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

This Article Full Text (PDF) All Versions of this Article: 18/10/1135 04-1493fjev1    most recent 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 NEMES, Z. Articles by FÉSÜS, L. Search for Related Content PubMed PubMed Citation Articles by NEMES, Z. Articles by FÉSÜS, L.

Cross-linking of ubiquitin, HSP27, parkin, and -synuclein by -glutamyl--lysine bonds in Alzheimer’s neurofibrillary tangles

ZOLTÁN NEMES^*,1, B. DEVREESE, P. M. STEINERT, J. VAN BEEUMEN and L. FÉSÜS

Departments of
^* Psychiatry and
Biochemistry and Molecular Biology and Signaling and Apoptosis Research Group, Hungarian Academy of Sciences, Research Center for Molecular Medicine, University of Debrecen, Debrecen, Hungary;
Laboratory of Protein Biochemistry and Protein Engineering, Ghent University, Gent, Belgium; and
Laboratory of Skin Biology, NIAMS, NIH, Bethesda, Maryland, USA

¹Correspondence: Department of Psychiatry, University of Debrecen Medical and Health Sciences Center, Nagyerdei krt. 98. H-4012 Debrecen, Hungary. E-mail: znemes{at}dote.hu

SPECIFIC AIMS

1) To analyze the frequency of the -glutamyl--lysine (GGEL) cross-linkages in Alzheimer’s disease (AD) brain tissue and isolate disease-related cross-linked protein aggregates to determine the density of GGEL cross-links in those. AD specimen were compared with age-matched nondemented and younger non-neurological controls.

2) To identify the proteins and specific amino acid residues involved in GGEL cross-linking and estimate their relative contribution to the total of cross-links in intraneuronal protein aggregates.

PRINCIPAL FINDINGS

1. Frequency of GGEL cross-links in AD cortex is higher than in controls
We measured the relative abundance of GGEL bonds in brain proteins. Control specimens from young and middle-aged adults contained 11–29 GGEL cross-links/10⁹ amino acids, and differences between hippocampus, frontal, and orbital cortex were not significant (P>0.05). In the age-matched (to the AD group) nondemented (AMND) control group, GGEL density was higher in tissue proteins than in either sample of the younger control group. The frequency of GGEL bonds was highest in the hippocampus and showed statistically significant (P<0.01) differences in the order hippocampus > frontal cortex > occipital cortex. AD specimen showed the same ranking, and the mean abundance of GGEL cross-links was >3-fold higher in the hippocampus and frontal cortex; occipital tissue showed only modest elevation of cross-link content (P=0.03) compared with AMND specimens.

2. Accumulation of GGEL cross-links in AD is confined to the chaotrope-insoluble brain proteins
The guanidine-thiocyanate (TriPure, Roche) insoluble fraction of brain proteins showed the same statistically significant differences between the sample groups as noted for the total (soluble+insoluble) tissue proteins, but the frequency of GGEL cross-links in the guanidine-HSCN soluble proteins was not statistically different between brain regions or between the young, AMND, and AD patients. This indicates that accumulation of GGEL is confined to the insoluble brain proteins accounting for 1.3–2.1% of total tissue proteins. The highest relative abundance of GGEL within the insoluble fraction reached 57/10⁶ amino acids in the hippocampus of an AD patient.

3. Neurodegeneration-associated cross-links copurify with ubiquitin immunopositive protein aggregates
Further fractionation of the chaotrope-insoluble particulate fraction of brain cortex specimens was accomplished by immunocapturing ubiquitylated protein particles (UPP) with anti-ubiquitin antibodies and separation of the ubiquitin-immunoreactive material by magnet-assisted sorting. From the younger control brains we could not purify sufficient quantities of UPP to subject them to further biochemical or immunological analysis. The frequency of GGEL cross-links in the nonubiquitylated residual proteins was not statistically different between brain regions or between young, AMND, and AD patients. Assuming that losses in GGEL during the purification of UPP were similar to that of nonubiquitylated particulate material, the accumulation of GGEL in AD and AMND brains is clearly confined to the UPP fraction (Fig. 1 ). The UPP from AD brains contained 1527–1955 (mean 1793±176 SD, n=4)/10⁶, and those from AMND controls 1969–2614 (mean 2311±285 SD, n=6)/10⁶ GGEL cross-links. No statistically significant differences were seen between UPP from hippocampi and (pooled) neocortex.

View larger version (29K): [in this window] [in a new window]   Figure 1. The insoluble particulate fraction remaining after exhaustive chaotropic extraction of brain proteins was immunoseparated by ubiquitin antiserum. Distribution of the total tissue GGEL content between solubilizable ubiquitylated protein particles (UPP) and other fractions reveals that the accumulation of GGEL cross-links in aging and AD is confined to the ubiquitylted insoluble proteins.

4. Antigenicity of UPP is similar to neurofibrillary tangles
We explored the immunophenotype of UPP using polyclonal antibodies to common neurofibrillary antigens. UPP from AD brains showed elongated, angular structures, reaching up to 2–3 µm and having sharp margins, which showed strong positivity with anti-tau and anti-neurofilament antibodies but were only weakly stained with anti--synuclein antiserum. UPP from AMND controls, on the contrary, were typically roundish particles of 0.5–1 µm diameter revealing strong immunopositivity to tau, neurofilament, and -synuclein antibodies. Such particles were also present in AD specimens. These findings indicate that Alzheimer’s neurofibrillary tangles and UPP harbor common epitopes and are likely to be related structures; however, association of -synuclein and tau positivity is more typical for Lewy bodies.

5. GGEL cross-linkages in UPP are identical in AD and AMND and connect ubiquitin and parkin to HSP27 and -synuclein
UPP from hippocampus and cortex were fragmented into soluble peptide fragments. The resulting mixture of straight and cross-linked peptides was fractioned by HPLC and time fractions showing GGEL content were further enriched by affinity purification using a monoclonal antibody against GGEL. GGEL cross-linked peptides were identified by their two amino termini using trimethylaminoacetyl modification of unprotonated amino groups. Six tryptic and six V8 peptide masses were identified as being cross-linked (Table 1 ). These were sequenced using electrospray ionization MS/MS. Remarkably, identical sequences were found in cross-linked peptides of both AD and AMND control specimen. Sequencing of GGEL cross-linked tryptic peptide pairs revealed two concurrent legible sequences in all cases. A search of the found sequences against human protein and EST databases yielded only one full match for each of the 12 GGEL cross-linked sequences if hypothetical proteins and entries containing fragments with high homology to fully sequenced proteins were ignored.

Our data show that two lysines of the ubiquitin (Ub) protein used to couple ubiquitins to each other, Ub-Lys²⁹ and Ub-Lys⁴⁸, are used for transglutaminase-mediated GGEL cross-linking. Lys⁴⁸ (P-Lys⁴⁸), a structural position equivalent of Ub-Lys⁴⁸ in the ubiquitin-homologous domain of parkin proteins, was cross-linked. Glutamine donor sequences for the cross-linking are provided by two sites of small heat shock protein HSP27 (HSP27-Gln³¹ and HSP27-Gln¹⁹⁰) and one of -synuclein (S-Gln⁹⁹).

6. The relative abundance of cross-links involving parkin and -synuclein is less in UPPs from AD brains than AMND controls
HSP27-Gln³¹ + Ub-Lys²⁹ bonds were most frequent in the four sample groups and accounted for more cross-links than all other types combined The second most frequent cross-link was HSP27-Gln¹⁹⁰ + Ub-Lys⁴⁸ in AD and controls, accounting for 26–38% of the cross-links. Bonds involving S-Gln⁹⁹ and P-Lys⁴⁸ were 7–12% and 5–8% of AMND controls but made up only 2–5% and 1–3% of GGEL cross-links in AD, respectively.

CONCLUSIONS AND SIGNIFICANCE

Cumulative evidence supports the notion that the impairment of the ubiquitin-proteasome system (UPS) is more likely to account for aggregate deposition and cellular pathology than the existence of unusually strong cohesion between specific misfolded proteins. Inhibition of the UPS appears downstream of ubiquitin attachment mechanisms ("quality control"), as both pathological and experimental inclusions carry massive ubiquitin labeling. The overload, block of access to or inhibition of proteasomes may be key features for the formation of ubiquitylated neuronal inclusions encountered in normal senescence as well as neurodegenerative diseases

Here we show that GGEL cross-linking of brain proteins is of magnitude fold increase in AD cortex. Our data also indicate that transglutamination of proteins is not specific to AD, as it is apparent in senile neuronal degeneration and shows only quantitative differences with the latter.

We found that the increased cross-link formation in brains affected by aging or AD is confined to organelle-sized, ubiquitylated protein inclusions bearing remarkable similar antigenicity to neurofibrillary tangles. Though these particles seemed bigger and more elongated in AD than in aged controls and -synuclein immunopositivity was markedly weaker in the elongated particles than in smaller (AMND-like) ones, we did not recognize any qualitative difference between these structures. Senile UPP might therefore mature into AD-like UPP with time. However, it is possible that small particles with strong -synuclein immunopositivity represent a different pool of UPP that might develop into bigger, Lewy body-like structures by further accumulation of -synuclein protein, as -synuclein is abundantly present in Lewy bodies.

We allocated six GGEL cross-links of UPP to four proteins: ubiquitin, HSP27, -synuclein, and parkin. Our most striking observation was that putative core proteins of neurofibrillary deposits, like tau or neurofilament components, were not directly cross-linked. Our data reveal that not only the protein sequence of aggregated proteins may branch by -glycyl--lysine isopeptide bonds, and ubiquitin itself by forming -glycyl⁷⁶-lysine^11/29/48/63 cross-links leading to polyubiquitin chains, but that also GGEL isopeptide cross-links are also present between ubiquitin and HSP27 or -synuclein.

The cross-linkage of P-Lys⁴⁸ and of S-Gln⁹⁹ may indicate a role for parkin and -synuclein as chain terminators, an end point in growth of cross-linked aggregates, as these proteins revealed only one potential cross-linkable residue.

The abundance of parkin and -synuclein cross-links was markedly less in AD compared with elderly controls, but the relative frequency of Ub + HSP27 cross-links was similar for AD and senile UPP. This may be an argument to postulate qualitative differences between AD and senile neuropathology.

Our data suggest that GGEL cross-linking of the (poly)-ubiquitin chains of waste proteins by transglutaminases are a decisive mechanism for the stabilization of these aggregates, possibly rendering them resistant to proteasomal degradation and causing neuronal pathology; covalently branched meshworks of proteins may be too large or tangled to enter the proteolytic cavity.

View larger version (38K): [in this window] [in a new window]   Figure 2. Model for the assembly of intraneuronal protein aggregates. Unfolded and aggregated waste proteins are ubiquitylated by N(-glycyl)lysine isopeptide bonds (blue); ubiquitin (yellow) is assembled into polyubiquitin chains by similar bonds. Transglutamination forms GGEL isopeptide bonds (red) between two polyubiquitin chains via HSP27 protein (green). The monovalent cross-linking partners parkin (lilac) and -synuclein (pink) act as chain terminators and limit the size of the isopeptide-bonded meshworks.

FOOTNOTES

To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.04-1493fje; doi: 10.1096/fj.04-1493fje

This article has been cited by other articles:

				D. Cripps, S. N. Thomas, Y. Jeng, F. Yang, P. Davies, and A. J. Yang Alzheimer Disease-specific Conformation of Hyperphosphorylated Paired Helical Filament-Tau Is Polyubiquitinated through Lys-48, Lys-11, and Lys-6 Ubiquitin Conjugation J. Biol. Chem., April 21, 2006; 281(16): 10825 - 10838. [Abstract] [Full Text] [PDF]

				R. A. Halverson, J. Lewis, S. Frausto, M. Hutton, and N. A. Muma Tau Protein Is Cross-Linked by Transglutaminase in P301L Tau Transgenic Mice J. Neurosci., February 2, 2005; 25(5): 1226 - 1233. [Abstract] [Full Text] [PDF]

This Article Full Text (PDF) All Versions of this Article: 18/10/1135 04-1493fjev1    most recent 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 NEMES, Z. Articles by FÉSÜS, L. Search for Related Content PubMed PubMed Citation Articles by NEMES, Z. Articles by FÉSÜS, L.