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

J. Biol. Chem., Vol. 277, Issue 37, 34462-34470, September 13, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/37/34462    most recent M205420200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map 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 Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Milewski, M. I. Articles by Cutting, G. R. Search for Related Content PubMed PubMed Citation Articles by Milewski, M. I. Articles by Cutting, G. R. Social Bookmarking                      What's this?

Aggregation of Misfolded Proteins Can Be a Selective Process Dependent upon Peptide Composition^*

Micha I. Milewski^§, John E. Mickle, John K. Forrest, Bruce A. Stanton^¶, and Garry R. Cutting

From the  Institute of Genetic Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland 21287, the ^§ Department of Medical Genetics, Institute of Mother and Child, Kasprzaka 17a, 01-211 Warsaw, Poland, and the ^¶ Department of Physiology, Dartmouth Medical School, Hanover, New Hampshire 03758

Received for publication, May 31, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Intracellular aggregation of misfolded proteins is observed in a number of human diseases, in particular, neurologic disorders in which expanded tracts of polyglutamine residues play a central role. A variety of other proteins are prone to aggregation when mutated, indicating that this process is a common pathologic mechanism for inherited disorders. However, little is known about the relationship between the sequence of aggregating peptides and the specificity of intracellular accumulation. Here we demonstrate that substitution of two residues eliminates aggregation of a 111-amino acid peptide derived from the C-terminal portion of the cystic fibrosis transmembrane conductance regulator (CFTR). We also show that fusion to a reporter protein considerably alters the subcellular distribution of aggregating peptide. When fused to green fluorescent protein, the peptide containing amino acids 1370-1480 of CFTR accumulates in large perinuclear or nuclear aggregates. The same CFTR fragment devoid of green fluorescent protein localizes predominantly to discrete accumulations associated with mitochondria. Importantly, both types of accumulation are dependent on the presence of the same two amino acids within the CFTR sequence. Co-expression studies show that both CFTR-derived proteins can co-localize in large cytoplasmic/nuclear aggregates. However, neither CFTR construct accumulates in intracellular inclusions formed by N-terminal fragment of huntingtin. In addition to unique accumulation patterns, each aggregating peptide shows differences in association with chaperone proteins. Thus, our results indicate that the process of intracellular aggregation can be a selective process determined by the composition of the aggregating peptides.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Two different pathological consequences of protein misfolding can contribute to the disease. The first is loss of protein function, which is often accompanied by improper localization and rapid degradation of defective product (1). Many genetic disorders, including familial hypercholesterolemia, Tay-Sachs disease, and maple syrup urine disease, are caused by mutations that affect the folding of an essential protein, leading to the loss of its function (reviewed in Ref. 2). A different mechanism of pathogenesis is associated with so-called conformational diseases (3-6). In this case, the disease is associated with toxic properties of aggregation-prone folding intermediates. However, it is not clear whether observed aggregation of misfolded proteins is a primary cause or merely a consequence of disease. The most intensively studied examples of conformational diseases are Alzheimer's disease (7), Parkinson's disease (8), Huntington's disease (9), and spongiform encephalopathies (10).

Abnormal protein folding is also a cause of most cases of cystic fibrosis (CF)¹ (11, 12), a common recessive genetic disorder characterized by aberrant transepithelial ion transport (13, 14). The most common CF-causing mutation, Phe-508, a deletion of a single phenylalanine, affects the folding of cystic fibrosis transmembrane conductance regulator (CFTR) (15). Protein bearing this mutation does not fold properly during an early step of biogenesis in the endoplasmic reticulum (ER). Consequently, the misfolded CFTR cannot be efficiently transported to the Golgi compartment, where wild type protein undergoes further maturation before being exported to the plasma membrane. As a result of a processing defect, the mutant protein is retained in ER and subsequently degraded by the proteasome complex. Interestingly, it has been recently shown that the Phe-508 CFTR, when overexpressed in cell culture, is able to aggregate and form large microtubule-dependent cytoplasmic inclusion bodies called aggresomes (16). Since this discovery, the formation of aggresomes has been described for a number of other proteins, including a GFP-p115 chimera (17), peripheral myelin protein 22 (18), Cu,Zn-superoxide dismutase (19), surfactant protein C (20), anti-Ras scFv fragments (21), Wilson disease-associated protein ATP7B (22), structural proteins of African swine fever virus (23), and mutant fragments of huntingtin (24). Although it has not been proven that aggregation of CFTR contributes to the pathogenesis of CF, the process of intracellular accumulation of this protein is often regarded as a suitable model for regulated deposition of misfolded polypeptides (25).

Our studies of CFTR biogenesis and trafficking have been focused on its carboxyl-terminal sequence. We have recently demonstrated that the C terminus of CFTR alone, or when fused to a green fluorescent protein (GFP), is able to localize to the correct apical domain of the plasma membrane in polarized epithelial cells (26). However, localization to the apical membrane is observed only after a region of nine amino acids (ag region) is removed from the C-terminal CFTR sequence. Presence of the ag region in the C-terminal sequence fused to GFP results in accumulation of the fusion protein in large perinuclear or nuclear aggregates. This suggests that the ag region is a part of larger folding domain and that disruption of this domain causes protein misfolding and aggregation.

To characterize the sequence responsible for aggregation of CFTR C-terminal peptides and to elucidate the relationship between amino acid sequence and the corresponding protein accumulation pattern, we have analyzed the cellular distribution and composition of aggregates formed by different CFTR-derived constructs. We have also compared the process of accumulation of CFTR-derived peptides with aggregation of the N-terminal portion of huntingtin, a protein mutated in Huntington's disease. Here we report that two conserved amino acids within the ag region are critical for the aggregation of the CFTR C terminus, and that different aggregating peptides show remarkable specificity with regard to their intracellular distribution and association with other proteins.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

DNA Constructs-- All constructs utilized the pRK5-SK mammalian expression plasmid containing the cytomegalovirus promoter. Construction of GFP-CFTR 1370-1480 hybrid and its ag version has been described previously (26). All other mutant versions of the GFP-tagged CFTR C-terminal construct were created using the site-directed mutagenesis system Transformer (CLONTECH, Palo Alto, CA). The Transformer mutagenesis system was also used to create the untagged and the hemagglutinin (HA)-tagged C-terminal constructs and their derivatives. The sequences of selection primers and mutagenic primers used to create point mutations, deletions, and insertions in the CFTR-derived constructs are available upon request. The N63-Q75-Myc-His construct containing the 63 N-terminal amino acids of huntingtin with 75 glutamine repeats has been described elsewhere (27).

Cell Culture and Transfection-- IB3-1 bronchial epithelial cells derived from a CF patient were cultured as described (28). For protein localization analysis, cells were grown on collagen-coated glass coverslips and transiently transfected using the Lipofectin reagent (Invitrogen) according to the instructions of the manufacturer. In the co-transfection experiments, a 1:1 DNA ratio was used for plasmids encoding different peptides. At 36-48 h after transfection, the cells were fixed with 4% paraformaldehyde for 20 min and, if not immunostained, mounted in SlowFade (Molecular Probes, Eugene, OR) containing 0.1 mg/ml DAPI (Sigma). To disrupt microtubules, cells were incubated for 3 h or overnight in fresh media containing 33 µM nocodazole (Sigma) and fixed as described above. The overnight treatment with nocodazole was started 24 h after transfection. The disruptive effect of nocodazole treatment on the structure of microtubular network was tested in parallel experiments by immunostaining microtubules with anti--tubulin antibody.

Antibodies-- Monoclonal antibodies against HA and c-Myc epitopes were purchased from Roche Molecular Biochemicals. Polyclonal anti-c-Myc antibody was from Santa Cruz Biotechnology (Santa Cruz, CA). Polyclonal anti-HA antibody and monoclonal anti-CFTR C-terminal antibody were from Zymed Laboratories Inc. (San Francisco, CA). Monoclonal antibodies against -tubulin and Golgi marker -adaptin were from Sigma. Monoclonal antibodies against mitochondrial marker GRP75, ER marker protein disulfide isomerase, heat shock proteins HSP90 and HSP27, as well as polyclonal antibodies against mitochondrial marker HSP60, ER marker calnexin, and molecular chaperones HSP70, HSC70, HSP40, and -B-crystallin were from StressGen (Victoria, British Columbia, Canada). Monoclonal anti-GFP antibody was from CLONTECH.

Immunofluorescent Staining-- Prior to immunostaining, the cells were fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100 for 5 min, and washed with PBS. Nonspecific binding sites were blocked with 2.5% goat serum in PBS. Staining was performed by two sequential incubation steps. Cells were first incubated with the primary antibody for 1.5 h, and then with appropriate secondary antibody conjugated to fluorescein isothiocyanate (green) or Cy3 (red) fluorescent dye (Sigma) for 30 min. The antibodies were used in dilutions recommended by the manufacturer. After staining, the coverslips with cells were washed in PBS and mounted in SlowFade with DAPI. For analysis of protein distribution, at least 100 transfected cells were examined in two separate experiments (150 cells in the case of co-expression analysis). The results were considered conclusive only if statistically significant number of cells showed characteristic pattern of distribution (p < 0.05).

Western Blotting Analysis-- The IB3-1 cells, transiently transfected with DNA constructs encoding different GFP fusion proteins, were lysed overnight in 0.5% Nonidet P-40, 150 mM NaCl, 1 mM EDTA, 20 mM Hepes, pH 7.0. The lysates were spun at 20,000 × g for 5 min to pellet the insoluble material. Both soluble and insoluble fractions were separated by the SDS-PAGE (12%) and electrophoretically transferred to polyvinylidene difluoride membranes (Amersham Biosciences). The probing with monoclonal anti-GFP antibody and subsequent detection with ECL+plus system (Amersham Biosciences) was performed according to the instructions of the manufacturer. After luminographic detection step, the films were scanned and the intensity of bands was measured, based on pixel count, using Adobe Photoshop.

Confocal Laser Microscopy-- The fluorescence label was examined with laser scanning confocal imaging system (LSM Carl Zeiss). Images were generated using 16-fold line averaging. Contrast and brightness settings were chosen to ensure that all pixels were within the linear range. Images were prepared for publication with LSM Carl Zeiss Software and Adobe Photoshop.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Two Conserved Amino Acids Are Required for the Aggregation of CFTR C Terminus-- While searching for localization signals in CFTR, we discovered that the GFP fusion protein containing the C-terminal portion of human CFTR formed aggregates (26). This aggregation was abolished by deletion of the ag region (ag) encompassing amino acids 1395-1403 of CFTR (Fig. 1, A and B). To characterize the sequence responsible for protein accumulation, we examined the effect of single amino acid substitutions on the aggregation ability of the GFP-tagged C terminus of CFTR. Residues that were either associated with CF-causing mutations or highly conserved in related proteins were selected for study. The only reported CF-causing mutation in the ag region was a substitution of valine at codon 1397 by glutamic acid (V1397E) (29). However, the aggregation of GFP CFTR 1370-1480 was not affected by introduction of this mutation into the fusion protein (Fig. 1C). Three of nine amino acids in the ag region, Thr-1396, His-1402, and Arg-1403, were conserved in corresponding C-terminal segments of several ATP-binding cassette (ABC) proteins closely related to CFTR (Fig. 1A). The replacement of Thr-1396 with alanine had no effect on aggregation of the GFP-tagged protein (Fig. 1D). However, the single amino acid substitutions of histidine at position 1402 or arginine at position 1403 significantly reduced the number of transfected cells showing intracellular protein aggregation. Only 25% (52/208) of cells expressing GFP fusion protein bearing single H1402A substitution and 27% (41/153) of R1403A mutants showed intracellular protein accumulation, whereas the corresponding number for the wild type sequence was 91% (132/145). Also, the size of aggregates formed by fusion proteins was reduced when single alanine substitutions at position 1402 or 1403 were introduced. The effect of double substitution of these two amino acids on protein aggregation was even more profound (Fig. 1E). Cells expressing the double mutant H1402A,R1403A showed no aggregation at all (0/108). Thus, these two conserved amino acids seemed to be critical for the aggregation process of CFTR-derived C-terminal peptides.

View larger version (58K): [in this window] [in a new window]   Fig. 1.   Amino acids critical for aggregation of the GFP-CFTR C-terminal fusion protein. A, a diagram of full-length CFTR illustrating the position of the C-terminal amino acids fused to GFP. The ag region and corresponding sequences of related ABC proteins are shown in the gray box, with conserved amino acids shown in bold. The CFTR amino acids within the ag region that are conserved across species are underlined. MSD, membrane-spanning domain; R, regulatory domain; MRP1, human multidrug resistance-related protein 1; MRP2, human multidrug resistance-related protein 2; SUR1, human sulfonylurea receptor 1; MDR1, human multidrug resistance protein 1. B-E, the effect of different mutations within the ag region, including ag (B), V1397E (C), T1396A (D), and double mutation H1402A,R1403A (E) on the aggregation of C terminus of CFTR fused to GFP (shown in green) in transiently transfected human airway epithelial (IB3-1) cells. Nuclei were stained blue with DAPI. Bars, 10 µm. F, solubility of different GFP-CFTR constructs assessed by separation of soluble and insoluble fractions of transiently transfected IB3-1 cells and subsequent SDS-PAGE and Western blot analysis. The upper panel is the longer exposure of the upper part of the blot. S, soluble fraction; I, insoluble fraction.

The role of His-1402 and Arg-1403 in the aggregation of CFTR-derived peptides was further confirmed by the SDS-PAGE and Western blot analysis, in which a relative amount of protein in soluble and insoluble fractions was assessed (Fig. 1F). In contrast to GFP alone that was found primarily (96%) in the soluble fraction, the GFP CFTR 1370-1480 fusion protein was detected mainly (74%) in the insoluble material. However, the introduction of the ag deletion or the H1402A,R1403A double mutation significantly decreased the amount of the fusion protein in the insoluble fraction to approximately 31 and 36%, respectively. The remaining portion of insoluble protein could correspond to micro-aggregates that were not detectable by fluorescence analysis. However, it is worth noting that only in the case of GFP CFTR 1370-1480 were SDS-insoluble complexes of high molecular weight observed (Fig. 1F), which suggested that amino acids His-1402 and Arg-1403 were responsible for induction of massive protein accumulation.

The C Terminus of CFTR Shows Two Different Pathways of Accumulation-- The ability of GFP to dimerize (30) suggested that GFP might contribute to the intracellular aggregation of the GFP fusion proteins. To test whether the aggregation of the GFP-CFTR 1370-1480 fusion protein could be attributed to the C-terminal sequence of CFTR, or was a phenomenon associated with the GFP molecule, we removed the GFP tag attached to the CFTR-derived peptide. The CFTR 1370-1480 construct was transiently expressed, and the subcellular distribution of this peptide was analyzed by immunostaining and confocal microscopy. Most of the transfected cells (83%, 99/120) showed a discrete pattern of accumulation (Fig. 2A). This pattern of distribution was distinct from the giant perinuclear or nuclear aggregates characteristic for the GFP-tagged C terminus, although large aggregates were also seen in 8% (9/120) of cells transfected with CFTR 1370-1480 (Fig. 2B). Attaching an HA epitope to the N terminus of CFTR 1370-1480 did not alter its intracellular distribution (Fig. 2C). These results indicated that the C terminus of CFTR was responsible for the aggregation process, and that fusion to GFP either facilitated formation of giant aggregates or prevented the formation of small discrete accumulations.

View larger version (47K): [in this window] [in a new window]   Fig. 2.   The effect of removal of the GFP tag on the distribution of the CFTR C terminus. Cellular distribution of untagged or HA-tagged CFTR-derived constructs transiently expressed in IB3-1 cells. Discrete accumulations (A) and occasional large nuclear and cytoplasmic inclusions (B) were formed by the CFTR 1370-1480 construct, detected with monoclonal anti-CFTR C-terminal antibody (red). C, discrete accumulations formed by HA-tagged CFTR C terminus detected with monoclonal anti-CFTR C-terminal antibody (green) and polyclonal anti-HA antibody (red). D, diffuse, cytoplasmic distribution of HA-tagged CFTR C-terminal construct containing the ag deletion. The cells were immunostained with monoclonal anti-CFTR C-terminal antibody (green) and polyclonal anti-HA antibody (red). E, lack of aggregation of HA-CFTR 1370-1480 construct carrying the H1402A and R1403A mutations. The protein was detected with monoclonal anti-HA antibody (red). F, intracellular accumulations formed by truncated HA-CFTR C-terminal construct containing amino acids 1370-1454 of CFTR, and detected with monoclonal anti-HA antibody (red). Nuclei were stained blue with DAPI. Bars, 10 µm.

To explore whether the different intracellular accumulation patterns of the different CFTR C-terminal constructs had a common molecular basis, we examined the role of the amino acids within the ag region in the aggregation process of HA-CFTR 1370-1480. The ag version of this construct showed a diffuse cytoplasmic distribution in all transfected cells (Fig. 2D). Additionally, the double mutation H1402A,R1403A prevented aggregation of HA-tagged CFTR C terminus (Fig. 2E). These results demonstrated that the same amino acids of the C terminus were responsible for the two patterns of aggregation. The aggregating HA-CFTR 1370-1480 and non-aggregating HA-CFTR 1370-1480 ag constructs were both detected by antibodies directed against their N-terminal or C-terminal amino acids (Fig. 2, C and D). This indicated that the aggregation process was not associated with post-translational truncation of the N-terminal or C-terminal sequences in CFTR-derived peptides. The very C-terminal sequence of CFTR contained a PDZ-binding motif required for the apical membrane localization in polarized epithelial cells (26, 31, 32). To examine whether this C-terminal region contributed to the unusual accumulation pattern of HA-CFTR 1370-1480, we deleted the last 26 amino acids in this construct by introducing the naturally occurring CFTR mutation S1455X (33), associated with abnormal localization of the full-length protein (31). This deletion did not affect the distribution of the HA-tagged construct (Fig. 2F), indicating that the very C-terminal amino acids of CFTR were not involved in the aggregation process.

CFTR-derived C-terminal Peptides Associate with Mitochondria-- We next determined whether the distribution of the small accumulations formed by the CFTR C-terminal peptides was a result of their association with the cytoskeleton or organelles. The discrete accumulations formed by these peptides occasionally appeared to be distributed in a linear fashion, which suggested their association with elements of the cytoskeleton (Fig. 3A). Formation of large perinuclear aggresomes was previously reported to be associated with the microtubule-dependent movement of small aggregates toward the center of the cell (16). To test whether discrete accumulations of HA-CFTR 1370-1480 were associated with the microtubular network, we examined the co-localization of both structures in the cell. Indeed, the position of small protein accumulations seemed to correspond with the orientation of microtubules (Fig. 3B). However, overnight exposure of cells transfected with HA- or GFP-tagged CFTR C-terminal constructs to microtubule-disrupter nocodazole caused minimal redistribution of both small and giant cytoplasmic aggregates from the perinuclear region to the periphery of the cell, and the formation of giant aggregates was not inhibited (data not shown).

View larger version (40K): [in this window] [in a new window]   Fig. 3.   Mitochondrial association of accumulations formed by CFTR C terminus in IB3-1 cells. A, B, and E, the regions within yellow boxes on the left side are shown magnified on the right. A, a linear distribution of accumulations (white arrow) formed by HA-CFTR 1370-1480 (red). B, co-localization of HA-CFTR 1370-1480 (green) with microtubular marker -tubulin (red). C, co-localization of HA-CFTR 1370-1480 (green) with mitochondrial marker GRP75 (red). D, co-localization of mitochondrial marker GRP75 (red) with microtubular marker -tubulin (green). E, co-localization of HA-CFTR 1370-1480 (green) with mitochondrial marker GRP75 (red) in cells treated with 33 µM nocodazole for 3 h. F, lack of co-localization between mitochondrial marker GRP75 (red) and GFP-CFTR 1370-1480 (green in the left picture) or N63-Q75-Myc-His (green in the right picture), The N63-Q75-Myc-His was detected using a polyclonal anti-c-Myc antibody. Nuclei were stained blue with DAPI. Bars, 10 µm.

Microtubules govern the localization of many membrane-bound organelles and other cellular components (34-36). Hence, the association of protein aggregates with the microtubular network could simply be a secondary effect, reflecting a primary relationship between the aggregating peptide and an unknown microtubule-associated component. Therefore, we tested whether the small accumulations formed by CFTR-derived peptides co-localized with organelles associated with microtubules. The intracellular distribution of HA-CFTR 1370-1480 did not match the distribution of Golgi and ER markers (data not shown). However, the co-localization analysis between the HA-CFTR 1370-1480 and mitochondrial marker GRP75 showed that aggregating protein was closely associated with mitochondria (Fig. 3C). Cellular distribution of mitochondria corresponded with the localization of microtubules (Fig. 3D) and showed sensitivity to nocodazole treatment. Importantly, the aggregates in cells treated with nocodazole retained a close association with mitochondria, indicating that the observed co-localization was not just a secondary effect related to the microtubular association of both mitochondria and aggregates (Fig. 3E). The mitochondrial association of CFTR C-terminal peptides was also observed when the co-localization between the untagged CFTR 1370-1480 peptide and another mitochondrial marker HSP60 was examined. However, no apparent association with mitochondria was seen in the case of giant aggregates formed by GFP-CFTR 1370-1480 or another aggregating peptide N63-Q75-Myc-His, containing the N-terminal portion of huntingtin (Fig. 3F).

Different CFTR-derived C-terminal Peptides Co-aggregate-- Because the aggregation of all CFTR-derived C-terminal constructs was dependent on the presence of the same CFTR sequence, we explored whether two related peptides showing distinct accumulation patterns were able to form common aggregates. Therefore, we co-expressed the GFP- and HA-tagged CFTR C-terminal peptides. Although the majority of cells (68%, 95/140), expressing both aggregating proteins, showed separate distribution (Fig. 4A), the remaining 32% of cells contained giant perinuclear or nuclear aggregates composed of both constructs (Fig. 4B). This suggested that, although the accumulation pattern of CFTR-derived peptides could be modified by fusion to reporter proteins, the misfolded proteins sharing the C-terminal sequences of CFTR could form common aggregates. On the other hand, the ability of HA-CFTR 1370-1480 to form the small accumulations seemed to partially prevent this peptide from forming giant aggregates and joining the aggregating GFP-tagged counterpart. Importantly, the non-aggregating derivatives of those proteins were also included in giant aggregates formed by their aggregating counterparts co-expressed in the same cell. The HA-tagged peptide devoid of the HR motif co-aggregated with the self-aggregating GFP-CFTR 1370-1480 construct (Fig. 4C). Similarly, the non-aggregating GFP-CFTR 1370-1480 H1402A,R1403A construct was included in giant aggregates formed by HA-CFTR 1370-1480 (Fig. 4D). The non-aggregating proteins were always included in giant aggregates, although their presence did not significantly alter the frequency of formation of giant aggregates by the self-aggregating GFP-tagged (87%, 131/150) or HA-tagged (11%, 16/150) CFTR C-terminal peptides. It is also worth noting that the GFP-tagged CFTR C-terminal constructs were always excluded from small accumulations formed by the HA-tagged CFTR-derived peptides, which confirmed our previous observation that the presence of GFP excludes the GFP-CFTR C-terminal peptides from this type of accumulation. Together, these results indicated that although the critical amino acids within the ag region were responsible for initiation of massive accumulation, the HR motif was not required for inclusion of related peptides into already existing protein aggregates.

View larger version (66K): [in this window] [in a new window]   Fig. 4.   Analysis of co-aggregation between different CFTR-derived peptides. Different CFTR-derived constructs were co-expressed in transiently transfected IB3-1 cells, and their co-localization was examined by fluorescence microscopy. A, different distribution of GFP-CFTR 1370-1480 (green) and HA-CFTR 1370-1480 (red) observed in most co-transfected cells. B, co-localization of GFP-CFTR 1370-1480 (green) and HA-CFTR 1370-1480 (red) in giant aggregates. C, co-localization of the aggregating GFP-CFTR 1370-1480 construct (green) and the non-aggregating HA-CFTR 1370-1480 H1402A,R1403A construct (red) in giant aggregates. D, co-localization of the non-aggregating GFP-CFTR 1370-1480 H1402A,R1403A construct (green) with the aggregating HA-CFTR 1370-1480 construct (red) in giant aggregates. Nuclei were stained blue with DAPI. Bars, 10 µm.

CFTR- and Huntingtin-derived Peptides Form Separate Aggregates-- To further explore the specificity of protein aggregation, we co-expressed the CFTR-derived peptides with N63-Q75-Myc-His, a huntingtin-derived construct containing an elongated polyglutamine sequence composed of 75 repeats. When expressed alone, N63-Q75-Myc-His accumulated in giant cytoplasmic or nuclear inclusions that resembled those formed by GFP-CFTR 1370-1480 (Fig. 3F). Surprisingly, when co-expressed in the same cell, these two peptides formed separate perinuclear or nuclear accumulations (Fig. 5A). Additionally, the HA-CFTR 1370-1480, found in both small and giant aggregates, did not co-localize with the huntingtin-derived construct (Fig. 5B). These data indicated that the aggregation process was characterized by remarkable specificity, which excluded unrelated peptides, even if they showed strong tendency to self-aggregate.

View larger version (64K): [in this window] [in a new window]   Fig. 5.   Analysis of co-aggregation between CFTR- and huntingtin-derived peptides. CFTR- and huntingtin-derived constructs were co-expressed in transiently transfected IB3-1 cells, and their co-localization was examined by fluorescence microscopy. A, lack of co-localization between giant aggregates formed by GFP-CFTR 1370-1480 (green) and the huntingtin-derived N63-Q75-Myc-His construct (red). B, lack of co-localization between small aggregates formed by HA-CFTR 1370-1480 (red) and giant aggregates containing the huntingtin-derived N63-Q75-Myc-His construct (green). Nuclei were stained blue with DAPI. Bars, 10 µm.

Different Aggregating Peptides Associate with Different Heat Shock Proteins-- To explore whether aggregating peptides showing different distribution patterns associate with different chaperone proteins, we examined the presence of several HSPs, functioning as molecular chaperones, in aggregates formed by GFP-CFTR 1370-1480, HA-CFTR 1370-1480, and N63-Q75-Myc-His. A stress inducible protein HSP70, known to associate with many aggregating proteins, was detected in giant aggregates formed by all three peptides studied. However, small accumulations containing HA-CFTR 1370-1480 were HSP70-negative (Fig. 6A). The association of chaperone protein HSP40 with aggregating peptides resembled that of HSP70. Again, only the small accumulations formed by HA-CFTR 1370-1480 did not contain HSP40 (data not shown). On the other hand, a heat shock cognate 70 (HSC70), a constitutively expressed protein related to HSP70, was found in aggregates formed by N63-75Q-Myc-His, but not in those composed of CFTR-derived peptides (Fig. 6B). However, it is worth noting that cells containing aggregates formed by CFTR-derived constructs showed a considerable decrease in the nuclear content of HSC70 (compare two cells in the middle panel of Fig. 6B), suggesting that the distribution of HSC70 was affected by these aggregates. Another common chaperone protein HSP90 was not detected in any aggregates formed by CFTR- or huntingtin-derived peptides.

View larger version (32K): [in this window] [in a new window]   Fig. 6.   Co-localization of aggregating peptides with different heat shock proteins. IB3-1 cells transfected with different CFTR- or huntingtin-derived constructs (green) were immunostained with antibodies directed against different heat shock proteins (red). A, co-localization of HSP70 in giant aggregates formed by GFP-CFTR 1370-1480 (upper panel), HA-CFTR 1370-1480 (middle panel), and N63-Q75-Myc-His (lower panel). B, co-localization of HSC70 in giant aggregates formed by N63-Q75-Myc-His (lower panel) but not in those formed by GFP-1370-1480 (upper panel) or HA-CFTR 1370-1480 (middle panel). C, accumulation of HSP27 in giant aggregates formed by GFP-CFTR 1370-1480 (upper panel) but not in those formed by HA-CFTR 1370-1480 (middle panel) or N63-Q75-Myc-His (lower panel). Nuclei were stained blue with DAPI. Bars, 10 µm.

-B-Crystallin, a member of the small heat shock protein family that was reported to be involved in protein aggregation (37), was absent from all aggregates tested (data not shown). However, a related small heat shock protein HSP27, also known to associate with protein aggregates (38, 39), was detected in accumulations formed by GFP-tagged CFTR C terminus but not in those containing HA-tagged CFTR C terminus or the N-terminal portion of huntingtin (Fig. 6C). Such exclusive association suggested that HSP27 could be involved in GFP-dependent protein aggregation. In summary, these results showed that the diversity of observed protein accumulations was reflected not only by differences in their distribution pattern, but also by their association with different molecular chaperones.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Protein aggregates have often been regarded as nonspecific associations of misfolded molecules. This concept has been challenged by in vitro studies showing that protein aggregation in a cell-free system occurs by specific interactions between folding intermediates (40, 41). Here, we demonstrate that specific intermolecular interactions are responsible for protein aggregation within cells. In addition, we identify the amino acids critical for aggregation of CFTR-derived peptides and show that fusion of an aggregating protein to GFP alters the cellular localization of aggregates and their association with molecular chaperones. Together, these results suggest that the intracellular accumulation of misfolded proteins is a selective process, and that the sequence of aggregating protein determines the composition and cellular distribution of aggregates. Indeed, this conclusion is supported by two recent reports emphasizing the specificity of formation of inclusion bodies by other aggregating proteins (42, 43).

In the first part of this study, we demonstrate that specific amino acids within the C terminus of CFTR facilitate the aggregation of CFTR-derived constructs. The misfolding and aggregation of the CFTR C terminus is likely to result from the disruption of a larger folding domain that encompasses the C-terminal sequences of CFTR. The analysis of recent models of nucleotide-binding fold (NBF) (44, 45), a domain characteristic for CFTR and other ABC proteins, shows that the N-terminal half of the CFTR 1370-1480 construct corresponds to the very C-terminal portion of NBF2 in full-length CFTR. We demonstrate that two amino acids in this region, His-1402 and Arg-1403, are critical for the aggregation of CFTR-derived peptides. However, although the HR motif is required for massive protein accumulation, the C-terminal constructs lacking this motif can also be included into protein aggregates formed by their self-aggregating counterparts. Moreover, a significant amount of non-accumulating peptides lacking the HR motif is found in the insoluble cellular fraction, suggesting that these proteins can form micro-aggregates that are not detectable by fluorescence analysis. Thus, our results suggest that the HR motif increases the protein aggregation rate, but is not absolutely required for the aggregation process. We speculate that these two residues prevent partial folding of neighboring sequences, when present in constructs lacking the remaining portion of the NBF domain. This, in turn, may lead to the formation of a stable aggregation-prone conformation with exposed hydrophobic residues and/or -sheet structures that are normally buried within NBF. We have applied the nearest-neighbor algorithm (PSSP/NNSP) (46) to predict the secondary structure of this particular region of CFTR (data not shown). However, the results have been inconclusive, as the relationship between acquiring the -sheet structure by the ag region and the presence of different mutations shown to abolish or retain the aggregation of the C-terminal peptide was not absolute. However, several regions in this C-terminal part of CFTR, including the sequence containing the HR motif, have been recently reported to affect the maturation and stability of the full-length protein (47, 48), which suggests that other structural motifs present in this portion of NBF2 may play an important role in CFTR folding.

The perinuclear localization of giant aggregates formed by CFTR-derived peptides resembles the distribution of aggresomes, large protein depositions formed by microtubule-dependent accumulation of small aggregates developed in the periphery of the cell (25). However, the presence of frequent nuclear inclusions of similar size and the insensitivity of both nuclear and cytoplasmic giant aggregates to microtubule disruption suggest that additional mechanisms are responsible for the formation of these accumulations. Furthermore, the small discrete accumulations formed by the untagged or HA-tagged CFTR-derived peptides have features that distinguish them from giant aggregates. Some of the CFTR-derived peptides may form both small and giant aggregates and both types of aggregates are dependent on the presence of the HR motif. However, not every aggregating CFTR-derived protein can be included into small aggregates, and none of the examined molecular chaperones associating with aggregating peptides is found in this type of accumulations. These observations, together with the association of aggregating protein with mitochondria, suggest that these small accumulations represent a unique pattern of aggregation that needs further investigation. The association between intracellular protein aggregates and mitochondria has been previously observed by others (23, 49, 50), and an increased rate of ATP metabolism has been recently suggested as a possible cause of concentration of mitochondria around the protein inclusions (24). Because mitochondrial dysfunction plays an important role in many neurodegenerative diseases (51), this association may offer a new theoretical basis for the explanation of the pathological role of protein aggregation in diseases associated with premature cell death.

We show that a reporter protein attached to an aggregating peptide can alter the distribution and composition of protein aggregates. In particular, fusion of the aggregating CFTR C-terminal construct to GFP excludes this protein from small discrete accumulations formed by untagged CFTR-derived peptides, and associates the protein with molecular chaperone HSP27. This indicates that GFP, although an important tool for studying the protein localization in the cell (52), has serious limitations when analysis of protein aggregation is considered. This conclusion finds additional support in reports describing an increase in protein aggregation rate (49), and interference with an ordered aggregation process (53), as the result of fusion to GFP.

Analysis of co-aggregation between related and unrelated self-aggregating proteins is an important test of specificity of aggregation process. We show that co-expression of CFTR- and huntingtin-derived peptides in the same cell results in formation of separate inclusion bodies, with each of them containing only one type of aggregating protein. In contrast, closely related peptides derived from the C-terminal portion of CFTR are able to co-aggregate. Interestingly, the inability of different misfolded proteins to co-aggregate or form a specific type of aggregates is accompanied by differences in association with molecular chaperones. For example, the aggregating huntingtin-derived construct associates with HSC70, a chaperone absent from aggregates formed by CFTR-derived C-terminal peptides. In contrast, HSP27 associates exclusively with the GFP-tagged CFTR C terminus, which suggests that this chaperone may be responsible for GFP-mediated alterations in aggregation process. It is not known whether association with specific chaperones can alter the pattern of aggregate distribution, although many heat shock proteins are known to prevent the aggregation of misfolded proteins (37, 49, 54, 55) and some have been reported to facilitate aggregation (56, 57). Therefore, future studies need to elucidate the relationship between these two processes.

In summary, our results demonstrate that intracellular protein aggregation results from selective association of misfolded molecules. This observation is critical for our understanding of the pathogenic role of protein aggregation. Although it is likely that certain metabolic alterations result from a nonspecific response to protein aggregation (58), sequence- or structure-specific intermolecular interactions may generate unique abnormalities that lead to the appearance of disease-specific symptoms. Characterization of these intermolecular interactions may reveal potential avenues for preventing aggregate formation in disease states.

	ACKNOWLEDGEMENTS

We thank Dr. Christopher A. Ross and Dr. Matthew F. Peters for the huntingtin-derived N63-Q75-Myc-His construct, and Dr. Abigail S. Hackam for comments on the manuscript.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grant DK44003 (to G. R. C.) and Komitet Badan Naukowych Grant PBZ/KBN/042/P05/06 (to M. I. M.).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.

To whom correspondence should be addressed: Inst. of Genetic Medicine, Johns Hopkins University School of Medicine, 600 N. Wolfe St., CMSC 9-123, Baltimore, MD 21287. Tel.: 410-614-0212; Fax: 410-614-0213; E-mail: gcutting@jhmi.edu.

Published, JBC Papers in Press, June 25, 2002, DOI 10.1074/jbc.M205420200

	ABBREVIATIONS

The abbreviations used are: CF, cystic fibrosis; CFTR, cystic fibrosis transmembrane conductance regulator; ER, endoplasmic reticulum; HA, hemagglutinin; DAPI, 4,6-diamidino-2-phenylindole; PBS, phosphate-buffered saline; GFP, green fluorescent protein; ABC, ATP-binding cassette; HSC, heat shock cognate; NBF, nucleotide binding fold; HSP, heat shock protein.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				H. Kawate, Y. Wu, K. Ohnaka, R.-H. Tao, K.-i. Nakamura, T. Okabe, T. Yanase, H. Nawata, and R. Takayanagi Impaired Nuclear Translocation, Nuclear Matrix Targeting, and Intranuclear Mobility of Mutant Androgen Receptors Carrying Amino Acid Substitutions in the Deoxyribonucleic Acid-Binding Domain Derived from Androgen Insensitivity Syndrome Patients J. Clin. Endocrinol. Metab., November 1, 2005; 90(11): 6162 - 6169. [Abstract] [Full Text] [PDF]

				A. M. van Heeckeren, M. D. Schluchter, M. L. Drumm, and P. B. Davis Role of Cftr genotype in the response to chronic Pseudomonas aeruginosa lung infection in mice Am J Physiol Lung Cell Mol Physiol, November 1, 2004; 287(5): L944 - L952. [Abstract] [Full Text] [PDF]

				S. M. Doyle, E. Anderson, K. N. Parent, and C. M. Teschke A Concerted Mechanism for the Suppression of a Folding Defect through Interactions with Chaperones J. Biol. Chem., April 23, 2004; 279(17): 17473 - 17482. [Abstract] [Full Text] [PDF]

				Z. Ignatova and L. M. Gierasch Monitoring protein stability and aggregation in vivo by real-time fluorescent labeling PNAS, January 13, 2004; 101(2): 523 - 528. [Abstract] [Full Text] [PDF]

				C. B. Pedersen, P. Bross, V. S. Winter, T. J. Corydon, L. Bolund, K. Bartlett, J. Vockley, and N. Gregersen Misfolding, Degradation, and Aggregation of Variant Proteins: THE MOLECULAR PATHOGENESIS OF SHORT CHAIN ACYL-CoA DEHYDROGENASE (SCAD) DEFICIENCY J. Biol. Chem., November 28, 2003; 278(48): 47449 - 47458. [Abstract] [Full Text] [PDF]

				A. Ben-Ya'acov and S. Ravid Epidermal Growth Factor-mediated Transient Phosphorylation and Membrane Localization of Myosin II-B Are Required for Efficient Chemotaxis J. Biol. Chem., October 10, 2003; 278(41): 40032 - 40040. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/37/34462    most recent M205420200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map 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 Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Milewski, M. I. Articles by Cutting, G. R. Search for Related Content PubMed PubMed Citation Articles by Milewski, M. I. Articles by Cutting, G. R. Social Bookmarking                      What's this?