Figure 1:
Bar diagrams showing the domain structure of tau (longest human isoform, accession nr. AF047863), MAP2 (three repeat isoform from rat, X51842) and MAP4 (mouse, M72414). The MAPs differ in size but contain regions of pronounced homology. The N-terminal projection domain protrudes from the MT surface, hence it can be clipped with chymotrypsin from the protein even when it is bound to MTs. It is of acidic character (red) and ends in a proline-rich region (P). The C-terminal MT binding domain is quite well conserved between the MAPs. It starts in the proline-rich region, is followed by 3 or 4 pseudorepeats of 31 residues each (R1-R4), which are of basic character (blue), followed by another region rich in prolines, and ends in a short acidic or neutral C-terminal tail. The second repeat is missing in some alternatively spliced isoforms, as is the extra repeat (1a) in MAP4. In case of tau and MAP2, the projection domain can contain inserts as a result of alternative splicing (hatched).
|
|
Figure 2:
(a) Microtubules nucleated on sea urchin sperm axonemes in the presence of tau and purified brain MARK without ATP as seen in dark field microscopy. (b) Upon addition of ATP to the system, tau is phosphorylated at Ser262, dissociates from the MT surface, and MTs depolymerize. Figure provided by Bernhard Trinczek, taken with permission from Drewes et al., 1995.
|
|
Figure 3:
All MARK/PAR kinases are composed of a divergent N-terminal header sequence (NH) followed by the catalytic domain in which all family members show considerable homology. Adjacent is a putative membrane-targetting motif (T, sequence EEEELKPY) followed by an ubiquitin-associated (UBA) domain. The large 'spacer' segment is the most divergent. The C-terminal tail shows very pronounced homology between the MARK sequences and par-1, but to a lesser extent with kin1+. The calculated isoelectric points for header, catalytic domain, UBA domain, spacer and tail are 6.3, 9.8, 3.9, 11.0 and 9.8, respectively, showing that the spacer domain is the most basic one. The non-catalytic domains are rather hydrophilic. The spacer domain displays very little predicted secondary structure. Numbering refers to the MARK1 sequence
|
|
Figure 4:
Overexpression of MARK in chinese hamster ovary cells causes microtubule breakdown. Epitope-tagged MARK2 (A-C) or an inactive MARK2 mutant (D-F) were transiently transfected into CHO cells. After 24 hr cells were fixed with formaldehyde and stained with anti-tag (A, D), anti-tubulin antibodies (B, E), and TRITC-phalloidin to visualize actin (C, F). Whereas microtubules are disrupted, the actin stress fibers remain intact. The mutant MARK2 causes no significant microtubule disruption. Bar, 20 µm
|
|
Figure 5:
MARK/PAR kinase family members. (accession numbers are given in parenthesis): MARK1 (Z83868) and MARK2 (Z83869, X97630), cloned from rat and human brain (Drewes et al., unpublished), homologs of MARK2 are canine mPAR and mouse EMK (X70764); MARK3, cloned from rat and human brain (Drewes et al., unpublished) and its homologue, p78 (M80359) from human pancreas; MARK4, cloned from rat and human brain (Drewes et al., unpublished); PAR1 from C. elegans (U22183, U40858), KIN1/KIN2 from S. cerevisiae (M69017, M69018); and kin1+ from S. pombe (M64799).
|
|
Figure 1:
Bar diagrams showing the domain structure of tau (longest human isoform, accession nr. AF047863), MAP2 (three repeat isoform from rat, X51842) and MAP4 (mouse, M72414). The MAPs differ in size but contain regions of pronounced homology. The N-terminal projection domain protrudes from the MT surface, hence it can be clipped with chymotrypsin from the protein even when it is bound to MTs. It is of acidic character (red) and ends in a proline-rich region (P). The C-terminal MT binding domain is quite well conserved between the MAPs. It starts in the proline-rich region, is followed by 3 or 4 pseudorepeats of 31 residues each (R1-R4), which are of basic character (blue), followed by another region rich in prolines, and ends in a short acidic or neutral C-terminal tail. The second repeat is missing in some alternatively spliced isoforms, as is the extra repeat (1a) in MAP4. In case of tau and MAP2, the projection domain can contain inserts as a result of alternative splicing (hatched).
Figure 2:
(a) Microtubules nucleated on sea urchin sperm axonemes in the presence of tau and purified brain MARK without ATP as seen in dark field microscopy. (b) Upon addition of ATP to the system, tau is phosphorylated at Ser262, dissociates from the MT surface, and MTs depolymerize. Figure provided by Bernhard Trinczek, taken with permission from Drewes et al., 1995.
Figure 3:
All MARK/PAR kinases are composed of a divergent N-terminal header sequence (NH) followed by the catalytic domain in which all family members show considerable homology. Adjacent is a putative membrane-targetting motif (T, sequence EEEELKPY) followed by an ubiquitin-associated (UBA) domain. The large 'spacer' segment is the most divergent. The C-terminal tail shows very pronounced homology between the MARK sequences and par-1, but to a lesser extent with kin1+. The calculated isoelectric points for header, catalytic domain, UBA domain, spacer and tail are 6.3, 9.8, 3.9, 11.0 and 9.8, respectively, showing that the spacer domain is the most basic one. The non-catalytic domains are rather hydrophilic. The spacer domain displays very little predicted secondary structure. Numbering refers to the MARK1 sequence
Figure 4:
Overexpression of MARK in chinese hamster ovary cells causes microtubule breakdown. Epitope-tagged MARK2 (A-C) or an inactive MARK2 mutant (D-F) were transiently transfected into CHO cells. After 24 hr cells were fixed with formaldehyde and stained with anti-tag (A, D), anti-tubulin antibodies (B, E), and TRITC-phalloidin to visualize actin (C, F). Whereas microtubules are disrupted, the actin stress fibers remain intact. The mutant MARK2 causes no significant microtubule disruption. Bar, 20 µm
Figure 5:
MARK/PAR kinase family members. (accession numbers are given in parenthesis): MARK1 (Z83868) and MARK2 (Z83869, X97630), cloned from rat and human brain (Drewes et al., unpublished), homologs of MARK2 are canine mPAR and mouse EMK (X70764); MARK3, cloned from rat and human brain (Drewes et al., unpublished) and its homologue, p78 (M80359) from human pancreas; MARK4, cloned from rat and human brain (Drewes et al., unpublished); PAR1 from C. elegans (U22183, U40858), KIN1/KIN2 from S. cerevisiae (M69017, M69018); and kin1+ from S. pombe (M64799).