Basic Protein Characterisation

The above-mentioned expression plasmids were used to transfect different mammalian cell lines (HEK293T, MRC5, HeLa, COS-7, etc), obtaining similar results in all cases. The addition of a tag (myc + His, EGFP, etc) at the C-terminus of chorein seems not to affect the protein. The apparent size of the bands detected by WB agrees with the expected size for the full-length proteins in all cases.

When assayed by IF, chorein presents several patterns, depending on individual cells (Fig. 2). The most characteristic is a vesicular-like pattern that is very easy to detect, although not all transfected cells present it. The second pattern suggests a cytoplasmic localisation. No general co-localisation with different sub-cellular markers was detected. The "vesicles" are not aggresomes [18], structures originat-

Fig. 2 Subcellular localisation of chorein. HEK293T cells overexpressing chorein with a myc + His tag at the C-terminus and detected with a monoclonal antibody against the myc tag. The cell on the right shows a typical vesicular-like pattern (see text)

ing after the aggregation of misfolded proteins (data not shown). In fact, similar structures can sometimes be detected in human cell lines with the chorein-specific antibody, suggesting that this result is not an artefact due to over-expression. For the rest of the human VPS13 proteins, a similar cytoplasmic localisation is detected with the over-expressed tagged proteins, and only with VPS13B are some vesiclelike structures detected, but at a much lower rate than for chorein.

The yeast homologue Vps13p has been reported to form high molecular weight complexes [2]. Co-IP experiments gave negative results for the detection of mul-timerisation of chorein and the other human VPS13 proteins, although this does not necessarily mean that such complexes are not formed.

One of the many unresolved issues about these proteins, the presence or absence of transmembrane domains (see Sect. 2.3), can now be addressed. Chorein can be detected in the soluble fraction of a protein lysate after high-speed centrifugation (data not shown), which indicates that it is a soluble and not an integral protein. It can also be detected in the precipitate, and it can be partially solubilised from this precipitate in several conditions, suggesting that it interacts with membranes. The same result was obtained with both over-expressed and endogenous chorein and with the other human VPS13 proteins, and it is consistent with the peripheral membrane-associated location reported for their yeast homologue Vps13p. This implies that the proposed model of COH1/VPS13B containing ten TM domains [21] would not be probable; this is an example of the care that must be taken with predictions obtained in silico.

Five missense ChAc mutations (1, 2, 3, 4, and 6 in Table 2) were introduced in a chorein expression plasmid by site-directed mutagenesis and the mutant proteins assayed by IF and WB. Protein of the normal expected size was detected in all cases by WB. The "vesicular" sub-cellular localisation pattern, however, was only detected with the p.Y2721C mutant but not with any of the other four mutants (data not shown). These results suggest that mutations p.I90K, p.A1095P, p.S1452P, and p.W2460R alter the ability of chorein to attach to membranes and, therefore, that this process is probably very important for the function of chorein. However, we have to be careful with this interpretation. As previously discussed (Sect. 2.4), chorein is detected at low levels in samples from patients homozygous for the c.4354T>C mutation (missense mutation leading to p.S1452P) and the effect of this mutation could affect the stability of chorein or maybe alter the normal splicing of the VPS13A gene. The result obtained with over-expressed p.S1452P may not therefore be the actual cause of disease. The fact that this mutant protein is easily detected suggests that stability is unaffected, implying that an effect on splicing may be more likely in vivo. In this case, the result will be absence of chorein or a deletion protein but not the S1452P substitution. The sub-cellular localisation defect found in p.S1452P could be due to a structural change introduced by the substitution, a common effect of proline residues. A similar explanation may apply for the p.A1095P mutation, i.e., an alteration of the secondary structure of chorein, while mutations p.I90K and p.W2460R might represent modification of key residues in the interaction of chorein with membranes. However, more experimental analyses would be needed to test these hypotheses and without patient cell samples we cannot know if these substitutions also lead simply to a reduction in chorein levels in vivo, as for p.S1452P.

0 0

Post a comment