Download Identification of cellular proteins that bind to the human

Journal of General Virology(1993), 74, 1581-1589. Printedin GreatBritain
1581
Identification of cellular proteins that bind to the human immunodeficiency
virus type 1 nef gene product in vitro: a role for myristylation
Mark Harris* and Karen Coates
MRC Retrovirus Research Laboratory, Department of Veterinary Pathology, University of Glasgow, Bearsden Road,
Glasgow G61 1QH, U.K.
The human immunodeficiency virus (HIV) type 1 nef
gene product was expressed as an N-terminal fusion
protein with glutathione-S-transferase (GST) in the
baculovirus system. The resulting nefGST fusion protein
was found to be authentically myristylated at the
N terminus and could be purified to homogeneity by
one-step affinity chromatography on immobilized glutathione. The high affinity of nefGST for glutathione was
exploited to develop an assay to identify cellular proteins
capable of interacting with nef Several such proteins
were identified in extracts from the Jurkat human T cell
line. The interaction between neJZbinding proteins and
immobilized nefGST could be specifically competed by
the addition of soluble nef Cell fractionation showed
that neJ:binding proteins were present in both cytosolic
and membrane-associated fractions. A non-myristylated
derivative failed to bind to the membrane-associated
proteins but was able to bind to the cytosolic group,
albeit with reduced affinity. In addition, a single protein
present in both soluble and membrane-associated fractions exhibited myristylation-independent binding to
nef By analogy with other myristylated proteins such as
MARCKS (myristylated alanine-rich C kinase substrate)
and the Rous sarcoma virus transforming protein, src,
the membrane-associated proteins that bind only to
myristylated nef may represent a specific membrane
target for nef The cytosolic proteins that interact with
nef may constitute soluble components of an as yet
unidentified signal transduction pathway which is the
target of nefaction in the HIV-l-infected cell.
Introduction
subsequently identified as important for this enhancement, including Cys143, Cysl70 and Leu182
(Zazapoulos & Haseltine, 1992). Luria et al. (1991)
examined the effect of nefon the induction of interleukin
2 receptor (1R2R) mRNA: Jurkat T cells constitutively
expressing nefwere unable to respond to a wide range of
stimuli that normally induced the synthesis of IL2R
mRNA. In contrast, cells expressing an alternative nef
allele that differed at only three amino acid residues
responded normally.
The nefgene product is myristylated at the N terminus
(Allan et al., 1985) and this acylation is required for the
location of nef on the cytoplasmic face of the plasma
membrane in infected cells (Yu & Felsted, 1992).
Myristylation of nef enhances HI¥-I replication
(Zazapoulos & Haseltine, 1992) and, conversely, suppresses HIV-1 LTR transcription (Yu & Felsted, 1992).
Myristylation is thus important for nef function but the
precise role of acylation remains to be elucidated.
Limited sequence similarity between nefand the GTPbinding domains of ras and G proteins has been reported
(Samuel et al., 1987; Guy et al., 1987). Furthermore,
bacterially expressed nef was shown to bind and
hydrolyse GTP (Guy et al., 1987). Other studies with
both bacterial and baculovirus-expressed nef gene pro-
The nef gene of human immunodeficiency virus (HIV)
type 1 encodes a polypeptide of between 200 and 205
amino acids, varying between isolates (Harris et al.,
1992a). Despite intensive work, the precise function of
the nefgene product remains the subject of controversy.
Initial studies identified the protein as a repressor of viral
replication (Fisher et al., 1986; Luciw et al., 1987;
Terwilliger et al., 1986) that acted by down-regulation of
transcription from the HIV-1 long terminal repeat (LTR)
(Ahmad & Venkatesan, 1988; Niederman et al., 1989;
Maitra et al., 1991), but other studies failed to reproduce
these results (Kim et al., 1989; Hammes et al., 1989;
Bachelerie et al., 1990). There is a significant degree of
amino acid sequence variation between nefisolates both
in vivo and in vitro (Harris et al., 1992a) and the
contradictions regarding nef function may represent the
consequences of this variation. This problem was
highlighted by a number of recent studies. Terwilliger
et al. (1991) demonstrated that the replication of the IIIB
provirus was suppressed by the corresponding IIIB nef
allele but that replacement of nefin the IIIB provirus by
the ELI nefallele enhanced the growth of the provirus. A
number of amino acid residues in the ELI nef gene were
0001-1605 @1993 SGM
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 04 Aug 2017 00:59:29
1582
M. Harris and K. Coates
ducts, however, failed to reproduce these findings,
suggesting that the activity was due to a contaminating
bacterial enzyme (Kaminchik et al., 1990; Nebreda et al.,
1991 ; Matsuura et al., 1991 ; Harris et al., 1992a). It has
been reported that nef is phosphorylated by protein
kinase C, although this observation was made following
the expression of nef in BHK21 cells infected with a
recombinant vaccinia virus (Guy et al., 1987) and has not
been demonstrated for nef expressed in HIV-l-infected
human cells. Nef also possesses limited sequence similarity with the human thyrotropin receptor (Burch et al.,
1991) and scorpion neuroactive peptides (Werner et al.,
1991). All these observations, together with the location
of nef on the plasma membrane and its potential ability
to modulate IL2R activation, suggest that its role in the
viral life cycle involves perturbation of cellular signal
transduction pathways. In this context, nef has been
reported to down-regulate cell surface expression of CD4
(Guy et al., 1987; Garcia & Miller, 1991) although even
this putative function is subject to controversy because
Cheng-Mayer et al. (1989) failed to observe this
phenomenon.
Unambiguous definition of the biochemical function
of nef in the life cycle of HIV will involve the precise
identification of the target(s) of nef action. We
previously described the expression of both laboratory
and primary isolates of HIV- 1 nef in Escherichia eoli as
fusions with glutathione-S-transferase (GST) (Harris et
al., 1992a) using the pGEX system (Smith & Johnson,
1988). By exploiting the high affinity of GST for
glutathione immobilized on agarose beads to generate
nef affinity reagents, these fusion proteins appeared to
provide the ideal method for identifying nef-binding
proteins. However, although GST-nef fusion proteins
linked to glutathione-agarose (GA) beads were able to
precipitate specifically a set of cytoplasmic proteins from
extracts of human T cell lines (Harriss et al., 1992b)
binding was inefficient and not reproducible. One
possible explanation for this might be that nef was not
expressed in its authentic eukaryotic conformation. In
the pGEX system proteins are expressed as C-terminal
fusions with GST and it is likely that the N terminus of
nef needs to be exposed. Moreover, myristylation does
not occur in prokaryotes and, by analogy with other
myristylated proteins such as poliovirus P1 capsid
precursor (Ansardi et al., 1992) and the Rous sarcoma
virus transforming protein src (Resh & Ling, 1990),
myristylation of nef is potentially important for its
interaction with cellular proteins. Nef was therefore
expressed in the baculovirus expression system as the Nterminal portion of a GST fusion molecule. Nefexpressed
in this system was myristylated at the N terminus and
could be efficientlypurified by one-step affinity chromatography on GA. This new reagent allowed the definition
of three classes of nef-binding proteins on the basis of
subcellular localization and myristylation dependence.
Methods
Plasmid construction. The coding sequences for GST were isolated by
PCR using the primer pair NNNGTCGACATGTCCCCTATACTAGGT (coding strand) and NNNGGTACCCCCGGGTTATTTTGGAGGATGGTC (non-coding strand) (where N is any
nucleotide) and were cloned into pBluescript as a SalI-KpnI fragment.
A KpnI EcoRI-HindIII-BamHI oligonucleotide linker adaptor was
ligated into the KpnI site to facilitate further cloning steps. This
construct was termed pGST.
The coding sequence of the HIV-1 nef gene (strain BH10) was
isolated by PCR using the primer pair NNGGATCCTATAAG A T G G G T G G C A A G T G G (coding strand) and NNNNATCGATGCAGTTCTTGAAGTACTCCGG (non-coding strand), and ligated
onto the ClaI BamHI fragment of pGST in the presence of BamHI.
The ligation product was gel-purified and ligated into the BamHI site
of the transfer vector pAcCL29 (Livingstone & Jones, 1989).
For GST expression, a BamHI linker-adaptor was ligated to the
5' end of the SalI-BamHI fragment of pGST and this was cloned into
pAcCL29. The myristylated GST derivative (myrGST) was constructed
by cloning oligonucleotides encoding the myristylation sequence of
HIV-1 gag p55 (Met-Gly-Ala-Arg-Ala-Ser) into the SaII site of pGST
to generate the plasmid pmGST.
For the production of non-myristylated neff the coding strand primer
N N G G A T C C T A T A A G A T G A G T G G C A A G T G G was used, replacing the glycine residue at position 2 with a serine. Neftagged with six
histidine residues at the C terminus (nef-6H) was produced by PCR
using the coding primer described above and NNGGATCCTTAGTGATGGTGATGGTGATGTCTGCAGTTCTTGAAGTACTC
as
the non-coding primer. This PCR product was cloned directly into
pAcCL29 as a BamHI fragment.
Isolation and propagation of recombinant baculoviruses. Sf9 cells were
grown in TC100 medium (Gibco) containing 10% fetal calf serum
(FCS) at 27 °C. Cells (1 x 106 per 35 mm dish) were transfected by the
calcium phosphate method with 5 to 10 ~tg of the appropriate pAcCL29
recombinants and 1 lag of AcRPIacZ viral DNA linearized with Bsu36I
(Kitts et al., 1990). After 2 days the medium was harvested, filtered
through a 0.22 gm filter unit and plated onto Sf9 cells. White plaques
were picked and underwent three rounds of plaque purification prior
to expansion.
Purification of recombinant proteins. Sf9 cells were infected with the
appropriate recombinant viruses at 10p.f.u./cell and harvested at
72 h post-infection (p.i.). For the production of GST and GST fusion
proteins, infected cells were pelleted, washed twice with ice-cold PBS
and lysed in PBS containing 1% Triton X-100, 5 mM-EDTA, 10 mMiodoacetamide and the protease inhibitors aprotinin (2 gg/ml), PMSF
(1 mM) and leupeptin (1 lag/ml). After incubation at 4 °C for 30 min the
lysate was clarified by centrifugation at 10000r.p.m. for 15 min at
4 °C. Clarified supernatant was incubated with GA beads at 4 °C for
2 h and the beads were then washed extensively in 50 mM-Tris-HC1 pH
8.0. Fusion protein was eluted by the addition of 50 mM-Tris-HC1 pH
8.0 containing 10 m~a reduced glutathione. Fractions containing protein
were pooled and dialysed overnight against two changes of 50 mMTris HC1 pH 8"0. Then, 1 mM-PMSF and 5 mM-EDTA were added and
purified protein was stored in aliquots at - 7 0 °C.
For the production of purified nef-6H, clarified lysate was incubated
with nickel nitrilotriacetic acid (Ni-NTA)-agarose, washed with PBS
and eluted with 100mM-citrate buffer pH 6.0. Protein-containing
fractions were pooled, dialysed against 20 mM-sodium phosphate
buffer pH 7.2 and stored at - 7 0 °C.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 04 Aug 2017 00:59:29
Nef binds to cellularproteins in vitro
Analysis of myristylation. Sf9 cells (5 x 106 per 25 cm 2 flask) were
infected with recombinant viruses at 10 p.f.u./cell. At 18 h.p.i, cells
were washed with TC100 medium containing 0-25% BSA and
incubated with 2.5ml (per flask) TC100/0.25% BSA containing
100 IxCi/ml [ZH]myristic acid (Amersham; 53.4 Ci/mmol) for 24 h. Cells
were harvested, washed once with ice-cold PBS and lysed as described
above. Labelling with [3SS]methionine was carried out similarly using
methionine-free TC100 containing 1% FCS and 100gCi/ml
[35S]methionine (Amersham; 1000 Ci/mmol).
nefGSr
Western blotting. Proteins were separated by SDS-PAGE, transferred to nitrocellulose and membranes were blocked for 30 min in
25 mM-Tris-HC1 pH 7.5, 150 mM-NaCI, 0'05 % Tween 20 and 2 % dried
skimmed milk. Blots were then incubated with primary antibodies
followed by goat anti-mouse alkaline phosphate conjugate in 25 mMTris-HC1 pH 7"5, 150 mM-NaCI, 0.5 % Tween 20, 4 % dried skimmed
milk and 20 % goat serum for 1 h at room temperature. Blots were
developed in 100 mM-Tris-HC1 pH 9"5, 100 mM-NaC1 and 5 mM-MgCI~
with 0"33 mg/ml nitroblue-tetrazolium and 0.17mg/ml 5-bromo-4chloro-3-indolylphosphate.
bacGST
[ GST
myrGST
Myr~ GST
Metabolic labelling and production of cell lysates. Log phase Jurkat
cells (clone E6-1) were pelleted, washed with PBS (37°C) and
resuspended in methionine-free RPMI 1640 containing 1% FCS at a
concentration of 107 to 2 x 107 cells/ml. After incubation at 37 °C for
1 h the cells were pelleted and resuspended in the same medium containing 100 laCi/ml [35S]methionine (Amersham; 1000 Ci/mmol) and
incubated for a further 4 to 5 h at 37 °C. Labelled cells were pelleted,
washed twice in PBS and lysed in 10 mM-PIPES-NaOH pH 7.2,
120mM-KCI, 30 mM-NaC1, 5 mM-MgC12, 1% Triton X-100, 10%
glycerol containing 10 mM-iodoacetamide, 1 mM-PMSF, 2 ~tg/ml
aprotinin and 1 I~g/ml leupeptin at a concentration of 5 x 107 cells/ml.
Lysates were incubated at 4 °C for 30 min, clarified by centrifugation
(12000g for 5 min) and stored at - 7 0 °C.
Cellfractionation. Labelled cells were washed twice in PBS, incubated
in hypotonic buffer (10 mM-PIPES-NaOH pH 7"2, 0.5 mM-MgCI2) for
I0 rain and lysed by 30 strokes of a tight-fitting Dounce homogenizer.
After adjustment of the salt concentration to 120 mM-KC1/30 mMNaC1, nuclei were pelleted (500 g/5 min) and then membranes were
pelleted from the clarified lysate (150 000 g for 45 min). The membrane
pellet was extracted with lysis buffer as above and the cytosol fraction
was adjusted to 5 mM-MgC12, 1% Triton X-100 and 10% glycerol.
Nef binding reactions. Lysates were precleared by sequential
incubations for 30 min at 4 °C firstly with 1/7 volume GA beads and
then with GA beads loaded with GST (expressed in E. colO. Then
100 ~tl cleared lysate was incubated for 1 h at 4 °C with 10 lal GA beads
loaded with the appropriate protein. All loaded beads contained
approximately 1 mg/ml protein. Beads were washed five times with 0-5
ml lysis buffer, boiled in sample buffer (0.8 % SDS, 8 % glycerol, 2 %
2-mercaptoethanol and 25 mM-Tris-HCl pH 6'8) and separated by
SDS-PAGE. Bound proteins were visualized by fluorography.
Results
Expression and purification of nef affinity reagents
Myristylated and non-myristylated nef gene products
were expressed as N-terminal fusion proteins with GST
[neJGST and nef(m-)GST respectively] using the
baculovirus vector system. Myristylated nef was also
expressed with a C-terminal tail of six histidine residues
to permit purification by metal chelate affinity chromato-
Myr~ NEF [ GST
nef(m-)GST
nef~H
1583
I
Myr~
NEF
I GST
NEF
]RHHHHHH
Fig. 1. Structure of recombinant baculovirus-expressed proteins for use
as affinity reagents. Myr, Myristic acid residue; R, arginine; H,
histidine.
graphy. GST was expressed alone and with an additional
N-terminal myristylation sequence (Met-Gly-Ata-ArgAla-Ser) to provide negative controls. The structure of
the recombinant fusion proteins is shown in Fig. 1, their
expression and purification in Fig. 2. Sf9 cells were
infected with 10p.f.u./cell of the appropriate recombinant viruses and cytoplasmic lysates were prepared
48 h p.i. Fig. 2 (a) demonstrates that all the recombinant
proteins were expressed in the cytoplasm as soluble and
largely undegraded polypeptides that could be efficiently
purified from lysates by one-step affinity chromatography on either GA or Ni-NTA-agarose. Fig. 2 (b), (e)
and (d) confirm that the proteins purified from infected
lysates were recognized on Western blots by the
appropriate monoclonal antibodies (MAbs). NefGST
and nef(m-)GST fusion proteins were recognized by
MAbs to both the N and C termini of nefand also by a
MAb to GST (vpg66). Nef-6H was recognized only by
the anti-nef MAbs and bacGST/myrGST were
recognized only by vpg66. The ability of nefGST fusion
proteins to bind strongly to GA indicates that the GST
portion of the fusion protein is in the native state.
Interactions between the two parts of the fusion protein
which could potentially interfere with the binding of
cellular proteins to nefare therefore unlikely, but cannot
be ruled out. Interestingly, the N terminus-specific
antibody recognizes a dimer of nefGST, nef(m-)GST
and nef~H that is not recognized by the C terminus
MAb or vpg66. This observation suggests that the
epitope bound by the C terminus MAb is involved in
dimerization and that the epitope on GST recognized by
vpg66 might be obscured in the dimer. The significance
of this apparent dimerization of nef is not clear.
To confirm the nefGST, nef-6H and myrGST were
myristylated, Sf9 cells were infected with the appropriate
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 04 Aug 2017 00:59:29
M. Harris and K. Coates
1584
(a)
1
(b)
2
3
4
5
6
7
8
9
1
10
2
3
4
5
6
7
8
9
10
2
3
4
5
6
7
8
9
10
- - 66K - - - 46K - - - 29K - - - 24K - -
(a)
(c)
1
2
3
4
5
6
7
8
9
1
10
66K
46K
o.
•
'
29K
- - 24K
Fig. 2. Expression and purification of recombinant protein affinity reagents. Infected Sf9 cell lysates (lanes 1, 3, 5, 7 and 9) or purified
proteins (lanes 2, 4, 6, 8 and 10) were separated by 12% S D S - P A G E and either stained with Coomassie brilliant blue R-250 (a),
or transferred to nitrocellulose and probed with an anti-GST M A b (b), M A b s directed against the N terminus (c) or C terminus (at) of
nef. Cells were infected with recombinant viruses expressing the following. Lanes 1 and 2, nefGST; lanes 3 and 4, neflm-)GST; lanes
5 and 6, nef-6H; lanes 7 and 8, m y r G S T ; lanes 9 and 10, bacGST.
recombinant viruses and labelled with either
[35S]methionine or [3H]myristic acid (Fig. 3). Whereas
[35S]methionine was incorporated into all the recombinant proteins (Fig. 3 a), only those containing an
N-terminal
myristylation sequence incorporated
[aH]myristic acid (Fig. 3 b). The presence of myristate at
the N terminus renders that protein refractory to Nterminal amino acid sequencing by Edman degradation,
so this method was used to determine the proportion of
myristylated product. Only 1% of a sample of nefGST
was accessible to cleavage by Edman reagent, indicating
that the protein was essentially completely myristylated
(data not shown). The correlation between the expected
and observed patterns of myristylation indicated that
correct initiation of translation was occurring at the first
methionine codon in nef. Initiation has been reported at
the second methionine (residue 20) in some expression
systems (Ahmad & Venkatesan, 1988; Kaminchik et al.,
1991). The data presented in Fig. 2 and 3 thus confirm
the physical integrity of the proteins used in this study.
Binding of cellular proteins to nef in vitro
Lysates were prepared from either unstimulated or
phorbol 12-myristate 13-acetate (PMA)-stimulated
Iurkat cells as described in Methods. Aliquots of the
lysates were sequentially precleared by incubation with
GA beads alone followed by E. col#expressed GST
bound to GA beads. Cleared lysates were then incubated
with baculovirus-expressed bacGST or nefGST bound to
GA beads. Following extensive washing of the beads
with lysis buffer, bound proteins were eluted with sample
buffer and analysed by SDS-PAGE on 6 to 20 % linear
gradient gels followed by fluorography (Fig. 4). A
number of proteins in the Jurkat extracts bound strongly
either to GST, glutathione or to agarose beads and thus
were retained by bacGST-GA beads (lanes 1, 2). In the
presence of nefGST (lanes 3, 4), an additional subset of
proteins, designated p280, p97, p75, p55/57 and p35
according to their apparent Mrs, were retained on the
beads. A number of tess abundant proteins of lower M~,
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 04 Aug 2017 00:59:29
Nef binds to cellularproteins in vitro
(a)
1
2
3
4
5
6
7
8
9
1
10 11 12
2
3
1585
4
- 69K
- 46K
30K
-
-21K
4
200K (b)
1
2
3
4
5
6
7
8
9
10 11 12
- 69K
- 46K
-
30K
97K -
4
-21K
69K Fig. 3. Baculovirus-expressed neJ~3STis myristylated at the N terminus.
Autoradiograph of lysates (lanes 1, 2, 3, 5, 7, 9 and 11) and purified
fusion proteins (lanes 4, 6, 8, 10 and 12), prepared from infected Sf9
cells labelled with either [35S]methionine (a) or [3H]myristic acid (b) and
separated by 12 % SDS-PAGE. Cells were infected with recombinant
viruses expressing proteins as follows. Lanes 1, Mock-infected; lanes 2,
13-galactosidase; lanes 3 and 4, nefGST; lanes 5 and 6, nef(m-)GST;
lanes 7 and 8, nef-6H; lanes 9 and 10, myrGST; lanes 11 and 12,
bacGST.
"4
46K -
4
30K -
21-5K -
p32, p28 and p26, were also observed. Only one difference
could be discerned between unstimulated and stimulated
extracts, a protein of 45K, that bound to nefGST but
not bacGST was present only in extracts of PMAstimulated Jurkat ceils (Fig. 4, lane 4).
The subcellular location of these nef-binding proteins
was determined by separating a Jurkat cell lysate into
cytosolic and membrane-associated protein fractions.
These fractions were precleared as described and
incubated with either bacGST- (Fig. 5, lanes 1 to 3) or
nefGST~GA beads (Fig 5, lanes 4 to 6). The results show
that nef-binding proteins could be separated into three
classes. The first is proteins that were exclusively
membrane-bound, including p280, p35, p32, p28 and
p26, and that were only observed in the total cytoplasmic
fraction (lane 4) or the membrane-bound fraction (lane
5). The second class is proteins that were present in the
total cytoplasmic fraction (lane 4) and the cytosol
fraction (lane 6, p57/55). A number of nef-binding
proteins, and in particular a doublet of p75 and a
previously unobserved protein of 180K, were present in
the cytosolic fraction but not in the total cytoplasmic
lysate. This may have been due to differences in
extraction procedures or it could be that the absence of
proteins such as p280 and p35 permitted the binding of
other proteins to nef. Thirdly, only one protein, p97, was
Fig. 4. NefGST interacts with cytoplasmic proteins from unstimulated
or PMA-stimulated Jurkat T cells. Autoradiograph of extracts from
unstimulated (lanes l, 3) or PMA-stimulated (50 riM) (lanes 2, 4) Jurkat
cells metabolically labelled with [35S]methionine, precleared as described, incubated with either b a c G S T - G A beads (lanes 1,2) or
nefGST GA beads (lanes 3, 4) and separated by 6 to 20 % SDS-PAGE.
Arrowheads, subset of proteins, p280, p97, p75 p55/57 and p35.
Arrow, 45K protein binding only to nefGST in PMA-stimulated
Jurkat cell extracts.
present in both cytosolic and membrane-associated
fractions (lanes 4 to 6).
The effect of N-terminal myristylation on the binding
of nefto cellular proteins was investigated using the nonmyristylated nef(m-)GST fusion protein. Precleared
Jurkat lysate was incubated with myrGST, bacGST,
nefGST or nef(m-)GST as described. The results are
shown in Fig. 6. A number of proteins, p280, p32 and
p28, that bound to myristylated nef(lane 3) failed to bind
to non-myristylated nef(lane 4). It is unclear whether this
is also the case for p35 and p26 because the higher
background obtained in the presence of myristylated nef
(lane 3) obscures these bands. Proteins p75 and p57/55
(migrating as a single band in Fig. 6) still bound to nonmyristylated nef but with markedly lower affinity. The
binding of p97 was independent of myristylation: p97
bound equally well to nefGST or nef(m-)GST. A
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 04 Aug 2017 00:59:29
1586
M. Harris and K. Coates
1
2
3
4
5
6
1
2
3
4
200K 200K -
97K -
97K .q
69K -
69K 46K 46K -
30K 30K -
21.5K 21.5K -
Fig. 5. Nef-binding proteins are b o t h cytosolic and m e m b r a n e associated. J u r k a t cells labelled with [aSS]methionine were fractionated
as described and incubated with b a c G S T G A beads (lanes 1 to 3) or
n e f G S T - G A beads (lanes 4 to 6). Lanes 1 and 4, total cytoplasmic
extract; lanes 2 and 5, m e m b r a n e fraction; lanes 3 and 6, soluble
fraction.
myristylated GST derivative (myrGST; lane 1), consisting of bacGST with the addition of the N-terminal six
amino acids from HIV-1 gag p55 (Met-Gly-Ala-ArgAla-Ser), failed to bind to any of the proteins that bound
only to myristylated nef. This confirmed that these
proteins were specifically binding to myristylated nef,
and were not simply interacting with the myristate
moiety at the N terminus.
To confirm the specificity of the binding of these
cellular proteins to nef, a competition assay was
performed (Fig. 7). Increasing amounts of soluble
myristylated nef (nef-6H) (lanes 3 to 5) or BSA (lanes 6
to 8) were added to aliquots of precleared Jurkat extract
prior to incubation with nefGST-GA beads. Addition of
exogenous soluble nefwas able to inhibit the binding of
p280, p97, p75, and p35 to nefGST-GA beads [compare
lane 2 (no competitor) with lane 5 (40 gg nef-6H)],
Fig. 6. Role o f myristylation in the binding of cellular proteins to nef
Extract f r o m J u r k a t cells labelled with [3~S]methionine was incubated
with m y r G S T G A beads (lane 1), b a c G S T - G A beads (lane 2),
nefGST G A beads (lane 3) or neflm-)GST-GA beads (lane 4).
presumably by itself binding to these proteins. Lack of
definition of low Mr proteins resulted in ambigious
visualization of competition for the p26, p28 and p32
proteins. Intriguingly, in the presence of 40 gg nef-6H
(lane 5) more p55/57 was bound by nefGST. This result
was reproducible and may be explained by the fact that
in the presence of high concentrations of soluble nef-6H
a small amount was found tightly associated with beads
(data not shown). By comparison of the intensity of the
bands the relative amounts of p55/57 were clearly higher
than those of the other nef-binding proteins and, thus, it
would not be as efficiently competed by soluble nef-6H.
The presence of nef-6H on the beads, whether by nonspecific association or specific dimerization with nefGST
as discussed above, would therefore result in the binding
of more p55/57. Addition of a non-specific protein
(BSA) to the assay had no effect on the binding of these
proteins to nefGST-GA beads. These proteins are thus
binding to nef amino acid sequences alone and are not
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 04 Aug 2017 00:59:29
Nef binds to cellular proteins in vitro
1
2
3
4
5
6
7
8
200K-
97K 69K 46K -
30K -
Fig. 7. Solublemyristylatednefcompeteswith immobilizednefGST for
the bindingof cellularproteins. Solubleprotein was addedto precleared
extract from Jurkat cells labelled with [zSS]methionine prior to
incubation with bacGST-GA beads (lane 1) or nefGST GA beads
0anes 2 to 8). Lanes 1 and 2, no solubleprotein; lanes 3 to 5, 10, 20 and
40 lag nef-6H; lanes 6 to 8, 10, 20 and 40 tag BSA.
Table 1. Classification of nef-binding proteins from
Jurkat T cells
Localization
Cytosol
Myristylationdependent
Partially dependent
on myristylation
Myristylationindependent
Membrane
and cytosol
Membrane
p280, p35, p32,
p28, p26
p55/57, p75
p97
binding to an artefactual combination of nef and GST
sequences.
Discussion
In this study, a baculovirus-expressed nefGST fusion
protein was used as an affinity reagent to identify cellular
proteins that specifically interact with nef. A number of
such proteins, both soluble and membrane-associated,
were identified in cytoplasmic extracts of the Jurkat
human T cell line. These proteins could be divided into
1587
three classes according to their subcellular location and
dependence on N-terminal myristylation of nef for
binding. This classification is summarized in Table 1.
Prokaryotic GST fusion proteins (expressed in the
pGEX system; Smith & Johnson, 1988) have previously
been used to identify proteins binding to the retinoblastoma gene product (Kaelin et at., 1991) and the SH2
domain of the actin-binding protein, tensin (Davis et
at., 1991). However, as discussed above, GSTnef expressed from pGEX proved to be an unsuitable reagent
for the identification of nef-binding proteins. One reason
for this is that N-terminal myristylation ofnefis required
for the interaction with three of the proteins identified in
this study. These proteins (p280, p32 and p28), which
failed to bind to non-myristylated nef, are all associated
with the cytoplasmic membrane, raising the possibility
that one or more of them represents a membrane target
for nef. Recent data for two other myristylated proteins,
the myristylated alanine-rich C kinase substrate
M A R C K S (Li & Aderem, 1992) and src (Resh & Ling
1990), show that association of myristylated proteins
with the membrane is not merely the result of insertion of
the fatty acid moiety into the lipid bilayer. Instead, a
combination of this fatty acid moiety and amino acid
sequences at the N terminus interact with specific
receptor molecules on the inner surface of the membrane.
In the case of src (Resh & Ling, 1990), a myristylated
peptide representing the N-terminal 11 amino acids, but
not the corresponding non-myristylated peptide, could
be cross-linked to a 32K plasma membrane protein. It is
unlikely that this 32K protein is identical to p32
described here because the N-terminal amino acid
sequences of nef and src are distinct, and myristylpeptides other than src did not cross-link to the 32K
protein (Resh & Ling, 1990).
Both nefGST and n e f ( m - ) G S T bound to p97 with
equal affinity (Fig. 6). p97 was detected in both cytosolic
and membrane-bound fractions (Fig. 5). p97 may be
only loosely associated with the membrane and readily
dislodged by the extraction procedure, or it may be
present in both soluble and membrane-bound forms.
Interestingly, p97 is the only protein reproducibly bound
by pGEX-expressed GSTnef (Harris et al., 1992b; data
not shown). These observations suggest that the interaction between nef and p97 is independent of the
membrane-bound class of nef-binding proteins and does
not involve the N terminus of nef.
A third group of proteins including p55/p57 and
possibly p75 bind to non-myristylated nef, but the
presence of a myristyl residue increases the affinity of
binding. These proteins appear to be exclusively cytosolic
although they could be very loosely associated with the
membrane and completely dislodged during extraction.
The fact that these proteins bind to non-myristylated nef
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 04 Aug 2017 00:59:29
1588
M. Harris and K. Coates
in the absence of the membrane-associated proteins
again suggests that, as with p97, their interaction with nef
is independent of this membrane-associated class. It is
not clear whether or not they are associated with p97.
The interactions of nef with both cytosolic and
membrane-bound proteins supports the hypothesis that
the mode of nefaction is in some way to perturb specific
signal transduction pathway(s) in infected cells. By
binding independently to both soluble and membranebound components of a particular signal transduction
pathway nefcould effectively prevent the protein-protein
interactions required for passage of the signal. Alternatively, nefmight alter the signal in some way, perhaps by
diverting it down an alternative pathway so as to elicit an
inappropriate response from the cell. Both of these
actions could contribute to the cytopathogenesis of viral
infection. Further speculation must await the
identification of these nef-binding proteins, a task that is
in progress.
Thanks are due to Dr Bob Possee for the transfer vector pAcCL29,
to Dr George Reid for plasmid pGST, to Celia Cannon for producing
the anti-nef MAbs and to Tom Dunsford for the anti-GST MAb.
Thanks also to Professor Jim Nell and Dr George Reid for critical
reading of the manuscript. This work was supported by the MRC
AIDS Directed Programme (ADP). Jurkat E6-1 cells were obtained
from the ADP Reagent Repository.
References
AHMAD, N. & VENKATESSAN,S. (1988). Nef protein of HIV-1 is a
transcriptional repressor of HIV-1 LTR. Science 241, 1481-1485.
ALLAN, J. S., COLIGAN, J.E., LEE, T.-H., MCLANE, M.F., KANKI,
P. J., GROOPMAN,J. E. & ESSEX,M. (1985). A new HTLV-III/LAVencoded antigen detected by antibodies from AIDS patients. Science
230, 810-813.
ANSARDI, D.C., PORTER, D.C. & MORROW, C.D. (1992).
Myristylation of poliovirus capsid precursor P1 is required for
assembly of subviral particles. Journal of Virology 66, 4556-4563.
BACHELERm, F., ALCAMI, J., HAZAN, U., ISRAEL, N., GOUD, B.,
ARENZANA-SEISDEDOSF. & VIRELIZIER,J.-L. (1990). Constitutive
expression of human immunodeficiency virus (HIV) nef protein in
human astrocytes does not influence basal or induced HIV long
terminal repeat activity. Journal of Virology 64, 3059-3062.
BURCH, H. B., NAGY, E. V., LUKES,Y. G., CAI,W. Y., WARTOFSKY,L.
& BtmMAN, K.D. (1991). Nucleotide and amino acid homology
between the human thyrotropin receptor and the HIV-1 nefprotein:
identification and functional analysis. Biochemical and Biophysical
Research Communications 181, 498-505.
CHENG-MAX'ER,C., IANNELLO,P., SHAW,K., Luclw, P. & LEVY, J. A.
(1989). Differential effects of nefon HIV replication: implications for
viral pathogenesis in the host. Science 246, 1629-1632.
DAVIS, S., Lu, M. L., LO, S. H., LIN, S., BUTLER,J. A., DRUKER,B. J.,
ROBERTS,T. M., AN, Q. & CHEN, L. B. (1991). Presence of an SH2
domain in the actin-binding protein tensin. Science 252, 712-715.
FISHER, A. G., RATNER,L., MITSUYA,H., MARSELLE,L. M., HARPER,
M.E., BRODER, S., GALLO, R.C. & WONG-STAAL, F. (1986).
Infectious mutants of HTLV-III with changes in the 3' region and
markedly reduced cytopathic effects. Science 233, 655-659.
GARCIA, J.V. & MILLER, A.D. (1991). Serine phosphorylationindependent downregulation of cell-surface CD4 by nef Nature,
London 350, 508-511.
Guy, B., KIENY, M. P., RiVtERE,Y., LE PEUCH, C. Do'r'r, K., GIRARD,
M., MONTAGNIER,L. & LECOCQJ.-P. (1987). HIV F/3'orf encodes a
phosphorylated GTP-binding protein resembling an oncogene
product. Nature, London 330 266-269.
HAMMES, S., DIXON, E. P., MALIM, M. H., CULLEN, B. R. & GREENE,
W. C. (1989). Nefprotein of human immunodeficiency virus type 1:
evidence against its role as a transcriptional inhibitor. Proceedingsof
the National Academy of Sciences, U.S.A. 86, 9549-9553.
HARRIS,M., HISLOP, S., PATSILINACOS,P. & NEIL, J. C. (1992a). In vivo
derived HIV-1 nef gene products are heterogeneous and lack
detectable nucleotide binding activity. AIDS Research and Human
Retroviruses 8, 53%543.
HARRIS, M., TALBOT, S., PATSILINACOS,P. & NEIL, J. C. (1992b) The
HIV-I nef gene product. Biochemical Society Transactions 20,
517-520.
KAELIN, W.G., PALLAS, D.C., DECAPRIO, J.A., KAYE, F.J. &
LIVINGSTON,D. M. (1991). Identification of cellular proteins that can
interact specifically with the T/E1A-binding region of the retinoblastoma gene product. Cell 64, 521-532.
KAMINCHIK, J., BASHAN, N., PINCHAS1, D., AMIT, B., SARVER, N.,
JOEINSTON,M. I., FISCHER,M., YAVIN,Z., GOI~CKI, M. & PANET,A.
(1990). Expression and biochemical characterization of human
immunodeficiency virus type 1 nefgene product. Journal of Virology
64, 3447-3454.
KAMINCHIK, J., BASHAN,N., ITACH, A., SARVER,N., GORECKI,M. &
PANET, A. (1991). Genetic characterization of human immunodeficiency virus type l nef gene products translated in vitro and
expressed in mammalian cells. Journal of Virology 65, 583-588.
KIM, S., IKEUCHI,K., BYRN,R., GROOPMAN,J. & BALTIMORE,D. (1989).
Lack of a negative influence on viral growth by the nef gene of
human immunodeficiency virus type 1. Proceedings of the National
Academy of Sciences, U.S.A. 86, 9544-9548.
KITTS, P. A., AYFmS,M. D. & POSSEE, R.D. (1990). Linearization of
baculovirus DNA enhances recovery of recombinant virus expression
vectors. Nucleic Acids Research 18, 5667-5672.
LI, J. & ADEREM,A. (1992). MacMARCKS, a novel member of the
MARCKS family of protein kinase C substrates. Cell 70, 791-801.
LIVINGSTONE, C. & JONES, I. (1989). Baculovirus expression vectors
with single strand capability. Nucleic Acids Research 17, 2366.
LUClW, P.A., CHENG-MAYER,C. & LEvY, J. A. (1987). Mutational
analysis of the human immunodeficiency virus: the orf-B region
down-regulates virus replication. Proceedings of the National Academy of Sciences, U.S.A. 84, 1434-1438.
LURIA, S., CHAMBERS,I. & BERG, P. (1991). Expression of the type 1
human immunodeficiency virus nef protein in T cells prevents
antigen receptor-mediated induction of interleukin 2 mRNA.
Proceedings of the National Academy of Sciences, U.S.A. 88,
5326-5330.
MAITRA,R. K., AHMAD,N., HOLLAND,S. M. & VENKAaXSAN,S. (1991).
Human immunodeficiency virus type 1 (HIV-1) provirus expression
and LTR transcription are repressed in NEF-expressing cell lines.
Virology 182, 522-533.
MATSUURA,Y., MAEKAWA,M., HATTORI,S., IKEGAMI,N., HAYASHI,
A., YAMAZAKI,S., MOR1TA,C. &;TAKEBE,Y. (1991). Purification and
characterisation of human immunodeficiency virus type 1 nef gene
product expressed by a recombinant baculovirus. Virology 184,
580-586.
NEBREDA,A. R., BRYAN,T., SEGADE,F., WINGFIELD,P., VENKATESAN,
S. & SANTOS,E. (1991). Biochemical and biological comparison of
HIV-1 NEF and ras gene products. Virology 183, 151-159.
NIEDERMA~, T. M. J., TrimLAN, B.J. & RATNER, L. (1989). Human
immunodeficiency virus type 1 negative factor is a transcriptional
silencer. Proceedings of the National Academy of Sciences, U.S.A. 86,
1128-1132.
RESH, M. D. & LING, H.-P. (1990). Identification of a 32K plasma
membrane protein that binds to the myristylated amino-terminal
sequence of p60 ..... Nature, London 346, 84-86.
SAMUEL,K. P., SETH,A., KONOPKA,A., LUTENBERGER,J. A. & PAPAS,
T. S. (1987). The 3'-orf protein of human immunodeficiency virus
shows structural homology with the phosphorylation domain of
human interleukin-2 receptor and the ATP-binding site of the
protein kinase family. FEBS Letters 218, 81-86.
SMITH, D.B. & JOHNSON, K.S. (1988). Single-step purification of
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 04 Aug 2017 00:59:29
N e f binds to cellular proteins in v i t r o
polypeptides expressed in Escherichia coli as fusions with
glutathione-S-transferase. Gene 67, 31-40.
TERWILLIGER,E., SODROSKI,J. G., ROSEN, C. A. & HASELTINE,W. A.
(1986). Effects of mutations within the 3'orf open reading frame
region of human T-lymphotropic virus type III (HTLV-III/LAV) on
replication and cytopathogenicity. Journal of Virology 60, 754-760.
TERVClLLIGER,E. F., LANGHOFF,E., GABUZDA,D., ZAZAPOULOS,E. &
HASELTINE,W. A. (1991). Allelic variation in the effects of the nef
gene on replication of human immunodeficiency virus type 1.
Proceedings of the National Academy of Sciences, U.S.A. 88,
10971-10975.
WERNER,T., FERRONI,S., SAARMARK,T., BRACK-WERNER,R., BANATI,
R. B., MAGER, R., STErNAA,L., KREUTZBERG,G. W. & ERFLE, V.
1589
(1991). HIV- 1 nefprotein exhibits structural and functional similarity
to scorpion peptides interacting with K ÷ channels. AIDS 5,
1301-1308.
Yu, G. & FELSTED, R. L. (1992). Effect of myristoylation on p27nef
subcellular distribution and suppression of HIV-LTR transcription.
Virology 187, 46-55.
ZAZAPOULOS,E. & HASELTINE,W. A. (1992). Mutational analysis of the
human immunodeficiency virus type 1 Eli nef function. Proceedings
of the National Academy of Sciences, U.S.A. 89, 6634~6638.
(Received 2 February 1993; Accepted 19 March 1993)
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 04 Aug 2017 00:59:29