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

This Article Abstract Full Text Full Text (PDF) All Versions of this Article: fj.06-5911fjev1 20/12/2168    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 Fontaine, J.-M. Articles by Benndorf, R. Search for Related Content PubMed PubMed Citation Articles by Fontaine, J.-M. Articles by Benndorf, R.

Abnormal small heat shock protein interactions involving neuropathy-associated HSP22 (HSPB8) mutants

Jean-Marc Fontaine^*, Xiankui Sun^*, Adam D. Hoppe, Stephanie Simon, Patrick Vicart, Michael J. Welsh^* and Rainer Benndorf^*,1

Departments of

^* Cell and Developmental Biology, and

Microbiology and Immunology, University of Michigan Medical School, Ann Arbor, Michigan, USA; and

EA 300 Stress et Pathologies du Cytosquelette, Université Paris 7, UFR de Biochimie, Paris, France

¹Correspondence: Department of Cell and Developmental Biology, University of Michigan Medical School, 109 Zina Pitcher Place, Ann Arbor, MI 48109-2200, USA. E-mail: rbenndo{at}umich.edu

SPECIFIC AIMS

Mutations in the small heat shock protein HSP22, ^K141EHSP22 and ^K141NHSP22, cause motor neuron degeneration in the human diseases distal hereditary motor neuropathy (dHMN) type II and Charcot-Marie-Tooth disease (CMT) type 2L. Small heat shock proteins (sHSP) interact with one another, and our aim was to determine whether changes occurred in the interaction properties of both forms of mutant (mu) HSP22 with themselves, with wild-type (wt) HSP22, and with the other sHSPs HSP27, HSP20, and B-crystallin (B-Cry), all of which are abundant in neurons.

PRINCIPAL FINDINGS

1. Disease-associated ^muHSP22 has an increased tendency to form intracellular protein aggregates
Mutant proteins, in general, have an increased tendency to form cytoplasmic protein aggregates. Formation of aggregates involving ^muHSP22 may result from aberrant interaction properties (see below), and it also may interfere with the applied methods to determine protein interactions. Therefore, the tendency of the two known ^muHSP22 proteins to form aggregates was examined in transfected COS-7 cells by determining the proportion of cells containing protein aggregates. In these assays, fusion proteins were used consisting of the various sHSP species fused to derivatives of the green fluorescent protein, citrine (CIT), or cyan fluorescent protein (CFP). Twenty-four hours after transfection, the proportion of cells containing aggregates was quantified. Expression of ^K141EHSP22-CIT and ^K141NHSP22-CIT resulted in a slight (21%) and great (54%), respectively, increase in the proportion of cells with aggregates, as compared to the ^wtHSP22-CIT control (17%), which defines the baseline level for this construct. Thus, both ^muHSP22 forms had, although to a different extent, an increased tendency to form aggregates as compared to ^wtHSP22.

Mutant proteins can recruit wild-type proteins into aggregates, while wild-type proteins can attenuate aggregate formation by mutant proteins. To evaluate this mutual relationship, experiments were conducted in which both ^muHSP22 forms and, for control also, ^wtHSP22 (all as CIT fusion proteins) were coexpressed with ^wtHSP22, HSP20, B-Cry, and ^wtHSP27 (all as CFP fusion proteins). Both ^muHSP22-CIT forms and also ^wtHSP22-CIT were found to recruit all tested other sHSP-CFP proteins (HSP20, B-Cry, ^wtHSP27, ^wtHSP22) into aggregates. These ^wtsHSPs also significantly attenuated the formation of aggregates by both ^muHSP22-CIT and ^wtHSP22-CIT. However, the obtained patterns were different for each tested sHSP. For example, aggregate formation by ^K141NHSP22-CIT was more effectively attenuated by HSP20 and B-Cry than by ^wtHSP22 or ^wtHSP27, while aggregate formation by ^wtHSP22-CIT was most effectively attenuated by B-Cry and ^wtHSP27.

Collectively, these data indicate that both ^muHSP22 proteins have an increased tendency to form aggregates and that aggregation can be attenuated by ^wtsHSPs. The two ^muHSP22 proteins differ in their aggregate formation tendencies and in their responsiveness to attenuation by ^wtHSP22 and other ^wtsHSPs, thus indicating different properties of the two ^muHSP22 forms.

These data also demonstrate that in all settings tested, a significant proportion of cells do not form aggregates by 24 h after transfection. In these cells, the sHSP-CIT/CFP fusion proteins show a relatively even cytoplasmic distribution. Such cells were selected for qFRET measurements, as described below.

2. Disease-associated ^muHSP22 shows abnormal interaction with itself and with ^wtHSP22
We have determined the ability of the two known ^muHSP22 forms to interact with themselves and with ^wtHSP22 using the yeast two-hybrid (TH) method. The TH experiments indicated activation of the reporter genes in all interactions tested (^wtHSP22/^wtHSP22; ^K141EHSP22/^wtHSP22; ^K141NHSP22/^wtHSP22; ^K141EHSP22/^K141EHSP22; ^K141NHSP22/^K141NHSP22). Thus, all HSP22 species interacted with one another. Within the limits of this method, no differences in the interaction intensities were observed as compared to the ^wtHSP22/^wtHSP22 control interaction.

To determine differences in the binding stoichiometry in these interactions, the more sensitive in vivo quantitative fluorescence resonance energy transfer (qFRET) method was applied. COS-7 cells were doubly transfected with pairs (CIT, CFP) of the various forms of HSP22 fusion protein cDNAs. The apparent average fluorescence resonance energy transfer efficiency (AAFE) values obtained for all tested interactions were significantly different from the negative control, thus indicating interaction (Fig. 1 ). The AAFE for the ^K141EHSP22/^wtHSP22 interaction was similar to that of the ^wtHSP22/^wtHSP22 interaction, while the AAFE of the ^K141EHSP22/^K141EHSP22 interaction was moderately, though significantly, increased as compared to the ^wtHSP22/^wtHSP22 interaction. In contrast, ^K141NHSP22 showed a strongly (approximately twofold) increased AAFE in the interactions with both ^wtHSP22 and itself, as compared to ^wtHSP22.

View larger version (17K): [in this window] [in a new window]   Figure 1. qFRET measurements of the interactions of K141EHSP22 and K141NHSP22 with wtHSP22 and with themselves in doubly transfected COS-7 cells 24 h after transfection. The AAFE was determined in cells expressing the various HSP22 species as indicated. The HSP22 species were used as CIT (top row of the abscissa legend) and CFP fusion proteins (bottom row of the abscissa legend). All sample values were significantly different from the negative control (expression of CIT and CFP). Significant differences from the wtHSP22/wtHSP22 (positive control) interaction (*), from the wtHSP22/K141EHSP22 interaction (+), and from the K141EHSP22/K141EHSP22 interaction (x) are indicated where appropriate. Abbreviations: C, control; wt, wtHSP22; E, K141EHSP22; N, K141NHSP22.

Cross-linking was also used to determine homodimer and oligomer formation of the various HSP22 species. HEK-293T cells were transfected with vectors to express myc-tagged ^wtHSP22, ^K141EHSP22, or ^K141NHSP22. Forty-eight hours later, cells were treated with disuccinimidyl suberate, a homobifunctional cross-linker. The analysis of the cross-linked proteins by SDS-PAGE/Western blotting revealed that ^wtHSP22, ^K141EHSP22, and ^K141NHSP22 form homodimers and possibly tetramers. No major differences between ^wtHSP22 and ^muHSP22 were observed.

Collectively, the TH, qFRET, and cross-linking data suggest that the ^muHSP22 proteins interact with ^wtHSP22 and themselves. Differences between the ^muHSP22 forms and ^wtHSP22 could be demonstrated by the qFRET method due to its greater sensitivity.

3. Disease-associated ^muHSP22 shows abnormal interaction with other sHSPs
HSP22 is known to interact with HSP20, B-Cry, and ^wtHSP27 that are known to be abundant in neuronal cells. To determine potentially aberrant interaction properties with these sHSPs, both ^muHSP22 forms were probed in TH and qFRET assays. The TH data suggested that both ^K141EHSP22 and ^K141NHSP22 interact with HSP20, B-Cry, and ^wtHSP27. Within the limits of this method, no differences in the interaction intensities between ^muHSP22 and ^wtHSP22 were observed.

The qFRET analysis in doubly transfected COS-7 cells was performed as described above using CIT- and CFP-sHSP fusion proteins. The AAFE values for both ^K141EHSP22 and ^K141NHSP22 with HSP20 as interacting partner were not significantly different from that of ^wtHSP22, suggesting that the mutations do not affect this interaction (Fig. 2 , HSP20 group). In contrast, the AAFE values for both ^K141EHSP22 and ^K141NHSP22 with B-Cry as interacting partner were moderately, although significantly, increased (B-Cry group) as compared with ^wtHSP22, suggesting that this interaction is affected by both mutations. The AAFE values for both ^muHSP22 proteins were similar when compared to each other. Finally, the AAFE values for ^K141EHSP22 and ^K141NHSP22 were moderately and strongly increased, respectively, with HSP27 as interacting partner (HSP27 group), as compared with ^wtHSP22. Thus, both mutations result in increased interaction with HSP27, although to a different extent.

View larger version (16K): [in this window] [in a new window]   Figure 2. qFRET measurements of the interactions of K141EHSP22, K141NHSP22, and wtHSP22 (positive control) with HSP20, B-Cry, and wtHSP27 in doubly transfected COS-7 cells 24 h after transfection. The AAFE was determined in cells coexpressing wtHSP22-CIT (open columns), K141EHSP22-CIT (gray columns), or K141NHSP22-CIT (solid columns) with HSP20-CFP, B-Cry-CFP, and wtHSP27-CFP. All sample values were significantly different from the negative control (expression of CIT and CFP). Significant differences within each group (*, muHSP22 vs. wtHSP22; +, K141NHSP22 vs. K141EHSP22) are indicated. Abbreviations are as in Fig. 1 .

4. Disease-associated ^S135FHSP27 shows abnormal interaction with ^wtHSP22
dHMN and CMT are clinically and genetically heterogeneous groups of neuropathies. In addition to mutations in HSP22, several mutations in HSP27 (^muHSP27) have been identified that are also associated with dHMN and CMT. Because ^muHSP27 may affect the HSP22/HSP27 interaction in a manner similar to^muHSP22, one of the known ^muHSP27 forms (^S135FHSP27) was also included in these experiments. The TH data showed that ^S135FHSP27 interacts with ^wtHSP22, in a manner similar to ^wtHSP27.

For the qFRET analysis, COS-7 cells were doubly transfected to express ^S135FHSP27-CIT or ^wtHSP27-CIT (positive control), together with ^wtHSP22-CFP. The AAFE values obtained for the interactions of ^S135FHSP27 and ^wtHSP27 with ^wtHSP22 were significantly different from the negative control, thus indicating interaction. The AAFE for the ^S135FHSP27/^wtHSP22 interaction was significantly greater than for the ^wtHSP27/^wtHSP22 interaction. Thus, this ^muHSP27 form affects the HSP22/HSP27 interaction in a similar way as the two studied ^muHSP22 forms do.

CONCLUSION AND SIGNIFICANCE

This work shows that both disease-associated forms of ^muHSP22 interacted with ^wtHSP22, with themselves, and with the other sHSPs HSP20, B-Cry, and HSP27. The stoichiometry of some of the analyzed interactions was increased, while others remained unchanged. None of the interactions was weakened. Each of the two ^muHSP22 forms had a characteristic pattern of aberrant interactions. The interaction characteristics of ^K141NHSP22 deviated more from ^wtHSP22 than those of ^K141EHSP22. Whether this correlates to the clinical phenotype, e.g., the severity of the associated diseases, remains to be determined. The aberrant interactions of ^muHSP22 proteins may relate to their increased ability to form aggregates. The more aberrant interaction properties of ^K141NHSP22 may explain the greatly increased formation of aggregates. One of the dHMN- and CMT-associated HSP27 mutants (^S135FHSP27) was also found to have an abnormally increased interaction with ^wtHSP22. A summary of the identified abnormal interactions of ^K141EHSP22, ^K141NHSP22, and ^S135FHSP27 is given in Fig. 3 A, B, and C, respectively.

View larger version (6K): [in this window] [in a new window]   Figure 3. Schematic of identified abnormal interactions at the dimer level involving disease-associated K141EHSP22, K141NHSP22, and S135FHSP27. Single interactions were found unchanged (???), moderately increased (???), or greatly increased (???) in comparison with the corresponding wild-type sHSPs in the same interaction. A) K141EHSP22 has moderately increased interactions with itself, B-Cry, and wtHSP27. B) K141NHSP22 has greatly increased interactions with itself, wtHSP22, and wtHSP27, and moderately increased interaction with B-Cry. C) S135FHSP27 has moderately increased interaction with wtHSP22. Note that each muHSP22 form has a characteristic, abnormal interaction pattern. That muHSP22 has increased interaction with wtHSP27 and that S135FHSP27 has increased interaction with wtHSP22 and mutations in HSP22 or HSP27 resulting in a similar disease suggest that both mutant proteins may act through the same mechanism.

The facts that HSP22 and HSP27 are interacting proteins and that either ^muHSP22 or ^muHSP27 result in similarly increased interaction between both proteins, support the hypothesis that mutations in either protein affect the same pathway or protein complex. All known mutations in HSP22 and HSP27 have dominant gain-of-function characteristics, and the presence of the ^wtsHSPs in the diseased tissues cannot prevent the development of the diseases. A plausible explanation is that insertion of only a few mutated protein molecules into the sHSP complexes may result in the formation of malstructured and thus malfunctioning complexes. Thus, the model based on aberrant sHSP interactions may provide the rationale for the genetic dominance as is seen in these diseases.

In summary, the data presented here may provide the rationale for the pathogenesis of the mutant sHSP-associated motor neuropathies dHMN and CMT. The molecular and cellular events by which the abnormal interactions and aggregate-forming properties of ^muHSP22 translate finally into the slow death of motor neurons remain to be elucidated.

FOOTNOTES

To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.06-5911fje

This article has been cited by other articles:

				H. Houlden, M. Laura, F. Wavrant-De Vrieze, J. Blake, N. Wood, and M. M. Reilly Mutations in the HSP27 (HSPB1) gene cause dominant, recessive, and sporadic distal HMN/CMT type 2 Neurology, November 18, 2008; 71(21): 1660 - 1668. [Abstract] [Full Text] [PDF]

				P A James, J Rankin, and K Talbot Asymmetrical late onset motor neuropathy associated with a novel mutation in the small heat shock protein HSPB1 (HSP27) J. Neurol. Neurosurg. Psychiatry, April 1, 2008; 79(4): 461 - 463. [Abstract] [Full Text] [PDF]

				J. Zhai, H. Lin, J.-P. Julien, and W. W. Schlaepfer Disruption of neurofilament network with aggregation of light neurofilament protein: a common pathway leading to motor neuron degeneration due to Charcot Marie Tooth disease-linked mutations in NFL and HSPB1 Hum. Mol. Genet., December 15, 2007; 16(24): 3103 - 3116. [Abstract] [Full Text] [PDF]

				S. Simon, J.-M. Fontaine, J. L. Martin, X. Sun, A. D. Hoppe, M. J. Welsh, R. Benndorf, and P. Vicart Myopathy-associated {alpha}B-crystallin Mutants: ABNORMAL PHOSPHORYLATION, INTRACELLULAR LOCATION, AND INTERACTIONS WITH OTHER SMALL HEAT SHOCK PROTEINS J. Biol. Chem., November 23, 2007; 282(47): 34276 - 34287. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text Full Text (PDF) All Versions of this Article: fj.06-5911fjev1 20/12/2168    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 Fontaine, J.-M. Articles by Benndorf, R. Search for Related Content PubMed PubMed Citation Articles by Fontaine, J.-M. Articles by Benndorf, R.