell line O are fibroblast cell lines. S is a lymphoblast line. Z, T, and S are previously described cell lines and are known to have normal amounts and sizes of DNA and mRNA (9). O is also previously described and is known to have a complete deletion of the STS gene (9). Cell line Z was previously called C. The name was changed in this paper to avoid confusion with another unrelated STS-deficient patient also named C (1). IMR90 and IMR91 are normal human female and male fibroblast lines, respectively. W is a normal lymphoblast cell line that is used as a control for S. A9 is a well described mouse fibroblast cell line that does not express STS activity. It was used as a recipient in all transfections and as a negative control. The fibroblasts were maintained in Dulbecco's modified Eagle's medium (Life Technologies, Inc.) supplemented with 10% fetal bovine serum (Irvine scientific). The lymphoblasts were maintained in RPMI and 20% fetal bovine serum.
Numbering of Nucleotide and Amino Acid SequenceIn the
interest of clarity, the numbering of the nucleotides in the STS
cDNA as well as the numbering of the predicted amino acids was
changed. Nucleotide +1 is positioned at the major STS transcription
start site (10) instead of at the beginning of the STS cDNA (4).
The major transcription start site is at position 221 with respect to
the AUG translation initiation codon (10). The amino acids have been
renumbered, making the initiating methionine +1. This agrees with the
numbering of some of the other sulfatases (5, 11, 12). The initiating
methionine was previously numbered
22 to reflect the predicted signal
sequence (4). These changes account for the discrepancy between the
nucleotide and residue positions in this paper and the positions
previously reported (1, 4, 13).
Total RNA from fibroblasts and lymphoblasts was isolated by the guanidine isothiocyanate procedure (14). The RNA was precipitated in a final concentration of 2.5 M LiCl in an ice bath for 4 h. For cDNA synthesis, 5 µg of total RNA were mixed with 100 pmol of random hexamers pd(N6) (Pharmacia) in 1 × Moloney murine leukemia virus reverse transcriptase reaction buffer (Life Technologies, Inc.) in a total volume of 18 µl containing 0.5 µl of RNasin (Promega), 1 µg of bovine serum albumin, and 0.5 mM each dATP, dCTP, dTTP, and dGTP and heated to 70 °C for 5 min. After cooling on ice, 2 µl (400 units) of Moloney murine leukemia virus reverse transcriptase (Life Technologies, Inc.) were added. The reactions were incubated at 42 °C for 60 min.
The cDNA reaction was mixed with 1 µg each of primers PCR 3 (forward; 5-ACAATCTCACCCAGAGGCTAACGG-3
) and PCR 4 (reverse; 5
-TTTCCCCAGTGCCACTCTCAGGCG-3
), corresponding to positions 994-1017 and 2053-2029, in a total volume of 100 µl (containing 1 × PCR buffer (Cetus), and 0.5 mM each dATP, dCTP, dTTP, and
dGTP). After initial denaturation for 10 min at 95 °C, 0.5 µl of
Taq polymerase was added. The reactions were subjected to an
initial incubation at 72 °C for 10 min and then 30 cycles of
denaturation at 95 °C for 30 s, annealing at 65 °C for 1 min, and extension at 72 °C for 2 min (15).
The resulting DNA fragments were gel-purified, blunt ended with the Klenow fragment of DNA polymerase, and subcloned into the SmaI sites of M13 mp19 and mp18. Positive clones were identified by hybridization with the wild type STS cDNA clone p331 (4) EcoRI fragment.
cDNA Synthesis and Amplification of the 55 µg of total RNA was mixed with 100 pmol of pd(N6) random hexamers and incubated at 70 °C for 10 min. After chilling the reactions on ice, the first strand was synthesized in a 20-µl reaction containing 1 × Moloney murine leukemia virus reverse transcriptase reaction buffer, 1 mM each dATP, dCTP, dTTP, and dGTP, and 200 units of SuperScript reverse transcriptase (Life Technologies, Inc.). The reaction mixtures were incubated at 42 °C for 1 h and 90 °C for 5 min and then chilled on ice. 2 units of RNase H was added, and the reactions were incubated for an additional 20 min at 37 °C before proceeding to PCR amplification (16).
PCR amplification was performed by adding 1 µg each of primers PGEM 1 (forward; 5-CTGGAGATGCCTTTAAGG-3
) and PCR 8 (reverse; 5
-CTCCCGAAGAGGCTGTGAAGAGG-3
), corresponding to positions 216-233 and
549-527, or primers PCR 7 (forward; 5
-CGATCAGGAATGGCATCTTGGTCC-3
) and PCR 2 (reverse; 5
-TCCGCTGTATGAACTGGGCCGCC-3
), corresponding to positions 490-513 and 1043-1021, to 7 µl of the first strand cDNA mixtures in a final volume of 100 µl (containing 1 × PCR buffer (Cetus), 0.5 mM each dATP, dCTP, dTTP, and dGTP,
and 1.25 units of Taq polymerase). After an initial
denaturation step at 94 °C for 5 min, the reactions were subjected
to 30 cycles of denaturation at 94 °C for 1 min, annealing at
65 °C for 2 min, and extension at 72 °C for 3 min. The PCR
products were gel-purified and reamplified as above except that the
annealing was performed at 55 °C and the number of cycles was
lowered to 25. All amplifications were run with controls in which no
reverse transcriptase had been added in the cDNA synthesis, and no
DNA had been added in the amplification.
Genomic DNA was
isolated from peripheral white blood cells (normal control), and
fibroblasts (patients) by standard techniques (17). Primers flanking
the sequence of interest were used to amplify 100 ng of the genomic
DNA. The buffers and concentrations of the reagents were the same as
those used for the cDNA amplification. Following an initial
denaturization at 94 °C for 5 min, the amplification conditions were
15 cycles of denaturing at 94 ° for 45 s, annealing at 65 °C
for 1 min, and extension at 72 °C for 1 min. 5 µl of amplified DNA
was reamplified using the same conditions as above for an additional 25 cycles. Primers flanking exon 8 are EX8-3 (forward;
5-AATCTCCCTTGTTGCCTCTTACC-3
), starting 41 nucleotides upstream of the
5
splice site of exon 8, and EX8-2 (reverse; 5
-ACCATTATGCAGGCAAACCAGAGC-3
), beginning 80 nucleotides downstream of
the 3
splice site of exon 8. The primers flanking exon 9 are EX9-1
(forward; 5
-CTGATTCACGTCTGAATGCCTGGC-3
), starting 38 nucleotides upstream of the 5
splice site of exon 9, and EX9-2 (forward; 5
-GCATAGGTTGCACGGAGCATGAGG-3
), starting 49 nucleotides downstream of
the 3
splice site of exon 9.
Initially, the resulting DNA fragments were gel-purified, blunt ended, and subcloned into the SmaI sites of M13mp18, M13mp19, and pUC 19. Subsequently, the amplified DNA was subcloned directly into the TA cloning vector pCR I (Invitrogen). The subcloned PCR products were sequenced by dideoxy chain termination using the Sequenase system (U.S. Biochemical Corp.) (18). Both single-stranded and double-stranded templates were used.
Construction of Expression PlasmidsThe full-length STS
cDNA p331 was modified to eliminate ATGs 5 of the initiating ATG.
The modified fragment was created by PCR amplification, using p331 as
the template and primers PGEM 1 (forward) and PCR4 (reverse). Following
a 5-min denaturation at 94 °C, the reaction was subjected to 25 cycles of denaturation at 94 °C for 1 min, annealing at 65 °C for
1 min, and extension at 72 °C for 2.5 min.
The modified STS fragment was inserted into the blunt ended EcoRI site of the expression vector pSG5 (Stratagene) (19). Orientation was determined by digestion with PstI/BamHI and HindIII/BamHI and by sequencing.
PSTA 404-bp EcoNI/SacI fragment
containing the 19-bp insertion was cut out of the cDNA clone T4-4.
T4-4 is a M13 vector containing the 3 half of patient T's STS
cDNA. The cDNA insertion fragment was directionally cloned into
the SacI/EcoNI site of PS10, replacing the wild
type fragment. The ligated DNA was transformed into JM109. Orientation
and presence of the insertion was confirmed by sequencing.
PS10 was mutagenized according to instructions using two slightly different systems, (CLONTECH and Invitrogen) to produce vectors PSS, PSZ, PSD, PSA, and PSC, which contain five of the six STS point mutations. Briefly, the mutagenic primer (the point mutation primer) and the "selection" primer, which mutates the NdeI site in PS10 to an AvrII site, were annealed to double-stranded PS10 DNA. The mutant DNA strand was synthesized, and the parental DNA was linearized by digestion with NdeI for 2 h. 2 µl of the reaction was used to transform either BMH 71-18 mut S cells or XLmut S cells, respectively. The cells were grown in 5 ml of 2 × YT medium overnight in the presence of Ampicillin to enrich for mutant plasmids. Miniprep DNA was isolated using the alkali lysis procedure (20), subjected to a second round of digestion with NdeI, and transformed into either JM109 cells or XL1-Blue cells, respectively, and grown on plates overnight. The resulting colonies were screened for the absence of the NdeI site and the presence of the AvrII site by restriction digest analysis and then sequenced to verify the presence of the point mutations.
Transfections1 × 106 A9 cells/60-mm dish were cotransfected with 10 µg of control or patient vector and 1 µg of CAT plasmid using 30 µl of LipofectAmine (Life Technologies, Inc.) (21). The cells were exposed to the DNA-lipid complexes for 24 h and then were supplemented with 3 ml of Dulbecco's modified Eagle's medium containing 20% fetal bovine serum. The cells were harvested 72 h post-transfection. All transfections were performed in triplicate. Each triplicate was assayed for protein concentration, STS and CAT activities, and STS cross-reacting material.
Steroid Sulfatase AssaySteroid sulfatase activity was assayed using a modification of the method of Ref. 22. The cell pellet from a 60-mm plate of transfected A9 cells was resuspended in 120 µl of 0.25 M Tris-HCl, pH 8.0. The cells were lysed by alternatively freezing/thawing in a methanol-dry ice bath and a 37 °C bath five times and homogenizing by hand five times. 30-µl duplicates were each added to Eppendorf tubes containing 70 µl of H2O, 100 µl of 0.25 M Tris-HCl, pH 7.2, and 50 µl of 200 µM [3H]estrone sulfate substrate (NEN Life Science Products). The reactions were incubated at 37 °C for 4 h. The reactions were terminated by the addition of 500 µl of benzene and mixing. The reactions were spun at 1500 rpm for 1 min to remove the bubbles from the interface. 250 µl of the benzene phase, containing the free steroid, were counted in 10 ml of 0.8% butyl-PBD primary fluor (Beckman) in toluene.
Chloramphenicol Acetyltransferase and Protein AssaysA portion of the cell lysate prepared for the STS assay was also used for the protein and CAT assays. 50 µl of homogenized lysate was incubated at 60 °C for 10 min to inactivate endogenous deacetylase activity. The extracts were spun at 14,000 rpm for 2 min. 15-µl duplicates of the supernatant were assayed for CAT activity using a Promega kit. 2 µl of lysate was diluted and assayed for protein using an Aldrich kit.
AntibodiesModified STS cDNA was amplified using
primers PGEM 1 (forward) and PCR 4 (reverse) and inserted into the
EcoRI sites of the pATH 1, 10, and 11 expression vectors
(23). The vectors were grown in the Escherichia coli RRI
strain. Clones were screened by growing single colonies in M9 media,
casamino acids, ampicillin, and 20 µg/ml tryptophan overnight at
37 °C. A 10-fold dilution of the overnight mixture was grown 1 h at 30 °C in the absence of tryptophan. The trpE gene
was induced by the addition of a 5 µl/ml concentration of 2 mg/ml
indoacrylic acid and incubation at 30 °C for 18 h. 1 ml of
cells were pelleted, resuspended in 50 µl of cracking buffer (10 mM sodium phosphate, pH 7.2, 1% -mercaptoethanol, 1%
SDS, and 6 M urea), and incubated at 37 °C for 1-2 h.
The expressed proteins were analyzed by SDS-PAGE. For preparative gels,
10 ml of induced cells were processed. The band representing the
TrpE-STS fusion protein was visualized by a Coomassie-water stain,
excised, and emulsified in Freund's adjuvant (complete for the initial injections, and incomplete for the subsequent boosts). Rabbits were
injected at 10 subcutaneous sites, 200 µl/site. The rabbits were
boosted 6 weeks after the initial inoculation and then every 4-6
weeks. Titers were checked 10 days postinjection starting after the
second boost. 5-10 ml of blood were drawn, allowed to clot, and spun
down. The aliquots of the sera were stored at
20 °C. A working
stock was stored at 4 °C (24).
Patient or transfected cells were washed with PBS and harvested by scraping. The cell pellets were resuspended in 200-400 µl of SDS-PAGE sample buffer and boiled for 5 min. 20 µl of transfected cell lysate or 40 µl of patient cell lysate (~1 × 105 cells) were loaded per lane and separated by SDS-PAGE. The gel was transferred to nitrocellulose by the wet transfer method, using 25 mM Tris, 190 mM glycine, and 20% methanol and transferring at 63 V for 16 h using a Bio-Rad apparatus (24).
The blots were first blocked with 10% nonfat dry milk in PBS and 0.1% Tween-20 (Sigma) for 2 h at room temperature and then incubated with a 1:10,000 dilution of rabbit anti-human STS antibody in PBS-Tween 20 containing 5% nonfat dry milk at room temperature overnight. After washing the blots several times in PBS-Tween, the blots were incubated in 5% NFDM containing a 1:5000 dilution of horseradish peroxidase-labeled donkey anti-rabbit Ig antibody (Amersham Corp.) for 1 h. The blots were washed as above, and horseradish peroxidase-labeled protein(s) were detected using ECL detection reagents (Amersham) as instructed. The blots were exposed to film for 20 s (transfected cell lysates) and 14 h (patient cell lysates).
In Vitro TranscriptionIn vitro transcription reactions containing 1 µg of XbaI-linearized template DNA were carried out as described in the RiboMax transcription kit (Promega).
In Vitro TranslationTranslation reactions containing 0.5 µg of in vitro transcribed RNA, 17.5 µl of rabbit reticulocyte lysate (Promega), 25 µM methionine-free amino acids, and 20 µCi of [35S]methionine (Amersham) were carried out according to instructions (Promega). 1.8 µl of nuclease-treated canine pancreatic microsomal membranes (Promega) were added as indicated. The reactions were incubated at 30 °C for 1 h. The translation products were analyzed by SDS-PAGE.
Protease protection experiments were performed as follows. 10 µl of the translation reactions were chilled in an ice water bath to 0 °C, and CaCl was added to 10 mM. A solution of 1 mg/ml of proteinase K (Boehringer Mannheim) in Tris-HCl (pH 7.5) and 10 mM CaCl was preincubated at 37 °C for 15 min to degrade contaminating lipases. 1 µl of treated proteinase K was added to the translation reactions in the presence or absence of 1% Triton X-100. The reactions were incubated at 0 °C for 30 min. The reactions were stopped by the addition of 2 µl of 2 mg/ml phenylmethylsulfonyl fluoride in ethanol and immediately transferred to boiling SDS-PAGE loading buffer (25).
Cell lines from three unrelated patients with typical XLI phenotypes were studied. All completely lacked STS enzymatic activity but showed normal Southern and Northern patterns when probed with the full-length STS cDNA p331 (9). Immunoprecipitation of cell extracts with anti-human native STS antibody (4) failed to show cross-reactive material (9).
The coding region was searched for mutations using RT-PCR. Total RNA
was isolated from patient lymphoblast and fibroblast cell lines. The
first strand cDNA was synthesized by reverse transcription. The STS
cDNA was amplified in three overlapping segments. Each amplified
fragment crossed at least two introns, eliminating the possibility that
contaminating genomic DNA was being amplified. The cDNA fragments
amplified from all three patients ran at the same mobility on a 1%
agarose gel as the corresponding amplified wild type fragments. After
subcloning the fragments, both strands of the STS coding region were
sequenced. All apparent mutations were confirmed by amplifying,
subcloning, and sequencing the corresponding regions of the patients'
genomic DNAs. Single point mutations were identified in the cDNA of
patients Z and S (Fig. 1, A
and B) and confirmed at the genomic level. A G C
substitution at nucleotide 1336 in patient Z results in a predicted
switch of a tryptophan to a proline at residue 372. An A
G
substitution at nucleotide 1552 in patient S causes a predicted switch
of a histidine to an arginine at residue 444. In patient T, a 19-bp insertion was found starting at nucleotide 1477 (Fig.
2A). A G
T substitution
was identified at the exon 8-intron 8 splice donor site and confirmed
in the genomic DNA (Fig. 2B). This splice junction mutation
results in the addition of 19 bp from intron 8 to the STS mRNA,
changing the reading frame. The predicted polypeptide prematurely
terminates at residue 427, 8 amino acids after the frameshift. As a
result, the mutant STS polypeptide is predicted to lose 156 residues
from its carboxyl terminus.
Basler et al. identified point mutations in an additional
three patients with XLI (A, C, and D) (1). We have studied these three
mutant STSs as well as the three mutations described above. All six
patients have unique single base pair substitutions on the genomic
level. All of the mutations are located within 105 residues in the
C-terminal half of the STS polypeptide. Fig.
3 shows the location of the mutations in
relationship to the rest of the protein. Table
I lists the actual nucleotide and amino acid changes. Mutations A and Z cause the same residue at position 372 to switch from a tryptophan to an arginine or a proline, respectively. Mutations C and S affect residues 444 and 446. C results in a predicted
switch of a cysteine to a tyrosine, and S results in a histidine being
replaced by an arginine. The close proximity of the mutations suggests
that this region of the STS polypeptide is important to structure or
function and thus worthy of further investigation.
|
To address whether
or not the six point mutations identified were responsible for the STS
deficiency observed in the patients, site-directed mutagenesis was used
to recreate five of the six mutations in STS cDNA (PS10). The
full-length cDNA p331 (4), containing 5 ATGs 5 of the initiating
ATG, was unable to produce protein in an in vitro
translation system. The coding region minus the 5
ATGs was subcloned
into the expression vector pSG5. The resulting vector, PS10b directs
protein expression both in an in vitro translation system as
well as in an in vivo system. To create the mutant vector
PST, a 404-bp fragment containing the 19-bp insertion was used to
replace the corresponding wild type sequence in PS10. The mutations and
the resulting vectors are listed in Table I. It is important to note
that, other than the single nucleotide switches in five of the
constructs and the 404-bp insertion in the sixth, all of the constructs
are identical to PS10.
The PS10 and mutant vectors were each cotransfected into STS-deficient mouse A9 cells along with a CAT expression vector. The parental vector pSG5 was transfected into A9 cells as a negative control. Cells were harvested 72 h post-transfection and assayed for CAT and STS activity as well as protein concentrations. All of the transfected cells exhibited comparable levels of CAT activity, whereas only the cells transfected with the wild type STS vector PS10 showed STS activity above background level (Table II). The failure of all six mutant STS vectors to produce STS enzyme activity confirms that these mutations do in fact cause STS deficiency and are responsible for the X-linked ichthyosis found in these patients.
|
To
investigate the effect of the amino acid substitutions on STS protein
expression, lysates were made from the transfected A9 cells and
analyzed by SDS-PAGE and immunoblotting. All six mutant plasmids
directed expression of steady state levels of STS polypeptide
comparable to wild type in transfected cells as shown by Western blot
experiments (Fig. 4). Neither the A9
cells alone nor the A9 cells transfected with the parental pSG5 vector produced cross-reacting protein. The polypeptides expressed from the
cells transfected with the mutant vectors PSS, PSA, PSC, and PSD ran at
the same mobility on an SDS-PAGE gel as the polypeptide expressed from
A9 cells transfected with the wild type STS PS10 vector (61 kDa). The
PSZ polypeptide, however, ran with a slightly higher mobility (63 kDa)
than PS10. To ensure that this difference in mobility was not an
artifact, the experiment was repeated several times. A gel was also run
in which the wild type PS10 lysate was loaded in alternating lanes next
to mutant lysates (data not shown), and the results were confirmed.
PSZ's shift in mobility could be due to the presence of a proline
residue not present in the normal or other mutant STSs or could
represent a difference in post-translational processing. Although the
PST polypeptide ran with a faster mobility than PS10, reflecting the
loss of the 156 C-terminal residues, its size (44 kDa) was larger than
the predicted 40.4 kDa.
To address the question of whether the vector-driven protein expression
accurately reflects STS expression in normal and deficient cells,
additional experiments were conducted with patient cell lines. Cells
from 75% confluent 100-mm plates or flasks were harvested and
resuspended in 200-400 µl of SDS sample buffer depending on the size
of the cell pellet. 40 µl of cell lysate was separated by SDS-PAGE.
The gel was blotted onto nitrocellulose and probed with the anti-human
STS antibody. STS cross-reacting protein appeared to be present in
patient cells Z, T, and S at substantially lower levels than seen in
normal IMR91 cells (Fig. 5). STS
cross-reacting material from patient cells S and Z ran with higher
mobilities than wild type. Background bands were present when the
nontransfected cells were examined. This is not unusual when human
whole cell extracts are immunoblotted using an anti-human protein
antibody. While most of these background bands correspond to those seen in the wild type extract, an unexplained band at <40 kDa was observed in the cells from patient Z. This band was not present in the cells
transfected with the corresponding construct PSZ. The Z STS clearly ran
with a higher mobility (66 kDa) than either normal or S STS. The 66-kDa
size observed in the patient Western blot is 3-kDa bigger than seen in
the transfection experiment (Fig. 4). T STS ran more rapidly, as
predicted. These results suggest that the mutant STS polypeptides are
expressed in the patient-derived cell lines as well as in the A9
transfected cells. We were not able to analyze the patient cell lines
A, C, and D of Basler et al., but they reported that no
cross-reacting material was detected when their Western blot containing
D and C was probed with anti-STS antibodies (1). It is not known if the
differences observed between these Western blot experiments are
technical or reflect the actual expression of the mutant STS
polypeptides.
Effects of Point Mutations on STS Processing
To investigate
the initial processing of the mutant STS polypeptides, the wild type
PS10 and the mutant PSA, PSC, PSD, PSS, PST, and PSZ vectors were used
as templates for the in vitro transcription of T7
polymerase-generated mRNA. The normal and mutant RNAs were translated in a rabbit reticulocyte lysate system in the presence (+)
and absence () of dog pancreas microsomes (Fig.
6). In the absence of microsomal
membranes, the primary translation products of five of the six mutants
were comparable with that of wild type STS, with an apparent
Mr of 52. PSZ ran with the same mobility as
PS10. The "unprocessed" PST had a predicted
Mr of 31. All six mutant STS polypeptides,
including the truncated PST, showed a shift in electrophoretic mobility
when translated in the presence of microsomes. The shift in size of the
PST polypeptide from 31 to 44 kDa agrees with the results observed in
the transfection experiments. PSZ ran with a slightly higher mobility
than PS10, PSD, PSA, PSC, and PSS, comparable with differences seen in
transfected cells. To confirm that the shift in mobility seen in the
translation products synthesized in the presence of microsomes is due
to processing, luciferase mRNA was translated under identical
conditions. Luciferase, which does not undergo post-translational
modifications, did not show a shift in mobility when translated in the
presence of microsomes (Fig. 6).
It has been previously shown that STS is completely protected from
protease digestion when translated in the presence of dog pancreas
microsomes, implying that most, if not all of the STS protein is
translocated into the luminal portion of the endoplasmic reticulum (5).
We wanted to determine if the mutant STS polypeptides were localized in
a similar fashion and protected to the same degree. Normal and mutant
STS polypeptides, synthesized in the presence and absence of microsomal
membranes, were subjected to protease digestion. The "processed"
translation products from all of the mutant mRNAs were completely
resistant to digestion with proteinase K, indicating that the resulting
polypeptides had been translocated into the microsomes and were
therefore protected from the protease (Fig.
7, A, B, and
C, +). The residual "unprocessed" polypeptides were
completely digested. When the detergent Triton X-100 was used to
disrupt the microsomes in the presence of proteinase K, the 44-, 61-, and 63-kDa "processed" bands disappeared as well. These results
show that all of the mutant polypeptides, including the truncated PST,
are correctly translocated into the lumen of the microsomal vesicles
and therefore protected from protease digestion.
Approximately 90% of patients with STS deficiency and XLI have deletions of the entire 146-kilobase pair STS gene and substantial amounts of flanking sequences (9, 26). Our group has studied the mechanism of these deletions in detail (27, 28). Most of the deletions are 1.9 megabase pairs and involve at least one other gene, GS1, whose function remains unknown (28). It is likely that these deletions result from unequal recombination between members of a family of low copy number repeat sequences flanking the STS locus (27-29). Two patients have also been identified with partial gene deletions (9, 26). The relative lack of small mutations in STS-deficient patients has hindered protein function studies. However, a small group of patients were identified who had normal Southern (1, 9) and Northern (9) hybridization patterns when probed with STS cDNA. We anticipated that these mutations would provide information about regions of the STS protein that are important for post-translational processing, activity, and possibly stability.
Unique point mutations were identified in three unrelated STS-deficient patients (S, T, and Z). Basler et al. (1) identified an additional three point mutations (A, C, and D). All six patients have unique single base pair substitutions in their genomic DNA. Five of the point mutations resulted in nonconservative amino acid changes, and the sixth resulted in a frameshift mutation. These alterations are listed in Table I. The mutations described were the only nucleotide alterations identified in the STS coding region and were confirmed in genomic DNA. Transfection of mutant STS expression plasmids into STS-deficient A9 cells showed that each of these alterations abolishes STS activity (Table II).
The six mutations are located within 105 residues in the C-terminal half of the polypeptide (Fig. 3). Two of the mutations, A and Z, involve the same amino acid, 372, changing a tryptophan to either arginine or proline, respectively. Mutations C and S result in substitutions of amino acids only 2 residues apart (444 and 446). The mutation in patient D converts a serine to a leucine at residue 341. The splice junction mutation in patient T results in the loss of 20% of the STS C terminus beginning at residue 419. This does not include the 8 additional mutant residues. The close proximity of these mutations suggests that this is an important region of the STS protein.
STS is a member of an evolutionarily conserved family of sulfatases (12). The importance of the sulfatases in human metabolism is stressed by the existence of eight distinct inherited disorders resulting from deficiency of these enzymes (30-32). Further evidence supporting the importance of the 105-residue region in STS is the fact that all six of the mutations involve amino acids that are conserved in both (33) and mouse (34) STS polypeptides. These proteins have little overall homology with human STS (4). Four of the five mutated residues are conserved in two sea urchin arylsulfatases (35, 36), which are thought to have diverged from human STS 500 million years ago (12). In addition, the affected residues in these patients are also conserved in 6-10 of the 11 sequenced mammalian sulfatases (4, 5, 11, 12, 37-41). The involved residues are part of peptide sequences that are themselves conserved to varying degrees among the family of sulfatases. The altered residues in patients C and S are located in the shortest of the homologous peptide sequences. Three residues are shared by five sulfatases (human STS, arylsulfatase D, and arylsulfatase E, as well as rat and mouse STS) (4, 5, 32, 34). The tryptophan that is altered in patients A and Z is the N-terminal amino acid in a sequence of 8 residues shared by seven mammalian sulfatases (human STS, ASB, arylsulfatase D, arylsulfatase E, arylsulfatase F, rat STS, and mouse STS) (4, 5, 12, 32, 34, 39). Four of the amino acids in this sequence are shared by nine of the sulfatases (all except G6S and iduronate sulfatase). Detailed mutation studies would be needed to determine whether it is the individual affected residues or the integrity of the conserved sequences that is essential to STS activity.
The predicted structure of the STS polypeptide is shown in Fig. 3. STS
contains two regions (residues 185-237 and 466-496) that are of
sufficient length and hydrophobicity to constitute membrane-spanning
domains (4, 5). Computer analysis predicts an -helical structure for
the first region and a
-sheet structure for the latter (5). The
first of these sequences contains a proline at residue 212. Of four
predicted possible N-linked glycosylation sites (residues
47, 259, 333, and 459), only residues 47 and 259 are thought to be
utilized (5). Based on the resistance of translocated STS to protease
digestion, on computer analysis, and on the location of the utilized
glycosylation sites, a model for the topology of STS was proposed (5,
12). In this model, the N- and C-terminal domains are located on the
luminal side of the membrane. The two domains are connected by a
membrane-spanning hydrophobic region. Proline 212, in the middle of the
hydrophobic sequence, enables the STS polypeptide to turn in the
membrane. The N- and C-terminal domains each contain a glycosylation
site. The six point mutations are found in the C-terminal region,
between the two hydrophobic regions and after the second
N-linked glycosylation site. It has been suggested that the
N-terminal domain, which shows the greatest degree of amino acid
conservation between the predicted sulfatase polypeptides, contains the
active site, while the less conserved C-terminal region could contain
the substrate binding site (12).
In an effort to understand how these mutations cause a loss of STS activity, the expression of the mutant STS proteins was examined in a rabbit reticulocyte lysate system, transfected A9 cells, and three of the patient cell lines. In contrast to previous reports, all six mutant STS proteins are detectable immunologically (Figs. 4, 5, 6). Previous immunoprecipitation and immunoblotting studies of patient cells failed to detect any STS cross-reacting material (1, 9). However, these utilized antibodies made against the native human STS protein (4, 42). Based on these results, it was originally thought that the mutations might result either in an unstable protein that is rapidly degraded or in a change in STS conformation that is necessary for antibody binding. The presence of detectable mutant protein now suggests that the latter is true. The antibody used to detect STS cross-reacting material in the current studies was made against a denatured TrpE- STS fusion protein. It is important to note that mutant protein was not only detected in the vector-driven systems but also appears to be present in the patient cells (Fig. 5). Although the immunoblotting studies did not include the patient cell lines A, C, and D, it is possible that they too produce protein that cross-reacts with the denatured human STS antibody. The discrepancy in results with the two types of antibodies suggests that the mutations are causing conformational changes in the STS proteins.
Although not studied in detail, no overall effect in stability was seen in any of the transfected or in vitro translated mutant polypeptides. All six mutant polypeptides had the same steady state level as did the wild type STS polypeptide in these systems. In the patients' cells (S, T, and Z), however, the steady state level of mutant protein was roughly one-tenth of the wild type protein level. This may reflect a possible difference in stability. Since the steady state levels of mRNA appear to be the same in patient and normal cells (9),2 it is unlikely that this variance is due to the effect of the mutations on transcription. It is also unlikely that protein translation is being affected, since the point mutations do not appear to affect the stability of the transfected and in vitro translated mutant proteins. One possible explanation is that the nascent polypeptides are more prone to degradation within the endoplasmic reticulum. This is not seen in the transfected cells due to overproduction or in the rabbit reticulocyte lysate system, which lacks other cellular components. Whatever the cause, the variances in steady state protein levels between the wild type and mutants are not so extreme as to explain the total loss of STS activity seen in the patients.
All of the mutant proteins, with the exception of PSZ and PST, show the same mobility as wild type STS. It is therefore unlikely that there is a major abnormality in post-translational processing. The processing of normal STS has been described (5, 7). STS is synthesized as a membrane-bound 63.5-kDa polypeptide with two N-linked oligosaccharide chains. Within 2 days, the protein is processed to its mature 61-kDa form. The decrease in size is due to processing of the oligosaccharide chains (7). We were interested in studying the effects that the point mutations and their subsequent amino acid substitutions had on the post-translational processing of STS. In the presence of dog pancreas microsomes, all six of the in vitro translated mutant STS polypeptides showed a shift in mobility on an SDS gel as well as complete protection from protease digestion. Thus, mutations, including the truncation seen in PST, do not appear to affect STS translocation. These results suggest that all of the mutant STS proteins reach the luminal side of the endoplasmic reticulum membrane, which is thought to be the major location of mature STS (5, 43, 44).
The mobility of the processed mutant proteins agrees with the transfection and intact patient cell results, with the exception of PSZ. The processed PSZ runs with a slightly higher mobility than does the wild type protein. In transfected cells and rabbit reticulocyte lysate plus membranes, the difference in mobility corresponds to a mass of 2 kDa. The difference in mobility between the patient and wild type cells appears to be greater, corresponding to a mass of 5 kDa. Interestingly, the processed PSA protein runs with the same mobility as wild type, suggesting that it is the addition of the proline and not the loss of the tryptophan that is affecting mobility. The replacement of an amino acid with a proline residue has been noted to change the electrophoretic mobility of resultant polypeptides. "Unprocessed" PSZ, however, runs with the same mobility as wild type STS, implying that PSZ may be processed differently.
The mutant PST is missing the last 20% of its carboxyl terminus,
including the -sheet hydrophobic region and the last two predicted
glycosylation sites, and yet appears to be fairly stable. It is
correctly translocated into microsomes, as shown by its resistance to
protease digestion, and it appears to be glycosylated. These findings
agree with the proposed model for STS discussed above (5, 12). The
-helical hydrophobic region and its flanking N-linked
glycosylation sites may be sufficient for translocation and
glycosylation of PST. While the electrophoretic mobility of the
unprocessed PST agrees with the predicted molecular weight, the
processed PST shows a shift in mobility that is slightly higher than
expected. The shift is comparable with 13 kDa as opposed to the 9-kDa
relative shift seen in the wild type, a difference of 4 kDa. The size
of the processed PST is the same in the transfected and patient cells
as well, arguing against an artifact. It is possible that the loss of
the C-terminal end of the STS polypeptide changes the conformation of
the protein in a way that prevents further processing of the
oligosaccharide chains. This would account for 2.5 kDa of the
difference (7). STS contains three sets of two negatively charged
residues in the region lost in PST in addition to three negatively
charged residues found in the N-terminal region. Any of these sets of
residues is sufficient for recognition by the signal peptidase (5). The
retention of the signal sequence would account for 2 kDa.
In conclusion, the data show that the mutations identified by Basler et al. (1) and our laboratory are responsible for the STS deficiency and XLI observed in the patients. The enzyme deficiency, however, is not due to a lack of STS protein expression or apparent incorrect localization. Further studies may address whether the mutations cause a shortened half-life in the mutant proteins or a loss of the substrate binding site. It would also be interesting to determine if the mutations affect the normal dimerization of STS. The results in this paper provide a preliminary insight into a region of the STS protein that is important for protein function and activity. Additional studies may help to elucidate not only important domains in the STS protein but those in the rest of the sulfatase family as well.
We thank Merry Passage for expert technical assistance with the cells. We appreciate the helpful discussions with Jay Ellison, Xiao Miao Li, and the rest of the members of the laboratory.