Full text of DOI:10.1093/glycob/11.4.283

Glycobiology vol. 11 no. 4 pp. 283–295, 2001
Analysis of Skp1 glycosylation and nuclear enrichment in Dictyostelium
Slim Sassi, Mark Sweetinburgh, John Erogul, Ping Zhang,
Patana Teng-umnuay, and Christopher M. West1
Introduction
Skp1 is a subunit of the SCF class of E3 ubiquitin ligases that
is responsible for targeting proteins for ubiquitination and
degradation (Williams et al., 1999; Laney and Hochstrasser,
1999; Koepp et al., 1999; Deshaies, 1999). Genetic and
biochemical studies show that Skp1 directly contacts the F-box
protein and cullin-1, which is in turn linked to an E2 enzyme.
Different SCF complexes, which vary in their F-box proteins,
are specific for the ubiquitination of individual cell cycle
regulatory proteins, transcriptional factors, and other signaling
proteins. F-box proteins and a cullin have recently been
identified in Dictyostelium (Chung et al., 1998; Ennis et al.,
2000; Nelson et al., 2000). The precise function of Skp1 has
yet to be elucidated.
Department of Anatomy and Cell Biology, Box 100235, 1600 SW Archer Road,
University of Florida College of Medicine, Gainesville FL 32610-0235, USA
Received on July 24, 2000; revised on October 12, 2000; accepted on
November 15, 2000
Skp1 is a subunit of SCF-E3 ubiquitin ligases and other
protein complexes in the nucleus and cytoplasm of yeast
and mammalian cells. In Dictyostelium, Skp1 is partially
modified by an unusual pentasaccharide O-linked to
hydroxyproline143. This modification was found to be
susceptible to known prolyl hydroxylase inhibitors based
on M -shift analysis using SDS–polyacrylamide gel electro-
r
phoresis/Western blotting. In addition, Dictyostelium Skp1
consists of 2 genetic isoforms, Skp1A and Skp1B, which
differ by a single amino acid and appear to be expressed
throughout the life cycle based on reverse-transcription
polymerase chain reactions. The significance of these structural
variations was examined by expressing myc-tagged Skp1s
and mutants that lacked the glycosylation site. Gel-based
Target substrates for SCF E3 ubiquitin ligases reside in both
the nucleus and the cytoplasm. Proteasomes, which are the
ultimate destination of proteins ubiquitinated by the SCF E3
complex, are also found in both the nucleus and cytoplasm
(Schauer et al., 1993). Skp1 is enriched in the nucleus of
cultured mammalian cells based on immunolocalization
studies (Freed et al., 1999; Gstaiger et al., 1999). Genetic and
biochemical studies show Skp1 in yeast kinetochores
M -shift studies showed that Skp1A and Skp1B are both
r
(
reviewed in Deshaies, 1999), where it apparently belongs to a
nearly completely glycosylated during growth and early
development, and mass spectrometry of glycopeptides
showed that they were glycosylated similarly. Skp1
expressed later in prespore cells was not glycosylated,
unlike bulk Skp1 persisting from earlier in development,
but became glycosylated after return to growth medium.
Skp1A and Skp1B were each concentrated in the nucleus and
regions of the cytoplasm, based on immunofluorescence
localization. However, when Skp1 glycosylation was
blocked by mutation, prolyl hydroxylase inhibitors, or
expression in prespore cells, nuclear concentration of Skp1
was not detected. Furthermore, nuclear concentration
occurred in a mutant that attached only the core disaccharide
to Skp1. Overall, there was no evidence for differential
Skp1 isoform expression, glycosylation variants in the bulk
Skp1 pool, or regulation of nuclear localization. However,
these studies uncovered evidence that the glycosylation
pathway is developmentally regulated and can function
posttranslationally, and that core glycosylation is required
for Skp1’s nuclear concentration.
protein complex distinct from the SCF complex. In the cyto-
plasm, Skp1 is concentrated in the mammalian centrosome
(
Freed et al., 1999; Gstaiger et al., 1999) and partially associ-
ated with the microsomal fraction in a salt-sensitive fashion in
Dictyostelium (Kozarov et al., 1995). A 2-hybrid screen has
suggested an association with the cytoplasmic actin-binding
protein coronin (Uetz et al., 2000). The mechanism by which
Skp1 is distributed between the nucleus and cytoplasm is
unknown. However, because the SCF complex is too large to
diffuse passively through nuclear pores, nuclear localization
might be an active process.
Multiple genetic isoforms of Skp1 are encoded in many
eukaryotes (West et al., 1997). Additional structural variation
of Skp1 has been seen in Dictyostelium (in which it was
previously referred to as FP21), where it is modified at
HyPro143 by a linear pentasaccharide with the sequence
Galα1,6Galα1,Fucα1,2Galβ1,3GlcNAc (Teng-umnuay et al.,
1
998). The Pro residue to which the pentasaccharide is
attached after hydroxylation is highly conserved in many Skp1
genes detected in fungi, invertebrates, and higher plants.
Structural variants of Skp1 may differentially localize in the
cell or enter into specific multiprotein complexes.
Key words: α,α′-dipyridyl/cytoplasmic glycosylation/ethyl
3,4-dihydroxybenzoate/hydroxyproline/prolyl hydroxylase/
To investigate the significance of Skp1 structural variation, we
examined the expression, glycosylation, and cellular localization
of the two genetic isoforms of Dictyostelium Skp1, Skp1A and
Skp1B. This has revealed that both Skp1s are stable and
expressed at the message level throughout the life cycle. They
ubiquitin ligase
1To whom correspondence should be addressed
©
2001 Oxford University Press
283

S. Sassi et al.
are similarly and mostly glycosylated, by a mechanism that
shows some developmental regulation and is not necessarily
cotranslational. They are each enriched in the nucleus and
regions of the cytoplasm, and nuclear enrichment appears to
depend on its glycosylation. These results provide the first clue
about the function of the unusual glycosylation of Skp1, which
is mediated by a novel glycosylation pathway that appears to
reside in the cytoplasm rather than the secretory pathway of the
cell (West et al., 1996; Teng-umnuay et al., 1999).
This yielded a predominant band of about 420 bp from HL-5
grown Ax3 cells (Figure 1, lane 16), similar to the 428-bp
length expected for Skp1 cDNA. No bands were detected in
the absence of RNA (data not shown). A control lacking
reverse transcriptase produced very faint bands at 420 and
5
80 bp (data not shown) attributable to Skp1A and Skp1B
DNA, because Skp1B DNA contains a 154-bp intron not
present in Skp1A (West et al., 1997).
Restriction enzyme digestions confirmed that the amplified
4
28-bp band derived from Skp1A and Skp1B RNA. BclI,
which cuts at nt 320 of only the Skp1A gene fragment, gave
bands corresponding to 320 and 108 bp and the full-length
Skp1B cDNA (data not shown). SspI, which cuts only Skp1B
DNA at position 361, produced the expected fragments
Results
Skp1A and Skp1B are each expressed throughout the life cycle
Analysis of Skp1 is complicated by the fact that it consists of
two different proteins that differ by only a single amino acid, A39
in Skp1A and S39 in Skp1B (West et al., 1997). This variation is
conserved among Skp1 genes in Drosophila (GenBank accession
numbers AF220066 and AF220067). First we asked whether
different Skp1s are expressed in specific cell types, or
expressed coordinately throughout the life cycle. A reverse-
transcription polymerase chain reaction (RT-PCR) approach
was used because the proteins are so similar.
The primer for reverse transcription of total cell RNA
hybridized to a region near the 3′ end of the coding region that
was identical between Skp1A and Skp1B. Subsequent PCR
was based on this and an upstream primer corresponding to a
second shared sequence near the 5′ end of the coding region.
(
361 and 67 bp) and intact Skp1A (data not shown). EcoRV,
which cuts both Skp1A (at bp 193 and 274) and Skp1B (at bp
2 and 193), yielded all of the expected fragments on digestion
Skp1A: 193, 154, 81 bp; Skp1B: 235, 141, 52 bp) while
leaving little DNA in the region of the original 428 bp band
Figure 1, lane 12). The residual DNA remaining at this position
5
(
(
is probably a contaminating DNA product of the original PCR
amplification, but, even if it consists of undigested Skp1
cDNA, it is present at such a low level that it does not affect the
interpretation based on analysis of the digestion products.
Relative ratios of the Skp1A- and Skp1B-derived bands were
not affected by the number of amplification cycles from 15–35
(data not shown). EcoRV bands derived from Skp1A and
6
6
Fig. 1. RT-PCR analysis of the expression of Skp1A and Skp1B mRNAs. Normal strain Ax3 cells were taken from HL-5 growth medium at 1 × 10 , 5 × 10 , and
7
7
2
(
× 10 cells/ml (lanes 10–12). For development, HL-5 grown cells were suspended in KP buffer at 2 × 10 cells/ml and incubated on a shaker for 3, 6, or 9 h
lanes 1–3). Alternatively, cells were plated on nonnutrient agar (NNA) for 9, 12, 15, 17, 18.5, and 20 h (lanes 4–9). Cells were also grown on K. aerogenes and
harvested before (lane 15), during (lane 13), or after (lane 14) consumption of all the bacteria, corresponding to growing and aggregating cells. In lanes 13 and 14,
strain TF125, a REMI-mutagenesis strain containing a spontaneous insertion of pyr5,6 900 bp upstream of the transcription start site of the Skp1A gene, was
examined. Total RNA was isolated from cells, and cDNA was amplified by RT-PCR using primer sequencs that were identical between the Skp1A (fpa1) and
Skp1B (fpa2) genes. The cDNA was digested with EcoRV and visualized on a 3% agarose gel with ethidium bromide. Band assignments with their predicted sizes
(
in bp) are listed on the left. Undigested cDNA from high density HL-5 grown Ax3 cells is in lane 16, and M standards are in lane 17.
r
284

Nuclear enrichment of glycosylated Skp1
Skp1B had similar intensities, indicating that Skp1A and Skp1B
message levels in the original RNA pool were comparable.
Similar levels of Skp1A and Skp1B cDNA were amplified
by RT-PCR at all stages of the life cycle tested, including cells
grown on bacteria or at different densities in HL-5 growth
medium and after various times of starvation either in suspension
or on filters (Figure 1). Total Skp1 mRNA levels were also
similar throughout the life cycle, except that cells from HL-5
exhibited slightly higher levels (data not shown). Both Skp1A
and Skp1B mRNAs are expressed at the protein level at least in
stationary phase cells (saturation density in HL-5 growth
medium), based on amino acid sequencing of total Skp1 from
these cells (West et al., 1997).
Skp1 glycosylation is affected by inhibitors of prolyl
hydroxylase
Fig. 2. M analysis of Skp1 from cells treated with the prolyl hydroxylase
inhibitor α,α′-dipyridyl. 10 cells of strain (A) Ax3, (B) HW302, or (C)
r
Prolyl hydroxylases that modify collagen and other secretory
6
proteins can be inhibited in vivo by α,α′-dipyridyl and ethyl
HW120, incubated in KP buffer for 17 h in the absence or presence of 50 µM
α,α′-dipyridyl as indicated, were subjected to SDS–PAGE on a 15–20%
gradient gel, Western blotted, and probed with either mAb 3F9 (upper panels),
which recognizes both endogenous and expressed Skp1s, or mAb 9E10 (lower
panels), which recognizes the C-terminal c-myc epitope tag on expressed Skp1
only. Bands marked with an E are endogenous Skp1 labeled only with mAb
3
,4-dihydroxybenzoate (Kivirikko and Pihlajaniemi, 1998).
Because the Skp1 glycan is attached via a HyPro residue, inhibition
of hydroxylation of this Pro residue would be expected to
block its glycosylation. Preliminary results suggested that
3
F9; bands marked with a B or A are expressed variants of Skp1B-myc or
1
mM α,α′-dipyridyl partially inhibits Skp1 prolyl hydroxy-
mutant Skp1A1-myc detected with both mAbs. Skp1B-myc has a lower
mobility owing to the c-myc tag, which adds a predicted 1185 Da to its mass.
Asterisked bands are unglycosylated variants that appear as the result of
treatment with α,α′-dipyridyl; these are calculated to be 851 Da less in mass.
The granular appearance of some bands results from reducing the amount of
protein loaded on the gel to close to the detection threshold, to maximize band
sharpness and separation.
lase activity in cell extracts (West, unpublished data). An
effect of α,α′-dipyridyl on Skp1 glycosylation in cells was
tested by M analysis of Skp1 by SDS–polyacrylamide gel
r
electrophoresis (PAGE) and Western blotting, as affected
Skp1 was expected to be reduced in mass by 851 Da. Strain
Ax3 cells attached to tissue culture plastic wells were starved
in KP buffer (17 mM KPO , pH 6.5; developmental conditions)
4
in 0.02–1 mM α,α′-dipyridyl for 17 h. The entire cell pellet
was solubilized in SDS–sample buffer and subjected to SDS–
PAGE/Western blotting with mAb 3F9 (Kozarov et al., 1995),
using 15–20% polyacrylamide gradient gels to maximize
separation of normal and nonhydroxylated Skp1. mAb 3F9
recognizes nonglycosylated Skp1 produced recombinantly in
Escherichia coli, indicating that its epitope does not require
glycosylation. After incubation in 0.05 mM α,α′-dipyridyl, the
inhibitory were similar to or less than those which affect
collagen hydroxylation in mammalian cells (Sasaki et al., 1987;
Eleftheriades et al., 1995). These results suggest that hydroxyla-
tion of Skp1′s Pro143 is partially sensitive to inhibitors of known
prolyl hydroxylases in vivo. Since negligible levels of normal
Skp1 were seen at the inhibited position, nearly all Skp1 is
usually glycosylated in the cell.
majority of Skp1 exhibited a lower lower M consistent with
r
Expressed Skp1A-myc and Skp1B-myc are similarly
glycosylated at HyPro143
the absence of hydroxylation and glycosylation of Pro143
(
Figure 2A, upper). The overall level of Skp1 was not greatly
35
To investigate any differences between Skp1A and Skp1B at
the protein level, it was necessary, owing to their high degree
of similarity, to express them separately with epitope tags.
Skp1A-myc and Skp1B-myc were initially expressed under the
discoidin 1γ promoter with a c-myc tag located one residue in
from their C-termini (see Table I for list of expression strains).
Although the endogenous discoidin 1γ promoter is regulated
in vivo (Blusch et al., 1992), it was active in these strains
throughout growth and early development (data not shown).
To determine whether the myc-tagged proteins were glyco-
altered. At this concentration, incorporation of [ S]Met and
[
3
H]Fuc into total acid precipitable material was reduced by
2
% and 9%, suggesting that the effect was not the result of
inhibiting overall protein synthesis via an effect on deoxy-
hypusyl hydroxylase (McCaffrey et al., 1995). A similar effect
of α,α′-dipyridyl on Skp1 M was seen when cells were
r
incubated in HL-5 growth medium, where incorporation of
35
3
[
S]Met and [ H]Fuc into total acid precipitable material was
reduced by 33% and 39%. A smaller fraction of Skp1 was
shifted to the lower M position at lower concentrations of
r
α,α′-dipyridyl or shorter time periods of treatment (data not
shown). The fraction of total unglycosylated Skp1 did not
increase at higher concentrations, and cells tended to detach
from the tissue culture plate. Similar partial effects on Skp1
glycosylation were observed using 200–400 µM ethyl 3,4-
dihydroxybenzoate (data not shown), an inhibitor that after
intracellular de-esterification is thought to function by a
mechanism distinct from that of α,α′-dipyridyl (Sasaki et al.,
sylated, their M values were compared in constructs in which
r
Pro143 was substituted with Ala, as occurs in certain Skp1
gene isoforms present in Caenorhabditis elegans (Z81084) and
Arabidopsis thaliana (AL138658). Whole-cell pellets of
colonies expressing high levels of Skp1-myc were analyzed by
SDS–PAGE followed by Western blotting with mAb 9E10,
specific for the c-myc epitope on expressed Skp1.
Skp1A3(P143A)-myc and Skp1B1(P143A)-myc were expressed
1987; Kivirikko and Pihlajaniemi, 1998). The concentrations of
at slightly lower M values than normal Skp1A-myc and
r
α,α′-dipyridyl and ethyl 3,4-dihydroxybenzoate that were
Skp1B-myc (Figure 3A, compare lanes 3 and 5 to lanes 2 and
2
85

S. Sassi et al.
To confirm this conclusion, Skp1A1-myc was examined by
SDS–PAGE/Western blotting as above. Skp1A1-myc predom-
inantly migrated at a single M position similar to that of native
r
Skp1 (Figure 2C), not at the higher positions of Skp1A-myc
with or without the P143A substitution. The reason for
anomolous migration is not understood. The mobility of the
main part of this band was not shifted by treatment of the cells
with α,α′-dipyridyl. However, labeling of the upper region of
this band appeared to be reduced, consistent with the presence
of a small proportion of glycosylated protein seen in the MS
and sugar composition studies. A second Skp1 mutant with
point mutations in its N-terminal half, Skp1A2-myc (Table I),
exhibited electrophoretic behavior that was indistinguishable
from that of Skp1A1-myc (data not shown). Thus, mutant
Skp1A1-myc and probably Skp1A2-myc consists predominantly
of the nonglycosylated form, consistent with demonstration of
the former as a substrate for the Skp1 GlcNAc-Tase (Teng-
umnuay et al., 1999). However, its low level of glycosylation
does not appear to be the result of overexpression (compare
Figure 2C with 2B) as originally surmised, but due to the
missense mutations, which also cause (possibly indirectly)
anomolous gel migration.
Fig. 3. M analysis of Skp1-myc mutants expressed under the discoidin
r
promoter. The normal strain Ax3 (lane 1) was transfected with cDNAs
encoding Skp1-myc under the control of the discoidin 1γ promoter. Lane 2,
normal Skp1A-myc; lane 3, Skp1A3(P143A)-myc; lane 4, Skp1B-myc; lane 5;
Skp1B1(P143A)-myc. Stationary phase (at maximum cell density in growth
medium) cells were collected and analyzed as in Figure 2 by Western blotting
6
with anti-Skp1 antibodies. (A) 3 × 10 growth-phase cells from high-level
expression level colonies were probed with mAb 9E10 against the c-myc tag on
6
To determine whether Skp1B is glycosylated in the same
way as are Skp1A and total Skp1, which were analyzed
previously (Teng-umnuay et al., 1998), Skp1B-myc was
purified to homogeneity from the high-level expressor strain
HW302, reduced, alkylated, and digested with endo-Lys-C.
The resulting peptides were separated on a reverse-phase high-
performance liquid chromatography (RP-HPLC) column and
analyzed by matrix-assisted laser desorption ionization–time
of flight–mass spectrometry (MALDI-TOF-MS). Two sequential
absorbance peaks were found to contain abundant ions with m/z
expressed Skp1. (B) 10 growth-phase cells from low-level expression clones
listed in Table I) were probed with mAb 9E10. (C) A parallel blot of low-level
(
expression strains was probed with mAb 3F9, which recognizes both
endogenous and expressed Skp1. Positions of Skp1 isoforms are identified at
the left; asterisks refer to nonglycosylated variants.
4
), comparable to the change induced by α,α′-dipyridyl
(Figure 2A). A similar M difference (Figure 3B) was seen in
r
low-expressing clones employed for some of the localization
studies described below. Skp1-myc was expressed in these
clones at substantially lower levels than endogenous Skp1, based
on comparisons of their levels using mAb 3F9 (Figure 3C).
Expressed and endogenous Skp1s were readily differentiated
+
ratios of 2487 and 1636, corresponding to the MH values of
the pentasaccharide-peptide139–155 and peptide139–155
detected previously in Skp1A-myc (Teng-umnuay et al., 1998).
In-source fragmentation of the m/z 2487 ion revealed a series of
daughter ions corresponding precisely to the sequence Hex-
Hex-deoxyHex-Hex-HexNAc-HyPro-peptide, suggesting that
the Skp1B oligosaccharide was identical to that of the Skp1A-
myc oligosaccharide. The presence of the unmodified peptide,
as indicated by the m/z 1636 ion, suggested the existence of
unmodified Skp1B-myc in the cell, but the level of this isoform
was low based on SDS–PAGE/Western blot analysis as
described above (Figure 2B).
by the increase in M of 1185 resulting from the c-myc epitope.
r
To verify that normal Skp1-myc was glycosylated, strain
HW302, which expresses a high level of Skp1B-myc, was
treated with α,α′-dipyridyl as described above. As seen for
endogenous Skp1, a fraction of Skp1B-myc migrated more
rapidly in the gel (Figure 2B) to a position similar to that of
Skp1B1(P143A)-myc. Altogether, these results showed that it is
possible to stably express Skp1A and Skp1B with C-terminal
c-myc tags at high and low levels relative to endogenous Skp1.
As seen for total Skp1, both expressed Skp1A and Skp1B are
nearly completely glycosylated even when expressed at high
levels. In addition, substitution of Pro143 with an Ala residue
blocks glycosylation, as expected, and nonglycosylated forms
of the proteins are also stably expressed.
A previous study (Teng-umnuay et al., 1998) examined the
glycosylation of a mutant version of Skp1A1-myc produced by
strain HW120 that contained two missense mutations distant
from the glycosylation site (Table I). MS analysis of attach-
ment-site peptides (139–155) showed substantial levels of
peptides lacking Pro143 hydroxylation or containing this
modification but lacking glycosylation, in addition to those
containing partial and complete pentasaccharides. Quantitative
sugar composition analysis suggested that only about 15% of
Skp1A1-myc was glycosylated (Teng-umnuay et al., 1998).
Skp1 expressed in prespore cells is not glycosylated
To investigate possible developmental regulation of glyco-
sylation, Skp1B-myc and Skp1B1(P143A)-myc were
expressed under the control of the cotB promoter, which is
absolutely specific for prespore cells (Fosnaugh and Loomis,
1993), which appear later in development. After 20 h of starvation,
at which time about 75% of the cells have differentiated into
prespore cells, Western blot analysis showed that each of these
Skp1s was expressed at a high level compared to endogenous
Skp1 (Figure 4B, lanes 2, 3). Their apparent M values were
r
indistinguishable (Figure 4A, lanes 1–3), indicating that
Skp1B-myc was not glycosylated. Although the capacity of the
glycosylation pathway might have been saturated by the high
level of Skp1B-myc expressed, there was no evidence for even
partial glycosylation of expressed Skp1.
286

Nuclear enrichment of glycosylated Skp1
Table I. Strains used in this report
Name
Skp1 protein expresseda
Promoter
self
Expression levelb
normal
low
% Glycosylationc
Ax3(A3)
HW348
HW120
HW351
HW349
HW350
HW302
HW352
HW358
HW360
TF125
parentala
> 90%
Skp1A-myc
discoidin
discoidin
discoidin
discoidin
discoidin
discoidin
discoidin
cotB
> 90%
Skp1A1(I34T,D71G)-myc
Skp1A2(K53A,V54S,L55S)-myc
Skp1A3(P143A)-myc
Skp1B-myc
high
20%
high
20%
low
none
low
> 90%
Skp1B-myc
high
> 90%
Skp1B1(P143A)-myc
Skp1B-myc
low
none
high
< 10%
Skp1B1(P143A)-myc
cotB
high
none
selfd
normal
normal
> 90%
HL250
self
> 90% (core disaccharide)
aAll strains express endogenous Skp1A and Skp1B (previously referred to as FP21), encoded by the fpa1 and fpa2 genes. Expressed Skp1s contained C-terminal
c-myc epitope tags.
b
Low level is equivalent to < 25% of the level of endogenous Skp1 (normal); high level refers to 1–3× the level of endogenous Skp1. Levels were estimated from
the ratio of cell extracts required to produce similar signals on the same Western blot.
c
Approximate percent glycosylation was estimated from the ratio of cell extracts required to produce similar signals from the lower mobility glycosylated form
and the higher mobility aglycoform on the same Western blot. An inequality means that an equivalence was not achieved at the indicated ratio. Glycosylated Skp1
appeared to express the full pentasaccharide chain based on SDS–PAGE and mass spectrometry, except for strain HL250, which expressed only the core
disaccharide.
d
Strain TF125 (provided by W.F. Loomis) contains an insertion of the pyr5,6 gene 900 bp upstream of the fpa1 start codon.
Fig. 4. M analysis of Skp1-myc mutants expressed under the prespore cell-specific cotB promoter. Strain Ax3 and clonal cell lines expressing Skp1B-myc or
r
Skp1B1(P143A)-myc (as indicated in bottom legend) under the control of the cotB promoter were developed to the slug stage (18–20 h). At this time (day 0), slugs
were dissociated into single cells and returned to HL-5 growth medium, which blocks cotB promoter activity. Whole cells were analyzed after 0 (immediately), 1,
3, and 5 days in this medium, as described above. Cell proliferation did not initiate until day 2. (A) Gel lanes were loaded to yield similar signal levels for
Skp1B-myc and Skp1B1(P143A)-myc from different days to facilitate M comparisons, and the blot was probed with mAb 9E10. The day 0 results are from
r
experiment 1 (same blot), and the day 1–5 results are from experiment 2. (B) Gel lanes were loaded with similar cell numbers (except lane 4, which is overloaded),
to show changes in absolute levels of expression, and probed with mAb 3F9 to compare levels of expressed and endogenous Skp1. Day 0 results are from
experiment 3, and the day 1–5 results are from experiment 4. Positions of Skp1 isoforms are identified at the right; asterisks refer to nonglycosylated variants.
Immunoreactive material that accumulated at higher M positions in lanes 4–9 is not shown.
r
To ask if Skp1B-myc expressed in prespore cells was
competent to be glycosylated, the 20-hour cells were dissociated
and restored to growth medium (HL-5) to determine whether
previously synthesized Skp1B-myc could be shifted up in Mr
as cells began to proliferate again. After 1 day in HL-5, Skp1B-myc
and Skp1B1(P143A)-myc persisted at high levels relative to
endogenous Skp1 as determined by probing with mAb 3F9
indistinguishable based on probing with either mAb 3F9 or
9E10 (Figure 4A,B, lanes 5, 6). Subsequently, the levels of
Skp1B-myc and Skp1B1(P143A)-myc declined and eventually
became undetectable by day 6, after about five population
doublings. By day 3, a higher M form of Skp1B-myc appeared
r
(Figure 4A,B, lanes 8, 9), and this became the predominant
isoform by day 5 (lanes 11, 12). Because the Mr of
Skp1B1(P143A)-myc did not change, this increase must have
(
Figure 4B, lanes 5, 6), and their apparent M values remained
r
2
87

S. Sassi et al.
been due to subsequent glycosylation at position 143. Thus
expressed Skp1B-myc was stable during this developmental
transition, whether or not it was glycosylated. Because Skp1B-myc
produced by prespore cells was glycosylated when cells were
returned to growth phase, this suggested that its failure to be
glycosylated in prespore cells was the result of down-regulation of
the hydroxylation/glycosylation pathway in prespore cells,
rather than an incompetence of Skp1 to be posttranslationally
modified.
unpublished data). Fluorescence was found throughout most of
the cytoplasm and was concentrated in a few patches in the
cytoplasm and in the nucleus, as determined by colabeling with
4′,6′-diamidino-2-phenylindole (DAPI) (Figure 5B, panels 1–3).
Only an indistinct haze was seen in the absence of primary
antibody (panel 4). Cells that had become multinucleated while
grown in suspension were labeled in all nuclei. Deconvolution
microscopy demonstrated that Skp1 was localized in the
nuclear interior (panels 1–3, insets). Similar results were
obtained using mAb 4E1, an independent anti-Skp1 mAb that
is also monospecific for Skp1 based on Western blot analysis
(Kozarov et al., 1995) (data not shown). Nuclear enrichment
was also observed when paraformaldehyde-fixed cells were
permeabilized with 0.1% NP-40 in place of acetone or MeOH,
but was not as pronounced when cells were fixed with only
cold MeOH or MeOH/HAc (data not shown), as previously
reported for mammalian Skp1 (Freed et al., 1999).
To determine if expressed Skp1B-myc was similarly localized,
growing cells of strain HW302 were immunoprobed with mAb
9E10, which recognizes a continuous epitope associated with
the C-terminal c-myc tag. mAb 9E10 is monospecific for
expressed Skp1B-myc based on Western blot analysis of
whole cells (Figure 5A, lanes 4, 3). Fluorescence labeling was
diffuse throughout the cytoplasm and was concentrated in the
nucleus (Figure 5B, panels 5–7). Only a low level of indistinct
fluoresence was seen in cells of the parental strain Ax3 labeled
Localization of Skp1 in the cell
Previous studies concluded that Skp1 is concentrated in the
nucleus and the centrosome, relative to the cytoplasm, based
on indirect immunofluorescence microscopy of cultured
mammalian cells (Freed et al., 1999; Gstaiger et al., 1999).
Localization of Skp1 in Dictyostelium was investigated based
on a similar approach using the mAbs described in the Western
blot studies above. Growing cells of the normal strain Ax3
were allowed to attach to cover slips in HL-5 growth medium,
fixed in 4% paraformaldehyde, permeabilized with cold
acetone, and immunoprobed using an indirect technique with
mAb 3F9. This antibody was induced against denatured, gel-
purified Skp1 and is monospecific for Skp1 based on Western
blot analysis (Figure 5A, lane 1), and immunoprecipitation of
either native or denatured cell extracts with mAb 3F9 failed to
enrich for any additional reactive bands (Larner and West,
Fig. 5. Immunofluorescence localization of total and expressed Skp1 in growing cells. (A) The specificity of mAb 3F9 for Skp1 (lanes 1, 2), and of mAb 9E10
lanes 3, 4) for expressed Skp1, was evaluated by Western blot analysis after SDS–PAGE on a discontinuous, 7–20% gradient polyacrylamide gel. The full length
(
6
of the blot is shown. 2 × 10 cells of strain Ax3 (lanes 1, 3) or strain HW302 (lanes 2, 4) was loaded on each lane, and probing was at a 1:100 dilution of primary
antibody. (B) Growth phase cells were allowed to attach to glass cover slips in HL-5 and fixed in paraformaldehyde and acetone. Normal strain Ax3 cells were
probed with mAb 3F9 to localize total Skp1 (panel 1), and counterstained with DAPI to show nuclei and mitochondria (panel 3). Panels 1 and 3 are overlaid in
panel 2. Insets: A single optical section derived from an image taken on a deconvolution microscope. Panel 4 shows strain Ax3 cells probed with the second Ab
only and photographed using the same exposure conditions as panel 1. Strain HW302 cells were probed with mAb 9E10 to selectively detect expressed
Skp1B-myc (panel 5) and counterstained with DAPI (panel 7); the overlay of panels 5 and 7 is in panel 6. Panel 8 shows strain Ax3 cells probed with mAb 9E10,
photographed using the same exposure conditions as panel 5.
288

Nuclear enrichment of glycosylated Skp1
with mAb 9E10 (panel 8), showing that fluorescence localization
was specific for expressed Skp1B-myc. Skp1A-myc was similarly
localized in strain HW348 (see Figure 7A,B below). The
congruent nuclear concentration of endogenous Skp1 detected
by mAbs 3F9 and 4E1 and expressed Skp1A-myc and Skp1B-
myc detected by mAb 9E10 strongly supports the inter-
pretation that Skp1 is normally enriched in this organelle. This
interpretation is in agreement with similar conclusions drawn
from different antibodies that Skp1 is nuclear-enriched in
mammalian cell lines (Freed et al., 1999; Gstaiger et al., 1999).
In an attempt to gain biochemical evidence for nuclear
enrichment of Skp1, previous experiments that failed to detect
Skp1 in purified nuclei (Kozarov et al., 1995) were repeated
with similar negative results (data not shown). Inefficient
recovery of Skp1 after purification of mammalian cell nuclei
was previously reported (Gstaiger et al., 1999). This suggests
that Skp1 belongs to the major class of nonbound nuclear
proteins that are rapidly lost from nuclei isolated in aqueous
buffers (Allfrey, 1959; Paine et al., 1983), which is consistent
with the lower level of Skp1 seen in alcohol-fixed compared to
aldehyde-fixed cells (see above).
shown). The varied range of distributions suggested that
Skp1’s cytoplasmic localization is dynamic, but that its enrich-
ment in the nucleus is constant.
Nuclear enrichment of Skp1 requires glycosylation
The potential role of glycosylation in Skp1 localization was
examined in the low-level Skp1-myc expression strains
described above in which the glycosylation site had been
mutated from P to A. When expressed in growing cells under
the discoidin promoter, no nuclear enrichment of
Skp1A3(P143A)-myc was observed in any of the hundreds of
cells examined in multiple trials (Figure 7C,D). In contrast,
total Skp1 was concentrated in the nucleus in the same
population of cells (Figure 7E), and normal Skp1A-myc was
concentrated in the nucleus (Figure 7A,B). Thus the P143A
substitution, which occurs naturally in some metazoan Skp1
genes, appeared to block both glycosylation and nuclear
concentration.
Two Skp1A-myc mutants described above, Skp1A1-myc
and Skp1A2-myc (Table I), are poorly glycosylated and thus
provided an opportunity to explore the relationship between
glycosylation and nuclear concentration further. Skp1A1-myc
was not enriched in the nucleus, but its localization in the
cytoplasm of cells developed in KP buffer appeared normal
The localization of Skp1 in early developing cells preparing
for chemotactic aggregation was investigated after transferring
cells into KP buffer for 2–8 h. The nuclear concentration of
endogenous and expressed Skp1 that was seen in growing cells
was conserved, but cytoplasmic labeling along the plasma
membrane and in cytoplasmic extensions was more prevalent
(
Figure 6E,F). Skp1A2-myc was similarly not concentrated in
the nucleus (data not shown). These results show that when
glycosylation is blocked by a different mechanism than the
P143A mutation, nuclear concentration relative to the cyto-
plasm still does not occur. Because most Skp1A1-myc is
(
Figure 6A–D). Cytoplasmic labeling was often (but not
always) correlated with F-actin as determined by colabeling
with FITC-phalloidin. Prespore cells, formed late in develop-
ment, also showed nuclear enrichment of Skp1 (data not
Fig. 6. Localization of total and expressed Skp1 in developing cells. Growth
phase cells were incubated in KP buffer for 8 h on cover slips, fixed, and
immunolabeled as shown in Figure 5 except that cells were counterstained with
FITC-phalloidin to localize F-actin. (A–D) Four examples of normal strain Ax3
labeled with mAb 3F9, showing the variable subplasmalemmal localization of
Skp1 that in some cases overlaps with F-actin. (E, F) Two examples of the
localization of expressed mutant Skp1A1-myc in strain HW120 using mAb
Fig. 7. Localization of the P143A mutant of Skp1A in growing cells. Growth
phase cells were attached to cover slips and imunoprobed as in Figure 5. (A, B)
Strain HW348 cells were probed with mAb 9E10 to localize expressed
Skp1A-myc and DAPI to localize nuclei. (C, D) Strain HW349 cells were
probed with mAb 9E10 to localize expressed Skp1A3(P143A)-myc, which is
not glycosylated, and DAPI, to localize nuclei. (E) Strain HW349 cells were
probed with mAb 3F9 to localize total Skp1.
9
E10. (G, H) Two examples of the localization of Skp1 in strain HL250 using
mAb 3F9, in which Skp1 is modified by only the core disaccharide. The
mAb-dependent focal cytoplasmic labeling in panels A–D,G and H correspond
to nuclei.
2
89

S. Sassi et al.
hydroxylated at Pro143 (Teng-umnuay et al., 1998), hydroxy-
lation alone was not sufficient for nuclear enrichment relative
to the cytoplasm.
Localization of Skp1B-myc expressed in prespore cells,
where it was not glycosylated, was also found to be not
enriched in the nucleus (Figure 8A). This result was confirmed
by deconvolution microscopy (Figure 8B). Its localization was
similar to expressed Skp1B1(P143A)-myc (Figure 8C).
Endogenous Skp1 was still nuclear-enriched in these cells
(
data not shown), showing they had not lost capacity of nuclear
concentration. Nucleoli were especially devoid of Skp1B-myc
Figure 8C), and in some cells, Skp1 appeared to accumulate at
(
the nuclear surface (Figure 8C).
Finally, the localization of total Skp1 in cells whose
glycosylation was partially inhibited using prolyl hydroxylase
inhibitors was examined. Both α,α′-dipyridyl and ethyl 3,4-
dihydroxybenzoate (DHB), at concentrations found to inhibit
Skp1 glycosylation, appeared to reduce the nuclear concentration
of Skp1. Shown in Figure 9 is an example from a concentration
series of DHB applied to HW302 cells in KP buffer for 17 h.
At 0.2 mM DHB, the fluorescence intensity of mAb 3F9
labeling of nuclei relative to cytoplasm was slightly reduced
relative to the untreated control (compare Figure 9B and A).
This effect was more pronounced at 0.4 mM and 0.6 mM DHB,
but a slight enrichment of Skp1 in nuclei relative to adjacent
cytoplasm was observed even at the highest concentration.
This was consistent with an inability of the inhibitors to block
all Skp1 hydroxylation (Figure 2). The effect of DHB was also
assessed by counting the number of DAPI-positive nuclei that
were not selectively immunolabeled relative to the surrounding
cytoplasm in Photoshop images. This fraction was 0.10, 0.17,
0.44, and 0.41 for the concentration series of 0, 0.2 mM,
Fig. 9. Localization of Skp1 when glycosylation was inhibited by ethyl
,4-dihydroxybenzoate. HW302 cells were incubated at the indicated
3
concentrations of ethyl 3,4-dihydroxybenzoate in KP buffer for 17 h on cover
slips. Cells were processed for immunofluorescence as described in Figure 5,
using mAb 3F9 for total Skp1 (left panels) and DAPI-staining to locate nuclei
(
right panels).
0
.4 mM and 0.6 mM shown in Figure 9 (N = 200 per data
point). These results show that when Skp1 glycosylation is
inhibited by a mechanism that is independent of altering Skp1
structure, the ability of Skp1 to be concentrated in the nucleus
is still compromised.
To examine whether the entire pentasaccharide modification
is important for nuclear accumulation, Skp1 was localized in a
GDP-Fuc synthesis mutant (HL250). In HL250, Skp1 contains
only the Galβ1,3GlcNAc core disaccharide at HyPro143
(
Teng-umnuay et al., 1998). Nevertheless, HL250 Skp1 exhib-
ited a high level of nuclear concentration (Figure 6G,H),
showing that only the core disaccharide is important for
nuclear concentration.
Fig. 8. Localization of Skp1B-myc expressed in prespore cells. Cells were
developed to the slug stage of development, dissociated into single cells, and
allowed to attach to cover slips. Fixation and immunoprobing were as in
Figure 5. (A) Strain HW358 was immunoprobed with mAb 9E10 to localize
expressed Skp1B-myc, and counterstained with DAPI to locate nuclei. (B) As
in A, except that this is a single optical plane from a deconvolution image.
Discussion
(
C) Strain HW360 was immunoprobed with mAb to localize expressed
Dictyostelium Skp1 is found in both the nucleus and cytoplasm
Skp1B1(P143A)-myc and counterstained with DAPI to locate nuclei.
Variations in nuclear staining occur because some are out of focus. Nucleoli are
evident as DAPI-negative zones in the nuclei.
Three independent monoclonal antibodies that are mono-
specific for Skp1 based on Western blot analysis were used to
290

Nuclear enrichment of glycosylated Skp1
localize Skp1 in cells by indirect immunofluorescence. One of
the mAbs is specific for epitope-tagged Skp1, and the other
two probably recognize continuous epitopes that are likely to
be faithfully displayed by the Western blot method used to
assess specificity. Fluorescence signals were substantially
higher than control levels in the nucleus and in a heterogeneous
pattern in the cytoplasm, and fluorescence signals over nuclei
were substantially greater compared to adjacent cytoplasm
owing to variations in cell shape. It is unlikely that the glyco-
sylation site mutation simply destabilized Skp1, as this amino
acid substitution occurs naturally in certain Skp1 genes of A.
thaliana and C. elegans, and Skp1B1-myc expressed in
prespore cells was as stable as normal expressed Skp1B-myc
(Figure 4).
A similar correlation between loss of nuclear enrichment and
lack of glycosylation was also seen in mutants with amino acid
substitutions at codons 34 and 71 (Skp1A1-myc), or codons
54–56 (Skp1A2-myc) (Figures 2C, 6E,F). In contrast, concen-
tration in the peripheral cytoplasm of developing cells was
unaffected (Figure 6). Nuclear enrichment was also lost when
Skp1B-myc was expressed in prespore cells (Figures 4, 8), in
which case the lack of glycosylation did not involve any
mutations in Skp1 itself. Finally, when Skp1 glycosylation was
partially inhibited by the independent method of treating cells
with the prolyl hydroxylase inhibitors α,α′-dipyridyl (Figure 2A)
or ethyl 3,4,-dihydroxybenzoate, the concentration of nuclear
Skp1 relative to nearby cytoplasm also appeared reduced
(Figure 9). Taken together, these results strongly indicate that
glycosylation of Skp1 is required for its enrichment in the
nucleus relative to its level in nearby cytoplasm. However, an
alternative explanation might be that absence of glycosylation
resulted in change in Skp1 binding in the nucleus, thereby
masking both the mAb 3F9 and 9E10 epitopes in the nucleus
selectively. Such a differential effect on nuclear and cyto-
plasmic forms of Skp1 seems unlikely, however, because the
effect was seen after acetone permeabilization and because
most cellular Skp1 is thought to be bound in similar SCF
complexes, but ruling out this formal possibility must await
future studies.
(
Figure 5). Fluorescence originated from within the nucleus
based on decovolution microscopy. No differences in nuclear
enrichment were observed between endogenous Skp1
and expressed Skp1A-myc or Skp1B-myc, indicating that the
C-terminal c-myc tag did not interfere with Skp1 localization.
Similar results were obtained when cells were fixed with para-
formaldehyde or organic solvents, and thus nuclear concentra-
tion of Skp1 is unlikely to be a stress response to the aldehyde.
Enrichment of Skp1 in the nucleus is consistent with similar
results obtained in cultured mammalian cells (Freed et al.,
1
999; Gstaiger et al., 1999) and, as previously noted, is
consistent with the compartmentalization of some target
substrates of SCF complexes there.
In developing cells, substantial fluorescence was also
concentrated over regions of peripheral cytoplasm either at the
cell margin or cell top, which sometimes superimposed over
actin-rich regions of the cell (Figure 6). Localization in the
cytoplasm is consistent with the compartmentalization of other
substrates of SCF complexes, and data that Skp1 interacts with
the actin-binding protein coronin (Uetz et al., 2000) and the
barbed ends of F-actin in vitro (Bubb et al., unpublished data).
Localization in the centrosome, seen previously using
polyclonal antisera in mammalian cells (Freed et al., 1999;
Gstaiger et al., 1999), was not observed in Dictyostelium using
our mAbs. Despite its glycosylation, Skp1 is not vesicular
consistent with the absence of targeting sequences or other
evidence for entering the secretory pathway (Kozarov et al.,
The core disaccharide was sufficient for nuclear enrichment,
as Skp1 was localized normally in the nuclei of the general
fucosylation-defective strain HL250. Pro143 hydroxylation
itself was insufficient, as Skp1A1-myc, which is mostly
hydroxylated but not glycosylated (Teng-umnuay et al., 1998),
was not enriched in the nucleus (Figure 6E,F).
1995).
Enrichment of Skp1 in the nucleus relative to the
surrounding cytoplasm was consistently observed at all stages
of development examined. This suggested that Skp1 might
also be enriched in isolated nuclei relative to other cellular
fractions, but this was not observed (Kozarov et al., 1995).
Loss of proteins from nuclei isolated in aqueous media has
been documented (Allfrey, 1959; Paine et al., 1983) and was
also observed for mammalian Skp1 (Gstaiger et al., 1999).
However, the interpretation that Skp1 is normally concentrated
in nuclei is strengthened by evidence, described below, that
this apparent localization is modulated by specific structural
changes in Skp1.
Possible mechanism of nuclear concentration
Although the great majority of Skp1 is glycoslyated, only a
fraction of it is present in the nucleus. Thus, glycosylation seems
to be necessary but not sufficient for nuclear accumulation. In
cell extracts, the majority of Skp1 is inaccessbile to immuno-
precipiation by any of the anti-Skp1 mAbs except after
denaturation, reduction, and alkylation (Larner and West,
unpublished data), suggesting that Dictyostelium Skp1 is in
protein complexes such as the SCF complex that would be too
large to enter the nucleus unless actively transported. Although
it is not known if complex formation occurs before or after
Skp1’s nuclear concentration, it is unlikely that glycosylation is
necessary for SCF complex formation as recombinant Skp1
from E. coli, which is not glycosylated, enters the SCF
complex (Deshaies, 1999). In addition, the glycosylation site
in Dictyostelium maps to the opposite side of the contract
interface between mammalian Skp1 and the F-box protein
(Schulman et al., 2000).
Nuclear concentration of Skp1 depends on its glycosylation
When the glycosylation site was mutated from Pro to Ala,
Skp1A3-myc was not enriched in the nucleus relative to the
nearby cytoplasm (Figure 7C,D). This did not appear to be the
result of saturating the nuclear concentration process, as
Skp1A3-myc was expressed at low levels in these strains
relative to endogenous Skp1 (Figure 3), and high levels of
expressed Skp1B-myc in other strains were still concentrated
in the nucleus (Figure 5B, panels 5–8). However, mutant
Skp1A3-myc did not appear to be absolutely excluded from
the nucleus, though the exact level is uncertain because it was
difficult to precisely establish the level of nonspecific labeling
Previous studies with neoglycoproteins have suggested that
sugars can potentiate nuclear uptake. When serum albumin is
derivitized with 20 or more mono- or disaccharides of Glc or
GlcNAc, it can be concentrated in permeabilized mammalian
cells by a process that requires ATP (Duverger et al., 1995,
2
91

S. Sassi et al.
1
996). If the Galβ1,3GlcNAc-disaccharide of Skp1 is a
nucleus. Current evidence suggests that Skp1 receives its
GlcNAc and Fuc residues in the cytoplasm (West et al., 1996;
Teng-umnuay et al., 1999); further work is required to localize
the Skp1 prolyl hydroxylase.
structural mimic of neoglycosylated serum albumin on at least
one site, then the nuclear enrichment of Skp1 might be related
to stimulated uptake of neoglycoproteins.
The effect of Skp1 glycosylation on nuclear concentration
might be mediated by a receptor. Eukaryotic cells contain
abundant levels of galectins in the cytoplasm and nucleus (for
example, Gaudin et al., 2000), including some that recognize
substituted Galβ1,3GlcNAc core disaccharides as seen in Skp1
Inhibitors of prolyl-4-hydroxylase have received considerable
interest for their potential to interfere with collagen deposition,
which depends on 4-hydroxylation of its Pro residues, in such
diseases as liver cirrhosis and atherosclerosis. Although there
is evidence for efficacy in cirrhosis, the mechanism seems to
involve inhibition of differentiation of collagen secreting cells
rather than a direct effect on collagen deposition per se
(Matsumura et al., 1997; Sakaida et al., 1999). A possible
explanation for this and other effects (McCaffrey et al., 1995)
is action on previously unknown prolyl hydroxylases that
modify cytoplasmic or nuclear rather than secretory substrates.
(Henrick et al., 1998). Dictyostelium expresses discoidins with
related binding specificities at high levels in the nucleus and
cytoplasm (Erdos and Whitaker, 1983; Alexander et al., 1992).
Thus, cytoplasmic/nuclear lectins are potential candidates for
participating in the mechanism of nuclear enrichment of Skp1.
Nuclear concentration of Skp1 might involve a shuttling
mechanism in which the balance of import over export (Feld-
herr, 1998; Kaffman and O’Shea, 1999) is influenced by glyc-
osylation. Skp1 harbors an amino acid sequence upstream of
its glycosylation site, 96-LFELILAANYL, that resembles
Crm1-dependent nuclear export signals in proteins of other
organisms (Nix and Beckerle, 1997). Phosphorylation can
promote or inhibit nuclear accumulation of some proteins
Summary
There is no evidence for differential expression or glyco-
sylation of the two Skp1 proteins expressed in Dictyostelium.
Both Skp1A and Skp1B are nearly quantitatively modified by
similar if not identical pentasaccharides at Pro143, except
when produced in prespore cells. Skp1 produced in this cell
type is not fully glycosylated but can be subsequently modified
when cells are returned to growth medium. The unusual
oligosaccharide-protein linkage of Skp1 is susceptible to
inhibitors of prolyl hydroxylases in vivo, which provides a new
method for detecting this type of complex glycosylation on
(
Kaffman and O’Shea, 1999), including inhibiting StatA
nuclear association in Dictyostelium (Ginger et al., 2000).
Interestingly, the Skp1 glycosylation site is adjacent to a string
of four Glu residues and might cover this acidic patch. Alterna-
tively, glycosylation may support nuclear localization indi-
rectly via an effect on Skp1 folding or SCF-complex assembly.
other cytoplasmic/nuclear proteins by M -shift analysis.
r
Prevention of glycosylation either by mutation of the attach-
ment site or other approaches inhibited concentration of Skp1
in the nucleus relative to the surrounding cytoplasm based on
immunofluorescence analysis. Although the mechanism of this
effect remains to be elucidated, it is apparent that this glycosyl
modification has a functional consequence on the compart-
mentalization of this protein.
Posttranslational modification of Skp1
Skp1 appeared to be constitutively expressed and glycosylated,
except that when it was expressed in prespore cells under the
cotB promoter it was not glycosylated (Figure 4). This
indicates that an early step in the pathway or a critical donor
substrate or cofactor is developmentally regulated. However,
this effect may not be physiologically significant because it is
not known whether Skp1 is normally produced by prespore
cells, and nonglycosylated variants of endogenous Skp1 were
not detected late in development (Figure 4). However, Skp1
expressed in these cells was stable and permitted an analysis of
nuclear concentration and potential reglycosylation (see
above). Both glycosylated and unglycosylated Skp1 were
stable and long-lived, as they were detectable for up to 5 days
when the prespore cells were returned to growth medium.
Furthermore, prespore-expressed Skp1B-myc became partially
glycosylated during this interval (Figure 4), showing that glyco-
sylation is not necessarily coupled to its biosynthesis. Another
conclusion that can be derived from the studies on mutant
Skp1A1-myc and Skp1A2-myc is that glycosylation depends
on the structure of Skp1, as their point mutations, which map to
the N-terminal half of the protein distant from the glycosylation
site, block its glycosylation (Table I).
Materials and methods
Cell strains and manipulations
Strain Ax3 was grown either axenically in HL-5 growth
medium on a gyratory shaker, or on SM agar plates in association
with Klebsiella aerogenes. For development, axenically grown
cells were washed in KP buffer (17 mM KPO , pH 6.5), and
4
either deposited on cover slips in KP buffer, maintained in
7
suspension at 2 × 10 /ml in KP buffer or deposited on nonnutrient
8
agar at 1.2 × 10 cells per 10-cm-diameter plate (West et al.,
1997). On cover slips or in suspension, cells enter the develop-
mental phase but did not aggregate, which normally occurs
after 6–9 h on nonnutrient agar surfaces. Cells develop into
slugs after 16–20 h on nonnutrient agar; about 80% of the slug
cells are prespore cells in which the cotB promoter is active
(Fosnaugh and Loomis, 1993). To dissociate prespore cells,
slugs were sheared through a 26-gauge needle in 20 mM
EDTA, 10 mM potassium phosphate, pH 7.5. The resulting single
cell suspension was washed in KP buffer by centrifugation and
resuspended in HL-5 medium. Strain TF125, isolated by W. F.
Loomis, contained an insertion of pyr5,6 about 900 bp
upstream from the Skp1A transcription start site. Strain HL250
cannot synthesize GDP-Fuc from GDP-Man and is a general
fucosylation-null strain (Gonzalez-Yanes et al., 1989).
The hypothetical Skp1 prolyl hydroxylase appears to be
related to known prolyl hydroxylases based on the sensitivity
of Skp1 glycosylation to two different inhibitors of prolyl
hydroxylases. Hydroxylation is thought to occur at the 4-position
of Pro143 as a synthetic peptide containing 4-OH-Pro is a
substrate of the Skp1-GlcNAc-transferase (Teng-umnuay et al.,
1
999). All known prolyl hydroxylases are residents of the
lumen of the rER (Kivirikko and Pihlajaniemi, 1998), which
contrasts with the localization of Skp1 in the cytoplasm and
292

Nuclear enrichment of glycosylated Skp1
RT-PCR
Mass spectrometry of glycopeptides
7
Aliquots of 10 cells were washed by centrifugation with KP
buffer and frozen at –80°C in 1.5-ml microcentrifuge tubes.
Samples were lysed with 1 ml TRIZOL™ Reagent (Life Tech-
nologies) by repetitive pipetting and incubated for 5 min at
room temperature. Chloroform (0.2 ml; Fisher HPLC grade)
was added, and the tubes were shaken vigorously and
incubated again for 5 min at room temperature. Tubes were
centrifuged at 12,000 × g for 10 min at 4°C. RNA was precipi-
tated from the upper, aqueous phase by mixing with 0.5 ml of
isopropanol, incubating for 10 min at room temperature, and
centrifuging at 12,000 × g at 4°C for 10 min. The pellet was
rinsed with 1 ml of 75% ethanol and dissolved in 50 µl of
water, yielding 100–150 µg of total RNA. One microgram
RNA was subjected to reverse transcription using 20 pmol of
the downstream primer (5′-TCTTCACACCATTCATTT-
TCTTTTCT), 1 mM of each dNTP, 20 U RNase inhibitor
Skp1B-myc was purified from strain HW302, digested with
endo-Lys-C, and fractionated on a C -reversed-phase column
as described previously (Teng-umnuay et al., 1998). All
fractions were analyzed by MALDI-TOF-MS on a Perceptive
Biosystems Voyager DE in the UF Protein Chemistry Core
Lab as also described in that report.
18
Skp1 expression constructs
The expression constructs were derived from pVEII∆ATG by
insertion of C-terminal myc-tagged Skp1-coding cDNA into
the KpnI and SacI sites of its MCS, positioned downstream of
the discoidin 1γ promoter intended to drive expression during
late growth phase and early development, as previously
described (Teng-umnuay et al., 1998, 1999). The original
strain HW120 expressed a version of Skp1 (Skp1A1-myc)
with two PCR-derived missense mutations (Table I). A
plasmid expressing normal Skp1A-myc was constructed by
repeating the procedure. The Skp1B-myc expression construct
was created by introducing an A39S substitution in the Skp1A-
myc plasmid using the QuikChange site-specific mutagenesis
kit (Stratagene, La Jolla, CA), with 5′-GGTGAATCAGAT-
GCGCCAATTCCATTA as a primer. A newly created HhaI
site was used to screen for the altered plasmid. P143A substi-
tutions were created by similar site-specific mutagenesis, using
(
Perkin Elmer), 5 mM MgCl , and 50 U MuLV reverse tran-
scriptase (Perkin Elmer), in a total vol of 20 µl of 1× PCR
buffer. The reaction was allowed to proceed for 15 min at
2
4
2°C, 5 min at 99°C, and 5 min at 5°C. The reaction mixture
was diluted fivefold in 1× PCR buffer to a final concentration of
®
2
mM MgCl and supplemented with 2.5 U AmpliTaq DNA
2
polymerase and 20 pmol of the upstream primer (5′-ATTG-
AAAAAGAAATCGCTTGTATGT). Primers corresponded to
sequences that were identical in fpa1 and fpa2 (Skp1A and
Skp1B genes). Amplification was carried out from 15 to 35
cycles (45 s, 94°C; 45 s, 49°C; 3 min, 72°C). DNA products
were resolved by electrophoresis in 1% agarose and visualized
with ethidium bromide. In some cases RT-PCR products were
digested with SspI, BclI, and EcoRV and visualized after electro-
phoresis in 3% NuSieve GTG Agarose (FMC Bioproducts).
5
′CAAGAACGACTTTACTGCAGAAGAAGAAGAAC as a
primer and screening for the newly created PstI site. Skp1A2-
myc, which contained amino acid substitutions at codons 53–55,
was similarly prepared using GCACTATTTTAGAAGCTTC-
TTCTGACTATTGCAGAC as a primer, and screening for the
introduced HindIII site. All sequences were confirmed
directly. Myc-tagged Skp1 was expressed throughout growth
and early development in these strains, and the expected
substantial suppression by folic acid (Blusch et al., 1992) was
not observed (data not shown). To direct expression of the
Skp1 constructs in prespore cells, the discoidin promoter was
replaced by the cotB promoter as previously described (Zhang
et al., 1999). Plasmids were transfected into Dictyostelium
Metabolic labeling
To test the effects of prolyl hydroxylase inhibitors on general
metabolism, growth phase cells were deposited at 10 cells/well of
7
a 24-well tissue culture plate in HL-5 growth medium. After
strain Ax3 as CaPO precipitates (Blusch et al., 1992) or by
4
cell attachment (15 min), the medium was replaced with 320 µl
electroporation (Pang et al., 1999), and transformants were
isolated in the presence of 15 µg/ml G418 and cloned.
0
–1000 µM α,α′-dipyridyl (Sigma) or 0–400 µM ethyl 3,4-
dihydroxybenzoate (Sigma) prepared in HL-5. After 30 min,
Electrophoresis
35
the medium was supplemented with [ S]Met (1000 Ci/mmol;
3
Cells were pelleted in a microcentrifuge tube, lysed in
Laemmli sample buffer, and separated on 15–20% linear
gradient SDS–polyacrylamide slab gels. Gels were electropho-
retically transferred on a semi-dry apparatus and immuno-
chemically probed as described elsewhere (West et al., 1997).
New England Nuclear) or [6- H]L-Fuc (85.2 Ci/mmol; New
England Nuclear), which had been dried down and recon-
stituted at 10× strength in HL-5, to a final concentration of
1
3
0 µCi/ml, and cells were allowed to incubate for an additional
h. For labeling of differentiating cells, HL-5 was replaced
with KP buffer after cell attachment, and cells were incubated
for 6 h in KP buffer before drug treatment and metabolic
labeling as above. Cells were transferred to 15-ml tubes,
diluted in KP, and centrifuged to recover the washed cells. The
pellet was resuspended in 1 ml of 0.2% NP-40 in KP buffer. An
aliquot of 0.25 ml was transferred to a microcentrifuge tube
and supplemented with 200 µl 0.125 mg/ml bovine serum
albumin in water. Protein was precipitated with 50 µl of 100%
trichloroacetic acid for 2 h on ice and collected on a GF/C glass
fiber filter, rinsed with cold 10% trichloroacetic acid and
acetone, and assayed by liquid scintillation counting.
Immunofluorescence localization
Growing cells from HL-5 medium were deposited on glass
cover slips and allowed to attach for 15 min. In some cases,
attached cells were either maintained in HL-5 or rinsed in KP
buffer and allowed to develop, for up to 17 h, in the absence or
presence of a range of concentrations of α,α′-dipyridyl or ethyl
3
2
,4-dihydroxybenzoate. Cells were fixed by treatment with
–4% paraformaldehyde in KP buffer, pH 7.4, for 15 min,
followed by treatment with either acetone (–20°C) or 0.1%
NP-40 in KP buffer for 5 min. Similar results were obtained
with both methods. Alternatively, cells were fixed in 50%
2
93

S. Sassi et al.
acetone:50% methanol at room temperature for 2 min,
followed by replacement with phosphate buffered saline
Chung, C.Y., Reddy, T.B., Zhou, K., and Firtel, R.A. (1998) A novel, putative
MEK kinase controls developmental timing and spatial patterning in
Dictyostelium and is regulated by ubiquitin-mediated protein degradation.
Genes Dev., 12, 3564–3578.
Deshaies, R.J. (1999) SCF and cullin/RING H2-based ubiquitin ligases. Annu.
Rev. Cell. Dev. Biol., 15, 435–467.
(
PBS), or in 95% ethanol:5% glacial HAc at –20°C for 5 min,
followed by PBS (Harlow and Lane, 1988). After rinsing in KP
or PBS, cells were blocked in 5% nonfat dry milk in TBS for
5
min, followed by incubation in 1°C antibody diluted in the
Duverger, E., Pellerin-Mendes, C., Mayer, R., Roche, A.-C., and Monsigny,
M. (1995) Nuclear import of glycoconjugates is distinct from the classical
NLS pathway. J. Cell Sci., 108, 1325–1332.
milk solution for 1 h, rinsing in TBS, and incubation in 1:200
dilution of affinity-purified Texas red conjugated rabbit anti-
mouse antibody (Sigma). Alternatively, fixed cells were
blocked in 3% (v/v) goat serum in PBS, incubated with anti-
bodies diluted in 1% goat serum, and probed with a 1:2,000
dilution of affinity-purified Alexa 568-conjugated goat anti-
mouse antibody (Molecular Probes). In some cases, cells were
subsequently incubated with FITC-phalloidin (Molecular
Probes) to detect F-actin. FITC-phalloidin was prepared by
drying an aliquot in a vacuum centrifuge, replacing with the
same volume of 10% Triton X-100, and diluting 1200-fold in
PBS. After final washing in TBS, cover slips were mounted in
Vectashield (Vector Labs) containing 0.1 µg/ml DAPI. Images
were collected digitally on a Zeiss Axiophot microscope
equipped with a SPOT-2 camera, or on an Applied Precision
deconvolution microscope equipped with Deltavision soft-
ware. Multiple color channel recordings were taken from the
same focal plane. Related images were processed similarly in
Adobe Photoshop. Conclusions were based on multiple trials
involving many cells viewed by independent observers. Unless
otherwise indicated, cells shown were representative of the
entire population.
Duverger, E., Roche, A.-C., and Monsigny, M. (1996) N-acetyl-Glucosamine
dependent nuclear import of neoglycoproteins. Glycobiology, 6, 381–386.
Eleftheriades, E.G., Ferguson, S.G., Spragia, M.L., and Sameral, A.M. (1995)
Prolyl hydroxylation regulates intracellular procollagen degradation in
culture rat cardiac fibroblasts. J. Mol. Cell Cardiol., 27, 1459–1473.
Ennis, H.L., Dao, D.N., Pukatzki, S.U., and Kessin, R.H. (2000) Dictyostelium
amoebae lacking an F-box protein form spores rather than stalk in chimeras
with wild type. Proc. Natl. Acad. Sci. USA, 97, 3292–3297.
Erdos, G.W., and Whitaker, D. (1983) Failure to detect immunocyto-
chemically reactive endogenous lectin on the cell surface of Dictyostelium
discoideum. J. Cell Biol., 97, 993–1000.
Feldherr, C.M. (1998) Macromolecular exchanges between the nucleus and
cytoplasm. J. Cell. Biochem. Suppl., 30–31, 214–219.
Fosnaugh, K., and Loomis, W.F. (1993) Enhancer regions responsible for
temporal and cell-type-specific expression of a spore coat gene in
Dictyostelium. Dev. Biol., 157, 38–48.
Freed, E., Lacey, K.R., Huie, P., Lyapina, S.A., Deshaies, R.J., Stearns, T., and
Jackson, P.K. (1999) Components of an SCF ubiquitin ligase localize to
the centrosome and regulate the centrosome duplication cycle. Genes
Develop., 13, 2242–2257.
Gaudin, J.C., Mehul, B., and Hughes, R.C. (2000) Nuclear localisation of wild
type and mutant galectin-3 in transfected cells. Biol. Cell, 92, 49–58.
Ginger, R.S., Dalton, E.C., Ryves, W.J., Fukuzawa, M., Williams, J.G., and
Harwood, A.J. (2000) Glycogen synthase kinase-3 (GSK-3) enhances
nuclear export of a Dictyostelium STAT protein. EMBO J., 119, 5483–5491.
Gonzalez-Yanes, B., Mandell, R.B., Girard, M., Henry, S., Aparicio, O.,
Gritzali, M., Brown, R.D., Erdos, G.W., and West, C.M. (1989) The spore
coat of a fucosylation mutant in Dictyostelium discoideum. Dev. Biol., 133,
Acknowledgments
5
76–587.
Gstaiger, M., Marti, A., and Krek, W. (1999) Association of human SCF
SKP2) subunit p19 (SKP1) with interphase centrosomes and mitotic
We are grateful to the ICBR Protein Chemistry Core Lab for
making the Voyager MALDI-TOF-MS available, to Alfred
Chung for his advice, and to Glenn Phillipsberg for carrying
out the analyses. Dr. W.F. Loomis generously provided strains
TF125 and HL250 and plasmid pCotBGal17, and Dr. M.R. Bubb
provided advice on the use of phalloidin. This work was
supported by a grant from the Florida Division of the American
Cancer Society and NIH RO1-GM-37539.
(
spindle poles. Exp. Cell Res., 247, 554–562.
Harlow, E., and Lane, D. (1988) Antibodies. A Laboratory Manual. Cold
Spring Harbor Laboratory, New York, p. 385.
Henrick, K., Bawumia, S., Barboni, E.A., Mehul, B., and Hughes, R.C. (1998)
Evidence for subsites in the galectins involved in sugar binding at the non-
reducing end of the central galactose of oligosaccharide ligands: sequence
analysis, homology modeling and mutagenesis studies of hamster galectin-
3. Glycobiology, 8, 45–57.
Kaffman, A., and O’Shea, E.K. (1999) Regulation of nuclear localization: a
key to a door. Annu. Rev. Cell. Dev. Biol., 15, 291–339.
Kivirikko, K.I., and Pihlajaniemi, T. (1998) Collagen hydroxylases and the
protein disulfide isomerase subunit of prolyl-4-hydroxylases. Adv.
Enzymol. Relat. Areas Mol. Biol., 72, 325–398.
Abbreviations
DAPI, 4′,6′-diamidino-2-phenylindole; DHB, dihydroxybenzoate;
KP, 17 mM KPO , pH 6.5; MALDI-TOF-MS, matrix-assisted
4
laser desorption ionization–time of flight–mass spectrometry;
MS, mass spectrometry; PAGE, polyacrylamide gel electro-
phoresis; PBS, phosphate buffered saline; RP-HPLC, reverse-
phase high-performance liquid chromatography; RT-PCR,
reverse-transcription polymerase chain reaction; TBS, Tris
buffered saline.
Koepp, D.M., Harper, J.W., and Elledge, S.J. (1999) How the cyclin became a
cyclin: regulated proteolysis in the cell cycle. Cell, 97, 431–434.
Kozarov, E., van der Wel, H., Field, M., Gritzali, M., Brown R.D. Jr., and
West, C.M. (1995) Characterization of FP21, a cytosolic glycoprotein from
Dictyostelium. J. Biol. Chem., 270, 3022–3030.
Laney, J.D., and Hochstrasser, M. (1999) Substrate targeting in the ubiquitin
system. Cell, 97, 427–430.
Matsumura, Y., Sakaida, I., Uchida, K., Kimura, T., Ishihara, T., and Okita, K.
(
1997) Prolyl 4-hydroxylase inhibitor (HOE 077) inhibits pig serum-
induced rat liver fibrosis by preventing stellate cell activation. J. Hepatol.,
7, 185–192.
References
2
Alexander, S., Sydow, L.M., Wessels, D., and Soll, D.R. (1992) Discoidin
proteins of Dictyostelium are necessary for normal cytoskeletal organization
and cellular morphology. Differentiation, 51, 149–161.
Allfrey, V. (1959) The isolation of subcellular components. In J. Brachet and
A.E. Mirsky, eds., The Cell, vol. I. Academic Press, New York, pp 193–290.
Blusch, J., Morandini, P., and Nellen, W. (1992) Transcriptional regulation by
folate: inducible gene expression in Dictyostelium transformants during
growth and early development. Nucleic Acids Res., 20, 6235–6238.
McCaffrey, T.A., Pomerantz, K.B., Sanborn, T.A., Spokojny, A.M., Du, B.,
Park, M, -H., Folk, J.E., Lamberg, A., Kivirikko, K.I., Falcone, D.J., and
others (1995) Specific inhibition of elF-5A and collagen hydroxylation by
a single agent. J. Clin. Invest., 95, 446–455.
Nelson, M.K., Clark, A., Abe, T., Nomura, A., Yadava, N., Funair, C.J.,
Jermyn, K.A., Mohanty, S., Firtel, R.A., and Williams, J.G. (2000) An F-box/
WD40 repeat-containing protein important for Dictyostelium cell-type
proportioning, slug behavior, and culmination. Dev. Biol., 224, 42–59.
294

Nuclear enrichment of glycosylated Skp1
Nix, D.A., and Beckerle, M.C. (1997) Nuclear-cytoplasmic shuttling of the
focal contact protein, zyxin: a potential mechanism for communication
between sites of cell adhesion and the nucleus. J. Cell Biol., 138, 1139–1147.
Paine, P.L., Austerberry, C.F., Desjarlais, L.J., and Horowitz, S.B. (1983)
Protein loss during nuclear isolation. J. Cell Biol., 97, 1240–1242.
Pang, K.M., Lynes, M.A,. and Knecht, D.A. (1999) Variables controlling the
expression level of exogenous genes in Dictyostelium. Plasmid, 41, 187–197.
Sakaida, I., Uchida, K., Hironaka, K., and Okita, K. (1999) Prolyl 4-hydroxy-
lase inhibitor (HOE 077) prevents TIMP-1 gene expression in rat liver
fibrosis. J. Gastroenterol., 34, 376–377.
Sasaki, T., Majamaa, K., and Uitto, J. (1987) Reduction of collagen production
in keloid fibroblast cultures by ethyl-3, 4-dihydroxybenzoate. J. Biol.
Chem., 262, 9397–9403.
Schauer, T.M., Nesper, M., Kehl, M., Lottspeich, F., Muller-Taubenberger, A.,
Erisch, G., and Gerisch, G. (1993) Proteasomes from Dictyostelium
discoideum: characterization of structure and function. J. Struct. Biol., 111,
novel pentasaccharide linked to hydroxyproline in Dictyostelium. J. Biol.
Chem., 273, 18242–18249.
Teng-umnuay, P., van der Wel, H., and West, C.M. (1999) Identification of a
UDP-GlcNAc:Skp1-hydroxyproline GlcNAc-Transferase in the cytoplasm of
Dictyostelium. J. Biol. Chem., 274, 36392–36402.
Uetz, P., Giot, L., Cagney, G., Mansfield, T.A., Judson, R.S., Knight, J.R.,
Lockshon, D., Narayan, V., Srinivasan, M., Pochart, P., and others (2000)
A comprehensive analysis of protein–protein interactions in Saccharomyces
cerevisiae. Nature, 403, 623–627.
West, C.M., Kozarov, E., and Teng-umnuay, P. (1997) The cytosolic glycoprotein
FP21 of Dictyostelium discoideum is encoded by two genes resulting in a
polymorphism at a single amino acid position. Gene, 200, 1–10.
West, C.M., Scott-Ward, T., Teng-umnuay, P., van der Wel, H., Kozarov, E.,
and Huynh, A. (1996) Purification and characterization of an α1, 2-L-fuco-
syltransferase, which modifies the cytosolic protein Skp1, from the cytosol
of Dictyostelium. J. Biol. Chem., 271, 12024–12035.
1
35–147.
Williams, A.R., Goh, T., Taylor, L., Chernushevich, I., Shevchenko, A., and
Tyers, M. (1999) SCF ubiquitin protein ligases and phosphorylation-
dependent proteolysis. Phil. Trans. R. Soc. Lond. B Biol. Sci., 354, 1533–1550.
Zhang, Y., Zhang, P., and West, C.M. (1999) A linking function for the
cellulose-binding protein SP85 in the spore coat of Dictyostelium
discoideum. J. Cell Sci., 112, 4667–4677.
Schulman, B.A., Carrano, A.C., Jeffrey, P.D., Bowen, Z., Kinnucan, E.R.E.,
Finnin, M.S., Elledge, S.J., Harper, J.W., Pagano, M., and Pavelick, N.P.
(
2000) Insights into SCF ubiquitin ligases from the structure of the Skp1-
Skp2 complex. Nature, 408, 381–386.
Teng-umnuay, P., Morris, H.R., Dell, A., Panico, M., Paxton, T., and
West, C.M. (1998) The cytoplasmic F-box binding protein Skp1 contains a
2
95