|
|
|
|||||||
|
|||||||||
(Received for publication, May 4, 1994; and in revised form, October 13, 1994) From the
The Src homology 2 (SH2) and Src homology 3 (SH3) domain are
approximately 50% conserved in various Src family kinase members.
Several lines of evidence suggest that in Src these domains are
sequence motifs that direct substrate recognition, regulate kinase
activity, or control subcellular localization. We sought to investigate
the function of the homology domains in human Lyn, and to determine
whether the differences between various SH3 domains affect function. To
do this, we generated variant forms of Lyn lacking SH2 and SH3 domains,
and created chimeras in which the SH3 domains in human c-Src and Lyn
were replaced with SH3 domains from other family members. In contrast
to similar deletions in Src, forms of Lyn lacking SH2 or SH3 had
decreased kinase activity. The SH3 chimeras all had individual
characteristics. Insertion of the Blk SH3 domain into Lyn restored
kinase activity, while insertion of the Fyn or Src SH3 into Lyn
enhanced the kinase activity 2-3-fold. Insertion of the Lyn SH3
into Src also doubled kinase activity. Expression of the Lyn-Src SH3
chimera in mammalian cells induced cell transformation. This study 1)
demonstrates that the regulation of Lyn is different than Src, and 2)
provides new evidence that despite their homology, there are important
functional differences between the SH3 domains of the various Src
family members. The Rous sarcoma virus contains the v-src gene which
encodes a 60-kilodalton plasma membrane-associated protein capable of
inducing neoplastic transformation(1, 2) . The Src
protein has protein tyrosine kinase activity which is essential for its
oncogenic capabilities(3) . Genetic screening has revealed nine
Src family members which are all normal cellular components: c-Src,
c-Yes, Lyn, Fyn, Lck, c-Fgr, Hck, Blk, Yrk(4, 5) .
Gene knockout experiments have demonstrated specific roles for some of
these proteins, and immunoprecipitation and receptor cross-linking
studies have implicated various potential signaling pathways. Each
of the Src family proteins can be divided into multiple domains with
distinct functional properties. The largest domain is located in the
carboxyl-terminal half of the protein and contains the kinase catalytic
activity and a short stretch of approximately 19 amino acids that
contains a kinase activity regulating tyrosine
residue(6, 7) . Deletion mapping has shown that amino
acids 1-14 are necessary for N-myristylation of
p60 Two domains situated between this unique
region and the catalytic kinase region are found in a wide variety of
intracytoplasmic signaling molecules(4) . These domains are
called Src homology region 2 (SH2) ( Specific
deletion or point mutations in SH2 or SH3 regions can activate or
inactivate kinase activity and alter the transforming potential of
tyrosine kinases in avian and mammalian
cells(19, 20, 21, 22, 23, 24, 25, 26) .
Interestingly, several mutations in the SH2 and SH3 domains create a
phenotype that varies according to the host cell, suggesting that
environment as well as protein sequence is
important(25, 26) . In addition, other work has shown
that SH3 domains may participate in substrate
recognition(14, 18, 24, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37) .
Genetic evidence demonstrates that mutations in the SH3 domain of sem5
severely impairs nematode vulva development, showing direct evidence
that SH3 domains are critical in cellular signaling (38) .
Taken together, these results imply that SH3 regions may regulate
kinase activity and direct interaction with specific cellular proteins. Lyn is a member of the Src family which is widely expressed in
hematopoietic cells and has been implicated in intracellular signaling
from receptors of the immunoglobulin supergene family(4) . To
determine the functional significance of the Src homology domains in
Lyn, we constructed a set of eukaryotic vectors containing domain
deletion variants of Lyn. Our results demonstrate that Lyn behaves
differently than Src with respect to specific kinase activity, sites of in vivo phosphorylation, and association with intracellular
tyrosine-phosphorylated proteins. We further addressed whether we could
restore wild-type phosphorylating activity of the Lyn SH3 deletion
polypeptide by substituting a SH3 domain from other Src kinases.
Expression of several SH3 chimerics all demonstrated individual in
vivo and in vitro characteristics. These results imply
that the regulation for Lyn is different than Src and suggest a
modulating role for the SH3 domain that is individually specific for
each Src family member.
The Src and Lyn chimeric
mutants were constructed using a two-step overlap extension technique,
first fusing the amino portion of Src or Lyn with a substituted SH3
domain, and then fusing this product with the carboxyl terminus of the
kinase. The template for human c-Src was a gift of Don Fujita, and the
template for human Fyn was a gift from Keith Robbins. Murine Blk SH3
cDNA was generated by reverse transcription-PCR from HL-60 RNA.
Lyn-SH3-Src was generated using PCR primers 5`
CAAGGAGACACCTTTGTGGCCCTCTAT 3`, 5` CACAAAGGTGTCTCCTTGTTCCTCTGG 3`, 5`
AGCAACTACGTGGCCAAACTCAACACC 3`, 5` TTTGGCCACGTAGTTGCTGGGGATGTA 3`. The
primers for Lyn-SH3-Fyn were 5` CAAGGAGACCTCTTTGTGGCCCTTTAT 3`, 5`
CACAAAGAGGTCTCCTTGTTCCTCTGG 3`, 5` AGCAATTATGTGGCCAAACTCAACACC 3`, 5`
TTTGGCCACATAATTGCTGGGAATGTA 3`. Lyn-SH3-Blk was generated using PCR
primers 5` CAAGGAGACGTGGTGGCTCTGTTTGAC 3`, 5`
AGCCACCACGTCTCCTTGTTCCTCTGG 3`, 5` AGCAACTTTGTGGCCAAACTCAACACC 3`, 5`
TTTGGCCACAAAGTTGCTGGGCACATA 3`. The primers for Src-SH3-Lyn were 5`
GGAGTGACCATTGTGGTAGCCTTGTAC 3`, 5` TACCACAATGGTCACTCCACCGGCCAG 3`, 5`
AGCAACTATGTGGCGCCCTCCGACTCC 3`, GGGCGCCACATAGTTGCTGGGGATGAA 3`. The
final product was then digested with EcoRI and BamHI
and cloned back into PCMV5. The full sequences of all cloned
PCR-generated products were correct as determined by the dideoxy method
using the Sequenase version 2.0 kit (U. S. Biochemical Corp.).
Metabolic labeling with [
Samples for in vitro kinase assays were
washed once with low salt buffer (100 mM NaCl, 10 mM Tris (pH 7.0), 2.5 mM MnCl Immunoblotting was performed to detect phosphotyrosine-containing
proteins(43) . Briefly, proteins fractionated on 7.5% PAGE were
transferred to nitrocellulose, incubated for at least 2 h with blocking
buffer (5% crystallized bovine serum albumin, 170 mM NaCl,
0.2% Nonidet P-40, 50 mM Tris (pH 7.5)). The blots were probed
with a mixture of murine anti-phosphotyrosine antibodies (4G10 and
PY20). A horseradish peroxidase-coupled secondary antibody was used and
detected by enhanced chemiluminescence (Amersham Corp.). Alternatively,
Lyn or Src protein was detected with AR1 or LA074 primary antibody.
This was followed with
Figure 1:
Amino acid sequence of
Src family SH3 domains and schematic diagrams of human p56 Lyn and
human p60 Src constructions. A, alignment of Src family SH3
domains. The 14 amino acids thought to interact with ligands are marked
with an asterisk(16) . The numbering on top corresponds to the residue from chicken c-Src. B,
schematic diagram of p56 Lyn, Lyn deletion mutants, Lyn SH3 chimerics,
c-Src and Src SH3 chimerics
Figure 2:
Determination of specific activity of
human Lyn and Lyn deletion mutant polypeptides. Lyn was
immunoprecipitated from [
We next
questioned whether we could restore wild-type phosphorylating activity
of the Lyn SH3 deletion polypeptide by substituting a SH3 domain from
another Src kinase. When the SH3 domain of Lyn was replaced with the
SH3 domain of Blk, Fyn, or c-Src, the mobilities of the metabolically
labeled proteins were faster (Lyn-SH3-Blk) or slower (Lyn-SH3-Fyn) than
predicted (Fig. 3). DNA sequencing confirmed that the mutations
were accurate, hence it is probable that the difference in apparent
size reflects folding or post-translational processing of the protein.
The in vitro kinase assays of these chimeric proteins revealed
that replacing the SH3 domain of Lyn with Blk essentially restores
wild-type activity (Fig. 3A). When the SH3 domain of
Lyn was replaced with the SH3 of Fyn or Src, the specific activity of
Lyn went up severalfold (Table 1). Paired comparison demonstrated
that the kinase activity of Lyn-SH3-Blk was significantly different
from Lyn-SH3-Fyn (p < 0.001), but experimental variability
prevented determining whether Lyn-SH3-Blk was different from
Lyn-SH3-Src.
Figure 3:
Determination of specific activity of
human Lyn and Lyn SH3 chimeric mutant polypeptides. Lyn was
immunoprecipitated from [
We questioned whether the reciprocal SH3 chimeric
mutation in Src would also alter its in vitro kinase activity.
When the SH3 of c-Src was replaced with the SH3 of Lyn, the apparent
mobility of the mutant protein did not change (Fig. 4). Like the
Lyn chimeras, replacement of the c-Src SH3 domain with the SH3 domain
of Lyn elevated the specific activity of Src 2-fold (Fig. 4A, Table 1).
Figure 4:
Determination of specific activity of
human Src and Src-SH3-Lyn chimeric mutant polypeptide. Src was
immunoprecipitated from [
In summary, we found that
deletion of the SH2, SH3, or unique domains caused a loss of Lyn
activity. This is in contrast to the reported gain in Src kinase
activity following SH2 or SH3 deletion. Replacing the deleted Lyn SH3
domain with the corresponding portion of Blk restored the kinase
activity to base line, but substituting the Fyn or Src SH3 domain
caused an increase in kinase activity.
As seen in Fig. 5A, cyanogen bromide digestion of c-Src labeled in vitro yields a product that is almost exclusively labeled
on Tyr-416 corresponding to the 10-kDa fragment. In contrast, cyanogen
bromide digestion of c-Src labeled in vivo produced a
phosphorylated 6- and 10-kDa fragment. This implies that overexpressed
c-Src is phosphorylated on both Tyr-416 and Tyr-527. This might be
attributable to the high levels of transiently overexpressed c-Src
which could overwhelm the regulating activity of Csk, the kinase that
phosphorylates Tyr-527 on Src. Cyanogen bromide digestion of in
vivo labeled Src-SH3-Lyn, which is the Src chimeric which contains
the Lyn SH3 domain, demonstrated a phosphorylated 10-kDa fragment and
no 6-kDa fragment. This indicates that, in contrast to c-Src, the
Src-SH3-Lyn variant is phosphorylated on Tyr-416, but not on Tyr-527.
The peptide map of Src-SH3-Lyn labeled in vivo also
demonstrated the predicted change in the size of the Ser-17 containing
a fragment from 32 to 12 kDa. This is attributable to the presence of a
methionine in the Lyn SH3 domain which produces an additional cyanogen
bromide cleavage site in Src-SH3-Lyn that is not found in c-Src.
Figure 5:
Cyanogen bromide cleavage of kinase
variant proteins. A, immunoprecipitated Src or Src-SH3-Lyn
polypeptide from
We
next examined the sites of phosphorylation on in vitro and in vivo phosphorylated Lyn. By analogy with Src, resting Lyn
should be phosphorylated on Tyr-508 and activated Lyn should be
phosphorylated on Tyr-397. These sites of phosphorylation are located
on cyanogen bromide digest fragments of size 4 and 8 kDa, respectively.
Lyn does not have a serine in its unique region corresponding to Ser-17
of Src, but does have several other potential phosphorylation sites in
its unique domain, all of which would be located on a cyanogen bromide
digest fragment of 10 kDa. Fig. 5B demonstrates that
cyanogen bromide digestion of in vivo and in vitro labeled Lyn produces three fragments of 4, 8, and 10 kDa. In
contrast to c-Src, there is no dramatic difference between the peptide
map of in vitro and in vivo phosphorylated Lyn.
Peptide mapping of the Lyn variants also did not reveal any change in
their sites in vivo phosphorylation (data not shown). Thus,
we found that substitution of the Lyn SH3 into Src results in a change
in the site of in vivo phosphorylation consistent with in
vivo activation. We also found that Lyn, in contrast to Src, does
not change its pattern of phosphorylation upon activation.
The morphology of multiple
cellular isolates expressing each of the mutants were studied. Typical
examples are shown in Fig. 6. Overexpression of c-Src or Lyn in
NIH 3T3 cells resulted in a morphology similar to untransfected NIH 3T3
cells or cells transfected with the neomycin resistant plasmid alone.
In contrast, cells expressing the Lyn SH3 deletion polypeptide or
Lyn-SH3-Src polypeptide displayed a weakly transformed morphology. The
cells were refractile, rounded, and displayed thin neuronal-like
processes. None of the other Lyn or Src variants displayed a
transformed morphology (Table 2).
Figure 6:
Cell morphology. NIH 3T3 cells expressing
roughly equivalent levels of various human Lyn or Src polypeptides were
photographed under light microscopy. A, neomycin plasmid with
empty expression vector; B, wild-type Lyn; C, Lyn SH3
deletion; D, Lyn-SH3-Blk; E, Lyn-SH3-Fyn; F,
Lyn-SH3-Src; G, c-Src; and H,
Src-SH3-Lyn.
To test each mutant for
anchorage independent growth, we assayed for the induction of colony
formation in soft agar. Several isolates of each mutant were assayed in
at least two separate experiments. Three out of three clones expressing
the Lyn SH3 deletion polypeptide and three out of three clones
expressing the Lyn-SH3-Src polypeptide formed approximately 200 small
colonies in soft agar (0.4% of cells plated). None of the other cells
expressing wild-type or mutant polypeptides formed colonies in soft
agar (data not shown). Thus, cells overexpressing Lyn lacking an SH3
domain, or Lyn containing the SH3 domain of Src, have a mildly
transformed phenotype. This is seen by alterations in cell morphology
and demonstration of anchorage-independent cell growth. In contrast,
substitution of the Blk or Fyn SH3 into Lyn or the Lyn SH3 into Src
failed to induce cell transformation.
Figure 7:
Anti-phosphotyrosine immunoblots. The
identification of phosphotyrosine-containing proteins, in NIH 3T3 cells
expressing mutant Src or Lyn polypeptides, was assessed by
anti-phosphotyrosine immunoblots as described under ``Materials
and Methods.''
As is shown in Fig. 7, overexpression of
wild-type Lyn in 3T3 cells caused only a mild increase in
phosphotyrosine-containing proteins compared to mock-transfected 3T3
cells. Like a similar deletions in Src or Lck, we found that deletion
of the SH2 in Lyn increased the tyrosine phosphorylation of cellular
proteins. Interestingly, although the level of intracellular
phosphorylation was increased with the Lyn SH2 deletion variant, this
was one of the mutations that lowered the specific activity of Lyn. In
addition to an overall increase in intracellular tyrosine
phosphorylation, several proteins appear to be particularly targeted
for phosphorylation with apparent molecular masses of 36 and 49 kDa. In
contrast to the SH2 domain, deletion of the unique or SH3 region of Lyn
did not alter its ability to phosphorylate other intracellular
proteins. We next tested the Lyn SH3 chimeric mutant polypeptides
for their ability to phosphorylate intracellular proteins. Although
several of the chimeric mutants phosphorylated enolase, in
vitro, more than wild-type Lyn, none of the chimeras were
associated with an elevated level of intracellular phosphorylation. In
addition, we could not detect any novel substrates associated with the
Lyn chimerics. As described by others, we found that cells
overexpressing c-Src had an increased amount of
phosphotyrosine-containing proteins compared to wild-type
3T3(47, 48, 49) . When the SH3 region of Src
was replaced with the SH3 region of Lyn, the phosphorylation of
cellular proteins went up dramatically. This effect does not appear to
be related to the level of kinase expression, since immunoblotting
indicated that there were roughly equivalent amounts of Src protein in
the cell lysates used in Fig. 7(data not shown). This increase
in phosphotyrosine-containing proteins was also seen in NIH 3T3 cells
stably expressing our polypeptides (data not shown). In addition to a
dramatic increase in the overall amount of various
phosphotyrosine-containing proteins, several proteins are
preferentially phosphorylated by Src-SH3-Lyn. The most prominent of
these phosphoproteins migrated at 36, 68, 80, and 110 kDa. Comparison
of the phosphorylation pattern of Src-SH3-Lyn-transfected 3T3 cells
with v-Src-transformed 3T3 cells demonstrates a prominent
phosphorylated 85-kDa band associated with v-Src and not with
Src-SH3-Lyn (data not shown).
Overexpression of the Lyn SH3
deletion polypeptide, gives subtle signs of cell transformation. The
cells acquired a neuronal morphology and produced evidence of anchorage
independence by some growth in soft agar. It is remarkable that the
cell transformation is associated with a Lyn variant polypeptide that
has a lower intrinsic specific kinase activity than the wild-type Lyn.
However, this is not without precedent since it has been shown that
some mutations in Src which lower kinase activity can still transform
murine 3T3 cells(26) . Of all the Lyn mutants examined, only
the SH2 deletion caused an increase of phosphotyrosine-containing
proteins in COS cells. This occurred despite its lower specific kinase
activity, suggesting that the SH2 region may play a role in the
repressing the ability of Lyn to phosphorylate substrates through a
mechanism independent of regulating the intrinsic kinase activity of
Lyn. This mechanism of regulation again appears different from the SH2
regulation of Src.
Second, the SH3
region may regulate kinase activity through an intermolecular
mechanism. This might occur by impairing the direct interaction with a
negative regulator. Imamoto and Soriano (53) have demonstrated
that the phosphorylation of the regulatory site, tyrosine 527, in c-Src
is mediated by Csk and at least one other cellular protein tyrosine
kinase. It is possible that in the Src family, the SH3 domain directs
interaction with other proteins that activate or inactivate kinase
activity. It has been proposed that the Abl kinase SH3 domain may
interact with an intracellular regulator(23) . Perhaps the Src
family SH3 domains direct interactions with specific regulators, and
substitution of a different SH3 domain alters this resultant
regulation. This is consistent with evidence of decreased intracellular
phosphorylating ability of the Lyn variants but elevated kinase
activity after immunoprecipitation under harsh lysis conditions. As
noted in Fig. 1A, the SH3 of Fyn has complete identity
with Src in 42 out of 52 amino acids. In contrast, Lyn and Blk share
only 29 and 28 amino acids, respectively, with Src. Of the 12 SH3
conserved residues that are proposed to be critical for interaction
with ligands, Lyn and Src share only 3 residues(26) . In this
context, it is not surprising that the Src family SH3 domains cannot
substitute for each other. Grandori (54) postulated that Src
kinase activity is regulated in part by Arg-95 of chicken Src. His
model proposed that kinase activity is inhibited by binding to
substrate-like sequences. Potts et al.(55) demonstrated that a single Arg-95 to Trp mutation is
sufficient to activated c-Src. It is remarkable that both Src and Fyn
contain an Arg at this site, but Lyn and Blk have an isoleucine and
valine, respectively. It is tempting to speculate that the substitution
of Src or Fyn SH3 domain into Lyn leads to a consistent increase in
kinase activity perhaps in part due to the substitution of the
isoleucine in Lyn for an arginine. Consistent with this,
Lyn-SH3-Blk's essentially identical kinase activity with
wild-type Lyn is possibly attributable to the relatively conservative
substitution of valine for isoleucine. Other studies have
demonstrated that the SH3 domain of Src can target
substrates(30, 31) . We are not able to detect binding
of the p85 subunit of PI3K to c-Src or Src-SH3-Lyn in our system. ( It is interesting that
Src-SH3-Lyn and Lyn SH2 deletion both phosphorylate a band of molecular
mass of 36 kDa. Potentially the Lyn SH3 domain is directing this
interaction, thus allowing Src-SH3-Lyn to phosphorylate it. Since the
Lyn SH2 deletion polypeptide is a more activated form of Lyn, as
demonstrated by the total cell lysate anti-phosphotyrosine blot, it may
phosphorylate p36 more readily than wild-type Lyn even though the
association could be present in both. Alternatively, this
phosphoprotein is calpactin I which becomes phosphorylated by activated
variants of Src. Further identification of this protein is ongoing. Since the protein kinase activity of Src family members is critical
for their transforming potential, determination of their target
proteins is important for understanding their mechanism of oncogenesis.
This study provides evidence for individual characteristics of
different Src family members as demonstrated by the contrasting
consequences of SH2 or SH3 deletions in Lyn or Src. This work also
demonstrates the inability of one SH3 domain to completely substitute
for another and implies that SH3-substituted forms of human c-Src and
Lyn may be useful for studying the regulatory role of the SH3 module.
Volume 270,
Number 1,
Issue of January 6, 1995 pp. 333-339
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
a process that appears necessary for membrane
association as well as anchoring to several receptors of the
immunoglobulin supergene family (8, 9, 10, 11) . Adjacent to this
site is a stretch of approximately 60 amino acids, termed the unique
domain, where the Src family members diverge most
dramatically(4) .
)and Src homology region
3 (SH3). Both have been implicated in protein-protein
interactions(12) . There is a substantial body of evidence that
SH2 domains, which are composed of a conserved stretch of approximately
100 amino acids, bind to phosphorylated tyrosine
residues(12, 13) . SH2 domains may bind to the
tyrosine in the regulatory tail of Src and modulate the kinase activity
of this protein. SH3 domains, composed of a module of approximately 60
amino acids, are less understood. Several SH3 domains have been found
to bind proline-rich sequences(14, 15) . Recently,
attention has been drawn to the fact that SH3-containing proteins are
usually membrane-associated, suggesting a role in subcellular
localization (16, 17, 18) .
DNA Constructs
The wild-type p56 construct was cloned into PCMV5 as described
previously(11) . The Lyn mutants were constructed by using the
PCR splice overlap extension technique to fuse the regions flanking the
deletions(39) . For all mutants, the external sense primer was
5` GCGGTAGGCGTGTACGGTGGGAGGTCT 3` and the external antisense primer was
5` CTGGAGTGGCAACTTCCAAGGCCAGGA 3`. The sense and antisense mutagenesis
primers for the Lyn unique (20-66) were 5`
GTAGATTTGATTGTGGTAGCCTTGTAC 3` and 5` TACCACAATCAAATCTACTCCATCGTC 3`.
The sense and antisense primers for the Lyn SH3 (67-117) were 5`
CAAGGAGACGTGGCCAAACTCAACACC 3` and 5` TTTGGCCACGTCTCCTTGTTCCTCTGG 3`,
and for the Lyn SH2 (118-229) were 5`
AGCAACTATAAGCCACAGAAGCCATGGG 3` and 5` CTGTGGCTTATAGTTGCTGGGGATGAAG 3`.
The final product was then digested with EcoRI and BamHI and cloned back into PCMV5.
Cell Culture, Transfections, and Metabolic
Labeling
Transient transfections of COS-1 and NIH 3T3 cells were
performed as described previously(40, 41) . All
transients were examined 2-3 days after transfection. NIH 3T3
stables were selected by adding G418 24 h after transfection and
selecting for single isolates approximately 2 weeks later. These
isolates were expanded and subsequently assayed for expression of the
individual proteins by immunoblotting or in vitro kinase
assays.S]methionine
of COS-1 cells was performed by changing the medium to 10% DMEM, 90%
methionine-free DMEM with 10% FCS and 50 µCi/ml
Tran
S-label (ICN) overnight. In vivo labeling of
cells with free
P was performed by starving COS-1 cells
for 20 min with phosphate-free DMEM with 10% fetal calf serum, and then
adding 1 mCi/ml orthophosphate for 3.5 h.
Immunoprecipitation, in Vitro Kinase Assays, and
Immunoblotting
Transiently or stably transfected NIH 3T3 or
COS-1 cells were washed twice with phosphate-buffered saline, scraped
off the plates, pelleted, and lysed in radioimmune precipitation buffer
(58 mM NaCl, 5 mM EDTA, 10 mM Tris (pH 7.2),
0.1% SDS, 1% sodium deoxycholate, and 1% Triton X-100) with 1%
aprotinin (Miles Pharm), 50 µg/ml leupeptin (Sigma), 1 mM phenylmethylsulfonyl fluoride, and 1 mM vanadate
(Fisher). After clarification at 18,000 
g, the lysates
were normalized arbitrarily for transfection efficiency, and saturating
quantities of AR1, rabbit polyclonal anti-Lyn peptide (amino acids
20-229), or LAO74, murine monoclonal anti-Src were added and
incubated at 4 °C overnight. Pansorbin (Calbiochem, catalog no.
507858) with 1% bovine serum albumin was added for 15 min, pelleted,
and washed twice with lysis buffer. Samples were then used for in
vitro kinase assays as described below. For immunoblotting or S-labeled protein gels, the samples were then boiled in
sample loading buffer, fractionated on a 7.5% SDS-PAGE, and either
transferred to nitrocellulose and immunoblotted as described below or
soaked in sodium salicylate for 1 h, dried, and exposed to XAR film at
-70 °C.
, 2.5 mM MgCl) and resuspended in 30 µl of low salt buffer
with 1 µM ATP and 25 µCi of
[-
P]ATP and 10 µg of acid-treated
exogenous enolase (Boehringer Manheim) and incubated at room
temperature for 15 min as described previously(42) . The
reaction was stopped by dilution with 1 ml of lysis buffer and then
pelleted. The pellet was resuspended in sample buffer and boiled prior
to fractionation on 7.5% SDS-PAGE, dried, and exposed at -70
°C to XAR film. Phosphorylation of enolase was quantitated on a
PhosphoImager or Densitometer (Molecular Dynamics). Significance was
determined by one-tailed Student's t test analysis.
I-conjugated secondary donkey
antibody, and the blot was exposed to XAR film.
CNBr Digestion of Proteins
Transiently transfected
COS-1 cells were labeled in vivo with P
or in vitro labeled in a kinase assay as described
above. Cyanogen bromide mapping was performed basically as described
previously except that the final fractionation was performed on a
Tricine gel(44, 45) .
Construction of Lyn/Src Deletion and Chimeric Mutants
We sought to investigate the function of the homology domains
in human Lyn and the qualitative differences in SH3 domains of human
Src family kinases. Three deletion mutants and three SH3 chimeric
mutations were introduced into Lyn. One SH3 chimeric mutation was
introduced into c-Src. The boundaries of the studied SH3 domains are
shown in Fig. 1A. The constructs are summarized in Fig. 1B.
In Vitro Kinase Assays
To determine the effects of
deletion of the unique, SH3, or SH2 domains on the kinase activity of
Lyn expressed in mammalian cells, we determined the specific activity
of each of the mutant proteins by an in vitro immunoprecipitation kinase assay with enolase as exogenous
substrate using electrophoresis to separate the phosphorylated
proteins. Shown in Fig. 2, the wild-type Lyn runs at 56 kDa, the
unique deletion mutant at 51 kDa, SH3 deletion at 50 kDa, and the SH2
deletion at 45 kDa. This is consistent with their predicted molecular
masses. A typical experiment is demonstrated in Fig. 2A. As summarized in Table 1, deletion of
the unique, SH3, or SH2 domains in human Lyn expressed in mammalian
cells decreased the activity for autophosphorylation and
phosphorylation of an exogenous substrate. Fig. 2B demonstrates that comparable amounts of proteins were
immunoprecipitated. This decrease in Lyn kinase activity is in contrast
to the findings of SH2 or SH3 deletion mutations in c-Src which have a
30-fold enhanced phosphorylating activity(24) .
S]methionine-labeled
COS cells which were transiently transfected with Lyn or Lyn mutant
proteins. A, one half of immunoprecipitate was used in an in vitro kinase assay with enolase as the exogenous substrate; B, the other half was used to quantitate the level of
[
S]methionine labeled Lyn protein by
fractionating on a 7.5% SDS-PAGE, treating with sodium salicylate, and
exposing to film.
S]methionine labeled
cells. A, one half of immunoprecipitate was used in an in
vitro kinase assay with enolase as the exogenous substrate; B, the other half was used to quantitate the level of
[
S]methionine-labeled Lyn
protein.
S]methionine-labeled
cells. A, One half of immunoprecipitate was used in an in
vitro kinase assay with enolase as the exogenous substrate; B, the other half was used to quantitate the level of
[
S]methionine-labeled Lyn
protein.
In Vivo Phosphorylation Mapping
Other
investigators have demonstrated that the sites of tyrosine
phosphorylation on Src and Lck reflect the state of activation. We
examined the tyrosine phosphorylation sites of the variant kinases by
performing peptide maps after in vivo phosphorylation.
Previous work has shown that cyanogen bromide digestion of Src produces
three phosphate-labeled fragments; a small 6-kDa fragment containing
Tyr-527, a 10-kDa fragment containing Tyr-416, and a 32-kDa fragment
containing Ser-17(46) . Wild-type c-Src is predominately
phosphorylated on Tyr-527 in vivo by Csk, while activated
variants of Src are phosphorylated on Tyr-416.
P
-labeled cells were purified
and digested with cyanogen bromide. The lane labeled Auto-P contains c-Src, which was labeled with P
in an in vitro kinase reaction and then cleaved with
cyanogen bromide as a marker for the Tyr-416-containing peptide. The
positions of peptides containing the three sites of Src phosphorylation
Ser-17, Try-416, and Tyr-527 are shown. B, immunoprecipitated
Lyn polypeptide from P
-labeled cells was
purified and digested with cyanogen bromide. The lane labeled Lyn
in Vitro contains Lyn which was labeled with P
in an in vitro kinase reaction and then cleaved with
cyanogen bromide.
Ability of Src or Lyn Variants to Transform NIH 3T3
Cells
In view of this increased kinase activity, we questioned
whether the SH3 chimeric mutations would transform 3T3 fibroblasts.
Mouse NIH 3T3 cells were transfected with plasmids encoding for the
appropriate human kinase polypeptide along with a neomycin resistant
plasmid, and selected for growth in G418-containing medium. Several
independent isolates of Lyn, Src, and each variant were studied.
Expression of c-Src, wild-type Lyn, the Lyn SH3 deletion polypeptide,
and the SH3 chimeric polypeptides were roughly comparable as determined
by immunoblotting (data not shown).
Phosphorylation of Cellular Substrates
Cells
expressing SH2 and SH3 deletion mutants of Src and Lck contain
increased phosphotyrosine-containing
proteins(22, 24, 26) . We tested the role of
the unique, SH3, and SH2 domains in directing critical protein-protein
interactions necessary for Lyn function by examining the level of
phosphotyrosine proteins in cells expressing the Lyn deletion mutants.
Whole cell lysates of transiently transfected NIH 3T3 cells (Fig. 7) were immunoblotted with anti-phosphotyrosine
antibodies.
Lyn Deletion Polypeptides
Deletion of the SH3 or
SH2 of Lyn causes a decrease in tyrosine kinase activity in
vitro. This is in contrast to the 15-30-fold increase in
kinase activity found with similar mutations in Src. Interestingly,
Seidel-Dugan et al.(24) and Hirai and Varmus (25) showed that deletion of the SH3 domain from activated
variants of c-Src, caused a similar decrease in specific activity in vitro and no change in the Src variant's
phosphorylation profile. In this regard, the role of the SH3 domain of
Lyn behaves more similarly to the SH3 domain of an activated Src
variant than to the SH3 of c-Src.Deletion Mutations in Lyn Demonstrate Intrinsic
Differences between Lyn and c-Src
SH2 and SH3 deletion mutations
in Src have been shown to elevate their specific kinase
activity(24, 25) . This report demonstrates that
similar mutations in Lyn decrease kinase activity and implies that the
regulation of one Src family member cannot necessarily be extrapolated
onto another. The findings of this study also demonstrate that Lyn, in
contrast to Src, does not change its sites of phosphorylation upon
activation. The tyrosine kinase activity of c-Src is negatively
regulated by phosphorylation of a tyrosine located at the carboxyl
terminus, tyrosine 527. Upon activation, Src undergoes a
dephosphorylation of tyrosine 527 and a phosphorylation on tyrosine
416. The phosphorylation pattern of Lyn remains constant regardless of
whether Lyn is activated in vitro or in vivo. Since
human Lyn is a kinase implicated in several hematopoietic signaling
pathways, knowledge that the lessons of Src cannot be applied
automatically to Lyn may have strong implications.SH3 Chimeric Mutations Demonstrate That Src Family SH3
Domains Cannot Completely Substitute for Each Other
This study
shows that the Src family SH3 domains, although highly homologous,
appear to have specificity in their function. We were struck by the
marked difference between in vitro kinase activity of our
chimeric kinase polypeptides. There are two explanations for why the
wild-type SH3 regulates a Src family kinase, while other closely
related SH3 regions fails to do so. First, the SH3 region might exert
its negative regulation via an intramolecular mechanism. There is a
substantial body of evidence that the SH2 domain interacts with the
phosphorylated regulatory carboxyl terminus to inhibit kinase activity.
It is possible that the SH3 region has a similar effect. The SH3 domain
could interact with the regulatory tail or elsewhere in the catalytic
region. Recent evidence suggests that the SH3 domain may cooperate with
the SH2 domain and the regulatory tail to inhibit kinase
activity(50, 51, 52) .)Since p85 is thought to interact only with activated Src,
it is interesting that the activated variant, Src-SH3-Lyn, does not
associate with p85. Perhaps the Lyn SH3 domain is not able to associate
with p85, even when substituted into Src. Although we were not able to
detect novel substrates of the Lyn chimeric proteins, we did observe
phosphorylation of currently unknown proteins with molecular mass of 36
and 49 kDa with the Lyn SH2 deletion.
))
We thank L. Brass, S. Shattil, and J. Brugge for their
helpful discussions and critical reading of this manuscript, and F. Yue
for his assistance in the preparation of the figures.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  N. Kato, M. Nakanishi, and N. Hirashima Flotillin-1 Regulates IgE Receptor-Mediated Signaling in Rat Basophilic Leukemia (RBL-2H3) Cells J. Immunol., July 1, 2006; 177(1): 147 - 154. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Kamei, S. R. Jones, B. M. Chapman, K. L. MCGonigle, G. Dai, and M. J. Soares The Phosphatidylinositol 3-Kinase/Akt Signaling Pathway Modulates the Endocrine Differentiation of Trophoblast Cells Mol. Endocrinol., July 1, 2002; 16(7): 1469 - 1481. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. F. Denny, H. C. Kaufman, A. C. Chan, and D. B. Straus The Lck SH3 Domain Is Required for Activation of the Mitogen-activated Protein Kinase Pathway but Not the Initiation of T-cell Antigen Receptor Signaling J. Biol. Chem., February 19, 1999; 274(8): 5146 - 5152. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Weijland, J. C. Williams, G. Neubauer, S. A. Courtneidge, R. K. Wierenga, and G. Superti-Furga Src regulated by C-terminal phosphorylation is monomeric PNAS, April 15, 1997; 94(8): 3590 - 3595. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Zhang, J. R. Falck, K. K. Reddy, C. S. Abrams, W. Zhao, and S. E. Rittenhouse Phosphatidylinositol(3, 4, 5) -Trisphosphate Stimulates Phosphorylation of Pleckstrin in Human Platelets J. Biol. Chem., September 29, 1995; 270(39): 22807 - 22810. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |