|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
J. Biol. Chem., Vol. 279, Issue 42, 43374-43377, October 15, 2004
Small Heat Shock Protein
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ABSTRACT |
|---|
|
|
|---|
B-Crystallin is a developmentally regulated small heat shock protein known for its binding to a variety of denatured polypeptides and suppression of protein aggregation in vitro. Elevated levels of
B-crystallin are known to be associated with a number of neurodegenerative pathologies such as Alzheimer disease and multiple sclerosis. Mutations in
B-crystallin gene have been linked to desminrelated cardiomyopathy and cataractogenesis. The physiological function of this protein, however, is unknown. Using discontinuous sucrose density gradient fractionation of post-nuclear supernatants, prepared from rat tissues and human glioblastoma cell line U373MG, we have identified discrete membrane-bound fractions of
B-crystallin, which co-sediment with the Golgi matrix protein, GM130. Confocal microscopy reveals co-localization of
B-crystallin with BODIPY TR ceramide and the Golgi matrix protein, GM130, in the perinuclear Golgi in human glioblastoma U373MG cells. Examination of synchronized cultures indicated that
B-crystallin follows disassembly of the Golgi at prometaphase and its reassembly at the completion of cytokinesis, suggesting that this small heat shock protein, with its chaperone-like activity, may have an important role in the Golgi reorganization during cell division. | INTRODUCTION |
|---|
|
|
|---|
B-Crystallin is a member of the small heat shock family of proteins whose expression accompanies a number of developmental programs and pathologies in various tissues (1, 2). Mutations in
B-crystallin lead to cataractogenesis (3) and desmin-related cardiomyopathy (4). The appearance of
B-crystallin in a temporally and spatially controlled fashion during development has lead to speculations that this small heat shock protein may have important roles in the regulation of cellular physiology and growth (59). Quite remarkably, the elevated presence of
B-crystallin has been linked with a number of neurodegenerative disorders such as Alexander and Alzheimer diseases, dementia, and scrapie (1, 2). In multiple sclerosis the presence of
B-crystallin in astrocytes has been directly implicated in the pathological immune response (10).
Confluent cultures of the human glioblastoma cell line, U373MG, express appreciable amounts of
B-crystallin (about 24 µg/mg soluble protein) (11), thus presenting an excellent paradigm for investigating the physiological function of this protein.
B-Crystallin is predominantly a cytoplasmic protein, but it has also been shown to be present in the nucleus (7, 9). Considering that its chaperone-like activity has been demonstrated in vitro with a large number of substrates, under a variety of denaturing environments (12, 13), we wanted to ascertain whether
B-crystallin showed any preferential association with, or within, a cellular compartment in vivo. Rat tissues and the human glioblastoma cell line U373MG, not exposed to any specific physical or chemical stress, were used for these studies.
| MATERIALS AND METHODS |
|---|
|
|
|---|
AntibodiesAn antiserum (anti-
B) against the conserved C terminus of
B-crystallin (cREEKPAVTAAPKK, the small case "c" in the sequence is for cysteine; it was added for coupling to keyhole limpet hemocyanin; there is no cysteine in this sequence (15)) was raised commercially in rabbits (Sigma-Genosys, The Woodlands, TX). This antiserum is as specific to
B-crystallin as the one made previously against the same peptide (15). Antibody GM130:FITC (anti-GM130) was a monoclonal antibody raised in mouse (BD Transduction Laboratories, Lexington, KY). For immunofluorescence, one of the following two secondary antibodies was used: 1) goat anti-rabbit TRITC or 2) goat anti-rabbit FITC (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). For immunoblotting one of the following secondary antibodies was used: goat-anti-rabbit-horseradish peroxidase (Pierce) and goat-anti-mouse-horseradish peroxidase (Santa Cruz Biotechnology).
ImmunofluorescenceU373MG cells seeded on a microscope coverslip (Fisher Scientific) were fixed with methanol (20 °C) for 6 min, washed, and incubated with the blocking buffer (5% normal goat serum, 1.0% bovine serum albumin, 0.20% Triton X-100 in phosphate-buffered saline) for 35 min. Subsequently, the cells were incubated with primary antibody (1:50 dilution) and then with the secondary antibody (1:200) following a standard protocol. The cells were stained in the dark with BODIPY TR ceramide (N-((4-(4, 4-difluoro-5-(2-thienyl)-4-bora-3a,4a-diaza-s-indacene-3-yl)phenoxy)acetyl)sphingosine), a selective stain for Golgi membranes (Molecular Probes Inc.) (16), and then counterstained with DAPI (4',6-diamidino-2-phenylindole) and mounted on slides using Prolong anti-fade mounting medium (Molecular Probes Inc.). For GM130 localization studies, instead of BODIPY, a second primary antibody (anti-GM130:FITC at 1:50 dilution) was used followed by DAPI. The pictures were taken using a Leica TCS-SP Multiphoton and Confocal Microscope with 100x oil immersion objective lenses (1.4). Z sections 0.5 µm thick at a 23-µm depth were taken and processed using Adobe Photoshop 6.0.
Cell FractionationThe post-nuclear supernatants were fractionated on discontinuous sucrose gradients (17, 18). The method was marginally modified for use with smaller volumes and tubes for centrifugation in a Sorvall Discovery M150 centrifuge with a swing out rotor (S55S) at 39,000 rpm for 30 min. These gradients were made in a 2.2-ml tube by sequential addition of the following solutions: 400 µl of buffer E (1.3 M sucrose), 700 µl of buffer D (0.86 M sucrose), 700 µl of post-nuclear supernatant (homogenate made in buffer C containing 0.5 M sucrose), and finally the tube was topped with buffer B (0.25 M sucrose). To further characterize the Golgi fraction obtained in this procedure, we followed the procedure described by Taylor et al. (18), which employs two centrifugations on two similar but non-identical gradients. The first centrifugation, done according to Slusarewicz et al. (17) provides the Golgi fraction, S2 (see Fig. 2). The second discontinuous sucrose gradient was made as follows. The Golgi fraction (S2) was collected, adjusted to 1.15 M with 2 M sucrose (total volume about 500 µl), placed at the bottom of a new tube, and filled successively with 500 µl each of 1.0 M, 0.86 M, and 0.25 M sucrose solutions. After centrifugation at 76,000 x g for 90 min stacked Golgi fractions 13 (SGF13) (18) were collected and analyzed by immunoblotting. All buffers contained 10 mM potassium phosphate and 5 mM MgCl2, pH 6.7. Different membrane fractions were collected by piercing the side of the centrifuge tube with a 22-mm gauge needle connected to a syringe and labeled as S1, S2, S3, and SGF13 (18).
|
ImmunoblottingSucrose gradient fractions were electrophoresed on SDS-PAGE (412% gradient NuPAGE bis-tris) gels and transferred to nitrocellulose membranes as per the manufacturer's instructions (Invitrogen). Immunoblotting was done using the SuperSignal West Dura protocol for chemiluminescence (Pierce). The same blots were used again, sometimes without stripping to immunodetect GM130.
| RESULTS |
|---|
|
|
|---|
B-CrystallinPost-nuclear homogenates from U373MG cells were fractionated on a discontinuous sucrose gradient (shown schematically in Fig. 2, left panel). This gradient produced two easily visible membrane fractions (whitish bands), one near the bottom of the gradient at the interface of 1.3 and 0.86 M sucrose, fraction S3, and the other around the middle of the gradient at the interface of 0.86 and 0.50 M sucrose, fraction S2 (Fig. 1A, top panel). The band around the middle of the gradient (fraction S2, Fig. 1A, top panel) contains the Golgi (17). The fractions from this gradient were collected and immunoblotted for determining the distribution of
B-crystallin. Interestingly, while most of the
B-crystallin is seen in the top fractions of the gradient, a small but significant portion of
B-crystallin is seen associated with the membrane fraction S3 (Fig. 1A, top panel).
|
B-crystallin, but the pattern of its presence is continuous with the rest of the soluble
B-crystallin in the top of the gradient (Fig. 1A, top panel). To gain insight into the identity of the two membrane fractions, S2 and S3, the same immunoblot was screened with anti-GM130 antibody for GM130, a well known Golgi matrix polypeptide (19) (Fig. 1A, middle panel). These data show that GM130 protein is concentrated in the same fractions (46) where
B-crystallin is found. GM130 is also present in fractions 13 and 14 (Fig. 1A, middle panel, asterisks) (although at much reduced levels), which includes the Golgi membrane fraction, S2 (Fig. 1A, top panel). Although
B-crystallin pattern in the S2 fraction does not show a discrete separation from the soluble pool, the distinct pattern of the presence of GM130 (higher concentration of GM130 in fractions 13 and 14 in comparison with surrounding gradient fractions) clearly indicates that they contain the Golgi membranes. By the same token, the S3 fraction that contains a high concentration of GM130 protein must also contain Golgi membranes (see below).
The data obtained in Fig. 1A (top panel) is confirmed by the pattern of
B-crystallin and GM130 presence obtained with the fractionation of the post-nuclear supernatants prepared from post-natal day 10 rat heart (Fig. 1B). The pattern in Fig. 1B is very similar to that shown in Fig. 1A; there is however a clear concentration of
B-crystallin in two fractions (Fig. 1B, brackets), one nearer the bottom of the gradient, fraction S3 (similar to that in Fig. 1A), and the other in the middle, fraction S2 (the Golgi fraction). The Golgi fraction (S2) is much higher in concentration in comparison with the lower membrane (S3 fraction) as assessed by the relative presence of
B-crystallin and GM130 (Fig. 1B, compare anti-
B panel with anti-GM130 panel). The difference in the two patterns (between U373MG and rat heart) is probably related to higher lability of Golgi (S2) fraction that degrades to S3 fraction in this preparation of U373MG gradients. This possibility was addressed in another experiment with these cells when different membrane fractions (S1S3) were collected directly, and the Golgi fraction (fraction S2) was refractionated on a second sucrose gradient (Fig. 2).
Fig. 2 shows that
B-crystallin is mostly in the fraction S2 in post-nuclear homogenates (Fig. 2, left panel). Upon refractionation of S2 (Fig. 2, right panel) most of the
B-crystallin is found in the fraction SGF3, the fraction comparable with S3 in the first gradient and Fig. 1A. Based on these data we conclude that the membrane fraction S3 (Fig. 1A) contains membranes and proteins derived from the Golgi fraction (S2, Fig. 1A). These data also confirm the lability of the Golgi membrane fraction (18). Extreme care is needed to keep the integrity of the Golgi fraction intact. The lability of the Golgi fraction isolated from the U373 MG cells (fraction S2) helped us identify Golgi as the origin of what fractionates as a discrete membrane fraction (S3) in heavier sucrose (Figs. 1 and 2). In summary the data presented in Fig. 1 indicate the presence of Golgi membrane-associated
B-crystallin in fractions S2 and S3.
B-Crystallin Localizes to the Perinuclear GolgiThe astrocytoma cell line, U373MG, was synchronized with a double thymidine block (14), and the G1/S phase cells were immunostained with anti-
B. Prominent labeling of the Golgi apparatus resident in the perinuclear area is seen in these cells (Fig. 3A). The anti-
B staining co-localizes perfectly (Fig. 3A, Control-MeOH panel) with BODIPY TR ceramide (BODIPY), which preferentially stains the Golgi membrane. Fig. 3A (BFA-MeOH panel) shows that brefeldin A (a potent antifungal macrocyclic lactone known to cause disassembly of the Golgi (20)) disperses the staining of the Golgi by anti-
B and its co-localization with BODIPY TR. When post-nuclear extracts from brefeldin A-treated cultures were analyzed on discontinuous sucrose gradients, no
B-crystallin was found in the S2 membrane (Golgi) fraction (Fig. 3B, arrows) further augmenting the observation of the presence of
B-crystallin in the Golgi. The co-localization of BODIPY and
B-crystallin and its disruption by brefeldin A treatment (Fig. 3A, lower panel), confirms the presence of
B-crystallin in the Golgi.
|
B-Crystallin Follows Assembly and Disassembly of the Golgi during Cell CycleGolgi biogenesis (or inheritance) is closely associated with its disassembly during mitosis and its re-assembly at the completion of cell division, which allows its equitable inheritance between the two daughter cells (21). It was of interest, therefore, to follow the presence of
B-crystallin in the Golgi as a function of its assembly and disassembly, as the cells go through the cell cycle. Golgi membrane components as well as the Golgi protein(s) independently (19) have been speculated to be the unit of inheritance. We, therefore, used the Golgi protein marker GM130 (Fig. 4A) and the Golgi membrane lipid marker, BODIPY TR (Fig. 4B), to guide us in the localization of
B-crystallin in U373MG cells as a function of the cell cycle. The co-localization of
B-crystallin and GM130 is evident in the perinuclear Golgi at the G1/S phase boundary (Fig. 4A, see Merged panel). It diffuses into the surrounding prometaphase and the metaphase cytoplasm as would be expected of disassembling Golgi components (similar data were obtained when the Golgi-specific stain, BODIPY, was used (data not shown)). The staining gradually reappears, first in the emergent perinuclear area of the two daughter nuclei (Fig. 4A, telophase) and then clearly during cytokinesis (Fig. 4, A and B).
|
| DISCUSSION |
|---|
|
|
|---|
B-crystallin has been very well established in vitro, but the physiological function of this protein remains conjectural (12, 13). This protein is considered to be a cytoplasmic soluble protein, yet it has been known for a very long time that
-crystallins associate with cellular membranes (2224). In this paper we demonstrate that a discrete fraction of
B-crystallin is associated with the Golgi membrane complex (Figs. 1 and 2).
An interesting aspect of the Golgi apparatus in mammalian cells is that it exists as a stack of flattened membranes positioned in a perinuclear location in the cell. The data obtained by immunostaining and confocal microscopy (Fig. 3) clearly established that
B-crystallin associates with Golgi membranes. Although
B-crystallin is also known to be present in the nucleus (7, 9), this is the first time that this protein has been reported to be associated with a specific cytoplasmic compartment such as the Golgi. The chaperone-like activity of
B-crystallin leads one to consider its role in well established functions of the Golgi complex as a major site of post-translational protein processing and sorting station for secreted proteins (25). In this connection the presence of
B-crystallin in the Golgi in astrocytes may be related to its implication in multiple sclerosis (10) and reactive gliosis and dementia (26).
A close examination of the mitotic panels in Fig. 4A reveals that staining obtained with either antibody (anti-
B or anti-GM130) is not amorphous but granular. It is particularly interesting that both bright red (
B-crystallin) as well as green granules (GM130) are seen around the newly formed nuclei (Fig. 4A, telophase). The presence of large granular co-localized
B-crystallin and GM130 is however unmistakable in the perinuclear area of the newly formed nuclei during cytokinesis (Fig. 4A, last panel, arrows). The timing of the
B-crystallin appearance in the nascent Golgi, immediately following cytokinesis, is noteworthy and suggestive of a role in the biogenesis of the Golgi in the daughter cells. This speculation is supported by the fact that Golgi reorganization (or Golgi inheritance) is closely interlinked with mitosis checkpoints; inhibition of this process is known to inhibit mitosis (27). Interestingly, lens epithelial cell cultures derived from
B-crystallin null mice show an increase in the proportion of polyploid cells (8, 28); this may be ascribed to the inability of the these cells to proceed unhindered through the cell cycle because of the absence of
B-crystallin.
The molecular mechanism of Golgi fragmentation and the molecular basis of its inheritance are a focus of intense and unresolved investigations (21, 29, 30). It is tempting to purport a role for
B-crystallin and other small heat shock proteins in Golgi inheritance by suggesting that the chaperone-like function of this protein protects and holds the Golgi components in the mitotic cytoplasm during cell division and then helps deliver these components to the appropriate site(s) in the daughter cells to template the biogenesis of the new Golgi. Future investigations should reveal the substrates for the
B-crystallin chaperone-like activity in the Golgi and during its mitotic transition.
| FOOTNOTES |
|---|
|| A Research to Prevent Blindness Inc., Wasserman Merit Scholar. To whom correspondence should be addressed: Vision Molecular Biology Laboratory, Jules Stein Eye Inst., BH 623, Geffen School of Medicine at UCLA, Los Angeles, CA 90095-7000. Tel.: 310-825-9543; Fax: 310-794-2144; E-mail: bhat{at}jsei.ucla.edu.
1 The abbreviations used are: MEM, minimum essential medium; NEAA, nonessential amino acid; FITC, fluorescein isothiocyanate; DAPI, 4',6-diamidino-2-phenylindole; SGF, stacked Golgi fraction; bistris, 2-[bis(2-hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol; TRITC, tetramethylrhodamine isothiocyanate. ![]()
| ACKNOWLEDGMENTS |
|---|
B. We also thank Drs. Joseph Horwitz, Dean Bok, and Michael Hall for reading the manuscript and for their useful suggestions. | REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. K. Gangalum and S. P. Bhat {alpha}B-crystallin: A Golgi-Associated Membrane Protein in the Developing Ocular Lens Invest. Ophthalmol. Vis. Sci., July 1, 2009; 50(7): 3283 - 3290. [Abstract] [Full Text] [PDF] |
||||
![]() |
C.-h. Xia, H. Liu, B. Chang, C. Cheng, D. Cheung, M. Wang, Q. Huang, J. Horwitz, and X. Gong Arginine 54 and Tyrosine 118 Residues of {alpha}A-Crystallin Are Crucial for Lens Formation and Transparency. Invest. Ophthalmol. Vis. Sci., July 1, 2006; 47(7): 3004 - 3010. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. den Engelsman, D. Gerrits, W. W. de Jong, J. Robbins, K. Kato, and W. C. Boelens Nuclear Import of {alpha}B-crystallin Is Phosphorylation-dependent and Hampered by Hyperphosphorylation of the Myopathy-related Mutant R120G J. Biol. Chem., November 4, 2005; 280(44): 37139 - 37148. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. Y. Lee, H. R. Kast-Woelbern, J. Chang, G. Luo, S. A. Jones, M. C. Fishbein, and P. A. Edwards {alpha}-Crystallin Is a Target Gene of the Farnesoid X-activated Receptor in Human Livers J. Biol. Chem., September 9, 2005; 280(36): 31792 - 31800. [Abstract] [Full Text] [PDF] |
||||
![]() |
K.-S. Chae, K.-S. Oh, and S. E. Dryer Growth Factors Mobilize Multiple Pools of KCa Channels in Developing Parasympathetic Neurons: Role of ADP-Ribosylation Factors and Related Proteins J Neurophysiol, August 1, 2005; 94(2): 1597 - 1605. [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 |