| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Basel Institute for Immunology, Basel, Switzerland
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ß T
lymphocytes (14, 15). Fyn kinase can efficiently couple both Ag
receptors to intracellular calcium mobilization (11, 12, 16, 17), which
may be mediated partially through direct tyrosine phosphorylation of
the inositol 1,4,5-trisphosphate receptor (18). Fyn is also involved in
the function of IL-7 receptors (19, 20) and IL-5 receptors (21) in
lymphocytes. Interestingly, Yasunaga et al. (22) have observed that the
kinase also plays a critical role in mitotic division (cytokinesis) of
proB cells. In addition, Ley et al. (23) have demonstrated significant
cytoplasmic localization of Fyn in lymphocytes. These results strongly
suggest that Fyn plays diverse cellular roles in addition to
receptor-proximal signal transduction, and defining these alternative
roles is an important challenge.
Functional diversity between members of the Src family of kinases is
mediated by four classical protein interaction domains. These Src
homology (SH) domains are the kinase domain (SH1), which binds and
phosphorylates substrate; SH2, which binds specific
phosphotyrosine-containing sequences; SH3, which selectively binds
proline rich sequences; and an amino (N)-terminal domain of 60 to 90
amino acids. Although the majority of N-terminal amino acids are unique
for each kinase, the first 10-residue segment contains common elements
and has been termed SH4 (24). This SH4 domain facilitates membrane
localization through myristoylation of a glycine and palmitoylation of
cysteines on most Src family members (24, 25, 26, 27, 28, 29). In addition, the SH4
domains of Fyn and p56lyn (LynA) have been
reported to mediate weak interactions with unphosphorylated
immunoreceptor tyrosine-based activation motifs (ITAMs) of lymphocyte
Ag receptors (3, 30, 31, 32) and the signaling effectors phospholipase
C-
, GAP (GTPase-activating protein), and MAP
(microtubule-associated protein) kinase (33), indicating that this
domain also facilitates important protein interactions.
Tctex-1 is the product of a gene within the t complex on chromosome 17 of mice, which was originally cloned from sperm cDNA as a potentially important distorter component of certain mutant t haplotypes (34). t haplotypes contain multiple conserved mutated genetic elements within the t complex that are responsible for various phenotypes of tail length, embryonic lethality, male sterility, and sperm cell transmission ratio distortion (35). Analysis of the Tctex-1 protein has been limited (36), until very recently, when King et al. (37) demonstrated that it is a light chain component of the cytoplasmic dynein complex. Cytoplasmic dynein is a large microtubule-based multicomponent ATP-dependent motor unit that plays important roles in intracellular retrograde organelle transport, membrane trafficking, mitotic spindle localization, and centrosome separation during mitosis (38, 39, 40). Tctex-1 has also recently been shown to be a component in inner arm I1 of flagellar dynein (41). A human homologue of the tctex-1 gene was reported to encode 94% protein homology to the murine protein (42), and a related human gene, "candidate RP3," encodes a protein with 55% amino acid identity to human Tctex-1 (43). A homologue of another mouse t complex-encoded protein, Tctex-2, which has limited sequence similarity to Tctex-1, was also recently identified as a light chain component of flagellar dynein in Chlamydomonas (44), and a Tctex-1 homologue in the same species exhibits 60% identity to the mouse protein (41).
To better define functionally important protein interactions with the
unique N-terminal domain of Fyn, we have utilized the yeast two-hybrid
technique. Using this system, we could not demonstrate direct Fyn
interactions with the ITAM-containing B cell Ag receptor components,
Ig- or Ig-ß. Upon screening a B cell cDNA library, however,
Tctex-1 was recognized as a strong Fyn-binding protein in the yeast
system, and this interaction was localized to the first 19 amino acids
of Fyn. The validity of this protein-protein interaction was confirmed
by the capacity of a Tctex-1 fusion protein to bind Fyn from cell
lysates. Fyn and Tctex-1 were found to colocalize during cytokinesis in
a T cell hybridoma, thereby suggesting that the interaction with
Tctex-1 can selectively recruit Fyn to the dynein motor complex during
mitosis of lymphocytes.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cell lines and their sources were the murine proB cell line 38B9 (Dr. A. Rolink, Basel, Switzerland); cytotoxic T cell clone CTLL-2 (Dr. J. García-Sanz, Madrid, Spain); COS-7 cells (Dr. B. Imhof, Geneva, Switzerland); the murine B cell lymphomas K46, A20, and WEHI-231.7; and NIH 3T3 cells transfected with the ZIP-Fyn plasmid to express high levels of p59fyn protein (45) (all from Dr. J. Cambier, Denver, CO). The fyn-transfected murine T cell hybridoma N17 (10) was generously provided by Dr. T. Yamamoto and colleagues (University of Tokyo, Japan). Cells were cultured as previously described (46). Geneticin (G-418; Life Technologies, Gaithersburg, MD) was supplemented in cultures of COS-7 transfectants (500 µg/ml) and N17 cells (770 µg/ml).
Antibodies
Anti-Tctex-1 polyclonal Abs (pAb) were generated in rabbits against purified full length murine Tctex-1, which was produced in Escherichia coli as a GST fusion protein (see below) and cleaved from the GST with human thrombin (Sigma). The Tctex-1-specific Ab was affinity purified using the cleaved protein on CNBr-activated Sepharose (Pharmacia, Uppsala, Sweden), eluted with 3.5 M MgCl2 (pH 7.2), dialyzed against PBS, and stored at -70°C. Anti-Fyn Abs (both from Santa Cruz Biotechnology, Santa Cruz, CA) were a mAb (sc-434, mouse IgG1; agarose conjugate used for immunoprecipitation or unconjugated for intracellular staining) and a rabbit pAb (sc-016, unconjugated, used for immunoblotting and intracellular staining). Anti-FLAG mAb (M2; mouse IgG1; Kodak, Rochester, NY) was conjugated to CNBr-activated Sepharose for immunoprecipitation. Anti-ß-tubulin mAb was from Boehringer Mannheim (Mannheim, Germany; mouse IgG2b). Anti-dynein 74-kDa intermediate chain Ab, clone 74.1 (47) (mouse IgG2b), was purchased from Chemicon (Temecula, CA). The anti-human MHC class I mAb, W6/32, was used as a hybridoma culture supernatant and was generously provided by Dr. Marina Cella (Basel Institute for Immunology). Rabbit anti-mouse IgG pAb, goat anti-rabbit IgG conjugated with Texas Red or FITC, and goat anti-mouse IgG1 conjugated with Texas Red were from Southern Biotechnology Associates (Birmingham, AL). Oregon Green 488-conjugated goat anti-rabbit IgG Ab was from Molecular Probes (Eugene, OR).
Yeast two-hybrid method
A Gal4-based yeast two-hybrid system was generously made
available by Drs. P. Chevray and D. Nathans (Johns Hopkins
University, Baltimore, MD) (48). The pPC62 plasmid encodes the
DNA-binding domain of Gal4 (amino acids 1–147) with 3' SalI
and NotI restriction sites, which were utilized to generate
"bait" fusion protein constructs. Bait constructs were generated by
PCR from a K46 murine B cell lymphoma cDNA library (described below);
these were the N-terminal domains of Fyn (amino acids 1–85, a mutated
version with lysines 7 and 9 mutated to alanines, and truncated
versions with indicated amino acids), p55blk
(Blk; amino acids 1–56), LynA (amino acids 1–68),
p53lyn (lynB; amino acids 1–47), the
cytoplasmic domain of Ig-ß (amino acids 186–228), and full length
Tctex-1 (amino acids 1–113). The pPC86 plasmid encodes the
Gal4 transcriptional activation domain (amino acids 767–881) and 3'
restriction sites for generation of "prey" fusion protein
constructs. A cDNA library was prepared from 5 µg of
poly(A)+-enriched RNA that had been isolated from the
K46 B cell lymphoma using the Poly(A) Quik Kit (Stratagene, La Jolla,
CA). One half of the K46 cells were resting, while the other half had
been stimulated with sheep anti-mouse IgG (Silenus, Hawthorne,
Australia; 1 µg/ml) and harvested at time points between 15 min and
6 h. Directionally cloned cDNA was generated with the ZAP-cDNA
synthesis kit (Stratagene) and inserted into
EcoRI/SalI-modified pPC86 plasmid (provided by
Dr. T. Watanabe, Kyushu University, Fukuoka, Japan). The library
contained about 1.2 million original clones with >95% inserts ranging
from 200 to 2000 base pair. Additional prey fusion protein constructs
that were generated by PCR into pPC86 were Tctex-1 truncations (amino
acid sequences described in the text), the cytoplasmic domain of Ig-
(amino acids 160–220), and the N-terminal domains of Fyn (amino acids
1–85), LynA (amino acids 1–68), and Lck (amino acids 1–67). The
yeast reporter strain was HF7c (generously provided by Dr. D. Beach,
Cold Spring Harbor Laboratory, Cold Spring Harbor, NY) with integrated
growth selection (HIS3) and ß-galactosidase
(lacZ) reporter genes (49). For cDNA library
screening, yeast were first transfected with the bait plasmid (in
pPC62) and subsequently transfected with a library (in pPC86) using
lithium acetate and polyethylene glycol (50). Colonies that grew in the
absence of histidine were secondarily screened for ß-galactosidase
activity (51). cDNA plasmids were released from positive yeast colonies
and cloned as described (52) and sequenced by PCR using the SequiTherm
cycle sequencing kit (Epicentre Technologies, Madison, WI).
Transfectants were streaked on agarose medium lacking histidine to test
for growth reporter activation or on medium containing histidine to
test for activation of the lacZ reporter.
Northern blotting
Total RNA (10 µg/sample) was isolated with guanidinium thiocyanate from cell lines or mouse tissues (C57BL/6 mice from Iffa Credo, L’Arbresle, France), separated on a 1% agarose gel, and transferred to a nylon membrane (Hybond-N; Amersham, Arlington Heights, IL). A full length tctex-1 cDNA sequence probe and a full length ß-actin sequence probe were generated by PCR and labeled with [32P]dCTP using the Prime-It random primer kit (Stratagene). The membrane blot was hybridized at 65°C and washed at 65°C.
Immunoblotting
Immunoprecipitates or whole cell lysates (lysed directly in Laemmli buffer) were separated on SDS-PAGE and electrophoretically transferred to Immobilon-P membranes (Millipore, Bedford, MA). Immunoblotting and stripping were performed as previously described (53). Abs were rabbit anti-Tctex-1 (10 µg/ml) or rabbit anti-Fyn pAb (1 µg/ml). Secondary reagent was 125I-labeled protein A (0.5 µCi/ml; Amersham), and proteins were visualized by autoradiography. It should be noted that detection of Tctex-1 can be obscured by a background band at 14 kDa in lymphocyte whole cell lysates when probed with horseradish peroxidase-conjugated secondary reagents; this result can be avoided by using 125I-labeled protein A.
Metabolic labeling, protein precipitations, and in vitro kinase reactions
COS-7 cells were metabolically pulse-labeled in some studies for 15 to 45 min with 5 mCi of [35S]cysteine/methionine (Amersham) in 20 ml of cysteine/methionine-free Dulbecco’s modified Eagle’s medium and chased for 2 to 9 h in normal medium. Cells were lysed for 30 min on ice in 1% digitonin buffer, 1% Triton X-100 (Surfact-Amps; Pierce, Rockford, IL) buffer, or RIPA buffer (1% Triton X-100, 0.1% sodium deoxycholate (Merck, Darmstadt, Germany)), and 0.1% SDS (Bio-Rad, Hercules, CA), each containing 75 mM NaCl, 10 mM Tris (pH 7.4), 10 mM NaF, 0.4 mM EDTA, 1 mM Pefabloc SC (Boehringer Mannheim), 2 mM sodium orthovanadate, and 1 µg/ml each of aprotinin, soybean trypsin inhibitor, and leupeptin (reagents were from Sigma unless otherwise noted). Lysates were microfuged at 14,000 rpm for 15 min and subjected to precipitation for 2 to 4 h with GST fusion proteins (10 µg/sample) or Abs (2–5 µg/sample) prebound to glutathione-Sepharose, GammaBindPlus Sepharose, or protein A-Sepharose 4 Fast Flow beads (Pharmacia). Protein precipitations were washed five times with 0.2% digitonin or RIPA buffer (as above). Some samples were subjected to in vitro kinase reactions with [32P]ATP, quenched with 1% Triton X-100 buffer (as above), and secondarily immunoprecipitated with anti-Fyn mAb-coupled agarose as previously described (46). Phosphoamino acid analysis was performed as previously described (54).
GST and epitope-tagged fusion proteins and protein expression
GST fusion protein constructs of the Fyn N-terminal domain
(amino acids 1–85) and Tctex-1 (amino acids 1–113) were generated by
excision of these cDNAs from pPC62 and pPC86 and ligation into
pGEX-4T-2 plasmid (Pharmacia). Other GST fusion constructs were human
platelet-derived growth factor receptor kinase insert region (amino
acids 698–797 in pGEX-3X; provided by Dr. A. Kazlauskas, Harvard
University) and cytoplasmic domains of murine Ig- and Ig-ß (3)
(provided by Dr. J. Cambier). E. coli (Top10F') were
transformed with these plasmids and induced for 2 to 4 h with 0.3
mM isopropyl ß-D-thiogalactopyranoside as described (3).
Cells were pelleted, and probe sonicated on ice in 10 ml of
Tris-buffered saline (TBS; 10 mM Tris, pH 7.4, 150 mM NaCl) containing
Complete protease inhibitor (Boehringer Mannheim; 1 tablet/50 ml TBS),
admixed to 1% Triton X-100 and pelleted at 12,000 rpm for 15 min.
Cleared lysate was precipitated for 2 h to overnight with
glutathione-Sepharose 4B beads (Pharmacia), and beads were washed with
1% Triton X-100 in TBS and stored at 4°C as a 50% slurry in the
same buffer. A C-terminal FLAG epitope-tagged Tctex-1 construct was
created by PCR encoding DYKDDDDK-stop after the 113-amino acid
murine Tctex-1 sequence and cloned into the mammalian expression
plasmid, pcDNA3 (InVitrogen, San Diego, CA). COS-7 cells (70%
confluent) were transfected in 100-mm petri dishes with 5 µg of
plasmid DNA using Dosper liposomal transfection reagent according to
the manufacturer’s protocol (Boehringer Mannheim), and protein
expression was assayed two days later. Stably integrated clones were
isolated by limiting dilution. An expression plasmid encoding
N-terminal FLAG-tagged Tctex-1 was also generated
(M-DYKDDDDK-), but did not produce significant levels of protein
expression in COS-7.
Immunofluorescence confocal microscopy
Some cells were grown on chamber slides (see Fig. 7
A) or allowed to adhere on poly-L-lysine coated
slides (Fig. 7B) before staining. Other cells were stained
in suspension after sorting and subsequently cytospun onto slides (Fig. 7C). The following staining protocol was utilized, since it
resulted in optimal retention of ß-tubulin structure in cells. Cells
were washed in 37°C serum-free Iscove’s modified Dulbecco’s
medium before fixing at 37°C for 10 min with 3%
paraformaldehyde (in PBS, pH 7.0, and 5.4% glucose). Subsequent steps
were performed at room temperature. After washing twice in PBS (pH
7.4), fixed cells were permeabilized for 30 min in PBS/saponin (PBS
with 0.1% saponin and 1 mM HEPES) and washed again in the same. Cells
were blocked for 30 min in PBS/saponin/BSA (PBS/saponin with 2% BSA)
and washed once in PBS/saponin. Cells were then incubated for 30 to 45
min with optimal concentrations of primary Ab in PBS/saponin/BSA and
washed three times with PBS/saponin. An optimal concentration of
secondary Ab in PBS/saponin/BSA was added for 30 to 45 min, and cells
were washed three times with PBS/saponin, followed by three washes with
PBS. Some cells were fixed, stained with Hoechst 33342 (Molecular
Probes), and sorted on a FACS Vantage into
G0/G1, S, and G2/M stages (55)
before intracellular Ab staining (Fig. 7C). Coverslips were
mounted with Mowiol (Hoechst, Frankfurt, Germany) or
Fluoromount-G (Southern Biotechnology) mounting solution and analyzed
through a 63x/1.4 Zeiss Plan-Apochromat lens on a Zeiss Axiovert
100 inverted microscope fitted with a Bio-Rad MRC 1024 laser scanning
confocal imaging system. Images were acquired in accumulation mode,
analyzed using LaserSharp software (Bio-Rad, version 2.1A), and
processed with Adobe Photoshop (version 3.0, Adobe Systems, San
Jose, CA). Compensation was rigidly controlled using single-stained
cell samples and/or independent sequential excitation of each
fluorescent dye to assure lack of bleedover between the green channel
and the red channel. Pretreatment of the anti-Tctex-1 pAb
preparation with recombinant Tctex-1 protein completely eliminated
immunofluorescent reactivity, thereby indicating specificity of the
staining.
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Previous experiments with GST fusion proteins have suggested that
the N-terminal domains of Fyn and LynA can interact directly with the
cytoplasmic domains of the B cell Ag receptor chains, Ig- and Ig-ß
(31). As demonstrated in Figure 1, we
could not detect any direct protein interactions between these kinase
N-terminal domains (or LynB; data not shown) and either Ig- or
Ig-ß cytoplasmic domains in a Gal4-based yeast two-hybrid system in
either orientation (DB or TA fusions), nor in a more sensitive
LexA-based system (data not shown; sensitivity described in 46 .
All fusion proteins of appropriate size were produced in the yeast
transfectants, however, as assessed by anti-GAL4 immunoblotting
(data not shown). Although these results suggest that Fyn cannot
directly interact with Ig- or Ig-ß, one must consider that many
factors can contribute to a negative result in this assay, and
therefore, this is not proof that these interactions cannot occur
in vivo.
|
, since
the tctex-1 gene product did not interact with the
N-terminal domains of LynA, LynB, Blk, or p56lck
(Lck; see Fig. 5) as measured by activation of either His3
(histidine-free growth selection) or lacZ
(ß-galactosidase; data not shown) reporters. In addition, the
N-terminal domain of Fyn was not observed to form homotypic
interactions (Fig. 1). Tctex-1 failed to interact with Ig- or Ig-ß
cytoplasmic domains in this assay (data not shown), and we could not
demonstrate any capacity of Tctex-1 to "couple" Ig- to Fyn when
expressed as a third protein in yeast two-hybrid experiments (data not
shown). The interaction of Fyn with Tctex-1 is clearly a strong
protein-protein interaction when compared with the lack of detectable
interactions between Fyn and Ig- or Ig-ß.
|
The tctex-1 gene had been cloned from sperm cDNA and
identified as a product of the t complex of mice (34).
Previous analysis by Northern blotting had determined that the gene was
strongly expressed in the testes and ovaries (34) and weakly in the
thymus of wild-type mice, but a detailed analysis in hemopoietic
tissues was lacking. As shown in Figure 2, Northern blot analysis demonstrated
ample expression of tctex-1 mRNA in all tissues examined
from BDF1 mice. Thymic expression was more pronounced than
that of spleen, while significant expression of tctex-1 mRNA
was observed in both B and T lymphocyte lines (Fig. 2). Less abundant
expression of tctex-1 was also evident in the brain,
suggesting that the protein is also available for interaction with Fyn
in neuronal cells (Fig. 2).
|
a, immunoprecipitation from a
35S metabolically labeled transfectant (clone C20)
with either anti-Tctex-1 or anti-FLAG Abs resulted in the
purification of a specific band of about 15 kDa. This apparent mass is
slightly higher than that predicted for Tctex-1 (
14 kDa) due to the
epitope tag. This band was the major specific protein
immunoprecipitated by the anti-Tctex-1 pAb preparation when
compared with control immunoprecipitations (anti-human MHC class I
and anti-tubulin mAbs in lanes 3 and 4 of
Fig. 3a). In addition, the pAb preparation and anti-FLAG
mAb reproducibly immunoprecipitated similar amounts of
Tctex-1.
|
b; data not shown). The specific reactive band was
observed at 14 kDa, as predicted, and this migration corresponded
exactly to that of recombinant Tctex-1 as shown in Figure 3b. No significant improvement in the amount of Tctex-1
immunoprecipited from lysates of K46 cells could be demonstrated upon
lysis with the addition of other nonionic detergents, harsher RIPA
buffer, 10 mM ATP, or nocodazole pretreatment of cells, and only slight
improvement was achieved using a monoclonal anti-dynein
intermediate chain Ab (data not shown). Therefore, both mRNA and the
protein product of the tctex-1 gene were detectably
expressed in hemopoietic tissues as well as T and B lymphocyte lines,
but protein levels were routinely low in these cells. Despite significant efforts, we could not detectably coimmunoprecipitate the two proteins from Fyn-transfectants of COS-7 cells, NIH 3T3 fibroblasts, or the T cell hybridoma N17 by immunoprecipitation of either protein or dynein intermediate chain (data not shown). Detergents tested in these studies included Triton X-100, digitonin, CHAPS (3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate), and RIPA, and detection methods employed were either 35S metabolic labeling, immunoblotting, or in vitro phosphorylation of immunoprecipitated kinase in the presence of [32P]ATP. These results suggest that the protein interaction is either transient in these cells or the off rate of the interaction is high, thereby limiting detection after washing of immunoprecipitates.
Fyn interaction with Tctex-1-GST fusion protein in vitro
Due to the difficulties in demonstrating coimmunoprecipitation, we
tested the capacity of recombinant Tctex-1-GST fusion protein to
interact with Fyn in cell lysates. Digitonin lysates of Fyn-transfected
NIH 3T3 cells (45) were adsorbed with various GST fusion proteins,
including Tctex-1, and precipitates were washed and subjected to in
vitro [32P]ATP phosphorylation reactions to detect
associated phosphoproteins. As presented in Figure 4a, GST fusion proteins of the
cytoplasmic domains of Ig-, and to a lesser extent, Ig-ß, were
significantly phosphorylated in this assay, while Tctex-1-GST fusion
protein was only weakly phosphorylated. The Tctex-1-GST fusion protein,
however, selectively coprecipitated a phosphoprotein band of about 59
kDa, which comigrated with p59fyn.
Reimmunoprecipitation from these phosphorylation reactions with an
anti-Fyn mAb confirmed that the kinase had associated with the
Tctex-1-GST fusion protein (Fig. 4a). The
[32P]p59 band from Figure 4a exclusively
contained phosphotyrosine, which further verified its identity as
autophosphorylated Fyn (Fig. 4b). Although Tctex-1-GST did
not provide a good substrate for the adsorbed Fyn in this assay, the
minimal 32P labeling of Tctex-1-GST was also found to be
exclusively incorporated onto tyrosine residues (Fig. 4b).
These results suggest that Tctex-1 can be weakly tyrosine
phosphorylated by Fyn kinase. Finally, Fyn binding to Tctex-1-GST was
confirmed by immunoblotting with anti-Fyn Ab as demonstrated
in Figure 4c. The Ig--GST fusion protein was also able to
adsorb Fyn in both of these assays as has previously been reported (3),
while Ig-ß, PDGFR-KI, and the N-terminal domain of Fyn did not bind
appreciable amounts of the enzyme. In summary, the Tctex-1-GST fusion
protein can selectively bind Fyn from digitonin lysates in vitro,
thereby confirming the yeast two-hybrid interaction.
|
To identify the specific interacting domains in both Fyn and
Tctex-1, truncation analysis studies were undertaken in the
yeast two-hybrid system. When truncations of the 85-amino acid
N-terminal domain of Fyn were tested as demonstrated in Figure 5a, elimination of just the
first 6 amino acids from the N terminus completely abrogated the
interaction. The C-terminal amino acids of this domain, however, did
not appear to contribute to the interaction with Tctex-1, yet amino
acids 1 through 19 still exhibited strong interaction. On the other
hand, further truncation to amino acids 1 through 10 completely
abolished the interaction (see Fig. 5a). Previous studies by
Timson Gauen et al. (32) have indicated that amino acids 1 through 10,
and in particular the lysine residues at positions 7 and 9 of Fyn (Fig. 5b), are critical elements in the interaction of this domain
with ITAM-containing sequences on lymphocyte Ag receptors. We also
tested for involvement of lysines 7 and 9 in interaction with Tctex-1
by mutating them to alanines in the complete N-terminal domain. As
shown in Figure 5c, this mutated fragment of Fyn did not
interact with Tctex-1, indicating that these lysine residues are
involved in interactions of the kinase with both ITAM-containing
sequences and Tctex-1. Thus, Tctex-1 interacts with the first 19 amino
acids of Fyn; lysine residues at positions 7 and 9 are critical
elements in this interaction domain, although they are clearly not the
only binding residues, as determined in the truncation analysis (Fig. 5a).
Tctex-1 was also truncated and tested for interaction with Fyn in the
two-hybrid system. As shown in Figure 6,
Tctex-1 was very sensitive to truncation from both ends, since
truncation of amino acids 105 through 113 completely abolished the
interaction with Fyn, and elimination of amino acids 1 through 12
significantly reduced the interaction. These results demonstrate that
the protein requires integrity of both termini to form the
Fyn-interacting structural domain.
|
Confocal immunofluorescence analysis was performed to determine
the intracellular localization of Tctex-1. Analysis of
C-FLAG-Tctex-1-transfected COS-7 cells demonstrated that anti-FLAG
mAb (Fig. 7A) and
anti-Tctex-1 pAb (identical pattern; data not shown) diffusely
stained the cytoplasm with the majority of staining concentrated in the
perinuclear region. Alternatively, microtubules (stained with
anti-ß-tubulin) emanated from the perinuclear Tctex-1
stained region and extended to the periphery of the cells (Fig. 7A). Tctex-1 staining was not detected in the nucleus or
plasma membrane.
Double-staining intracellular immunofluorescence studies were also
performed to determine whether Tctex-1 and Fyn are colocalized in
lymphocytes. The fyn-transfected murine T cell hybridoma,
N17 (10), provided detectable levels of kinase for these studies.
Elevated Fyn expression in N17 cells was clearly evident in immunoblots
of whole cell lysates and immunofluorescent staining when compared with
other lymphocyte cell lines (data not shown). Most N17 cells exhibited
Fyn staining that was predominantly concentrated to the plasma
membrane, although some cells exhibited significant foci of Fyn
staining within the cytoplasm (Fig. 7B) as previously
reported by Ley et al. (23). Although Tctex-1 staining was concentrated
within the perinuclear/cytoplasmic region and generally distinct from
Fyn-stained regions, a subset (25%) of the T hybridoma cells
demonstrated distinct cytoplasmic foci of Fyn, some of which were
clearly colocalized with Tctex-1 (Fig. 7B). We surmised that
this inconsistent colocalization of Fyn with Tctex-1 within the
population of T hybridoma cells might represent colocalization only at
a distinct stage(s) of the cell cycle. To subdivide the cells into
distinct stages of cycle, the N17 cells were sorted by DNA content on
FACS into G0/G1, S, and G2/M stages
before intracellular Ab staining. Although no consistent colocalization
was identified in the G0/G1 or S stage
populations (data not shown), the G2/M stage population
exhibited a unique overlap in staining that was observed in all
dividing cells. As shown in Figure 7C, this population is
enriched in cells undergoing cytokinesis, which exhibited a striking
colocalization of Fyn and Tctex-1 at the developing cleavage furrow
that forms as daughter cells begin to divide. In addition,
colocalization was evident in these cells at the mitotic spindles. It
is interesting to note that Tctex-1, but not Fyn or ß-tubulin, seems
to interdigitate through the segregated chromosomes and connects the
mitotic spindles with the cleavage furrow via fibrous arrays.
ß-Tubulin staining, although a major component of the mitotic
spindles, did not colocalize with Fyn and Tctex-1 at the center of the
cleavage furrow in the cell cross-sections as presented in Figure 7C. Overall, this stage of cell division was the only
interval during which we consistently observed intracellular
colocalization of Fyn with Tctex-1. Tctex-1 staining was always
perinuclear throughout all other stages of the cell cycle, and we never
observed staining of the nucleus or plasma membrane. In conclusion, we
observed highly reproducible colocalization of Fyn and Tctex-1 during
cytokinesis at both the cleavage furrow and mitotic spindles in this T
cell hybridoma.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Amino acids 1 through 19 of Fyn (see Fig. 5b), which were
identified as interacting with Tctex-1, encompass the domain that
interacts with unphosphorylated ITAM motifs in lymphocyte Ag receptors
(SH4, amino acids 1–10 (31, 32)) and contain three lipid modification
sites (Fig. 5b). Cotranslational myristoylation of the
glycine at position 2 is believed to be a permanent alteration, while
posttranslational palmitoylation of cysteines at positions 3 and 6 is a
reversible modification (24, 25, 26, 27, 28, 29, 57). Interestingly, the two lysines at
positions 7 and 9 within the Fyn interactive sequence are critical
elements for interactions of Fyn with both Tctex-1 (Fig. 5c)
and ITAM sequences (32). These basic residues have previously been
implicated in enhancing plasma membrane association of Src family
kinases by interacting with negatively charged phospholipid head groups
(24). Our results and those of Timson Gauen et al. (32) indicate that
these lysine residues and presumably other residues in the extreme
N-terminal domain of Fyn are also important for protein-protein
interactions. One attractive hypothesis would be that a lymphocyte
activation or cell cycle-related event might promote depalmitoylation
of Fyn and thereby release it from the membrane, exposing the domain
that interacts with Tctex-1. It is unlikely that the Fyn fusion
proteins are palmitoylated in our yeast experiments, since
palmitoylation of this sequence is considered to be dependent upon
nearby myristoylation in Src family kinases (24, 26, 28), which is not
possible with the initiating methionine of Fyn fused to Gal4. Mutation
of the two lysines of Fyn has been shown to alter its localization from
the plasma membrane to the cytoplasm (32), although our mutational
studies indicate that this localization is not due to Tctex-1 binding
(Fig. 5c).
The colocalization of Fyn with Tctex-1 during cytokinesis is a particularly intriguing result in light of other recently published reports. Of particular interest is a study by Yasunaga et al. (22), which determined that proB cells cultured from fyn-deficient mice grew essentially normally until transferred to defined serum-free conditions, at which point these cells arrested during cytokinesis at telophase. In striking contrast, proB cells from normal animals continued through the cell cycle in these serum-free conditions. Their results indicate that Fyn plays a critical role in cell division, although its requirement can be overcome by serum-derived growth factors that presumably bypass the Fyn deficiency block. The same report also demonstrated the localization of Fyn in the cleavage furrow at anaphase of normal proB cells, which we have reproduced in our studies of a T cell hybridoma. Taken together with this genetic evidence, our results suggest that Tctex-1 might provide the crucial scaffold link that tethers Fyn to cytoskeletal motor structures in lymphoid cells, where it functions during cytokinesis. Previous observations of decreased proliferative capacity of thymocytes from fyn-deficient mice (11, 12) might, in fact, be partially explained by this requirement and further reinforces the importance of this newly identified function for the kinase.
Ley et al. (23) have reported that Fyn is almost exclusively localized
at the centrosome and mitotic spindles of interphase and mitotic T
cells, respectively. Since dynein mediates retrograde transport toward
these structures, the association of Fyn with Tctex-1 in the
dynein complex could clearly mediate this localization of Fyn in
lymphocytes. Although we have observed predominantly plasma membrane
localization of Fyn in interphase cells using two different Abs in the
murine T cell hybridoma, N17 (and other T lymphomas; data not shown), a
subpopulation of cells demonstrated distinct cytoplasmic foci of Fyn
staining, some of which colocalized with Tctex-1 (Fig. 7C).
Ley et al. used a polyclonal Ab directed to the C-terminal domain of
Fyn (23), while our Abs were N-terminal reactive, which may account for
the predominantly plasma membrane staining pattern in our studies.
Roche et al. (58) have demonstrated a requirement for Src kinases,
including Fyn, at an earlier stage of the mitotic cell cycle. They
reported increased activity of Fyn and other Src kinases in
G2/M phase-blocked fibroblasts. In addition, fibroblasts
were arrested before prophase, in the same report, by microinjection of
an anti-Fyn/Src/Yes Ab (the same Ab used by Ley et al.) or a GST
fusion protein of the Fyn SH2 domain. Finally, Marie-Cardine et al.
(59) reported that T cell activation results in tyrosine
phosphorylation of -tubulin and that this phosphorylated -tubulin
can bind the Fyn SH2 domain. Katagiri et al. (60) have also noted
tyrosine phosphorylation of tubulin during monocyte differentiation of
HL-60 cells and concomitant association of Fyn and Lyn with tubulin.
These results taken together suggest potentially important roles for
Fyn recruitment to and phosphorylation of microtubule cytoskeletal
elements during cellular activation and mitosis and suggest that
Tctex-1 might be an important mediator of the Fyn recruitment.
What is the role of this interaction during the cell cycle? Although we can only speculate, colocalization of Fyn with Tctex-1 at the cleavage furrow/mitotic spindles and fyn requirements for cytokinesis in proB cells (22) suggest specific roles for these proteins in the division of lymphocytes. Alternatively, the binding of Fyn to Tctex-1 may occur as a capture mechanism for specific cargo in dynein-mediated protein sorting during mitosis. Our understanding of the complex molecular events occurring at the cleavage furrow is only currently unfolding. The role of actin filaments and associated structures in this process is clear, but microtubules and even dynein appear also to play roles (38, 61). Cytoplasmic dynein has also previously been shown to play roles in centrosome separation, anaphase B spindle elongation, and positioning of mitotic spindles during mitosis (38, 39, 40). Recruitment of Fyn to dynein by Tctex-1 might affect some or all of these mitotic events.
Although Tctex-1 does not seem to serve as a major substrate for Fyn in
our in vitro phosphorylation studies (Fig. 4a and direct
mixing in in vitro phosphorylation reactions, data not shown), it was
nevertheless detectably tyrosine phosphorylated. Other proteins within
the dynein complex may, however, be more efficacious substrates once
the kinase is recruited. Ley et al. (23) have described protein
tyrosine phosphorylation surrounding the Fyn at microtubule organizing
centers, and as previously mentioned, -tubulin is a potential
tyrosine phosphorylated substrate for Fyn at this location (59, 60).
Phosphorylation events identified to date during cytokinesis have
predominantly focused on the myosin chains, which undergo
serine/threonine phosphorylation (reviewed in 61 . Components of
the cytoplasmic dynein complex have been reported to be phosphorylated
during cell cycle and transport processes, but again, only serine and
threonine phosphorylation has been identified (47, 62, 63, 64). Karki et
al. (65) recently reported the association of casein kinase II with
dynein, which appears to mediate some of this protein phosphorylation
and thereby appears to affect function of the motor complex.
Determination of the functional roles of recruited Fyn during
cytokinesis in lymphocytes and the possibility of consequent tyrosine
phosphorylation of the dynein complex and associated structures should
allow for many enlightening future investigations.
In summary, accumulating evidence is implicating Fyn kinase as an important effector within the cytoskeleton of lymphocytes, particularly during mitosis. The identification of a direct interaction of the Fyn N-terminal domain with the cytoplasmic dynein motor complex light chain, Tctex-1, provides a novel mechanism for the recruitment of Fyn to this distinct intracellular location.
Note Added in Proof. Two potential Fyn substrates that might be important during cytokinesis are the cleavage furrow-associated protein PSTPIP (66) and the inositol 1,4,5-trisphosphate receptor (67).
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Kerry S. Campbell, Fox Chase Cancer Center, 7701 Burholme Avenue, Philadelphia, PA 19111. E-mail address:
3 Current address: Department of Human Retrovirology, Academic Medical Center, University of Amsterdam, Meibergdreef 15, 1105 AZ Amsterdam, The Netherlands.
4 Abbreviations used in this paper: Fyn, protein tyrosine kinase p59fyn; GST, glutathione S-transferase; Lck, protein tyrosine kinase p56lck; SH, Src homology; LynA, p56lyn; LynB, p53lyn; ITAMs, immunoreceptor tyrosine-based activation motifs; Blk, p55blk; pAb, polyclonal Ab; DB, DNA-binding (protein domain); TA, transcriptional activation (domain).
Received for publication November 14, 1997. Accepted for publication April 15, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and Ig-ß with distinct cytoplasmic effectors. Science 258:123.ß T cell development is abolished in mice lacking both Lck and Fyn protein tyrosine kinases. Immunity 5:429.[Medline]
. Proc. Natl. Acad. Sci. USA 91:4268.2, microtubule-associated protein kinase, GTPase-activating protein, and phosphatidylinositol 3-kinase. Mol. Cell. Biol. 13:5877.-tubulin and its association with the SH2 domain of the p59fyn protein tyrosine kinase. Eur. J. Immunol. 25:3290.[Medline]
This article has been cited by other articles:
|
|
 |
|
  J. C. Williams, P. L. Roulhac, A. G. Roy, R. B. Vallee, M. C. Fitzgerald, and W. A. Hendrickson Structural and thermodynamic characterization of a cytoplasmic dynein light chain intermediate chain complex PNAS, June 12, 2007; 104(24): 10028 - 10033. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Song, G. Benison, A. Nyarko, T. S. Hays, and E. Barbar Potential Role for Phosphorylation in Differential Regulation of the Assembly of Dynein Light Chains J. Biol. Chem., June 8, 2007; 282(23): 17272 - 17279. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. C. Williams, H. Xie, and W. A. Hendrickson Crystal Structure of Dynein Light Chain TcTex-1 J. Biol. Chem., June 10, 2005; 280(23): 21981 - 21986. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. A. MacCorkle and T.-H. Tan Inhibition of JNK2 Disrupts Anaphase and Produces Aneuploidy in Mammalian Cells J. Biol. Chem., September 17, 2004; 279(38): 40112 - 40121. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M.-g. Li, M. Serr, E. A. Newman, and T. S. Hays The Drosophila tctex-1 Light Chain Is Dispensable for Essential Cytoplasmic Dynein Functions but Is Required during Spermatid Differentiation Mol. Biol. Cell, July 1, 2004; 15(7): 3005 - 3014. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. S. Campbell, S.-i. Yusa, A. Kikuchi-Maki, and T. L. Catina NKp44 Triggers NK Cell Activation through DAP12 Association That Is Not Influenced by a Putative Cytoplasmic Inhibitory Sequence J. Immunol., January 15, 2004; 172(2): 899 - 906. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. P. Aumais, S. N. Williams, W. Luo, M. Nishino, K. A. Caldwell, G. A. Caldwell, S.-H. Lin, and L.-y. Yu-Lee Role for NudC, a dynein-associated nuclear movement protein, in mitosis and cytokinesis J. Cell Sci., May 15, 2003; 116(10): 1991 - 2003. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Mueller, X. Cao, R. Welker, and E. Wimmer Interaction of the Poliovirus Receptor CD155 with the Dynein Light Chain Tctex-1 and Its Implication for Poliovirus Pathogenesis J. Biol. Chem., March 1, 2002; 277(10): 7897 - 7904. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. W. Tai, J.-Z. Chuang, and C.-H. Sung Cytoplasmic Dynein Regulation by Subunit Heterogeneity and Its Role in Apical Transport J. Cell Biol., June 25, 2001; 153(7): 1499 - 1510. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Porter and W. S. Sale The 9 + 2 Axoneme Anchors Multiple Inner Arm Dyneins and a Network of Kinases and Phosphatases That Control Motility J. Cell Biol., November 27, 2000; 151(5): 37 - 42. [Full Text] [PDF]
|
|
|
|
 |
|
 W. Wang and R. E. Chapin Differential Gene Expression Detected by Suppression Subtractive Hybridization in the Ethylene Glycol Monomethyl Ether-Induced Testicular Lesion Toxicol. Sci., July 1, 2000; 56(1): 165 - 174. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Perrone, S. H. Myster, R. Bower, E. T. O'Toole, and M. E. Porter Insights into the Structural Organization of the I1 Inner Arm Dynein from a Domain Analysis of the 1beta Dynein Heavy Chain Mol. Biol. Cell, July 1, 2000; 11(7): 2297 - 2313. [Abstract] [Full Text]
|
|
|
|
|
|
A. Takesono, M. J. Cismowski, C. Ribas, M. Bernard, P. Chung, S. Hazard III, E. Duzic, and S. M. Lanier Receptor-independent Activators of Heterotrimeric G-protein Signaling Pathways J. Biol. Chem., November 19, 1999; 274(47): 33202 - 33205. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. B. McCabe and L. G. Berthiaume Functional Roles for Fatty Acylated Amino-terminal Domains in Subcellular Localization Mol. Biol. Cell, November 1, 1999; 10(11): 3771 - 3786. [Abstract] [Full Text]
|
|
|
|
|
|
Y.-K. Mok, K. W.-H. Lo, and M. Zhang Structure of Tctex-1 and Its Interaction with Cytoplasmic Dynein Intermediate Chain J. Biol. Chem., April 20, 2001; 276(17): 14067 - 14074. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Yano, F. S. Lee, H. Kong, J.-Z. Chuang, J. C. Arevalo, P. Perez, C.-H. Sung, and M. V. Chao Association of Trk Neurotrophin Receptors with Components of the Cytoplasmic Dynein Motor J. Neurosci., February 1, 2001; 21(3): RC125 - RC125. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
