HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Xi, J.-h. Articles by Andley, U. P. Search for Related Content PubMed PubMed Citation Articles by Xi, J.-h. Articles by Andley, U. P.

Alpha-crystallin expression affects microtubule assembly and prevents their aggregation

Jing-hua Xi^*, Fang Bai^*, Rebecca McGaha^* and Usha P. Andley^*,,1

^* Department of Ophthalmology and Visual Sciences and of

Biochemistry and Molecular Biophysics, Washington University School of Medicine, St. Louis, Missouri, USA

¹Correspondence: Department of Ophthalmology and Visual Sciences, Washington University School of Medicine, 660 S. Euclid Ave., Campus Box 8096, St. Louis, MO 63110, USA. E-mail: andley{at}vision.wustl.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The molecular chaperones A- and B-crystallins are important for cell survival and genomic stability and associate with the tubulin cytoskeleton. The mitotic spindle is abnormally assembled in a number of A–/– and B–/– lens epithelial cells. However, no report to date has studied the effect of -crystallin expression on tubulin/microtubule assembly in lens epithelial cells. In the current work we tested the hypothesis that the absence of A- and B-crystallins alters microtubule assembly. Microtubules were reconstituted from freshly dissected explants of wild-type, A–/–, B–/–, and (A/B) –/– (DKO) mouse lens epithelia and examined by electron microscopic and biochemical analyses. The wild-type microtubules were 4 µm long and 25 nm wide and had a characteristic protofilament structure, but B–/– microtubules were 2.5-fold longer. Microtubule-associated proteins (MAPs) extracted from microtubules by washing with salt included transketolase, -enolase, and ßB2-crystallin. In DKO lens epithelial microtubules but not in wild-type, A–/– or B–/– microtubules, extraction of the MAPs gave very long (14–20 µm) "polyfilament" assemblies that were tightly bundled. Addition of exogenous -crystallin (A+ B) was ineffective in preventing polyfilament formation. However, normal microtubule structure could be restored by including MAPs derived from wild-type lens epithelial cells during microtubule reconstitution. Intriguingly, these data suggest that -crystallin may interact with MAPs to inhibit aggregation of microtubules in lens epithelial cells. Sedimentation analysis and 90° light scattering measurements showed that -crystallin suppressed tubulin assembly in vitro. -Crystallin did not have a strong effect on the GTPase activity of purified tubulin. SDS-PAGE analysis showed that -crystallin prevented heat-induced aggregation of tubulin, suggesting that -crystallin may affect microtubule assembly by maintaining the pool of unassembled tubulin. Xi, J.-h, Bai, F., McGaha, R., Andley, U. P. Alpha-crystallin expression affects microtubule assembly and prevents their aggregation.

Key Words: tubulin • lens • cytoskeleton • chaperone

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
LENS -CRYSTALLIN is a dynamic polymeric complex of two closely related polypeptides, A and B. -Crystallin functions as a molecular chaperone in an ATP-independent manner in which it can prevent protein aggregation by binding early unfolding intermediates of denaturing proteins (1 2 3 4 5) . The expression of -crystallin in a variety of tissues outside the lens suggests that A- and B-crystallin have distinct regulatory roles in cells. Additional studies have shown that -crystallin may function as a chaperone for cytoskeletal components such as actin and intermediate filaments (6 7 8 9) . B-crystallin associates with intermediate filaments and prevents their undesired aggregation (7) .

Recent work suggests that the microtubule cytoskeleton may be an additional physiological target of -crystallin (10 , 11) . B-Crystallin can suppress the aggregation and precipitation of denatured tubulin in vitro and displays an affinity for tubulin. Cellular B-crystallin expression increases in the presence of microtubule-destabilizing drugs, and decreases with microtubule-stabilizing drugs (12) . These results suggest that tubulin dimers may be a substrate of B-crystallin. Overexpression of B-crystallin protects cardiac cells from ischemic injury and this phenotype correlates with enhanced microtubule stability (13 , 14) . Furthermore, phosphorylated B-crystallin has been immunolocalized in midbodies and centrosomes of dividing cells (15) . It has been suggested that B-crystallin binds to microtubules via microtubule-associated proteins and gives microtubules resistance to disassembly (10) . A coincident loss of tubulin and B-crystallin expression occurs with muscle atrophy, and the expression of B-crystallin appears to be important for protecting tubulin in muscle cells undergoing differentiation (11) .

The ocular lens is composed of two types of cells: epithelial cells that make up the monolayer that covers the anterior surface of the lens, and the differentiated fiber cells that are not replaced throughout life (16) . The molecular chaperones A- and B-crystallin serve as structural proteins of the lens fibers and also protect the lens epithelium (17 , 18) . Using A–/– lens epithelial cells, it has been demonstrated that A-crystallin protects lens epithelial cells from apoptosis in vivo (18) . These studies further suggest that some of the A–/– cells die in pairs during the late stages of mitosis (18) . Moreover, the tubulin cytoskeleton is not properly assembled in some of the A–/– lens epithelial cells during mitosis in vivo. B–/– lens epithelial cells also show numerous spindle abnormalities and genomic instability (19 , 20) . A-Crystallin is concentrated in centrosomes (microtubule organizing centers) and midbodies in mitotic lens epithelial cells (18) . One of the functions of molecular chaperones is protecting the cytoskeleton, and it has been suggested that there is an association between B-crystallin and the microtubule cytoskeleton (10) . Our previous results suggest that A- and B-crystallin may be directly involved in interacting with tubulin and microtubules during mitosis (18 19 20) . Although microtubules have been described in the lens (23 24 25) , there are no reports on the effect of -crystallin expression on microtubule assembly.

Microtubules perform important functions in cells including the assembly of the mitotic spindle. Microtubule assembly involves the addition of -ß tubulin (50 kDa) heterodimers to the positive ends of microtubules (26) . Elongated microtubule protofilaments form at 37°C in the presence of GTP and later close into hollow cylindrical polymers. Microtubule growth often originates at microtubule organizing centers (centrosomes) in cells. Microtubules exhibit the property of dynamic instability that has been established using elegant studies of living cells (27) .

Microtubules interact with cytoplasmic effectors, including motor proteins and structural microtubule-associated proteins (MAPs). Microtubules form faster in the presence of MAPs, which protect microtubules from depolymerization (28) . In the brain, MAPs include tau (55–62 kDa), which protrudes from the microtubule surface, and high MW MAPs (200 and 220 kDa), while in HeLa cells, MAP1 has been identified (29) . A defective microtubule assembly is the basis of several neurological diseases (30) .

De novo tubulin synthesis is autoregulated by free tubulin dimer concentrations (31 , 32) . Unassembled tubulin denatures relatively easily, and the denatured tubulin can inhibit the assembly itself (33) . Molecular chaperones associate with tubulin in the cell and are an important element of tubulin quality control (34) . -Crystallin may function as a chaperone for cytoskeletal proteins including tubulin (34) . The cytoplasmic chaperone TCP-1 plays an important role in microtubule nucleation and in cell growth by assisting in the folding of tubulin and other proteins (35) . Mutations in TCP-1 genes cause microtubule abnormalities and aberrant chromosome segregation (35) . HSP70 may regulate tubulin assembly-disassembly and suppress microtubule formation by acting as an antagonist of MAPs, which are able to promote tubulin assembly and stabilize microtubules (34) . Inhibition of tubulin polymerization could be necessary for cell division and differentiation, times when a highly dynamic cytoskeleton is favored (34) .

Here we sought to analyze the contribution of -crystallin to microtubule organization, with cells from genetically modified animals lacking A-, B-, or both A- and B-crystallin isoforms (36 37 38) . Our data suggest that -crystallin is important for maintaining the pool of unassembled tubulin, and it may interact with microtubule-associated proteins to prevent microtubule aggregation in lens epithelial cells.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Animals
129SvEv (wild-type) mice were purchased from Taconic Farms (Germantown, NY, USA). A–/–, B–/–, and (A/B)–/– (DKO) mice (36 , 37) were generously provided by Dr. Eric Wawrousek (National Eye Institute, NIH, Bethesda, MD, USA). Prior to use, animals were routinely genotyped to assess the expression of A- and B-crystallin mRNA by polymerase chain reaction (PCR)-based methods. A- and B-crystallin protein expression was routinely assessed by immunoblot analysis (18 19 20 21 22) .

Materials
Microtubules were prepared from freshly isolated lens epithelial fractions from wild-type, A–/–, B–/–, and A/B–/– (DKO) mice. Brain tubulin was purchased from Cytoskeleton, and glycerol from Fluka (St. Louis, MO, USA). Bovine lens A-crystallin, B-crystallin and the -crystallin hetero-aggregate (A:B in 3:1 ratio) were obtained from Stressgen (San Diego, CA, USA). A monoclonal antibody (mAb) to ß-tubulin was obtained from Sigma (St. Louis, MO, USA). A mAb to B-crystallin was obtained from NovoCastra (Newcastle upon Tyne, UK). All other chemicals were purchased from Sigma. Stock solutions of ATP (30 mM) and GTP (25 mM) were prepared in water and kept frozen at –80°C.

Assembly of microtubules
Microtubules were reconstituted from mouse lens epithelial cells using Taxol, a microtubule polymerization-inducing agent (11 , 29) . Mouse lens epithelial cell fractions freshly isolated from 10–12 lenses were homogenized with 100 µl PME buffer (80 mM PIPES, 1 mM MgCl₂, 1 mM EGTA, pH 6.8) at 4°C, followed by centrifugation at 30,000 g for 15 min and the resulting supernatants were centrifuged at 100,000 g for 60 min in a Beckman Optima TLX ultracentrifuge. Taxol (20 µM) and GTP (1 mM) were added to the supernatants. The drug Taxol binds tightly to microtubules and stabilizes them. After incubation for 15 min at 37°C, the supernatants were ultracentrifuged at 37°C for 15 min through a 12% sucrose cushion to pellet the microtubules (29) . The sucrose cushion removes any contaminating soluble cytoplasmic proteins entrapped in the microtubules. Cellular fractions were analyzed by electron microscopy (EM). The expression of tubulin in different cellular fractions, including the large cellular debris, the vesicles and filaments fraction, the 37°C microtubule pellet and the supernatant was analyzed by SDS-PAGE and immunoblotting with antibodies to ß-tubulin. Microtubule reconstitution experiments were performed at least three times and representative data are shown in the figures.

Extraction of microtubule-associated proteins (MAPs)
Microtubule pellets were washed in 20 µl of PME buffer plus Taxol (20 µM) and the microtubules were centrifuged at 30,000 g for 25 min. To dissociate any MAPs from microtubules, the microtubule pellet was resuspended in 20 µl PME buffer plus Taxol at 37°C, and NaCl was added to 0.35 M concentration (29) . The solution was centrifuged again at 30,000 g for 25 min, leaving the MAPs in the supernatants. The procedure was adapted from that described for the isolation of brain microtubule-associated proteins (29) . A repeat wash was performed. Proteins extracted in the supernatant as well as in the microtubule pellet were analyzed by SDS-PAGE, MALDI mass spectrometric analysis, and immunoblotting. The microtubule pellet after extraction of MAPs was negative stained and analyzed by EM.

EM
Microtubules reconstituted from lens epithelial cell fractions of wild-type, A–/–, and B–/– and DKO lenses were analyzed. The microtubule fraction was taken up in a pipette once and was absorbed onto copper grids floating on water, and stained with 1% aqueous uranyl acetate (Ted Pella Inc., Redding, CA, USA) to give the specimen a little contrast. Samples were viewed on a JEOL 1200EX electron microscope at an accelerating voltage of 80 keV at the Washington University Imaging Core laboratory (St. Louis, MO, USA). In each sample, microtubules were examined in random 14.5 µm x17 µm fields.

GTPase assay
The GTPase activity of tubulin was measured using the Enzcheck Free Phosphate assay kit (Molecular Probes, Eugene, OR, USA) according to the manufacturer’s instructions. This assay takes advantage of the shift from a maximum absorbance of 330 nm to 360 nm that occurs when the substrate 2-amino-6-mercapto-7-methylpurine riboside (MESG) is converted to ribose-1-phosphate and 2-amino-6-mercapto-7-methylpurine through the purine nucleoside phosphorylase (PNP)-dependent addition of P_i to MESG. Tubulin was at 3 µM in PME buffer and A- or B-crystallin was at 1.5 µM for these experiments. GTP (100 µM) was incubated in the presence of the kit reagents, tubulin, or the tubulin plus -crystallin. Optical density (OD) was measured every 1 min at 360 nm in a Perkin-Elmer HTS 2000 plate reader. The average GTPase rate was calculated using a best fit function and P_i standards.

Microtubule assembly by sedimentation assay
The effect of A-, B-, or -crystallin hetero-aggregate on microtubule assembly was determined by a sedimentation assay and SDS-PAGE analysis (39, 40 ). Bovine brain tubulin (Cytoskeleton, Denver, CO, USA) at a concentration of 25 µM was mixed with A-, B-, or total -crystallin (0 to 40 µM) in PME buffer at 37°C for 15 min and centrifuged at 30,000 g for 30 min. The supernatant and pelleted microtubule fractions were analyzed by SDS-PAGE and the proteins were stained with Coomassie blue. Gels were analyzed by densitometric scanning.

Microtubule assembly kinetics by 90° light scattering
Tubulin assembly was followed by 90° light scattering. A 100 µl reaction mixture containing bovine brain tubulin (15 µM) was mixed with different concentrations of A-, B-, or total -crystallin at 0°C (0–2.5 µM) in PME buffer containing 25% glycerol (40) . Polymerization was initiated by raising the temperature to 37°C, and OD was recorded every 40 s at 37°C. The rate and extent of polymerization were measured for 30 min.

Thermal denaturation of tubulin
Bovine brain tubulin (5 µM) was incubated in the absence or presence of 0, 8, 20, and 40 µM -crystallin for 5 min at 37, 42, 55, and 80°C in a water bath/incubator. The samples were centrifuged for 5 min at 10,000 g and pellet and supernatants obtained were analyzed by SDS-PAGE and Coomassie blue staining. The intensity of the bands was quantified by densitometric analysis.

Immunoblot analysis
Proteins were separated using SDS-PAGE as described previously (19 20 21 22) . Immunoblotting with a mAb to ß-tubulin (Sigma, St. Louis, MO, USA) was used to examine the expression of tubulin in different cellular fractions of mouse lens epithelium. The same amount of total protein (20 µg) was loaded from each fraction. The antibody (Ab) was used at a dilution of 1:1000. Immune complexes were detected using a horseradish peroxidase labeled secondary Ab and luminol (Santa Cruz Biotechnology, Santa Cruz, CA, USA). We also used immunoblotting to determine the total amount of tubulin expressed in lens epithelium of wild-type, A–/–, B–/–, and DKO lenses. The amount of tubulin from the same amount of total cell lysate protein (30 µg) was compared.

Densitometric analysis
Immunoblots or Coomassie stained gels were quantitatively analyzed by densitometric scanning using the AlphaInnotech Photoimaging system according to the manufacturer’s instructions.

MALDI mass spectrometric analysis
Proteins were visualized by staining gels with Sypro Ruby (Molecular Probes) and stained bands were excised from the polyacrylamide gel and digested with trypsin using a protocol developed in the Washington University Protein Chemistry Laboratory as described (41) . The recovered peptides were dried and dissolved in 5 µl of 50% acetonitrile made in 0.1% trifluoroacetic acid. 0.5 µl of this solution was cocrystallized on the MALDI target plate with 0.5 µl of a saturated solution of -cyano-4-hydroxycinnamic acid. The MALDI spectrum was acquired on a PerSeptive Biosystems (Framingham, MA, USA) Voyager DE-PRO MALDI-TOF mass spectrometer running Voyager version 5.1 software. The peptide masses from the resultant mass spectrum were submitted to Protein Prospector MS-Fit peptide mass fingerprinting analysis and searched against the NCBI nonredundant protein database.

Statistical analysis
All data are presented as mean ± SE, and the differences between wild-type (control) and A–/–, B–/–, or DKO cells were analyzed using Student’s t test.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Altered morphology of microtubules derived from A–/– and B–/– lens epithelium
Molecular chaperones including the -crystallins stabilize the cytoskeleton and are essential to maintaining its integrity. We demonstrated abnormal microtubule spindle formation in mitotic lens epithelial cells derived from A–/– or B–/– mice (18 , 20) . To determine whether A- and B-crystallin affect microtubule assembly in lens epithelial cells, we reconstituted microtubules in lysates of freshly dissected wild-type, A–/–, or B–/– lens epithelial fractions in the presence of Taxol, and examined their structure by EM. Microtubules had a distinct protofilament structure (Fig. 1 A, B). The average length of wild-type microtubules was 4 µm and the width was 27 ± 3 nm (Table 1 ). Lens epithelial microtubules isolated from the A–/– lens epithelial fractions were significantly thinner than the wild-type and protofilaments peeling away from the microtubule could be seen (Fig. 2 ). Microtubules of B–/– lens epithelial fractions were 2.5-fold longer than the wild-type microtubules (Fig. 3 ). In B–/– microtubules, clumps of proteins appeared to decorate the microtubule surface (Fig. 2) .

View larger version (113K): [in this window] [in a new window]   Figure 1. Electron micrographs and biochemical analysis of microtubules reconstituted from wild-type lens epithelial fractions. Lens epithelial fractions were freshly dissected from wild-type adult mice in PME buffer, and fractionated serially into large cellular debris, vesicles and filaments and microtubule pellet and supernatant. Microtubules were reconstituted at 37°C in the presence of Taxol and GTP. The microtubule fraction was analyzed by negative staining and EM. (A, B) Microtubules reconstituted from wild-type mouse lens epithelial fractions showed a distinct protofilament structure, with individual protofilaments clearly evident (B). C) SDS-PAGE and immunoblot analysis of the proteins in different cellular fractions during microtubule reconstitution procedure with an Ab to ß-tubulin. Fractions containing 20 µg proteins were applied to the gel. (P1), large cellular debris; (P2), vesicles and filaments; (P3), microtubule pellet; (S), supernatant. Note that tubulin was detected only in the microtubule fraction (P3). D) Proteins in the microtubule pellet were separated by SDS-PAGE. Proteins were visualized by Sypro Ruby staining. Protein bands were excised from the gel and analyzed by mass spectrometric analysis. The bands were identified as transketolase, tubulin (-tubulin and ß-tubulin), -enolase, ßB2-crystallin, and A-crystallin. For more details, see Table 2 . B-Crystallin was identified by immunoblot analysis. Bar = 0.5 µm (A) and 3 µm (B).

View this table: [in this window] [in a new window]   Table 1. Width and length of microtubules reconstituted from wild-type, A–/– and B–/– lens epithelial microtubule fractions

View larger version (150K): [in this window] [in a new window]   Figure 2. Electron micrographs of microtubules reconstituted from wild-type, A–/– and B–/– lens epithelial fractions. Microtubules were reconstituted from lens epithelial fractions of wild-type (WT), A–/– and B–/– mice and their morphology was compared. Note that the width of the A–/– microtubules was significantly smaller than the wild-type or B–/– microtubules. For more details, see Table 1 . Note that peeling protofilaments could be observed in some microtubules, suggesting that A-crystallin may affect lateral interactions between the protofilaments. Note also the association of proteins along the microtubule in the B–/– microtubules. Bar = 0.1 µm. Data are representative of three independent experiments.

View larger version (102K): [in this window] [in a new window]   Figure 3. Electron micrographs of microtubules reconstituted from dissected lens epithelial fractions of wild-type and B–/– mice at low magnification (x7500). Note that B–/– microtubules (B) were significantly longer (2.5-fold) than wild-type microtubules (A). For more details, see Table 2 . Data are representative of three independent experiments.

The cellular fractions obtained from lens epithelial cells were analyzed by SDS-PAGE and immunoblotting with an Ab to ß-tubulin. Tubulin was detected only in the microtubule pellet (Fig. 1C ). To further analyze proteins in the microtubule fraction, microtubule pellets were separated by SDS-PAGE and the protein bands were stained with Sypro Ruby (Fig. 1D ). Bands were excised from the gel and identified by MALDI mass spectrometric analysis (Table 2 ). Proteins in the microtubule pellet were identified as transketolase, - and ß-tubulin, -enolase, ßB2-crystallin. A-Crystallin was identified by mass spectrometric analysis. Immunoblotting with an Ab to B-crystallin confirmed the identity of this protein in the microtubule fraction (data not shown).

View this table: [in this window] [in a new window]   Table 2. Proteins identified by MALDI-TOF mass spectrometry in wild-type lens epithelial microtubule fractions

Microtubules derived from DKO lens epithelial fractions
Taxol-stabilized microtubules reconstituted from DKO lens epithelial fractions were, on average, smaller than the wild-type microtubules (Fig. 4 A, E). The distribution of microtubule length in wild-type and DKO lens epithelial fractions showed that 80% of DKO microtubules were 1 to 3.4 µm in length, whereas 60% of wild-type (WT) microtubules were 4 µm long (Fig. 4E ). Moreover, at a higher magnification, breaks in microtubules and the unraveling or loosening of the microtubule hollow cylindrical structure were observed (Fig. 4B-D ). This suggests that microtubules were more fragile in the DKO fractions and may break during isolation. Proteins associated with DKO microtubules were observed all along the length (Fig. 4D ).

View larger version (104K): [in this window] [in a new window]   Figure 4. Electron micrographs of microtubules reconstituted from DKO lens epithelial fractions. A) Low magnification (x5000) image of DKO microtubules. B, C) Higher magnification images of DKO microtubule showing a break along the microtubule length (white arrowhead). Regions of normal protofilament assembly are indicated by the black arrows. Some regions of the microtubules had a looser structure (white arrow) in panel B. C) Higher magnification image of the break (white arrowhead) and looser region (white arrow) of the microtubule shown in panel B. D) A DKO microtubule with normal protofilament structure (black arrows) and a region where the hollow cylinder appears to have opened up (white arrow). Note the proteins associated along the length of the DKO microtubule (black arrowheads). E) Comparison of microtubule length distribution in wild-type (WT) (filled bars) and DKO (open bars) microtubules. Note that the DKO microtubules were significantly shorter than the wild-type. >60% of the wild-type microtubules were 4 to 14 µm long. In contrast, 80% of DKO microtubules were shorter than 3.4 µm. B) Bar = 100 nm. Data are representative of three independent experiments.

Effect of extracting MAPS from microtubule pellets
Microtubule fractions obtained with Taxol are known to contain microtubule-associated proteins (MAPs) that promote tubulin assembly and other factors (10 , 11 , 29) . MAPs dissociate from Taxol-stabilized microtubules at high ionic strengths (29) . Taxol-stabilized microtubules were treated with assembly buffer containing Taxol and 0.35 M NaCl, followed by centrifugation. This procedure has been shown to remove brain MAPs (29) . We determined the effect of removing MAPs on microtubule morphologies. We observed distinct microtubule morphologies in wild-type and DKO lens epithelial fractions after this treatment (Fig. 5 ). Whereas the washed wild-type microtubules displayed microtubules with an average length of 4 µm (Fig. 5A, C, E ), the washed DKO microtubules lacking the expression of both A and B-crystallin, were organized into long bundles with 40% of the microtubules 14–20 µm in length (Fig. 5B, D, F ). Whole protofilaments seem to have unraveled, then formed long polymers with aggregated "polyfilaments" lacking the characteristic protofilament structure. In contrast, individual protofilaments characteristic of normal microtubule structure remained in washed microtubules derived from wild-type cells (Fig. 5C ). These tight microtubule associations observed exclusively in the DKO microtubules persisted in the presence of exogenously added -crystallin (Fig. 6 A, B). However, the addition of wild-type MAPs to the DKO cell lysates restored the normal microtubule morphology in washed DKO microtubules (Fig. 6C ). This suggests that -crystallin interacts with microtubule polymers via MAPs, and prevents their aggregation. Like the washed wild-type microtubules, A–/– and B–/– microtubules did not form strong lateral associations and polyfilament structures when MAPS were extracted (Fig. 7 A, B).

View larger version (95K): [in this window] [in a new window]   Figure 5. Electron micrographs of wild-type and DKO lens epithelial microtubules extracted by salt to remove MAPs (washed microtubules). Microtubule pellets of wild-type and DKO lens epithelial fractions were extracted twice with 0.35 N NaCl containing PME buffer, GTP, and Taxol and the pellets were analyzed by negative staining and EM. A) Low magnification (x5000) image of wild-type microtubules after extracting MAPs. B) Low magnification image of DKO microtubules after extracting MAPs. C) High magnification image of wild-type microtubules after extracting MAPs. D) High magnification image of DKO microtubules after extracting MAPs. Note the extremely long microtubules (14–20 µm) (B, F) and the tight associations between microtubules (D) isolated from DKO lens epithelial fractions. C, D) Bar = 0.5 µm. Microtubule length distribution isolated from wild-type (E) and DKO lens epithelial fractions (F). Open bars: microtubule length before salt extraction; filled bars: microtubule length after salt extraction. Note that the microtubule length distribution of wild-type microtubules did not change appreciably after washing with 0.35 N NaCl containing PME buffer. Note also that 40% of the washed DKO microtubules were 14–20 µm in length. Data are representative of three independent experiments.

View larger version (78K): [in this window] [in a new window]   Figure 6. Electron micrographs of DKO microtubules washed with salt to extract MAPs. Microtubules were reconstituted from DKO lens epithelial fractions in the absence or presence of exogenous -crystallin or the MAPs extracted from wild-type microtubules. Microtubules were washed with salt and their morphologies were examined to determine whether -crystallin or wild-type MAPs restore and normal microtubule morphology and inhibit polyfilament formation. All other conditions were the same as in Fig. 5 . A) Washed DKO microtubules reconstituted in the absence of exogenous -crystallin. B) Washed DKO microtubules in the presence of exogenous -crystallin (10 µM). Note that the addition of -crystallin to DKO cell lysates did not affect the polyfilament formation, bundling and longer lengths of washed DKO microtubules. C) Washed DKO microtubules reconstituted in the presence of exogenous wild-type MAPs. Note the complete restoration of normal protofilament structure (inset) and microtubule length and width of DKO microtubules. Bar = 0.5 µm (A, B); 0.1 µm (C). Data are representative of three independent experiments.

View larger version (122K): [in this window] [in a new window]   Figure 7. Electron micrographs of A–/– and B–/– lens epithelial microtubules washed with salt to extract MAPs. Microtubules were reconstituted from A) A–/– and B) B–/– lens epithelial fractions and extracted by 0.35 N NaCl containing PME buffer (washed) and their morphologies were examined by EM. These microtubules demonstrated normal protofilament structure and did not form polyfilament aggregates. Note that the effect of washing with salt on the morphology of A–/– and B–/– microtubules was the same as that on wild-type microtubules (see Fig. 5C ) but different from DKO microtubules (Fig. 5D ), which appeared to undergo bundling and polyfilament formation by washing with salt. Data are representative of three independent experiments.

Effect of -crystallin on tubulin in the absence of Taxol
Since A–/– and B–/– microtubules were longer than wild-type microtubules, we next determined whether addition of -crystallin alters total tubulin assembly. The effect of -crystallin on the assembly of brain tubulin was determined in the absence of Taxol by a 90° light scattering assay performed at 37°C in the presence of 25% glycerol. Brain tubulin (15 µM) was polymerized in the absence or presence of bovine lens A-, B-crystallin, or -crystallin hetero-aggregate, and the increase in 90° light scattering was measured (Fig. 8 ). At this concentration of tubulin (15 µM), light scattering at 360 nm increased immediately and continued to rise for 30 min. Other samples contained A-crystallin at concentrations between 0.05 and 3 µM. We observed that the addition of 0.5 µM A-crystallin to 15 µM tubulin resulted in a 60% inhibition of tubulin assembly, with an extended lag phase (Fig. 8) . The light scattering at 360 nm was completely suppressed in the presence of 3 µM A-crystallin (molar ratio of 5:1 between tubulin and A-crystallin), suggesting that it inhibits tubulin assembly. B-Crystallin or total -crystallin had a similar effect on the 90° light scattering by purified tubulin (data not shown). The requirement for glycerol may affect the results of the tubulin assembly described above. We therefore used a sedimentation assay to study total tubulin assembly in the presence of -crystallin. Tubulin was assembled from 25 µM brain tubulin in the absence and presence of A-crystallin at 37°C. The assembly was performed in the absence of Taxol, and microtubule pellet and supernatant fractions were separated by centrifugation at 30,000 g and analyzed by Coomassie blue staining and densitometric analysis. As shown in Fig. 9 , low concentrations of A-crystallin ( 2 µM, corresponding to a molar ratio of 12.5:1 of tubulin and A-crystallin) marginally affected the amount of tubulin in the microtubule pellet. In contrast, 10 µM A-crystallin, at a molar ratio of 2.5:1 of tubulin and A-crystallin decreased the amount of tubulin in the pellet by 60%.

View larger version (21K): [in this window] [in a new window]   Figure 8. 90° light scattering assay for assembly of purified tubulin. Purified bovine brain tubulin (15 µM) was incubated at 37°C in the absence of Taxol with different concentrations of A-crystallin. A) Tubulin assembly of bovine brain tubulin was measured for 30 min in the presence of 25% glycerol by 90° light scattering at 360 nm in the absence of Taxol. —, tubulin alone; —, tubulin + 0.1 µM A-crystallin; —, tubulin + 3 µM A-crystallin. Note that in the absence of A-crystallin, light scattering of tubulin begins to increase immediately. At low concentrations of A-crystallin (0.08 to 0.5 µM), the increase in light scattering of 15 µM tubulin due to tubulin polymerization showed an initial lag phase, then increased to a concentration lower than in its absence. The increase in light scattering due to tubulin polymerization was completely suppressed by 3 µM A-crystallin. B) Quantitative analysis of percent tubulin assembly at different concentrations of A-crystallin measured by the assay in panel A. Data are mean ± SE of three independent measurements (tubulin alone vs. tubulin+0.1 µM A-crystallin, P < 0.005; tubulin alone vs. tubulin+0.5 µM A-crystallin, P < 0.05; tubulin alone vs. tubulin+3 µM A-crystallin, P < 0.05).

View larger version (33K): [in this window] [in a new window]   Figure 9. Sedimentation assay for purified tubulin assembly at 37°C in the absence of Taxol. Purified bovine brain tubulin (25 µM) was incubated at 37°C in the absence of Taxol with different concentrations of A-crystallin. Polymerized tubulin was separated from unassembled tubulin by sedimentation as described in Materials and Methods. The amount of assembled tubulin in pellets was analyzed by SDS-PAGE and Coomassie blue staining (A). Note the concentration-dependent decrease in tubulin in the pellet in the presence of A-crystallin. B) Quantitative analysis of the bands shown in panel A by densitometric scanning. Data represent mean ± SE of three independent experiments (tubulin alone vs. tubulin + A-crystallin, P < 0.005).

We next considered the possibility that the loss of a molecular chaperone such as A-crystallin may decrease the amount of tubulin in lens epithelial cells. The amount of tubulin protein in the same amount of total cell lysate protein was compared. In our initial attempts to quantify tubulin expression in the A–/–, B–/–, and DKO cellular fractions, we found that tubulin in microtubule pellets derived from A–/– and DKO cells was barely detectable by immunoblotting when samples were heated to 90°C in sample buffer containing SDS for gel electrophoresis (Fig. 10 A). We therefore heated the samples at temperatures 80°C and found that tubulin in A–/– and DKO became detectable (Fig. 10B ). We established that the total amount of tubulin detected by immunoblotting and densitometric analysis of proteins in lysates of A–/–, B–/–, and DKO cells was 40–50% lower than that in wild-type cells (Fig. 10C ).

View larger version (32K): [in this window] [in a new window]   Figure 10. SDS-PAGE and immunoblot analysis of tubulin expression in wild-type, A–/–, B–/–, and DKO lens epithelial cells. Lens epithelial fractions of adult mice were lysed in PME buffer and protein concentrations were determined. 30 µg proteins of total cell lysates were loaded on the gel and tubulin was detected by immunoblotting with an Ab to ß-tubulin. A) Lysates were treated with SDS-PAGE sample buffer and the samples were heated at 90°C for 5 min prior to loading on the gel. Note that the tubulin band was barely detectable in the A–/– and the DKO lens epithelial samples. B) Lysates were treated with SDS-PAGE sample buffer and the samples were heated at 80°C for 5 min prior to loading on the gel. Note that the tubulin bands were now clearly detectable by immunoblotting. C) Quantitative analysis by densitometric analysis of the bands in panel B. For each group, n = 5. Mean ± SE are shown. Note the 35% decrease in total tubulin levels in all the -crystallin knockout lens epithelial samples.

To study this effect in more detail and determine whether -crystallin prevents tubulin denaturation, we analyzed the precipitation of 10 µM tubulin at different temperatures in the absence and presence of a range of -crystallin concentrations. Tubulin was heated to different temperatures and soluble and insoluble fractions were separated by centrifugation. Figure 11 shows the distribution of tubulin in soluble and insoluble fractions after heating at 37, 42, 55, and 80°C. More than 90% of tubulin was in the soluble fraction in the absence or presence of -crystallin at 37, 42 and 55°C (Fig. 11B, C ). However, at 80°C, tubulin when heated alone precipitated and > 90% was detected in the insoluble pellet (Fig. 11A ). Addition of -crystallin to tubulin prior to heating prevented the insolubilization of tubulin at 80°C (Fig. 11B ). These data suggest that -crystallin can maintain the unassembled pool of tubulin in a native state.

View larger version (47K): [in this window] [in a new window]   Figure 11. SDS-PAGE and Coomassie blue staining of tubulin incubated at different temperatures in the absence and presence of -crystallin. Bovine brain tubulin (10 µM) was heated for 5 min at 37, 42, 55, and 80°C in the absence or presence of 8, 20, and 40 µM bovine lens -crystallin. A) Samples separated by centrifugation into supernatant and pellet fractions were analyzed by Coomassie blue staining. Data for 37 and 80°C are shown. B) Quantitative analysis of percent soluble tubulin at different temperatures in the absence or presence of different concentrations of -crystallin by densitometric scanning of gels shown in panel A. Note that the percent of soluble tubulin is very low at 80°C in the absence of -crystallin (black bars), but increases in the presence of -crystallin in a concentration-dependent manner. C) Quantitative analysis of percent soluble tubulin at different temperatures in the presence of 20 µM -crystallin. Data are representative of three independent experiments.

We also measured tubulin’s GTPase hydrolysis rate in the presence or absence of -crystallin by a spectrophotometric method to quantify free phosphate (Molecular Probes). At a concentration of 3 µM, bovine brain tubulin was able to hydrolyze GTP at a rate of 0.50 ± 0.04 GTP/tubulin min^–1. The addition of -crystallin did not have a strong effect on GTP hydrolysis. The average rate of GTP hydrolysis by tubulin in the presence of 1.5 µM A-crystallin, a molar ratio of 2:1 of tubulin to A-crystallin, was 0.46 ± 0.07 GTP/tubulin min^–1. Similar results were obtained at a 2:1 molar ratio of tubulin to B-crystallin (0.48±0.06 GTP/tubulin min^–1). Taking into account the margin of error, these values are essentially equivalent to the rate of GTP hydrolysis by tubulin alone.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The small heat shock protein and molecular chaperone -crystallin can be isolated as a hetero-oligomer of A- and B-crystallin polypeptides from lens cells (17) . A-Crystallin colocalizes with intercellular bridge microtubules during mitosis, and the absence of A- or B-crystallin in lens epithelial cells of gene knockout mice produces abnormal anaphase spindles, cell death, or genomic instability (18 19 20) . No information currently exists on the modulation of tubulin assembly by the molecular chaperones A- and B-crystallin. It was recently proposed that B-crystallin is a key molecule suppressing tubulin aggregation in muscle cells and provides resistance to microtubule disassembly in vitro (10 , 11) . The availability of genetically modified animals lacking either A- or B-crystallin, or both (36 37 38) , affords a valuable system to study the biochemistry and assembly of microtubules within the context of cells specifically lacking one or both isoforms. The present results provide new insight into the relationship between -crystallin expression and microtubule assembly in lens epithelial cells. One of the main findings of this study is that microtubules reconstituted from wild-type, A–/–, and B–/– lens epithelial cells had different assembly characteristics with the microtubules for the B–/– cells being 2.5-fold longer than their wild-type counterparts. Microtubule length is associated with nucleation, and longer lengths reflect poor nucleation and fewer microtubules (42) . This is complicated by higher concentrations of tubulin that grow rapidly off any nuclei formed. In this model, fewer microtubules would be formed in the B–/– cell lysates as compared with wild-type cells. This is supported by our results which show that the number of microtubules in the B–/– cells was 40% lower than in the wild-type cells (Table 1) .

A–/– Microtubules appear to be more fragile than wild-type and protofilaments peeling away laterally were observed. Microtubules grow by elongation of open sheets that later close into a cylinder. Because the stability of a microtubule is thought to be largely governed by the lateral interactions between adjacent protofilaments (43) , the thinner A–/– microtubules observed in our study suggests that the A–/– microtubules were more fragile and protofilaments were lost laterally during the isolation procedure. This result is consistent with the findings of Fujita et al. (10) , which showed that B-crystallin expression enhances the resistance of microtubules to disassembly in vitro. Microtubules isolated from DKO lens epithelial fractions were shorter than wild-type, suggesting that these microtubules may be very fragile and break along their length into shorter fragments during the isolation procedure. Indeed, DKO microtubules showed breaks and unraveling along the length of the microtubule (Fig. 4) . This might indicate that the absence of -crystallin leads to microtubule instability. Our results are consistent with the studies by Atomi and coworkers who concluded that the loss of B-crystallin leads to microtubule instability in vitro and the loss of tubulin protein in muscle cells in vivo (10 , 11) . The degeneration of some skeletal muscles in B–/– mice raises the possibility that it may be associated with loss of tubulin in these cells (37) .

It is known that MAPs stabilize microtubule polymers (29 , 44) . To gain insight into the molecular basis of the differences in assembly of the wild-type and DKO microtubules, we compared the effect of removing MAPs from the Taxol stabilized microtubules. The results of our study indicate that extraction of MAPs from DKO cells that lacked both A- and B-crystallin yielded lengthy microtubules. Reduced tubule size could not be restored with the addition of -crystallin alone to washed DKO microtubules, suggesting that some additional factors are required. When MAPs from wild-type cells were added, normal microtubule structure was restored, suggesting that factors in addition to -crystallin contribute to microtubule aggregation. These results suggest that -crystallin may interact with MAPs to prevent microtubule aggregation. In the current work, ßB2-crystallin, transketolase (S-crystallin) (45) and -enolase (-crystallin) (46) were detected in the microtubule fraction of lens epithelial cells, suggesting that they may function as MAPs. MAPs protect microtubules from depolymerization. Microtubule-associated protein Dam1 in yeast (41) increases tubulin polymerization and decorates microtubules in rings that stabilize the polymers, while tau laterally associates with the protofilaments and increases microtubule assembly in neuronal cells (47) .

It is believed that microtubule structure obtained in vitro is dependent on many factors including protein concentration and accessory proteins (29 , 42) . Addition of factors before assembly can cause complete reorganization of the length and width distribution. Therefore, there is considerable interest in the in vitro effects of -crystallin on total tubulin assembly. 90° light scattering assay and sedimentation analysis in the absence of Taxol showed a decrease in tubulin assembly by -crystallin. A very similar decrease in assembly of the bacterial tubulin-like protein Ftz by the molecular chaperone ClpX has been shown recently (48) . The decrease in total assembly of tubulin by -crystallin suggests that -crystallin decreases microtubule mass by raising the critical concentration for polymerization. Taken together, these results suggest that -crystallin suppresses microtubule assembly and further indicate that stabilizing MAPs must compete with -crystallin in overcoming this inhibition.

Tubulin expression in a cell is tightly regulated by an autoregulatory mechanism through which an increase in the intracellular concentration of tubulin subunits leads to specific degradation of mRNA (31 , 32) . We compared total tubulin expression in lens epithelial cells by immunoblotting, and found that it was reduced by 40% in the absence of A-crystallin, B-crystallin, or both. The decrease in tubulin levels in the -crystallin knockout cells may be due to degradation of tubulin. During preparation of samples for SDS-PAGE we observed that tubulin derived from A–/– or DKO cells was extremely sensitive to heat (Fig. 10) . To study this further, we determined whether tubulin denaturation and aggregation is altered in the presence of -crystallin. Our data indeed suggest that the thermal denaturation and aggregation of tubulin is suppressed by -crystallin (Fig. 11) . Our data support a model in which -crystallin helps maintain the cytoplasmic pool of unassembled tubulin that is required for correct microtubule assembly, and imply that the effect of -crystallin on microtubule formation is primarily governed at the level of maintaining the unassembled tubulin pool. The inability of -crystallin to alter tubulin’s GTPase activity at 2:1 ratio of tubulin to -crystallin suggests that -crystallin does not substantially disrupt the interaction between tubulin monomers despite its ability to inhibit tubulin assembly at these concentrations in both the sedimentation and the 90° light scattering assays (Figs. 8 , 9) .

Work has shown that -crystallin associates with the cytoskeleton and is essential for maintaining microfilament and intermediate filament integrity (7 , 49) . -Crystallin dramatically inhibits the in vitro assembly of intermediate filament proteins GFAP and vimentin, and increases the soluble pool of GFAP when added to preformed filaments (7 , 49) . Until recently very little information was available on the effect of expression of A- and B-crystallin on the microtubule cytoskeleton. Our previous work with A–/– and B–/– lens epithelial cells had demonstrated that A- and B-crystallin have important but distinct effects on the integrity of the mitotic spindle microtubules (18 , 20) . A loss of microtubules was observed in some of the anaphase spindles of A–/– and B–/– cells between the separating chromosomes (18 , 20) . The current work suggests that this phenotype may be the result of poor microtubule nucleation and/or destabilization of microtubules. It was found that the lack of -crystallin expression did not depolymerize microtubules in interphase (18 , 39) . Because microtubule assembly and disassembly increases 20-fold during mitosis, the expression of -crystallin in lens epithelial cells may improve their ability to conduct mitosis. Based on our results, we can propose that -crystallin may prevent the abnormal anaphase spindle observed in A–/– and B–/– cells by maintaining the unassembled pool of tubulin. This model predicts that the lack of A- or B-crystallin should give poor nucleation and microtubule destabilization in vivo. Further studies with live cell imaging are underway to test this hypothesis.

In summary, we have demonstrated that -crystallin protects unassembled tubulin and that the lack of -crystallin decreases tubulin levels in lens epithelial cells. Reduced tubulin levels may lead to poor nucleation and longer but fewer microtubules, and/or microtubule destabilization. Intriguingly, our work also shows that -crystallin may interact with microtubule-associated proteins to prevent microtubule aggregation. Although the specific nature of an interaction between -crystallin and MAPs remains to be determined, the present studies show that lens epithelial fractions derived from the genetically modified -crystallin knockout animals used here can serve as a useful model for understanding the effects of -crystallin on the microtubule cytoskeleton under physiological conditions.

	ACKNOWLEDGMENTS

 
The authors thank Dr. Eric Wawrousek (National Eye Institute, NIH) for generously providing A–/–, B–/–, and DKO mice. We thank Dr. Wandy Beatty (Molecular Microbiology Imaging Core Laboratory at WUSM) for performing the EM work and Mark Crankshaw (Protein and Nucleic Acid Core Laboratory) for protein identification by MALDI mass spectrometry. This work is supported by grants from NIH (EY05681 to U.P.A. and Core grant EY02687), and by an unrestricted grant to the Department of Ophthalmology at Washington University from the Research to Prevent Blindness.

Received for publication December 12, 2005. Accepted for publication December 28, 2005.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Horwitz, J. (2003) Alpha-crystallin. Exp. Eye Res. 76,145-153[CrossRef][Medline]
2. Andley, U. P., Mathur, S., Griest, T. A., Petrash, J. M. (1996) Cloning, expression and chaperone-like activity of human A-crystallin. J. Biol. Chem. 271,31973-31980[Abstract/Free Full Text]
3. Sun, T. X., Liang, J. J-N. (1998) Intermolecular exchange and stabilization of recombinant human alphaA- and alphaB-crystallin. J. Biol. Chem. 273,286-290[Abstract/Free Full Text]
4. Das, K. P., Choo-Smith, L-P., Petrash, J. M., Surewicz, W. K. (1999) Insight into the secondary structure of non-native proteins bound to a molecular chaperone -crystallin. J. Biol. Chem. 274,33209-33212[Abstract/Free Full Text]
5. Wang, K., Spector, A. (1995) Alpha-crystallin can act as a chaperone under conditions of oxidative stress. Invest. Ophthalmol. Vis. Sci. 36,311-321[Abstract/Free Full Text]
6. Wang, K., Spector, A. (1996) Alpha-crystallin stabilizes actin filaments and prevents cytochalasin-induced depolymerization in a phosphorylation-dependent manner. Eur. J. Biochem. 242,55-66
7. Nicholl, I. D., Quinlan, R. A. (1994) Chaperone activity of -crystallin modulates intermediate filament assembly. EMBO J. 13,945-953[Medline]
8. Muchowski, P. J., Valdez, M. M., Clark, J. I. (1999) B-crystallin selectively targets intermediate filament proteins during thermal stress. Invest. Ophthalmol. Vis. Sci. 40,951-958[Abstract/Free Full Text]
9. Gopalakrishnan, S., Takemoto, L. (1992) Binding of actin to lens alpha-crystallins. Curr. Eye Res. 11,929-933[Medline]
10. Fujita, Y., Ohto, E., Katayama, E., Atomi, Y. (2004) B-Crystallin-coated MAP microtubule resists nocodazole and calcium-induced disassembly. J. Cell Sci. 117,1719-1726[Abstract/Free Full Text]
11. Sakurai, T., Fujita, Y., Ohto, E., Oguro, A., Atomi, Y. (2005) The decrease of the cytoskeleton tubulin follows the decrease of the associating molecular chaperone B-crystallin in unloaded soleus muscle atrophy without stretch. FASEB J. 19,1199-1201[Abstract/Free Full Text]
12. Kato, K., Ito, H., Inaguma, Y., Okamoto, K., Saga, S. (1996) Synthesis and accumulation of B crystallin in C6 glioma cells is induced by agents that promote the disassembly of microtubules. J. Biol. Chem. 43,26989-26994
13. Martin, J. L., Mestril, R., Hilal-Dandan, R., Brunton, L. L., Dillmann, W. H. (1997) Small heat shock proteins and protection against ischemic injury in cardiac myocytes. Circulation 96,4343-4348[Abstract/Free Full Text]
14. Bluhm, W. F., Martin, J. L., Mestril, R., Dillmann, W. H. (1998) Specific heat shock proteins protect microtubules during simulated ischemia in cardiac myocytes. Am. J. Physiol. (London) 276,H2243-H2249
15. Inaguma, Y., Ito, H., Iwamoto, I., Saga, S., Kato, K. (2001) B-crystallin phosphorylated at Ser-59 is localized in centrosomes and midbodies during mitosis. Eur. J. Cell Biol. 80,741-748[CrossRef][Medline]
16. Bhat, S. P. (2001) The ocular lens epithelium. Biosci. Rep. 21,537-563[CrossRef][Medline]
17. Bloemendal, H. (1977) The vertebrate eye lens. Science 197,127-138