PROYECTOS PEDAGÓGICOS ADICIONALES

Besides neuronal cells, the baker’s yeast Saccharomyces cerevisiae was used as model systems in this work to study polyQ protein interaction, aggregation and toxicity as described previously in several other studies (Krobitsch and Lindquist, 2000; Muchowski et al., 2000; Meriin et al., 2002). Several independent studies demonstrate that at a basic level, the yeast model faithfully recapitulates the molecular basis of polyQ length-dependent aggregation and toxicity (Colby et al., 2004; Cashikar et al., 2005; Vacher et al., 2005; Zhang et al., 2005). The yeast system offers the unique opportunity to dissect modulators of polyQ aggregation and toxicity in a defined and uniform cellular environment. In fact, an advantage of yeast cells

Results 58 is that they are not confounded by the perplexing variations in cellular proteome that characterize distinct cell types in whole organisms or in cultured mammalian cells. In addition, numerous genetic tools available in yeast make it a powerful instrument to explore the intra- and intermolecular factors that govern polyQ aggregation and toxicity.

To confirm the recruitment of human TBP into Htt aggregates in the yeast system, myc-tagged Htt and GST-Htt fusion constructs as well as HA-tagged human TBP were cloned into yeast expression vectors under control of copper- and galactose-inducible promoters, respectively, and coexpressed in the wild-type strain YPH499 (Sikorski and Hieter, 1989).

Figure 16: Recruitment of human TBP into Htt aggregates in wild-type yeast cells.

(A) myc-tagged Htt20Q, Htt53Q and Htt96Q with or without N-terminal GST fusion under CUP1 control were coexpressed with HA-tagged human TBP under control of GAL1 in wild-type cells for 12 h at 30 o_{C. Lysates}

were analyzed by Western blot (WB) and by filter assay (FA). Htt constructs and recruited human TBP were immunodetected with anti-myc and anti-HA antibody, respectively. * indicates specific protein band, < GST cleavage products and > stacking gel. (B) Cells derived from (A) were analyzed by indirect immunofluorescence with anti-myc or anti-HA antibody coupled either to Cy3- or FITC-conjugated secondary antibodies, respectively. Nuclei were counterstained with DAPI.

Small amounts of Htt20Q were detected as a protein band on SDS-PAGE, while Htt53Q and Htt96Q did not migrate into the gel and were only detectable in the stacking gel (Figure 16A). Consistently, GST-Htt20Q and GST-Htt53Q were produced similarly in substantial amounts and almost exclusively in SDS soluble form. To a minor extent, GST- Htt96Q was detected as a protein band but additionally also in the stacking gel (Figure 16A). Htt53Q and Htt96Q displayed polyQ-length dependent formation of SDS-insoluble

Results 59 aggregates and recruitment of TBP. Htt20Q did neither aggregate nor recruit TBP (Figure 16A). Fusion to GST inhibited aggregation of Htt53Q completely and of Htt96Q to some residual extent. Recruitment of TBP could not be observed for any GST-Htt construct (Figure 16A). These results were confirmed by immunofluorescence analysis: TBP resided in the nucleus when coexpressed with Htt20Q or any of the GST-Htt constructs and redistributed from the nucleus to cytoplasmic Htt aggregates when coexpressed with Htt53Q and Htt96Q (Figure 16B).

Expression of mutant Htt fused to GST was without any effect on the recruitment of TBP (Figure 16A and Figure 16B). These proteins did not form aggregates by themselves, although their polyQ segments were exposed, based on the observation by filter assay that GST-Htt53Q but not GST alone co-aggregated with Htt96Q (Figure 17A). Consistently, GST- Htt53Q coalesced with Htt96Q in aggregates observable by immunofluorescence (Figure 17B). Hence, since mutant Htt when fused to GST is not per se aggregation-incompetent, a conformational rearrangement of Htt occurring upon its release from the GST moiety might be a prerequisite for initiating aggregation and recruitment of TBP. In vitro FRET experiments with GST-Htt53Q indicated that a rapid conformational compaction occurs in the Htt fragment upon its proteolytic release from a protective sequence context (Schaffar et al., 2004).

Figure 17: Accessibility of the Q stretch in GST-Htt fusions protein

(A) myc-tagged Htt96Q under control of CUP1 was coexpressed with HA-tagged GST or GST-Htt53Q under control of GAL1 in wild-type cells for 12 h at 30 °C. SDS-resistant aggregates in lysates were detected by filter assay and immunostaining for Htt with anti-myc and for recruited GST-Htt with anti-GST antibody. (B) Cells derived from (A) were analyzed by indirect immunofluorescence with anti-myc or anti-HA antibody coupled either to Cy3- or FITC-conjugated secondary antibodies, respectively. Nuclei were counterstained with DAPI.

In order to verify whether liberation of mutant Htt from its GST fusion partner initiates its aggregation and recruitment of TBP, GST-Htt fusions were constructed that harbor an internal, highly specific tobacco etch virus (TEV) protease cleavage site between the GST part and the myc-tagged mutant Htt exon 1 fragment. TEV protease was cloned under a

Results 60 doxycycline-controllable promoter and coexpressed with GST-TEV-myc-Htt constructs in wild-type yeast. Expression and proteolytic cleavage of GST-TEV-myc-Htt constructs were analyzed by SDS-PAGE and immunoblotting. GST-TEV-myc-Htt53Q and GST-TEV-myc- Htt96Q were produced to similar levels and migrated according to their nominal size in the gel. Only minor amounts of unspecific cleavage products could be detected in the absence of TEV protease (Figure 18A). Upon induction of TEV protease, protein bands corresponding to GST-TEV-myc-Htt constructs diminished and substantial amounts of cleaved GST accumulated. Concurrently, myc-Htt53Q and myc-Htt96Q were almost exclusively detectable in the stacking gel (Figure 18A). GST-TEV-myc-Htt53Q and GST-TEV-myc-Htt96Q did not form SDS-insoluble aggregates detectable by filter assay unless TEV protease was induced (Figure 18B). Thus, analogous to in vitro experiments (Scherzinger et al., 1997; Schaffar et al., 2004), aggregation of formerly soluble mutant Htt fused to GST can be initiated in vivo in yeast cells by cleavage of the Htt part from the GST moiety. Similarly, aggregation of full- length polyQ expanded Ataxin-3 was demonstrated to be initiated by proteolytical cleavage at an engineered TEV protease cleavage site (Haacke et al., 2006).

Figure 18: Initiation of Htt aggregation upon cleavage of the GST-Htt fusion protein.

(A) GST-TEV-myc-Htt53Q and GST-TEV-myc-Htt96Q were expressed under control of CUP1 in wild-type cells for 12 h at 30o_{C with and without expression of TEV protease from a doxycycline-regulated promoter. Cell}

lysates were analyzed by Western blot, and cleaved and uncleaved GST-Htt constructs were detected with anti- GST and anti-myc antibodies, respectively. * indicates specific protein band, < GST cleavage products and > stacking gel. (B) SDS-resistant aggregates in lysates from (A) were detected by filter assay and immunostaining for Htt constructs with anti-myc antibody.

To determine whether recruitment of TBP could be initiated similarly, proteolytic cleavage of GST-Htt in vivo was analyzed in parallel by filter assay and immunofluorescence. GST-TEV-myc-Htt96Q and HA-tagged human TBP were coexpressed in the absence of TEV protease. Formation of SDS-insoluble aggregates in lysates could be detected neither for

Results 61 Htt96Q nor for TBP. Remarkably, continuous coexpression of GST-TEV-myc-Htt96Q and TBP in the presence of TEV protease but not in its absence led to substantial aggregation of Htt96Q and TBP (Figure 19A). In the absence of TEV protease, GST-TEV-myc-Htt96Q was diffusely distributed throughout the cytosol and TBP located to the nucleus. Consistently, TBP was dislocated from the nucleus and coalesced with Htt96Q in aggregate in the presence of TEV protease (Figure 19B). Hence, mutant Htt acquires, presumably by a structural rearrangement, the ability to aggregate and consequently to recruit TBP upon its release from the GST moiety.

Figure 19: Initiation of TBP recruitment into aggregates upon cleavage of the GST-Htt fusion protein. (A) GST-TEV-myc-Htt96Q and HA-tagged human TBP under control of CUP1 and GAL1,respectively, were coexpressed in wild-type cells for 4 h at 30 °C followed by 2 h with and without induction of TEV protease. SDS-resistant aggregates in lysates from samples taken after 0 and 2 h induction of TEV protease were detected by filter assay and immunostaining for Htt96Q with anti-myc and recruited TBP with anti-HA antibody. (B)

Cells derived form (A) were analyzed by indirect immunofluorescence with anti-myc or anti-HA antibody coupled either to Cy3- or FITC-conjugated secondary antibodies, respectively. Nuclei were counterstained with DAPI.