Hrs, a Tyrosine Kinase Substrate with a Conserved Double Zinc Finger Domain, Is Localized to the Cytoplasmic Surface of Early Endosomes*

(Received for publication, March 24, 1997)

Masayuki Komada Dagger §, Ryuichi Masaki , Akitsugu Yamamoto and Naomi Kitamura par **

From the Dagger  Institute for Liver Research and the  Department of Physiology, Kansai Medical University, Moriguchi, Osaka 570, Japan and the par  Department of Life Science, Faculty of Bioscience and Biotechnology, Tokyo Institute of Technology, Nagatsuta, Midori-ku, Yokohama 226, Japan

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES


ABSTRACT

Hrs is a 115-kDa double zinc finger protein that is rapidly tyrosine phosphorylated in growth factor-stimulated cells. However, its function remains unknown. Here we show that Hrs is localized to early endosomes. Intracellular localization of endogenous Hrs and exogenously expressed Hrs tagged with the hemagglutinin epitope was examined by immunofluorescence staining using anti-Hrs and anti-hemagglutinin epitope antibodies, respectively. Hrs was detected in vesicular structures and was colocalized with the transferrin receptor, a marker for early endosomes, but only partially with CD63, a marker for late endosomes. A zinc finger domain deletion mutant of Hrs was also colocalized with the transferrin receptor, suggesting that the zinc finger domain is not required for its correct localization. Immunoelectron microscopy showed that Hrs was localized to the cytoplasmic surface of these structures. By subcellular fractionation, Hrs was recovered both in the cytoplasmic and membrane fractions. The membrane-associated Hrs was extracted from the membrane by alkali treatment, suggesting that it is peripherally associated with early endosomes. These results, together with our finding that Hrs is homologous to Vps27p, a protein essential for protein traffic through a prevacuolar compartment in yeast, suggest that Hrs is involved in vesicular transport through early endosomes.


INTRODUCTION

Hrs (hepatocyte growth factor (HGF)1-regulated tyrosine kinase substrate) was originally identified as a tyrosine phosphorylated protein in cells stimulated with HGF (1). In the cells expressing the HGF receptor tyrosine kinase, c-Met, which was mutated in the cytoplasmic domain and could not mediate the cellular responses triggered by HGF, Hrs was not tyrosine phosphorylated in response to ligand stimulation (1). This result suggested that Hrs plays an important role in the signaling pathway of HGF.

Hrs was purified from HGF-treated B16-F1 mouse melanoma cells by anti-phosphotyrosine immunoaffinity chromatography, and its cDNA was subsequently cloned based on the partial amino acid sequences obtained from the purified protein (2). The cDNA sequence revealed that mouse Hrs is a novel 775-amino acid protein with a double zinc finger domain that is structurally conserved in several other proteins. Northern blotting showed that Hrs mRNA is ubiquitously expressed in mouse tissues. In addition, experiments using anti-Hrs antibody, which was raised against bacterially expressed Hrs, showed that 1) tyrosine phosphorylation of Hrs reaches a maximal level 5 min after exposing cells to HGF and disappears within 2 h; 2) Hrs does not form a stable complex with the HGF receptor, c-Met, even when tyrosine phosphorylated; and 3) tyrosine phosphorylation of Hrs is also induced in cells stimulated with epidermal growth factor or platelet-derived growth factor (2).

There was no domain in the Hrs sequence that allowed its function to be predicted. However, the double zinc finger domain of Hrs was 40-50% identical to those of several other proteins including FGD1, EEA1, Fab1p, Vps27p, and ZK632.12. The consensus sequence is Cys-Xaa2-Cys-Xaa3-Phe-Xaa4-5-Arg-(Arg/Lys)-His-His-Cys-(Arg/Lys)-Xaa-Cys-Gly-Xaa-(Val/Ile)-Val/Phe-Cys-Xaa2-Cys-Ser-Xaa14-16-Arg-Val-Cys-Xaa2-Cys-(Tyr/Phe). FGD1 is a product of the human faciogenital dysplasia gene and is considered to be a guanine nucleotide exchange factor of small G proteins (3). EEA1 (early endosome antigen 1) is a hydrophilic human protein that is peripherally associated with early endosomes (4). Fab1p is a phosphatidylinositol 4-phosphate 5-kinase required for normal vacuole morphology and function in Saccharomyces cerevisiae (5). Vps27p is required for vacuolar and endocytic traffic through a prevacuolar compartment in S. cerevisiae (6). The function of the Caenorhabditis elegans protein ZK632.12 (Swiss Prot accession number P34657) is not known. In addition, the NH2-terminal half of the zinc finger domain of Hrs was 57% identical to that of Vac1p, a protein required for vacuolar inheritance and vacuolar protein sorting in S. cerevisiae (7). These facts suggest that the zinc finger domains of these proteins play the same or similar important role(s) in vesicular functions in cells and, more importantly, that Hrs is also involved in vesicular function.

In this study, we showed that the amino acid sequence of Hrs is 23% identical to that of Vps27p, a protein required for vesicular traffic in yeast. Moreover, we showed by immunofluorescence and immunoelectron microscopy that Hrs is localized to the cytoplasmic surface of early endosomes. These results suggested that, like Vps27p in yeast, Hrs is important for vesicular transport through early endosomes in mammalian cells.


EXPERIMENTAL PROCEDURES

Construction of Hrs Expression Vectors

To construct the cDNA encoding Hrs that is COOH-terminally tagged with the human influenza virus hemagglutinin (HA) epitope (Hrs-HA), four oligonucleotide primers were synthesized: 5'-ACAGACCAGCAACATAGGCT-3' (primer 1), 5'-CCCTCTAGAGTCAAAGGAGATGAGCTG-3' (primer 2), 5'-GGGTCTAGATCTAGCTATCCTTATGAC-3' (primer 3), and 5'-GGGGAGCTCAAGCTTTCATCCTCCCAGGCTGGCATA-3' (primer 4). Using primers 1 and 2 and mouse Hrs cDNA as a template, a 274-base pair fragment was amplified by polymerase chain reaction, digested by SmaI and XbaI, and purified by polyacrylamide gel electrophoresis (PAGE) (fragment 1). Using primers 3 and 4, and pCG-HA, which contains the HA epitope sequence (8), a 72-base pair fragment was amplified, digested by XbaI and SacI, and purified by PAGE (fragment 2). In addition, the pBluescript including the full-length mouse Hrs cDNA at the HindIII site was digested by SmaI and SacI, and the fragment containing the vector sequence and the partial Hrs sequence (nucleotides 1-2193) was purified from an agarose gel (fragment 3). Fragments 1, 2, and 3 were ligated together to construct the cDNA for Hrs-HA. The cDNA was excised from the vector by digestion with HindIII and cloned into the HindIII site of the mammalian expression vector pmiw (9).

The sequence of the zinc finger domain of the Hrs-HA cDNA was deleted using the Transformer Site-Directed Mutagenesis Kit (CLONTECH) according to the manufacturer's instructions. We used 5'-CTTGTTCAGCTGCTCATATTCCTCAGCATCCACCCA-3' as a mutagenic primer and 5'-CTTCCTTTTTCGATATCATTGAAGCATT-3' as a selection primer.

Cell Culture and DNA Transfection

HeLa cells were provided by Dr. T. Kimura (Kansai Medical University) and cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. DNA transfection was performed by the standard calcium phosphate precipitation method. Cells were examined 2 days after transfection.

Immunoprecipitation and Immunoblotting

Immunoprecipitation and immunoblotting were performed as described previously (2). For immunoprecipitation, 7.5 µl of anti-Hrs antiserum (2), 2 µg of anti-HA epitope monoclonal antibody 12CA5 (Boehringer Mannheim), 2 µg of anti-phospholipase C-gamma (Santa Cruz), or 2 µg of anti-c-Met (Santa Cruz) was used. For immunoblotting, 100 times diluted anti-Hrs antiserum, 1 µg/ml of anti-phospholipase C-gamma , or 1 µg/ml of anti-c-Met was used.

Indirect Immunofluorescence Staining

Immunofluorescence staining was performed as described previously (10) except for the fixation conditions for staining with anti-Hrs. In this case, cells were fixed and permeabilized in 100% methanol for 5 min at -20 °C. The primary antibodies were the rabbit anti-Hrs polyclonal antiserum (1:2,000) (2), a rabbit anti-HA polyclonal antibody (5 µg/ml) (MBL), a mouse anti-human transferrin receptor monoclonal antiboy (10 µg/ml) (Oncogene Science), and a mouse anti-human CD63 monoclonal antibody (1:50) (Immunotech S.A.). The secondary antibodies were the Cy3-conjugated goat anti-rabbit immunoglobulin G antibody (1:1,000) (Amersham Corp.), rhodamine-conjugated anti-rabbit immunoglobulin G antibody (1:100) (Protos Immunoresearch), and fluorescein-conjugated anti-mouse immunoglobulin G antibody (1:100) (American Qualex Antibodies and Immunochemicals). Cells were examined by confocal immunofluorescence microscopy.

Immunoelectron Microscopy

Immunoelectron microscopic detection was performed as described by Mrini et al. (11) with a slight modification. Cells were transfected with or without the Hrs-HA cDNA on collagen-coated plastic sheets (Sumitomo Bakelite), fixed in 4% paraformaldehyde for 30 min, and treated with 1% sodium borohydride for 20 min. The cells were cryo-protected in 35% sucrose and 14% glycerol for 15 s, frozen in liquid nitrogen, and thawed. After incubation in 0.25% saponin, 5% bovine serum albumin, and 5% normal goat serum for 20 min, the cells were incubated with rabbit anti-HA epitope antibody (10 µg/ml) for 2 h and processed for the horseradish peroxidase-labeled streptavidin-biotin method using the LSAB kit (DAKO) according to the manufacturer's instructions, followed by incubation with 0.1% diaminobenzidine and 0.002% H2O2. The cells were post-fixed with 1% OsO4 for 30 min, incubated with 50% ethanol for 10 min, and block-stained with 2% uranyl acetate for 2 h. The cells were dehydrated with a graded series of ethanol and embedded in epoxy resin. Ultrathin sections were observed under an electron microscope.

Subcellular Fractionation

Near confluent cells in 90-mm dishes were homogenized by a Teflon-glass homogenizer in 1 ml of homogenizing buffer (10 mM Tris-HCl, pH 7.4, 1 mM phenylmethylsulfonyl fluoride, 2 µg/ml aprotinin, and 1.4 µg/ml pepstatin A), with or without 1 mM EDTA and 1 mM CaCl2, then centrifuged at 1,000 × g for 5 min at 4 °C. The cytoplasmic fraction was obtained from the post-nuclear supernatant by centrifugation at 105,000 × g for 1 h at 4 °C and collecting the supernatant. The pellet (membrane fraction) was solubilized by an incubation in 1 ml of homogenizing buffer containing 0.5% Nonidet P-40 for 1 h on ice and centrifuged at 105,000 × g for 1 h. The supernatant was collected as membrane proteins. The cytoplasmic and membrane proteins were examined by immunoprecipitation and immunoblotting.

Membrane Extraction

The membrane fraction, prepared as described above, was resuspended in 1 ml of 100 mM Na2CO3, pH 11.5, incubated for 30 min on ice, and centrifuged at 105,000 × g for 1 h at 4 °C. The supernatant was collected as the Na2CO3-extractable fraction and neutralized by HCl. Proteins in the resultant pellet were solubilized by Nonidet P-40 as described above. The Na2CO3-extractable fraction and the resultant Nonidet P-40-solubilized fraction, as well as the cytoplasmic fraction, were examined by immunoprecipitation and immunoblotting.


RESULTS

Homology between Hrs and Vps27p

The S. cerevisiae gene, VPS27, which is required for the normal morphology and function of a prevacuolar compartment has recently been identified (6). In VPS27-deficient cells, transport of endocytosed proteins and newly synthesized vacuolar proteins to vacuoles was disrupted, and these proteins accumulated in a prevacuolar compartment. The gene encoded a novel protein with a double zinc finger domain that is conserved in several proteins including Hrs, and the gene product was localized to the prevacuolar compartment. A further search of the homology between Vps27p and Hrs showed that they are 23% identical throughout the sequences (Fig. 1). This homology suggested that in mammalian cells, Hrs has a function similar to that of Vps27p in yeast. Thus, we examined the intracellular localization of Hrs.


Fig. 1. Alignment of the amino acid sequence of Hrs with that of Vps27p. Identical amino acid residues are shown by asterisks, and gaps in the sequences are shown by hyphens. The zinc finger domains are boxed, and the amino acid numbers are indicated on the right.
[View Larger Version of this Image (44K GIF file)]

Expression of Hrs and Transfection of HA-tagged Hrs in HeLa Cells

Expression of Hrs in HeLa cells was examined by immunoblotting. When the cells were lyzed, immunoprecipitated by anti-Hrs, and immunoblotted by the same antibody, a single 115-kDa band was detected (Fig. 2, human Hrs). It had the same apparent molecular mass as mouse Hrs immunoprecipitated from B16-F1 cells (Fig. 2, mouse Hrs). Immunoblotting with anti-phosphotyrosine antibody showed that Hrs was tyrosine phosphorylated in HeLa cells stimulated with HGF or epidermal growth factor (data not shown), suggesting the involvement of Hrs in the growth factor signaling pathway also in HeLa cells. Furthermore, to determine the intracellular localization site of exogenously expressed Hrs, we constructed a cDNA for Hrs that is tagged by the HA-epitope at the COOH terminus (Hrs-HA). Because the zinc finger domain of Hrs is homologous to those of several proteins that are involved in vesicular functions, we also constructed a zinc finger domain deletion mutant (Delta ZF) to determine whether the zinc finger domain is required for its intracellular localization. A region from Cys166 to Cys215 (Fig. 1) was deleted in the mutant. The Hrs-HA and Hrs-HADelta ZF cDNAs were inserted into the mammalian expression vector, pmiw, then transfected into HeLa cells. When the cells transfected with the Hrs-HA cDNA were lysed, immunoprecipitated by anti-HA antibody, and immunoblotted by anti-Hrs, a single band with a molecular mass of about 115 kDa was detected (Fig. 2, Hrs-HA). It was not detected in the immunoprecipitate of cells transfected with the expression vector alone (Fig. 2, control), indicating that the 115-kDa band was derived from the cDNA. It migrated slightly slower on a SDS gel than the endogenous mouse and human Hrs (Fig. 2). This is in accord with the fact that Hrs-HA is 15 amino acids larger than the endogenous protein. Hrs-HADelta ZF migrated faster than Hrs-HA by SDS-PAGE (Fig. 2, Hrs-HADelta ZF), because Hrs-HADelta ZF is 50 amino acids smaller than Hrs-HA.


Fig. 2. Immunoblot analysis of endogenous Hrs, Hrs-HA, and Hrs-HADelta ZF in HeLa cells. Mouse Hrs and human Hrs were immunoprecipitated by anti-Hrs antibody from B16-F1 and HeLa cells, respectively. Hrs-HA and Hrs-HADelta ZF were immunoprecipitated by anti-HA epitope antibody from HeLa cells transfected with the respective cDNAs. Immunoprecipitated proteins were separated by SDS-PAGE under reducing conditions and immunoblotted with anti-Hrs. Molecular mass standards are indicated in kDa.
[View Larger Version of this Image (71K GIF file)]

Indirect Immunofluorescence Staining of Hrs

To elucidate the intracellular localization site of Hrs, we performed indirect immunofluorescence staining of Hrs. When HeLa cells were fixed with methanol and stained by anti-Hrs, vesicular structures in the cytoplasmic and perinuclear regions were stained (Fig. 3, A and B). These structures were not stained by a preimmune serum or in the presence of excess of the bacterially expressed Hrs that was used as an immunogen to raise the antibody (data not shown). To identify these structures, the cells were double stained with anti-transferrin receptor or anti-CD63. The transferrin receptor is a marker of early endosomes and plasma membrane (12, 13), whereas CD63 is that of late endosomes and early lysosomes (14). Most of the vesicles that were stained with anti-Hrs were also stained with anti-transferrin receptor, although some of the small vesicles that were positive for transferrin receptor were not stained with anti-Hrs (Fig. 3, A' and A"). In addition, the plasma membrane, which was stained with anti-transferrin receptor, was not stained with anti-Hrs (Fig. 3, A, A', and A"). In contrast, most of the Hrs-positive vesicles were not stained with anti-CD63 (Fig. 3, B, B', and B"). These results indicated that Hrs is localized to early endosomes.


Fig. 3. Immunofluorescence staining of endogenous Hrs in HeLa cells. HeLa cells were fixed and incubated with rabbit polyclonal anti-Hrs (A and B), mouse anti-transferrin receptor (A'), or mouse anti-CD63 (B') antibody followed by incubation with Cy3-conjugated anti-rabbit immunoglobulin G antibody (A and B) or fluorescein-conjugated anti-mouse immunoglobulin G antibody (A' and B'). Overlays of A and A' and B and B' are shown in A" and B", respectively. The yellow stain indicates colocalization. Bar, 10 µm.
[View Larger Version of this Image (129K GIF file)]

Indirect Immunofluorescence Staining of Hrs-HA

To confirm the early endosomal localization of Hrs by examining the localization of exogenously expressed Hrs, HeLa cells were transfected with the Hrs-HA cDNA, fixed with paraformaldehyde, and stained with anti-HA. Similar to the staining with anti-Hrs, anti-HA also detected vacuolar or vesicular structures in transfected cells (Fig. 4, A and B). In addition, Hrs-HA was completely colocalized with the transferrin receptor (Fig. 4, A' and A") but only partially with CD63 (Fig. 4, B' and B"), further confirming that Hrs is localized to early endosomes. In this case, the plasma membrane was not stained with anti-transferrin receptor (Fig. 4A'), as observed in Fig. 3A'. This may be due to the difference of the conditions to fix cells. It should be noted that the early endosomes in Hrs-HA-transfected cells were larger than those in untransfected cells (Fig. 4A', arrow and arrowheads).


Fig. 4. Immunofluorescence staining of Hrs-HA and Hrs-HADelta ZF in transfected cells. HeLa cells expressing Hrs-HA (A, A', A", B, B', and B") and Hrs-HADelta ZF (C, C', and C") were fixed, permeabilized, and incubated with rabbit anti-HA epitope (A, B, and C), mouse anti-transferrin receptor (A' and C'), or mouse anti-CD63 (B') antibody followed by incubation with rhodamine-conjugated anti-rabbit immunoglobulin G antibody (A, B, and C) or fluorescein-conjugated anti-mouse immunoglobulin G antibody (A', B', and C'). Overlays of A and A'; B and B'; and C and C' are shown in A", B", and C", respectively. The yellow stain indicates colocalization. The arrows and arrowheads in A' and C' indicate the transfected cells and untransfected cells, respectively. Bar, 10 µm.
[View Larger Version of this Image (134K GIF file)]

It was possible that the structurally conserved double zinc finger domains in several proteins involved in vesicular function play a role in their intracellular localization. However, Hrs-HADelta ZF was also colocalized with the transferrin receptor (Fig. 4, C, C', and C"), suggesting that the zinc finger domain is not required for the correct localization of Hrs.

Immunoelectron Microscopic Detection of Hrs-HA

To investigate the more detailed localization of Hrs in early endosomes, we detected Hrs-HA in transfected cells by means of immunoelectron microscopy. Unfortunately, we could not use anti-Hrs for immunoelectron microscopic detection of Hrs, because this antibody could detect Hrs only in cells fixed with methanol and methanol-fixed cells cannot be adapted for immunoelectron microscopy. HeLa cells transfected with the Hrs-HA cDNA were fixed, permeabilized, and incubated with rabbit anti-HA epitope antibody. The cells were then successively incubated with biotin-labeled anti-rabbit immunoglobulin G antibody and horseradish peroxidase-conjugated avidin, and localization of Hrs was visualized using diaminobenzidine and H2O2. Consistent with the results of immunofluorescence staining shown in Fig. 4, vesicular structures were again stained (Fig. 5A, arrows). No dense background staining was detected in untransfected cells (Fig. 5A, cells without arrows). When the ultrathin sections of the Hrs-HA-expressing cells were examined by an electron microscope, Hrs-HA was detected on the cytoplasmic surface of the vesicular structures (Fig. 5B). The size of these structures (roughly 0.2-0.5 µm in diameter) was similar to that of early endosomes (15). This immunoreactivity was not detected on vesicular structures of untransfected cells (Fig. 5A and not shown), indicating that this staining is specific for transfected Hrs. Together with the results of immunofluorescence staining, this finding indicated that Hrs is localized to the cytoplasmic surface of early endosomes. The Hrs-HA-positive endosomes in transfected cells often aggregated with each other, which was not detected in untransfected cells (Fig. 5B and not shown).


Fig. 5. Immunoelectron microscopic localization of Hrs-HA. A, HeLa cells transfected with the Hrs-HA cDNA were fixed, permeabilized, and incubated with rabbit anti-HA epitope antibody. The cells were then successively incubated with biotin-labeled anti-rabbit immunoglobulin G antibody, with horseradish peroxidase-conjugated avidin, and with diaminobenzidine and H2O2. The arrows indicate the cells expressing Hrs-HA. Bar, 10 µm. B, ultrathin sections of the cells that express Hrs-HA were examined using an electron microscope. The arrowheads indicate the vesicular structures of which cytoplasmic surface is stained by anti-HA epitope antibody. N, nucleus; G, Golgi apparatus; M, mitochondria. Bar, 1 µm.
[View Larger Version of this Image (140K GIF file)]

Subcellular Fractionation

In the case of Vps27p, it has been shown by subcellular fractionation that about 60% of the protein in whole yeast cell lysates sediments with the membrane fraction only in the presence of divalent cations (6). We therefore performed the subcellular fractionation of endogenous Hrs, Hrs-HA, and Hrs-HADelta ZF in untransfected or transfected HeLa cells in the presence of EDTA or Ca2+. The cells were homogenized in the presence or the absence of 1 mM EDTA and 1 mM Ca2+, and then the post-nuclear supernatants were separated into cytoplasmic and membrane fractions. When endogenous Hrs was immunoprecipitated from each fraction by anti-Hrs and immunoblotted by the same antibody, the protein was detected both in the cytoplasmic and membrane fractions irrespective of the presence of EDTA or Ca2+ (Fig. 6A, top panel). The amount of Hrs that was recovered in the cytoplasmic fraction was severalfold more than that recovered in the membrane fraction. When Hrs-HA and Hrs-HADelta ZF were fractionated from transfected cells, these proteins were equally detected in both of the fractions (Fig. 6A, second and third panels). In contrast, the cytoplasmic protein, phospholipase C-gamma , and the transmembrane protein, c-Met receptor, were recovered in the cytoplasmic and membrane fractions, respectively, under all conditions, verifying the integrity of the fractions (Fig. 6A, fourth and bottom panels).


Fig. 6. Subcellular fractionation of endogenous Hrs, Hrs-HA, and Hrs-HADelta ZF. A, cells were homogenized in the presence or the absence of EDTA and Ca2+, and then post-nuclear supernatants were separated into cytoplasmic (C) and membrane (M) fractions. Cytoplasmic proteins and proteins eluted from the membrane fractions by Nonidet P-40 were analyzed. Hrs, phospholipase C-gamma (PLCgamma ), and c-Met were immunoprecipitated from untransfected HeLa cells by anti-Hrs, anti-phospholipase C-gamma , and anti-c-Met, respectively, and then immunoblotted with the respective antibodies. Hrs-HA and Hrs-HADelta ZF were immunoprecipitated by anti-HA epitope antibody from HeLa cells transfected with the respective cDNAs and immunoblotted with anti-Hrs. B, the cytoplasmic (C) and membrane fractions of untransfected HeLa cells or cells transfected with the Hrs-HA or Hrs-HADelta ZF cDNA were prepared as described in A. The membrane fractions were extracted with 100 mM Na2CO3 and separated into supernatants (M-1) and pellets. Proteins in the pellets were eluted with Nonidet P-40 (M-2). Endogenous Hrs, Hrs-HA, and Hrs-HADelta ZF were detected by immunoprecipitation and immunoblotting as described in A.
[View Larger Version of this Image (35K GIF file)]

To characterize the nature of the association of Hrs, Hrs-HA, and Hrs-HADelta ZF, with the membrane, we treated the membrane fractions with 100 mM Na2CO3, pH 11.5, which extracts peripherally associated membrane proteins from membranes (16). As shown in Fig. 6B, the membrane-associated portions of the three proteins were extracted into the supernatants of the Na2CO3-treated membranes (M-1) but were undetectable in the treated membranes (M-2), suggesting the peripheral association of Hrs with early endosomes.

These results suggest that the membrane association of Hrs, which was observed by immunofluorescence and immunoelectron microscopy, was not so tight and the proteins easily dissociated from the membrane during fractionation. This was in agreement with the fact that Hrs has no hydrophobic region that may act as a membrane spanning domain (2). A similar type of peripheral membrane association has also been reported for EEA1, a hydrophilic early endosomal protein with a zinc finger domain highly homologous to that of Hrs (4). In addition, the membrane association of Hrs depended on neither Ca2+ nor its zinc finger domain.


DISCUSSION

In this study, we found that Hrs is localized to early endosomes in HeLa cells. In addition, the amino acid sequence of Hrs was 23% identical to that of Vps27p throughout the sequences. VPS27 is one of the genes whose disruption in S. cerevisiae interrupts traffic of endocytosed proteins and newly synthesized vacuolar proteins to vacuoles (a counterpart of mammalian lysosomes) and causes the accumulation of these proteins in a prevacuolar compartment termed the "class E" compartment (6), which is considered to be a counterpart of mammalian endosomes (17). Moreover, Vps27p is localized to the class E compartment (6). Considering from these observations, it is possible that Hrs plays a role in mammalian cells that is similar to that of Vps27p in yeast. Namely, Hrs may be involved in protein traffic through early endosomes. Although it is not known whether Hrs is a mammalian homolog of Vps27p, it is of interest to examine whether Hrs complements the phenotype of the VPS27 mutation in yeast.

The zinc finger domain of Hrs is highly homologous not only to that of Vps27p but also to those of EEA1 (4), Fab1p (5), and Vac1p (7), which are all involved in the function of endosomes or vacuoles. Thus, the zinc finger domains in these proteins must be important for their function. However, the zinc finger domain of Hrs was not required for its correct localization to early endosomes. It is reported that mutations in the zinc finger domain of Vps27p also do not affect its membrane association, although these mutant proteins do not complement the phenotypes of the VPS27 mutation, indicating that the zinc finger domain is essential for its function but not for its membrane localization (6). In contrast to these observation, Stenmark et al. (18) recently reported that the zinc finger domain of EEA1 is essential for its early endosomal localization and membrane association. In the case of Fab1p, it is known that the zinc finger domain is not required to complement the phenotypes of FAB1 mutation such as enlargement of vacuoles and the formation of aploid and binucleate cells (5). Therefore, further investigation is necessary to clarify the function of the zinc finger domain in each of these proteins.

Immunoelectron microscopy showed that in cells expressing Hrs-HA, early endosomes often aggregated with each other. This may be the reason why Hrs-HA-positive endosomes that were observed by immunofluorescence staining looked exaggerated. In addition, the ratio of the amounts of Hrs-HA and Hrs-HADelta ZF that were subcellularly fractionated into the membrane fractions was higher than that of endogenous Hrs. This may also be due to the aggregation of early endosomes in cells expressing Hrs-HA or Hrs-HADelta ZF. It remains unknown whether it is a functional consequence of overexpression of Hrs or simply an artificial structure caused by overexpression irrespective of its function. However, the latter case is likely because it was also observed in cells expressing Hrs lacking the structurally conserved double zinc finger domain.

After binding of growth factors, the growth factor-receptor tyrosine kinase complexes are internalized and targeted to lysosomes through early endosomes and late endosomes (19, 20). In this pathway, they are delivered to early endosomes within a few minutes after growth factor binding and receptor tyrosine kinases remain active after delivery to early endosomes (21-25). Considering that the tyrosine phosphorylation of Hrs is induced in a few minutes after exposing cells to growth factors (2) and that Hrs is localized on the cytoplasmic surface of early endosomes, it is most likely that Hrs is tyrosine phosphorylated on early endosomes by internalized receptor tyrosine kinases that are exposed on the cytoplasmic surface of early endosomes (Fig. 7). Taken together, we propose the following hypothesis for the function of Hrs (Fig. 7). The internalized receptor tyrosine kinase phosphorylates Hrs on early endosomes. Tyrosine phosphorylation activates Hrs and, as speculated from the function of Vps27p in yeast, it stimulates vesicular transport of the growth factor-receptor complex from early endosomes to late endosomes and lysosomes for degradation.


Fig. 7. Hypothetical model for the function of Hrs. The growth factor-bound receptor tyrosine kinase is internalized and delivered to the early endosome, where it phosphorylates Hrs residing on the cytoplasmic surface of the early endosome. Tyrosine phosphorylation activates Hrs, and the activated Hrs stimulates the vesicle transport of the growth factor-receptor complex to the late endosome and lysosome.
[View Larger Version of this Image (16K GIF file)]

Tyrosine kinase activity of growth factor receptors is required for the internalization and/or degradation of growth factor-receptor complexes (26). However, the mechanism of the process is not well understood. Several proteins are considered to be implicated in the process. A platelet-derived growth factor receptor mutant that cannot associate with phosphatidylinositol 3-kinase is not endocytosed into and not degraded in cells, suggesting a role for phosphatidylinositol 3-kinase in the down-regulation of the receptor (27). Annexin I and annexin II may also be involved in the process, because they are localized to the endosomal compartments, promote vesicle fusion, and are phosphorylated by receptor tyrosine kinases (28, 29). Benmerah et al. (30) and Okabayashi et al. (31) have reported that the receptor tyrosine kinase substrates, eps15 and Shc, are constitutively associated with the plasma membrane adaptor, AP-2. Because AP-2 participates in the recruitment of growth factor-receptor complexes into clathrin-coated pits, they proposed roles for eps15 and Shc in endocytosis of the complexes. In this study, we propose that Hrs can be one of the factors that regulate this process, especially after delivering the complexes to early endosomes.

Very recently, it was reported that a splice variant of Hrs, which has additional 150 amino acid residues at its COOH terminus, 1) interacts with SNAP-25, 2) hydrolyse ATP, and 3) inhibits Ca2+-triggered noradrenaline release from permeabilized PC12 cells, implicating that the protein is involved in secretion of neurotransmitters through Ca2+- and ATP-dependent modulation of vesicle-trafficking protein complexes (32). Therefore, Hrs may be a regulator of several kinds of vesicular functions (such as endocytosis and exocytosis) in a variety of cells.


FOOTNOTES

*   This work was supported in part by research grants from the Ministry of Education, Science, Sports and Culture of Japan.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§   Present address: Division of Basic Sciences, Fred Hutchinson Cancer Research Center, 1100 Fairview Ave. N., Seattle, WA 98109.
**   To whom correspondence should be addressed. Tel.: 81-45-924-5701; Fax: 81-45-924-5771.
1   The abbreviations used are: HGF, hepatocyte growth factor; HA, hemagglutinin; PAGE, polyacrylamide gel electrophoresis.

ACKNOWLEDGEMENTS

We thank I. Kobayashi and K. Yasaka for excellent technical assistance and T. Kimura for providing HeLa cells and helpful advice.


REFERENCES

  1. Komada, M., and Kitamura, N. (1994) J. Biol. Chem. 269, 16131-16136 [Abstract/Free Full Text]
  2. Komada, M., and Kitamura, N. (1995) Mol. Cell. Biol. 15, 6213-6221 [Abstract]
  3. Pasteris, N. G., Cadle, A., Logie, L. J., Porteous, M. E. M., Schwartz, C. E., Stevenson, R. E., Glover, T. W., Wilroy, R. S., and Gorski, J. L. (1994) Cell 79, 669-678 [Medline] [Order article via Infotrieve]
  4. Mu, F.-T., Callaghan, J. M., Steele-Mortimer, O., Stenmark, H., Parton, R. G., Campbell, P. L., McCluskey, J., Yeo, J.-P., Tock, E. P. C., and Toh, B.-H. (1995) J. Biol. Chem. 270, 13503-13511 [Abstract/Free Full Text]
  5. Yamamoto, A., DeWald, D. B., Boronenkov, I. V., Anderson, R. A., Emr, S. D., and Koshland, D. (1995) Mol. Biol. Cell 6, 525-539 [Abstract]
  6. Piper, R. C., Cooper, A. A., Yang, H., and Stevens, T. H. (1995) J. Cell Biol. 131, 603-617 [Abstract]
  7. Weisman, L. S., and Wickner, W. (1992) J. Biol. Chem. 267, 618-623 [Abstract/Free Full Text]
  8. Hirai, H., Suzuki, T., Fujisawa, J., Inoue, J., and Yoshida, M. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 3584-3588 [Abstract]
  9. Suemori, H., Kadodawa, Y., Goto, K., Araki, I., Kondoh, H., and Nakatsuji, N. (1990) Cell Differ. Dev. 29, 181-186 [CrossRef][Medline] [Order article via Infotrieve]
  10. Masaki, R., Yamamoto, A., and Tashiro, Y. (1994) J. Cell Biol. 126, 1407-1420 [Abstract]
  11. Mrini, A., Moukhles, H., Jacomy, H., Bosler, O., and Doucet, G. (1995) J. Histochem. Cytochem. 43, 1285-1291 [Abstract/Free Full Text]
  12. Hopkins, C. R. (1983) Cell 35, 321-330 [Medline] [Order article via Infotrieve]
  13. Hopkins, C. R., and Trowbridge, I. S. (1983) J. Cell Biol. 97, 508-521 [Abstract]
  14. Metzelaar, M. J., Wijngaard, P. L. J., Peters, P. J., Sixma, J. J., Nieuwenhuis, H. K., and Clevers, H. C. (1991) J. Biol. Chem. 266, 3239-3245 [Abstract/Free Full Text]
  15. Griffiths, G., Back, R., and Marsh, M. (1989) J. Cell Biol. 109, 2703-2720 [Abstract]
  16. Fujiki, Y., Hubbard, A. L., Fowler, S., and Lazarow, P. B. (1982) J. Cell Biol. 93, 97-102 [Abstract]
  17. Raymond, C. K., Howald-Stevenson, I., Vater, C. A., and Stevens, T. H. (1992) Mol. Biol. Cell 3, 1389-1402 [Abstract]
  18. Stenmark, H., Aasland, R., Toh, B.-H., and D'Arrigo, A. (1996) J. Biol. Chem. 271, 24048-24054 [Abstract/Free Full Text]
  19. Beguinot, L., Lyall, R. M., Willingham, M. C., and Pastan, I. (1984) Proc. Natl. Acad. Sci. U. S. A. 81, 2384-2388 [Abstract]
  20. Gruenberg, J., and Howell, K. E. (1989) Annu. Rev. Cell Biol. 5, 453-481 [CrossRef]
  21. Carpentier, J.-L., White, M. F., Orci, L., and Kahn, R. C. (1987) J. Cell Biol. 105, 2751-2762 [Abstract]
  22. Lai, W. H., Cameron, P. H., Doherty, J.-J., II, Posner, B. I., and Bergeron, J. J. M. (1989) J. Cell Biol. 109, 2751-2760 [Abstract]
  23. Sorkin, A., Eriksson, A., Heldin, C.-H., Westermark, B., and Claesson-Welsh, L. (1993) J. Cell. Physiol. 156, 373-382 [Medline] [Order article via Infotrieve]
  24. Burgess, J. W., Wada, I., Ling, N., Khan, M. N., Bergeron, J. J. M., and Posner, B. I. (1992) J. Biol. Chem. 267, 10077-10086 [Abstract/Free Full Text]
  25. Baass, P. C., DiGuglielmo, G. M., Authier, F., Posner, B. I., and Bergeron, J. J. M. (1995) Trends Cell Biol. 5, 465-470 [CrossRef]
  26. Ullrich, A., and Schlessinger, J. (1990) Cell 61, 203-212 [Medline] [Order article via Infotrieve]
  27. Joly, M., Kazlauskas, A., Fay, F. S., and Corvera, S. (1994) Science 263, 684-687 [Medline] [Order article via Infotrieve]
  28. Futter, C. E., Felder, S., Schlessinger, J., Ullrich, A., and Hopkins, C. R. (1993) J. Cell Biol. 120, 77-83 [Abstract]
  29. Biener, Y., Feinstein, R., Mayak, M., Kaburagi, Y., Kadowaki, T., and Zick, Y. (1996) J. Biol. Chem. 271, 29489-29496 [Abstract/Free Full Text]
  30. Benmerah, A., Gagnon, J., Bègue, B., Mégarbané, B., Dautry-Varsat, A., and Cerf-Bensussan, N. (1995) J. Cell Biol. 131, 1831-1838 [Abstract]
  31. Okabayashi, Y., Sugimoto, Y., Totty, N. F., Hsuan, J., Kido, Y., Sakaguchi, K., Gout, I., Waterfield, M. D., and Kasuga, M. (1996) J. Biol. Chem. 271, 5265-5269 [Abstract/Free Full Text]
  32. Bean, A. J., Seifert, R., Chen, Y. A., Sacks, R., and Scheller, R. H. (1997) Nature 385, 826-829 [CrossRef][Medline] [Order article via Infotrieve]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.