Identification of MYO4, a second class V myosin ... - Semantic Scholar

0 downloads 0 Views 468KB Size Report
devised by Cheney et al. (1993a ..... originally identified by Chris Davies (C. Davies, J. Trgovcich ... insensitive calmodulin-binding sites (Cheney and Mooseker,.
1055

Journal of Cell Science 107, 1055-1064 (1994) Printed in Great Britain © The Company of Biologists Limited 1994

Identification of MYO4, a second class V myosin gene in yeast B. K. Haarer*, A. Petzold, S. H. Lillie and S. S. Brown† Department of Anatomy and Cell Biology, University of Michigan Medical School, Ann Arbor, MI 48109-0616, USA *Present address: Department of Zoology, University of Texas, Austin, TX 78712-1064, USA †Author for correspondence

SUMMARY We have isolated a fourth myosin gene (MYO4) in yeast (Saccharomyces cerevisiae). MYO4 encodes a ~170 kDa (1471 amino acid) class V myosin, using the classification devised by Cheney et al. (1993a; Cell Motil. Cytoskel. 24, 215-223); the motor domain is followed by a neck region containing six putative calmodulin-binding sites and a tail with a short potential ‘coiled-coil’ domain. A comparison with other myosins in GenBank reveals that Myo4 protein is most closely related to the yeast Myo2 protein, another class V myosin. Deletion of MYO4 produces no detectable phenotype, either alone or in conjunction with mutations in myo2 or other myosin genes, the actin gene, or secretory genes. However, overexpression of MYO4 or MYO2 results INTRODUCTION The family of identified myosins is growing. In addition to the extensively studied conventional myosin, which dimerizes via a long α-helical tail, there are an increasing number of myosins that resemble conventional myosin in the motor domain but not in the rest of the molecule. Cheney et al. (1993a) have divided the myosins into classes based on the amount of sequence similarity in the motor domains. So far, myosins from three of these classes have been identified in yeast: Myo1 protein (Myo1p) (Watts et al., 1987; Rodriguez and Paterson, 1990) is a member of the myosin II class (= conventional myosin), and like the Dictyostelium myosin II has been implicated in cytokinesis. Myo3p (Goodson et al., 1990) is a myosin I (‘minimyosin’) whose function is unknown (deletion of MYO3 has no discernible phenotypic effect). Myo2p (Johnston et al., 1991) is a myosin V; studies of the temperature-sensitive myo2-66 mutant have implicated Myo2p in polarized growth (i.e. the budding process in yeast). Myosin V’s have three characteristic domains; a ‘head’ or motor domain, a ‘neck’ with six repeats that putatively bind calmodulin, and a long ‘tail’ containing stretch(es) predicted to form coiled coils. Here we report the identification of a fourth yeast myosin, Myo4p, which, like Myo2p, is a member of the myosin V class. MATERIALS AND METHODS Growth and handling of yeast Yeast strains (Table 1) were grown routinely at room temperature

in several morphological abnormalities, including the formation of short strings of unseparated cells in diploid strains, or clusters of cells in haploid strains. Alterations of MYO4 or MYO2 indicate that neither the motor domains nor tails of these myosins are required to confer the overexpression phenotype, whereas the neck region may be required. Although this phenotype is similar to that seen upon MYO1 deletion, we provide evidence that the overexpression of Myo4p or Myo2p is not simply interfering with Myo1p function. Key words: myosin, MYO4, yeast

(~22°C) in the rich liquid medium YM-P (Lillie and Pringle, 1980) or on YEPD plates (Sherman et al., 1986). Standard techniques were used for transformation (by lithium acetate) and genetic manipulations (Sherman et al., 1986). Selection for plasmids was maintained by growth in synthetic complete medium with uracil or leucine (SCuracil, SC-leucine), or on the corresponding plates (essentially as described by Sherman et al., 1986). Stability tests for plasmids were performed by inoculating cells from selective medium into YM-P. Cultures were grown to stationary phase (~8 generations) and plated on YEPD (Sherman et al., 1986). Resulting colonies were tested for the presence or absence of plasmid by streaking on selective plates. DNA manipulations Standard procedures were used for DNA manipulations (Sambrook et al., 1989). DNA fragments were subcloned in pBluescript (Stratagene, La Jolla, CA) for sequencing by the dideoxy method (Sanger et al., 1977) as described by the Sequenase DNA Sequencing Kit (United States Biochemical, Cleveland, OH). In some cases, oligonucleotides were synthesized for use as internal sequencing primers. Both strands were sequenced in and near the coding region for MYO4, and sequencing was carried out across cloning junctions on at least one strand. In addition, a single strand of flanking DNA was sequenced as indicated in Fig. 1. Isolation of the complete MYO4 gene The ~5.5 kbp EcoRI fragment (see Fig. 1, line b) obtained from C. Davies was used as a probe to screen, by colony hybridization, a YCp50 plasmid-based yeast genomic library (Rose et al., 1987). Several hybridizing plasmids were obtained, whose inserts overlapped or encompassed the probe DNA as judged by restriction mapping and Southern blot analysis. These inserts were used for subcloning and DNA sequence analysis.

1056 B. K. Haarer and others Table 1. Saccharomyces cerevisiae strains used in this study Strain 22AB myo4∆U5 myo4∆U5-2A myo4∆U5-2B myo4∆U5-2C myo4∆U5-2D myo4∆U5 (2A+2D) SLY55 SLY34 DBY1993 DBY1999 HA1A HA1C BGE1 myo1∆-12B NY3 NY17 NY57 NY61 NY130 NY402 NY405 NY410 NY432 NY760

Relevant genotype MATa ura3 trp1 lys2 leu2 his3 MATα ura3 trp1 lys2 leu2 his3 MATa ura3 trp1 lys2 leu2 his3 MYO4 MATα ura3 trp1 lys2 leu2 his3 myo4::URA3 MATα ura3 trp1 lys2 leu2 his3 myo4::URA3 MATa ura3 trp1 lys2 leu2 his3 MYO4 MATα ura3 trp1 lys2 leu2 his3 MYO4 MATa ura3 trp1 lys2 leu2 his3 myo4::URA3 MATa ura3 trp1 lys2 leu2 his3 myo4::URA3 MATα ura3 trp1 lys2 leu2 his3 myo4::URA3 MATa ura3 leu2 his4 smy1∆-2::LEU2 MATα ura3 leu2 myo2-66 MATα ura3 act1-2 MATa ura3 act1-1 MATa ura3 trp1 leu2 his3 ade2 can1 myo3::HIS3 MATα ura3 trp1 leu2 his3 ade2 can1 myo3::HIS3 MATα ura3 leu2 ade1 myo1::LEU2 MATa ura3 leu2 ade1 myo1::LEU2 MATa ura3 sec1-1 MATa ura3 sec6-4 MATa ura3 sec9-4 MATa ura3 sec10-2 MATa ura3 sec2-41 MATa ura3 sec5-25 MATa ura3 sec4-8 MATa ura3 sec8-9 MATα ura3 sec18-1 MATα ura3 sec7-1

Source

MYO4

(SA)

b b c c d d e a,f g g g g g g g g g g

aThis study. bS. Lillie, unpublished. cProvided by D. Botstein. dProvided by H. Goodson and J. Spudich. eRodriguez and Paterson, 1990. fSegregant from cross of BGE1 by myo4∆U5-2D. gProvided by P. Novick.

Mapping of the MYO4 gene A 3.5 kbp SalI-BglII fragment (contained entirely within the MYO4 coding region) was used to probe a yeast chromosome blot (generously provided by K. Corrado) on Genescreen Plus (New England Nuclear, Boston, MA). The DNA probe was labeled with [32P]dATP by random primer labeling (Feinberg and Vogelstein, 1983) and was hybridized to the chromosome blot as described by New England Nuclear. Results suggested that MYO4 was on chromosome I; examination of transcript maps of chromosome I (Diehl and Pringle, 1991) suggested that MYO4 might correspond to the transcribed region reported as FUN22 (function unknown). Comparison of downstream DNA sequence to sequence of the FUN20 region (accession no. X62577) confirmed this location of MYO4. (However, the direction of the MYO4 open reading frame is opposite the reported direction of transcription for FUN22 (Diehl and Pringle, 1991).) Deletion of MYO4 The majority of the MYO4 gene was deleted as follows (refer to Fig. 1, line c): (i) an 881 bp SpeI fragment (which includes the C-terminal 25 codons of MYO4) was cloned into the SpeI-XbaI sites of pBluescript KS+; (ii) an ~320 bp KpnI-SalI fragment (containing codons 11-118 of MYO4) was cloned into the KpnI-SalI sites of the polylinker region of the plasmid from (i); (iii) the 1170 bp URA3containing HindIII fragment from plasmid YEp24 was inserted in the HindIII site of the plasmid from (ii); the resulting plasmid was then digested at the KpnI and SacI sites of the polylinker region to generate an ~2.4 kbp fragment that contained the URA3 gene within MYO4 coding sequence. This KpnI-SacI fragment was used to

R

EE X

H X

X

HB

Sp (SA)

*

Sp

X

* E

b K

c

a a a a a

S

a

a a

K

SNC1

S d

S S

E

5.5 kb URA3

Sp

Sp H3

1 kb

Fig. 1. MYO4 physical map and DNA manipulations. (a) Restriction map of the region of chromosome I containing MYO4. Heavy arrows denote the coding regions of MYO4 and SNC1 (the latter includes intron sequences; Gerst et al., 1992). The DNA segment is oriented such that the centromere of chromosome I is to the left of the Figure. Asterisks denote boundaries of double-stranded sequence. The sequence of the entire region (including some single-stranded sequence) has been deposited in GenBank under accession number M90057; note that some of this sequence overlaps with SNC1 sequence reported by Gerst et al. (1992; accession number M91157). (b) The original 5.5 kbp EcoRI fragment isolated by C. Davies; accession number M87364. (c) The KpnI-SalI and SpeI fragments used for disruption of MYO4 by replacing the intervening sequence with the URA3 gene. (d) Fragments used in the construction of YEp351-MYO4 (see Materials and Methods); stippled boxes denote YCp50-derived DNA. B, BglII; E, EcoRI; H, HpaI; H3, HindIII; K, KpnI; R, EcoRV; S, SalI; Sp, SpeI; X, XbaI; (Sa), Sau3A sites that occurred at the boundaries of the library clone, YCpMYO4; the restriction map shown at (a) is complete for all sites shown except Sau3A and KpnI. transform the diploid yeast strain 22AB to Ura + (by one-step replacement (Rothstein, 1983) of one of the two copies of the MYO4 gene). Subsequent sporulation and tetrad dissection produced haploid segregants that lacked an intact MYO4 gene. DNA hybridization analysis of chromosomal DNA from the diploid heterozygote (myo4∆U5) and its haploid progeny indicated that the MYO4 gene deletion had occurred as expected. Construction of full-length and altered MYO4-containing plasmids The complete MYO4 gene was cloned into YEp351 (a LEU2-based yeast high-copy-number vector; Hill et al., 1986) by sequentially inserting a 4.7 kbp SalI-HindIII fragment from a library clone (YCpMYO4; the insert is indicated in Fig. 1, line a) into the SalI and HindIII sites in the polylinker region of YEp351, and then a 1.4 kbp SalI fragment into the SalI site, to form plasmid YEp351-MYO4 (Fig. 1, line d). Truncated forms of the MYO4 gene (Fig. 3B) were cloned into YEp351 as follows: (i) a 4.9 kbp BamHI-BglII fragment of YEp351-MYO4 was inserted into the BamHI site of YEp351 to create plasmid YEp351-MYO4(Bgl); (ii) a 3.6 kbp BamHI-HpaI fragment of YEp351-MYO4 was inserted into the BamHI-SmaI sites of YEp351 to create plasmid YEp351-MYO4(Hpa); (iii) a 3.2 kbp XbaI-XbaI fragment of YEp351-MYO4 was inserted into the XbaI site of YEp351 to create plasmid YEp351-MYO4(Xba); (iv) a 2.4 kbp BamHI-EcoRV fragment of YEp351-MYO4 was inserted into the BamHI-SmaI sites of YEp351 to create plasmid YEp351-MYO4(EcoRV). (In all cases, the upstream restriction site was in the polylinker region of YEp351 into which MYO4 had been inserted.) It was also necessary to construct these plasmids with a different selectable marker in order to use them in a MYO1 disruptant (Table 4). Complete and truncated forms of MYO4 were cloned into the highcopy-number URA3-based vector YEp352 (Hill et al., 1986) as

Identification of MYO4 1057 follows: Full-length MYO4 was moved as a 5.8 kbp BamHI-HindIII fragment from YEp351-MYO4 to YEp352 to create plasmid YEp352MYO4. YEp352-MYO4(Bgl) and YEp352-MYO4(Hpa) were constructed in essentially the same manner as YEp351-MYO4(Bgl) and YEp351-MYO4(Hpa); YEp352-MYO4(Xba) was made by inserting the BamHI-HindIII fragment from YEp351-MYO4(Xba) into the BamHI and HindIII sites of YEp352; YEp352-MYO4(EcoRV) was made by inserting the BamHI-EcoRI fragment from YEp351MYO4(EcoRV) into the BamHI and EcoRI sites of YEp352. Deletions within the MYO4 gene (Fig. 3B) were created as follows: (i) YEp351-MYO4 was cut with NruI and NcoI, the ends were filled using Klenow fragment, and religated to create YEp351MYO4(N,Nc). (ii) YEp352 was cut with BamHI in the polylinker region and ends were filled using Klenow fragment; then it was cut with HindIII in the polylinker region, and ligated to a ~2kbp SspIHindIII fragment from YEp351-MYO4 (the HindIII site is shown in Fig. 1, line d). Into the KpnI site in the polylinker region of this construct was inserted a 1 kbp KpnI fragment from YEp351-MYO4 (the upstream KpnI site was in the polylinker region of YEp351 into which MYO4 had originally been inserted), to create YEp352MYO4(K,Ss). (iii) A 0.6 kbp XbaI-XbaI fragment of YEp351-MYO4 was inserted into the XbaI site in the polylinker region of YEp352. Into the KpnI site in the polylinker region of this construct was inserted a 1 kbp KpnI fragment (the same fragment as described in (ii)) from a YEp351-MYO4 whose BamHI site had first been destroyed (by cutting, filling with Klenow fragment, and religating). Finally, the BamHI site in the polylinker region of this construct was cut, filled using Klenow fragment, and religated to place the fragments of the MYO4 gene in frame, producing YEp352-MYO4(K,X,X). Construction of full-length and altered MYO2-containing plasmids YEpMYO2-40, which contains the complete MYO2 gene, was made by inserting the ~6 kbp ClaI fragment from pJP10-2B (Johnston et al., 1991) into the ClaI site of YEp102 (a high-copy-number yeast vector; Vojtek et al., 1991). (The ClaI fragment includes vector sequence upstream of MYO2.) YEpmyo2 was constructed in the same manner as YEpMYO2-40, except that the ClaI fragment contained the temperature-sensitive myo2-66 allele (S. H. Lillie and S. S. Brown, unpublished data). YEpMYO2-42, which contains a truncated MYO2 gene, was made by inserting the ~4kbp EcoRI fragment from pJP102B into the EcoRI site of YEp352. (The EcoRI fragment includes vector sequence upstream of the MYO2 gene and truncates the MYO2 gene at codon 1116.) Antibody production, western blotting and immunofluorescence Fusion proteins were produced containing the tail portion of Myo4p (amino acids 975-1448; indicated in Fig. 2) fused to either trpE protein (see Haarer and Pringle, 1987) or β-galactosidase (Rüther and Müller-Hill, 1983). A 1.4 kbp SspI-SpeI fragment of the MYO4 gene was ligated into pATH1 cut with SmaI and XbaI, or into pUR289 cut with EcoRV and XbaI. Polyclonal antibodies were raised in rabbits against these fusion proteins (in the manner described by Kim et al., 1991), and the antibody was affinity-purified against the heterologous fusion protein (in the manner described by Haarer and Pringle, 1987). Western blots of yeast total protein were prepared: cells were grown in YM-P or SC-leucine (for strains harboring LEU2-based plasmids) to mid-log stage and were harvested by centrifugation, then broken in cold water by vortexing in the presence of 0.5 mm glass beads. Solubilization buffer was added to samples, which were then placed in a boiling water bath for 2 minutes and loaded on a 5% polyacrylamide gel (with 3% stack). Gels were blotted to nitrocellulose at ~400 mAmp for 24 hours. Blots were probed with affinity-purified anti-Myo4p antibodies at ~1:60 dilution; horseradish peroxidase-conjugated goat anti-rabbit secondary antibody was used at 1:250. Parallel staining of blots with anti-Myo2p antibody (S. H. Lillie and

S. S. Brown, unpublished data) confirmed that the Myo4p antibodies do not crossreact with Myo2p. Cell staining with anti-yeast actin antibodies, anti-yeast tubulin monoclonal antibody, Calcofluor (stains cell-wall chitin), and DAPI (stains DNA) were as described previously (Haarer et al., 1990, and references therein).

RESULTS

MYO4 sequence A DNA fragment with sequence similarity to myosins was originally identified by Chris Davies (C. Davies, J. Trgovcich and C. A. Hutchison, unpublished data). We used this fragment (Fig. 1, line b) to obtain the complete MYO4 gene from a yeast genomic library. Sequence analysis revealed a predicted product of 1471 amino acids, with a calculated molecular mass of 169,342 and isoelectric point of 7.63. We feel confident that we have determined the correct limits of the open reading frame, for the following reasons: There are stop codons in all three reading frames within 131 bp upstream and 119 bp downstream of the open reading frame. The SNC1 gene (Gerst et al., 1992) is found just 102 bp downstream, reading in the opposite direction. A search for splice recognition sequences revealed two TACTAAC sequences (the motif employed by yeast); however, these are in regions of conserved amino acid sequence (Fig. 2) and are therefore unlikely to indicate the existence of introns. We have classified Myo4p as a class V myosin based on primary sequence and overall predicted structure (Fig. 3A). The other class V myosins that have been identified so far include yeast Myo2p (Johnston et al., 1991), a 190 kDa myosin from chick brain (p190; Larson et al., 1988, Espindola et al., 1992), and the protein encoded by the dilute locus in mice (Mercer et al., 1991). Dilute protein and p190 are probable homologs of one another, as they show about 90% identity throughout (Espreafico et al., 1992). A comparison of the primary amino acid sequence of Myo4p to that of other myosins showed that yeast Myo2p has the most closely related sequence, followed by dilute protein and p190 (Figs 2 and 3A). Other, non-class V, myosins are less similar to Myo4p in the head, and show no similarity in the tail. This can be seen in the dot matrix comparisons of yeast myosins shown in Fig. 4; while Myo1p (a conventional, class II myosin) resembles Myo4p only in the head, Myo2p shows similarity to Myo4p throughout. Myo4p has several structural motifs that are characteristic of class V myosins. First, Myo4p has a ‘neck’, consisting of six 23-25 amino acid repeats, immediately downstream of the head (marked in Figs 2 and 3A and can be seen in Fig. 4, see legend). These repeats, referred to as ‘IQ domains’, each resemble sequences found in proteins containing calciuminsensitive calmodulin-binding sites (Cheney and Mooseker, 1992). Class V myosins have six of these repeats, whereas other myosins have fewer such repeats (see the alignment in Espreafico et al., 1992). Espreafico et al. (1992) have mapped the calmodulin-binding sites of p190 to the neck by expressing fragments of p190 in bacteria. Another characteristic that Myo4p shares with other class V myosins is a stretch predicted to be capable of coiled-coil formation (Lupas et al., 1991) in the first part of the ‘tail’ (Fig.

1058 B. K. Haarer and others

Fig. 2. Sequence comparison between Myo4p, Myo2p, dilute protein and p190. The PILEUP program (Devereux et al., 1984) was used for alignment. Identities between Myo4p and any of the other myosins are boxed. Brackets below the sequences mark the six IQ domains. Filled circles above Myo4p indicate locations of two TACTAAC potential splicing motifs (see text). Filled squares indicate the portion of Myo4p used for antibody production (amino acids 975-1448).

Identification of MYO4 1059

A

Table 2. Effects of MYO4 deletion and overproduction on the frequencies of budded and abnormal cells*

*

Myo2p

+Control vector 73%

35%

27%

Myo4p 50%

25%

22%

neck

tail

X Nc H X Ss

B

B K

N

R

Myo4p B

Bgl

R

EcoRV

Nc

N

N,Nc

K,X,X

-

X

Xba

K,Ss

+ +

H

Hpa

K

+

Ss

K X

+/+

+/∆

∆/∆

+/+

+/∆

∆/∆

Unbudded Budded Abnormal

27 72 1

25 75 0

24 74 2

47 40 13

42 50 8

37 48 15

*Wild-type cells (+/+; strain 22AB) and cells that were heterozygous (+/∆; strain myo4∆U5) or homozygous (∆/∆; strain myo4∆U5(2A+2D)) for deletion of MYO4 were transformed with a high-copy-number plasmid expressing MYO4 (YEp351-MYO4) or control vector (YEp351). †Expressed as % of 200 cells for each strain.

p190 motor

+MYO4 plasmid

Cell phenotype†

-

X

100 aa

Fig. 3. (A) Structure of class V myosins. Shown are the % amino acid identity of the head (or motor domain; amino acids 1-777), neck (comprised of six IQ domains; amino acids 778-921), and tail regions of Myo4p, compared to Myo2p and p190 (calculated by BESTFIT; Devereux et al., 1984). Black boxes indicate regions with a greater than 90% probability of forming a coiled-coil domain (amino acids 994-1023 for Myo4p), as predicted by the analysis of Lupas et al. (1991). An asterisk indicates the position of truncation of MYO2, at codon 1116 (see text). (B) Deletions in MYO4. Restriction sites (N, NruI; Nc, NcoI; Ss, SspI; otherwise, as abbreviated in Fig. 1) used in making deletions in the MYO4 coding region are indicated. The pluses and minuses indicate whether (+) or not (−) transformation of wild-type yeast with these partially deleted genes on high-copy-number plasmids resulted in the overexpression phenotype seen with full-length MYO4. Note that small amounts of polylinker sequence that put the MYO4 fragments in frame are not shown for YEp352-MYO4(K,Ss) and YEp352-MYO4(K,X,X).

Deletion of MYO4 The MYO4 gene was deleted by replacing the majority of the gene with the URA3 gene (Fig. 1, line c). This deletion removes all but 118 N-terminal and 25 C-terminal codons of MYO4. No phenotypic effects of the deletion were observed when diploid yeast were transformed with this construct and haploid segregants examined (cells containing the deletion could be identified as Ura+ progeny following tetrad analysis). We confirmed that the expected deletion had occurred by Southern blotting. (The Southern blots also revealed that there were no closely related, crosshybridizing sequences.) Segregants with the deletion grew at the same rate (Fig. 5) as the wild-type sister segregants when examined at several temperatures (16°C, 22°C, 36°C), and consistent with these results, showed no differences in the frequency of budded cells (Table 2, under ‘+control vector’). Furthermore, no morphological abnormalities were detected by light microscopy (Table 2, under ‘+MYO4 plasmid’). MYO4 deletion strains were stained with anti-actin and anti-tubulin antibodies, with Calcofluor to localize chitin in the cell wall (Sloat et al., 1981), and with DAPI to examine the behavior of the nucleus during the cell cycle. Localization of all of these components appeared normal (for example, compare chitin localization and cell shape in Fig. 6A and B). MYO4 deletion strains also had a normal polarity of budding (Hicks et al., 1977) in both haploids and diploids (Table 3), and were respiratory competent, as they grew on plates containing glycerol, a nonfermentable carbon source. Table 3. Effect of MYO4 deletion on budding polarity Polarity*

3A). It has been shown by electron microscopy that p190 does in fact dimerize (Cheney et al., 1993b). However, it should not be concluded that Myo4p is necessarily capable of dimerization, as the putative coiled-coil region in Myo4p is considerably shorter (30 amino acids) than that in p190 (379 amino acids). This may be a marginal number of amino acids to make a stable coiled-coil (Cohen and Parry, 1990). Mapping MYO4 MYO4 was mapped to chromosome I by probing a yeast chromosome blot. We have determined that MYO4 is the same as FUN22, an open reading frame of unknown function on a transcript map of chromosome I (Diehl and Pringle, 1991), although we report an orientation of the coding region that is opposite to the reported direction of transcription.

Strain Wild-type diploid (+/+)† MYO4 deletion homozygote (∆/∆)‡ Haploid segregant: +§ +¶ ∆㛳 ∆**

Unipolar

Bipolar

Random

54 54 99 100 100 98

40 40 1 0 0 1

6 6 0 0 0 1

*Cells were stained with Calcofluor to visualize bud scars. Those having > one bud scar were scored as either unipolar (all scars at one pole of the cell), bipolar (scars at both poles), or random. Polarity was expressed as % of the 32-48 cells scored per strain. †22AB. ‡myo4∆U5(2A+2D). §myo4∆U5-2C (MYO4). ¶myo4∆U5-2B (MYO4). 㛳myo4∆U5-2A (myo4∆). **myo4∆U5-2D (myo4∆).

1060 B. K. Haarer and others A

B

Myo4p

0

1000

500

Myo4p

0

1000

500

1500

1500

1000

Myo1p

Myo2p

1000

500 500

0

HEAD

NECK

0

TAIL

HEAD

NECK

Finally, these strains mated and sporulated normally: We found that 40% of the cells sporulated from a cross of MYO4 deletion sister segregants (myo4∆U5-2A × myo4∆U5-2D), whereas 44% sporulated from a cross of the corresponding wild-type sister segregants (myo4∆U5-2B × myo4∆U5-2C), and 54% of wild-type diploid control cells (22AB) sporulated (n=600 for all three samples). We have found by blotting that wild-type cells do indeed express a protein of the predicted molecular mass that labels with anti-Myo4p antibodies (Fig. 7). This protein is overex10

A OD 660

pressed in cells carrying the gene on a high-copy-number plasmid, but is not present in MYO4 deletion strains. In addition, expression of a truncated MYO4 gene results in the appearance of a smaller protein of the predicted size. We looked for evidence of functional redundancy or other interaction between Myo4p and the other known yeast myosins by looking for a synthetic (more than additive) phenotype in double mutants. MYO4 deletion strains (myo4∆U5-2A or -2D) were crossed with MYO1 deletion strains (BGE1 and myo1∆12B), MYO3 deletion strains (HA1A and HA1C), or a myo2-

10

C

1

1

.1

, ., .01

.1

2A Wild Type 2B

,

.,

2C Deletion 2D

.01

B

2A Wild Type 2B Deletion 2C 2D

D 1

1

OD 660

TAIL

Fig. 4. Dot-matrix comparison of Myo4p to other yeast myosins. The COMPARE and DOTPLOT programs (Devereux et al., 1984) were used. In the comparison of Myo4p to Myo2p (A), note the multiple lines in the neck region, which are due to the similarity between IQ domains. In the comparison to Myo1p (B), note the similarity of the predicted coiled-coil domain at the beginning of the Myo4p tail to the entire Myo1p tail, which is all coiled-coil.

.1

.1

(2A+2D) Wild Type

Wild Type (2A+2D) 22ABDeletion

22AB Deletion .01

.01

.001

.001 0

5

10

15

Time (h)

20

25

0

5

10

Time (h)

15

Fig. 5. Effect of MYO4 deletion on growth rates. (A) Growth rates of the four haploid segregants of myo4∆U5 at 22°C. myo4∆U5-2B (circles) and -2C (filled diamonds) contain wild-type MYO4, whereas myo4∆U5-2A (squares) and -2D (open diamonds) contain the MYO4 deletion. (B) Growth rates of diploid strains homozygous for MYO4 (22AB; circles) or deletion of MYO4 (myo4∆U5(2A+2D); triangles) at 22°C. (C and D) The same strains as in A and B, respectively, grown at 36°C.

Identification of MYO4 1061

Fig. 6. Micrographs of yeast transformed with MYO4 on a high-copy-number plasmid (YEp351-MYO4), or with control vector (YEp351). (A) Wild-type diploid strain (22AB) + YEp351 (control vector), stained with Calcofluor. Bar, 5 µm, for all micrographs. (B) Diploid MYO4 deletion strain (myo4∆U5(2A+2D)) + YEp351, stained with Calcofluor. (C) Strain 22AB + YEp351-MYO4, stained with Calcofluor. (D-F) Strain myo4∆U5(2A+2D) + YEp351-MYO4, stained with (D) Calcofluor or (E,F) DAPI. (G) Haploid wild-type strain (myo4∆U5-2B) + YEp351-MYO4, stained with Calcofluor. (H) Haploid MYO4 deletion strain (myo4∆U5-2A) + YEp351-MYO4, stained with Calcofluor.

66 mutant strain (SLY34). In addition, crosses were performed with strains carrying the act1-1 and act1-2 actin mutations (DBY1999 and DBY1993, respectively), with strains carrying mutations in genes whose products are involved in the secretory pathway (all NY strains listed in Table 1) and with strain SLY55 carrying a SMY1 deletion (which is synthetically lethal with myo2-66; Lillie and Brown, 1992). Double mutants were obtained by tetrad dissection. In no case did deletion of MYO4 alter the phenotype of any of these other mutants. We also looked for evidence of overlapping function by asking whether extra copies of MYO4 could compensate for the defect in the myo2-66 mutant. Introducing MYO4 on either

a high-copy-number plasmid (YEp351-MYO4) or low-copynumber plasmid (YCpMYO4) did not overcome the temperature sensitivity of the myo2-66 mutant. (YEp351-MYO4 did cause the phenotypic changes described below for wild-type cells.) Overexpression of Myo4p (and Myo2p) As another way of identifying potential function, we have overexpressed Myo4p in yeast, using either wild-type diploid yeast (strain 22AB) or strains heterozygous (strain myo4∆U5) or homozygous (strain myo4∆U5(2A+2D)) for deletion of MYO4. Overexpression of Myo4p in all three strains gave

1062 B. K. Haarer and others Table 4. Stability of MYO4-containing plasmids % Stability* Plasmid† Control vector Full-length MYO4 Bgl Hpa Xba EcoRV N,Nc K,Ss K,X,X

Wild type‡ 65±11 13±19¶ 30±3 22±8 66±8 72±11 0±0 65±18 75±21

Table 5. Stability of MYO2- and myo2-66-containing plasmids % Stability* at

MYO1 disruptant§ 58±18 3±4 3±4 3±4 88±4 60±7

*Plasmid stability is presented as the average percentage ± s.d. of colonies that scored positive for the plasmid-borne marker out of 25-40 tested for each of 2-8 independent transformants (after ~8 generations in the absence of selection). †The plasmid designations are the same as used in Fig. 3B. ‡Strain 22AB §Strain myo1∆-12B ¶The large range here is due to one aberrant transformant. An occasional transformant carrying either MYO4 or MYO2 high-copy-number plasmids will reproducibly show normal plasmid stability, and will not display the overexpression phenotype. This is presumably due to some infrequent event that ‘inactivates’ these genes. If the aberrant transformant is omitted, the mean ± s.d. for the other seven transformants is 6±6.

indistinguishable results: an increased frequency of unbudded cells and morphological abnormalities (Table 2), including short strings of somewhat elongate cells (Fig. 6C-E). Haploid yeast tended to form clusters rather than strings of cells (Fig. 6G and H); this difference is presumably due to the fact that haploid yeast tend to bud in a unipolar pattern and diploid yeast in a bipolar pattern (Hicks et al., 1977). Some cells were observed to have more than a single nucleus (Fig. 6F), or to lack a nucleus. Calcofluor staining revealed random chitin localization in some cells, and some cells were larger than normal (Fig. 6D). However, a large fraction of the cells in the population were normal-appearing (Table 2, and the single cell in Fig. 6C). We suspect that these cells may be normal because they have a reduced copy number of the MYO4 plasmid. There are two indications that the plasmid is selected against; first, not much overexpression of Myo4p is seen on western blots (Fig. 7), and second, there is a high rate of plasmid loss when selection is removed (Table 4, compare control vector to fulllength MYO4). Similar effects were seen when Myo4p was overexpressed in haploid or diploid MYO4 deletion strains. Overexpression of Myo2p gave a similar phenotype to that seen with Myo4p, including strings of cells and plasmid instability (Table 5). To determine what domain of Myo4p causes the overexpression phenotype, we have made a series of truncations and partial deletions (Fig. 3B), and introduced the altered genes on high-copy-number plasmids into the wild-type diploid strain 22AB. We found that truncation of the MYO4 gene at a HpaI site (near the end of the fourth IQ domain in the neck) produced the abnormal phenotype, as did another, less truncated MYO4 (at a BglII site). A more severe truncation (at an XbaI site), encoding a Myo4p missing ~5% of the head domain, did not cause the abnormal phenotype. These results rule out tail function (for example, heterodimerization) as a cause of the effect. Removing most of the head as well as a little of the neck (deleting between NruI and NcoI sites) did not abolish the effect, whereas removing all of the neck as well as a little more

Plasmid

22°C

36°C

Control vector† YEpMYO2-40 YEpmyo2

72±4 0±0 0±0

71±1 0±0 0±0

*Wild-type strain 22AB was transformed with the indicated plasmids and maintained at either 22°C or 36°C throughout. Plasmid stability is presented as the average percentage ± s.d. of colonies that scored positive for the plasmid-borne marker out of 25 tested for each of two independent transformants (after ~8 generations in the absence of selection). †YEp102.

Fig. 7. Western blot of yeast total protein labeled with antibody against Myo4p tail region. (A) Antibody raised against trpE-Myo4p fusion protein and affinity-purified against β-galactosidase-Myo4p fusion protein. (B) Antibody raised against β-galactosidase-Myo4p fusion protein and affinity-purified against trpE-Myo4p fusion protein. Lanes 1, MYO4 deletion strain (myo4∆U5(2A+2D)) +YEp351 (control vector); lanes 2, wild-type strain (22AB) + YEp351; lanes 3, wild type + YEp351-MYO4; lanes 4, wild type + YEp351-MYO4(Bgl). Bands of ~170 kDa (lanes 2-4) and ~155 kDa (lane 4) were observed, which corresponded to the calculated molecular masses of full length and truncated Myo4p’s, respectively. Although other bands were seen, especially with the antibody used in (A), these were judged to be nonspecific, as they were seen in all four lanes.

head (deleting between KpnI and SspI sites) did abolish the effect. The simplest interpretation of these results is that the neck region may be responsible for the effect. However, we were unable to cause the effect by expressing essentially only the neck region (contained in an XbaI fragment). One plausible explanation is that this latter construct is either not expressed or not stable; we cannot look at the expression of Myo4ps lacking the tail domain on western blots, because our antibody is directed only against this portion of the molecule. The same limitation applies to the most severely truncated Myo4p’s; we cannot be certain that they are incapable of conferring an altered phenotype, as we cannot confirm that they are present. For constructs encoding tail regions, we have demonstrated bands of the correct molecular mass on western blots (YEp351-

Identification of MYO4 1063 MYO4(Bgl) in Fig. 7 and YEp351-MYO4(N,Nc), data not shown). The results in Table 4 show that the instability of the altered plasmids correlates with their ability to confer an abnormal phenotype. YEp351-MYO4(Hpa), YEp351-MYO4(Bgl), and YEp351-MYO4(N,Nc), which affect the phenotype, are lost at a higher rate than the other altered plasmids, which appear to have no effect on phenotype. Similarly, truncation of MYO2 (resulting in loss of about 1/2 of the tail; see Fig. 3A) did not abolish either its overexpression phenotype or its elevated instability. We have obtained further evidence that the head domain is not involved in the abnormal phenotype from an examination of the effects of overexpressing a mutated myosin, the myo266 allele. We have determined that a single point mutation in the head domain is both necessary and sufficient for the temperature sensitivity of the myo2-66 mutant (S. H. Lillie and S. S. Brown, unpublished data). This mutation lies near an actinbinding face (Rayment et al., 1993), and thus is likely to perturb the ability of Myo2p to bind actin. A high-copynumber plasmid containing the myo2-66 allele (YEpmyo2) was introduced into wild-type cells (strain 22AB) and transformants were maintained at either 22°C or 36°C. (At >30°C, myo2-66 mutants cease normal growth and die.) These transformants displayed the overexpression phenotype not only at 22°C, but also at 36°C, and the YEpmyo2 plasmid showed elevated instability even in cells maintained at 36°C (Table 5). In fact, there was no detectable difference in either the abnormal phenotype or plasmid instability of cells overexpressing wild-type Myo2p or mutated Myo2p at either temperature, indicating that a fully functional head domain is not required. Because the abnormal phenotype produced by Myo4p or Myo2p overexpression is reminiscent of that produced by disruption or partial deletion of MYO1 (Watts et al., 1987; Rodriguez and Paterson, 1990), we explored the possibility that the overexpression of these myosins is interfering with Myo1p function. However, introduction of YEp352-MYO4 or YEpMYO2 into a MYO1 disruption strain (myo1∆-12B) caused the disruption strain to grow more slowly than when transformed with vector alone. As Myo4p or Myo2p overexpression is more deleterious to growth than MYO1 disruption alone, the effects of overexpression cannot be completely accounted for by interference with Myo1p function. Furthermore, YEp352-MYO4 shows the same elevated instability in a MYO1 deletion strain as it does in a wild-type background (Table 4); we would have predicted normal stability in the deletion strain if the only deleterious effect of the plasmid were interference with Myo1p function. DISCUSSION We report the existence of a fourth myosin, Myo4p, in yeast. Myo4p is a class V myosin, based on its sequence similarity to other members of the class, especially in the head domain, and also based on the presence of six IQ domains (putative calmodulin binding sites) in the neck, and a short putative coiled-coil domain in the tail. We find that deletion of the MYO4 gene produced no detectable phenotypic change. Because it is so similar to

Myo2p, another class V myosin in yeast, we initially thought that Myo4p might be redundant with Myo2p. However, Myo4p cannot substitute for Myo2p, even when overexpressed, and MYO4 deletion has no synthetic (more than additive) phenotype with myo2 in double mutants. While myo2 is synthetically lethal with mutations in a subset of genes involved late in the secretory pathway (Govindan et al., 1991), we have found that MYO4 deletion is not synthetically lethal with these mutations, or with mutations in other secretory genes. Thus, redundancy between MYO4 and MYO2 seems unlikely, and perhaps the structural similarity instead reflects some aspect of their function that is similar. For example, Myo2p and Myo4p might be involved in transport of different classes of vesicles or other organelles. Several lines of evidence indicate that Myo4p is present and active as a myosin in the cell. First, it can be visualized with antibody on western blots. Second, overexpression of Myo4p gives a phenotype similar to that caused either by overexpression of another class V myosin, Myo2p, or by disruption of the MYO1 gene, which encodes a conventional class II myosin. The phenotype in all three cases includes strings or clusters of cells, the increased diameter of some cells, and defects in chitin localization and nuclear segregation. These similarities in phenotype raised the possibility that overexpressed Myo2p or Myo4p might be competing with Myo1p. However, we know that Myo4p or Myo2p is doing more than simply blocking Myo1p function, since MYO1 disruptants grow more slowly when either of these myosins is overexpressed. Perhaps they are also blocking the function of a second, partly redundant myosin. We can’t judge whether overexpression interferes with any of the other known myosins; defects in Myo3p and Myo4p produce no identifiable phenotypic change, and a defect in Myo2p gives changes (an increase in cell size and chitin delocalization) that partially overlap those produced by defects in Myo1p. Attempts to identify the domain(s) in Myo4p that are causative indicate that neither the head nor the tail domain is solely responsible for the overexpression phenotype. On the other hand, it seems rather likely that the neck domain plays a role. This domain has been demonstrated to bind calmodulin in p190 (another class V myosin; Espreafico et al., 1992). We therefore postulated that Myo2p or Myo4p overexpression might lower calmodulin levels. However, we have been unable to cure the overexpression phenotype by introducing the calmodulin gene on a high-copy-number plasmid (data not shown), nor is the overexpression phenotype obviously reminiscent of that of a calmodulin mutant (Davis, 1992). While we have not altogether ruled out the possibility that the overexpression phenotype is due to effects on calmodulin, it is also possible that the neck region is important in some other, unsuspected function. Attempts to immunolocalize Myo4p using the antibodies we’ve described have been unsuccessful to date, despite the use of a number of different fixation and permeabilization procedures and antibody amplification techniques (Pringle et al., 1991; Roberts et al., 1991). We have also attempted to immunolocalize the construct (encoded by YEp352MYO4(K,Ss)) that lacks head and neck sequences (and can therefore be overexpressed without perturbing morphology) but retains most of the tail (so that it should react with antibody). However, this approach was also unsuccessful.

1064 B. K. Haarer and others In summary, studies of a newly identified myosin have revealed that it can function in the cell, but it is not yet clear what its normal function is. Further immunofluorescence (e.g. using an epitope tag) and genetic approaches (e.g. hunts for other genes that are lethal in combination with MYO4) should shed further light on this question. We thank Chris Davies for originally identifying a fragment of the MYO4 gene and giving it to us, Kathy Corrado for yeast chromosome blots, Jose Rodriguez for the MYO1 disruption strain, Holly Goodson and Jim Spudich for the MYO3 deletion strains, and Trisha Davis for the plasmid-borne calmodulin gene. This work was supported by NIH grant GM46745 and in part by NIH grant no. MO1 RR00042.

REFERENCES Cheney, R. E. and Mooseker, M. S. (1992). Unconventional myosins. Curr. Opin. Cell Biol. 4, 27-35. Cheney, R. E., Riley, M. A. and Mooseker, M. S. (1993a). Phylogenetic analysis of the myosin superfamily. Cell Motil. Cytoskel. 24, 215-223. Cheney, R. E., O’Shea, M. K., Heuser, J. E., Coelho, M. V., Wolenski, J. S., Espreafico, E. M., Forscher, P., Larson, R. E. and Mooseker, M. S. (1993b). Brain myosin-V is a two-headed unconventional myosin with motor activity. Cell 75, 13-23. Cohen, C. and Parry, D. A. D. (1990). α-helical coiled coils and bundles: how to design an α-helical protein. Proteins 7, 1-15. Davis, T. N. (1992). A temperature-sensitive calmodulin mutant loses viability during mitosis. J. Cell Biol. 118, 607-617. Devereux, J., Haeberli, P. and Smithies, O. (1984). A comprehensive set of sequence analysis programs for the VAX. Nucl. Acids Res. 12, 387-395. Diehl, B. E. and Pringle, J. R. (1991). Molecular analysis of Saccharomyces cerevisiae chromosome I: identification of additional transcribed regions and demonstration that some encode essential functions. Genetics 127, 287-298. Espindola, F. S., Espreafico, E. M., Coelho, M. V., Martins, A. R., Costa, F. R., Mooseker, M. S. and Larson, R. E. (1992). Biochemical and immunological characterization of p190-calmodulin complex from vertebrate brain: a novel calmodulin-binding myosin. J. Cell Biol. 118, 359368. Espreafico, E. M., Cheney, R. E., Matteoli, M., Nascimento, A. A. C., De Camilli, P. V., Larson, R. E. and Mooseker, M. S. (1992). Primary structure and cellular localization of chicken brain myosin-V (p190), an unconventional myosin with calmodulin light chains. J. Cell Biol. 119, 15411557. Feinberg, A. P. and Vogelstein, B. (1983). A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal. Biochem. 132, 6-13. Gerst, J. E., Rodgers, L., Riggs, M. and Wigler, M. (1992). SNC1, a yeast homolog of the synaptic vesicle-associated membrane protein/ synaptobrevin gene family: genetic interactions with the RAS and CAP genes. Proc. Nat. Acad. Sci. USA 89, 4338-4342. Goodson, H. V., Titus, M. A. and Spudich, J. A. (1990). Disruption of a yeast unconventional myosin. J. Cell Biol. 111, 168a. Govindan, B., Bowser, R. and Novick, P. (1991). Role of the unconventional myosin gene MYO2 in the yeast secretory pathway. J. Cell Biol. 115, 185a. Haarer, B. K. and Pringle, J. R. (1987). Immunofluorescence localization of the Saccharomyces cerevisiae CDC12 gene product to the vicinity of the 10nm filaments in the mother-bud neck. Mol. Cell. Biol. 7, 3678-3687. Haarer, B. K., Lillie, S. H., Adams, A. E., Magdolen, V., Bandlow, W. and Brown, S. S. (1990). Purification of profilin from Saccharomyces cerevisiae and analysis of profilin-deficient cells. J. Cell Biol. 110, 105-114.

Hicks, J. B., Strathern, J. N. and Herskowitz, I. (1977). Interconversion of yeast mating types III. Action of homothalism (HO) gene in cells homozygous for the mating type locus. Genetics 85, 395-405. Hill, J. E., Myers, A. M., Koerner, T. J. and Tzagoloff, A. (1986). Yeast/E. coli shuttle vectors with multiple unique restriction sites. Yeast 2, 163-167. Johnston, G. C., Prendergast, J. A. and Singer, R. A. (1991). The Saccharomyces cerevisiae MYO2 gene encodes an essential myosin for vectorial transport of vesicles. J. Cell Biol. 113, 539-551. Kim, H. B., Haarer, B. K. and Pringle, J. R. (1991). Cellular morphogenesis in the Saccharomyces cerevisiae cell cycle: localization of the CDC3 gene product and the timing of events at the budding site. J. Cell Biol. 112, 535544. Larson, R. E., Pitta, D. E. and Ferro, J. A. (1988). A novel 190 kDa calmodulin-binding protein associated with brain actomyosin. Brazilian J. Med. Biol. Res. 21, 213-217. Lillie, S. H. and Pringle. J. R. (1980). Reserve carbohydrate metabolism in Saccharomyces cerevisiae: responses to nutrient limitation. J. Bacteriol. 143, 1384-1394. Lillie, S. H. and Brown, S. S. (1992). Suppression of a myosin defect by a kinesin-related gene. Nature 356, 358-361. Lupas, A., Van Dyke, M. and Stock, J. (1991). Predicting coiled coils from protein sequences. Science 252, 1162-1164. Mercer, J. A., Seperack, P. K., Strobel, M. C., Copeland, N. G. and Jenkins, N. A. (1991). Novel myosin heavy chain encoded by murine dilute coat colour locus (see correction Nature 352, 547). Nature 349, 709-712. Pringle, J. R., Adams, A. E. M., Drubin, D. G. and Haarer, B. K. (1991). Immunofluorescence methods for yeast. Meth. Enzymol. 194, 565602. Rayment, I., Rypniewski, W. R., Schmidt-Base, K., Smith, R., Tomchick, D. R., Benning, M. M., Winkelmann, D. A., Wesenberg, G. and Holden, H. M. (1993). Three-dimensional structure of myosin subfragment-1: a molecular motor. Science 261, 50-58. Roberts, C. J., Raymond, C. K., Yamashiro, C. T. and Stevens, T. H. (1991). Methods for studying the yeast vacuole. Meth. Enzymol. 194, 644661. Rodriguez, J. R. and Paterson, B. M. (1990). Yeast myosin heavy chain mutant: maintenance of the cell type specific budding pattern and the normal deposition of chitin and cell wall components requires an intact myosin heavy chain gene. Cell Motil. Cytoskel. 17, 301-308. Rose, M. D., Novick, P., Thomas, J. H., Botstein, D. and Fink, G. R. (1987). A Saccharomyces cerevisiae genomic plasmid bank based on a centromerecontaining shuttle vector. Gene 60, 237-243. Rothstein, R. J. (1983). One-step gene disruption in yeast. Meth. Enzymol. 101, 202-211. Rüther, U. and Müller-Hill, B. (1983). Easy identification of cDNA clones. EMBO J. 2, 1791-1794. Sambrook, J., Fritsch, E. F. and Maniatis, T. (1989). Molecular Cloning, a Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sanger, F., Nicklin, S. and Coulson, A. R. (1977). DNA sequencing with chain-terminating inhibitors. Proc. Nat. Acad. Sci. USA 74, 5463-5467. Sherman, F., Fink, G. R. and Hicks, J. B. (1986). Methods in Yeast Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sloat, B. F., Adams, A. and Pringle, J. R. (1981). Roles of the CDC24 gene product in cellular morphogenesis during the Saccharomyces cerevisiae cell cycle. J. Cell Biol. 89, 395-405. Vojtek, A., Haarer, B., Field, J., Gerst, J., Pollard, T. D., Brown, S. and Wigler, M. (1991). Evidence for a functional link between profilin and CAP in the yeast S. cerevisiae. Cell 66, 497-505. Watts, F. Z., Shiels, G. and Orr, E. (1987). The yeast MYO1 gene encoding a myosin-like protein required for cell division. EMBO J. 6, 3499-3505. (Received 10 November 1993 - Accepted 19 January 1994)