Schizosaccharomyces pombe
3.6 ScCdc6p induced re-replication requires cdcISp function but does not increase its endogenous protein levels
In order to differentiate betw een these alternatives I expressed Rep41-cdcl8 and Rep41-ScCDC6 w ith either Rep42, Rep42-cdtl or Rep82-cdtl in a cdcl8-K46 m utant background. Cells were grow n at the permissive tem perature of 25°C for 18 hours following derepression of the prom oter by thiam ine wash-out. Cell cultures were split and half was shifted to the restrictive tem perature of 36°C. Cells were grown for a further 2.5 generations which am ounts to 10 hours at 25°C and 5 hours 50 m inutes plu s 1 h our for tem perature shift recovery at 36°C. Cells w ere then analysed by FACS (Fig. 3.7 A), DAPI staining (Fig. 3.7 B) and W estern blotting (Fig. 3.7 C and D). I show ed that Rep41-cdcl8 is able to prom ote re-replication to a sim ilar extent at both 25°C and 36°C. H ow ever, although Rep41-ScCDC6 can prom ote re-replication at 25°C, the re-replication is abolished at 36°C. This suggests th at the endogenous c d c l8p provides an essential function for this ScCdcbp induced re-replication. However, c d c l8p levels do not accumulate w hen Rep41-ScCDC6 is overexpressed (Fig. 3.7 D u p p er panel) and in fact the protein level rem ains sim ilar to w ildtype (Fig. 3.7 C lanes 1 an d 2). W ildtype c d c l8p function is therefore required for the re-replication induced w hen ScCdc6p is co overproduced w ith c d tlp but this re-replication is not caused by an increase in the endogenous protein levels.
Chapter 3 A nalysis of cdc18 h o m o lo g u es
D iscussion
Proteins that are both structurally and functionally hom ologous to cdcl8 have been id en tified in several organism s. In experim ents to exam ine the ability of hom ologues to com plem ent the cdcl8-K46 tem p era tu re sensitive m u ta n t I identified two situations in which a cdcl8 hom ologue was able to complem ent the cell cycle defect. The first w as w hen the Saccharomyces cerevisiae hom ologue, ScCDCô, w as expressed at low levels using the nmt81 prom oter. The second was w hen the Xenopus laevis homologue, XICDC6, was co-expressed w ith cdtl using the m edium strength prom oters in the Rep41 and Rep42 vectors respectively (Table 3.1 an d Fig. 3.3). H ow ever, I have been unable to rep eat this experim ent w hich suggests that complem entation is very sensitive to specific conditions.
W ork from several different labs has given conflicting results about the ability of cdcl8 hom ologues to functionally rescue cdcl8 m utants. It has been reported that ScCDC6 is unable to functionally substitute for cdcl8 (Dutta and Bell, 1997; Wolf et al., 1999a). H ow ever an o th er recent stu d y states th at ScCDCô is able to complem ent both a cdcl8-K46 m utant and a cdcl8A strain (Sanchez et al., 1999).
O ne possibility for the discrepancies could be the p ro tein level of the cdcl8 hom ologues. It is know n that overexpression of cdcl8 results in re-replication of the DNA and is lethal to the cells. For this reason only the m edium and low level prom oters w ere used in our experiments. H ow ever, the w ork by Sanchez et al. states th at the n m tl prom oter w as used and th at high levels of ScCdc6p are required to achieve complem entation. The problem w ith trying to compare these results is that different ScCDCô cDNAs m ay have been used w hich m ay affect expression level even from the same prom oter. A n example of how small changes in the cDNA sequence can have dram atic consequences for gene expression is that of the original cdcl8 cDNA isolated (Kelly et al., 1993) com pared to one isolated later that contained an extra 89 nucleotides upstream from the start site (Nishitani an d N urse, 1995). This additional sequence w as sufficient to allow re-replication w hen expressed using the n m tl prom oter. An alternative explanation for w hy the full strength n m tl prom oter was required for com plem entation of cdcl8 could be
(hlûo
l i ^ L 9 H J A c U v ^ i ^ J ^ U ( d i l l i j y j ü (jA ^ d C c ( jL l^ ') ( 4 t ['Sf'XdGiVÜ (/(litOf
a) U ^ o it ^
6
Chapter 3 A nalysis of cdc18 h o m o lo g u es
because an integrated copy of nmtl:ScCDC6 was used, w hereas I used a m ulticopy plasm id. Several low er level expressing plasm ids could allow the sam e level of protein production as can be achieved from one integrated higher expression level plasm id. The only w ay to come to a firm conclusion about this w ould be to tag both constructs and directly com pare the protein expression levels w ithin a cell population that is either able or unable to com plem ent the cdcl8 gene. The cDNAs could also be exchanged and the experiments repeated to see if the same results are obtained. A t present it is impossible to conclude w hether the ScCDCô and XICDC6 gene products are able to fully complement the cdcl8-K46 m utant.
In terestin g ly , the p resen ce of c d tlp w as of no b en efit w h en looking for com plem entation w ith ScCDCô, an organism that does not appear to possess a cdtl homologue. However, even though XlCDCô w as only found to com plem ent once, it only did so w hen co-expressed w ith cd tl. As a hom ologue of the fission yeast cd tl gene has recently been identified in Xenopus laevis (Maiorano et al., 2000) it w ould be interesting to see w hether co-expression of the Xenopus laevis cdcl8 and cd tl hom ologues w ould result in com plem entation of the cdcl8-K4ô m utation in a m ore consistent m anner. Furtherm ore, as the hum an cdcl8 hom ologue w as also unable to complement, it m ay be interesting to repeat the experim ent in the future if a hum an cdtl hom ologue is cloned. The sequence of a 5' truncated hum an cDNA encoding a putative cdtl hom ologue can be found in the databases.
The m echanism by w hich ScCdc6p is able to prom ote re-replication has also been som ew hat controversial (Sanchez et al., 1999; Wolf et al., 1999a). Wolf et al. state that ScCdc6p induced re-replication requires only am ino acids 2-73. This p art of
the protein is sufficient to allow an interaction w ith both cdc2p and p o p2p, thus overexpression results in titration of the proteolytic m achinery th at degrades r u m lp and cdcISp, allow ing bo th proteins to accum ulate. H ow ever, the re replication still occurs in a rum lA strain and the cdcISp levels appear to increase only by about 2 fold w hen the N -term inus of ScCDCô is overexpressed. This level of increase w ould not be inconsistent w ith our results (Fig. 3.7 C lane 1 and D lane 1) w hen ScCDCô is overexpressed. It seems unlikely that a doubling of cdcISp caused by ScCDCô overexpression could produce the same level of re-replication as that found w hen cdcl8 is overexpressed (Fig. 3.7 C lane 3). Wolf et al. discuss
Chapter 3 A nalysis of cdc18 h o m o lo g u es
this point, stating that both popl rum l and pop! ru m l double m utants are able to m aintain a norm al genome ploidy despite levels of cdcISp as high as those in cells overexpressing ScCdc6p. They therefore conclude th at accum ulation of ru m lp appears to be the key event to perm it re-replication. As re-replication can still occur in a rum lA strain, they suggest that in this case, re-replication could be due to sequestration of cdc2p by cdcISp or ScCdc6p, p rev en tin g the kinase from phosphorylating substrates required for inhibiting re-replication. A lthough I can not rule out this possibility, I w ould propose a m ore direct role for ScCdc6p in prom oting re-replication as I have show n in m y w ork that overexpressing the N- term inus of cdcISp does not result in re-replication, despite being able to inhibit the cd c2 p /cd cl3 p kinase (Fig. 2.3), bind to cdc2p (Fig. 2.4) and allow accumulation of ru m lp (Fig. 2.5).
In com paring the w ork of Sanchez et al. w ith th at of m y ow n the question of expression levels is, once again, param ount. In order to achieve significant re replication, an integrated nmtl:5cC0Cb strain had to also possess a m ulticopy vector carrying the transcription factor n tfl, know n to regulate expression from the n m tl prom oter (Tang et al., 1994). This increased ScCdc6p levels sufficiently to allow re replication. The fact that the nmtlScCOCb strain is unable to re-replicate alone is presum ably related to the fact that ScCOib is integrated and also perhaps that they could be using a different cDNA, as discussed previously.
The m ain difference betw een my results and those of Sanchez et al. arises w hen the ability of ScCdc6p to induce re-replication w hen the endogenous cdcl8 function is
com prom ised is examined. Sanchez et al. exam ined this in both a cdcl8-K46 and cdcl8A background and found in both cases th at re-replication occurred to the sam e extent as w hen cdcISp was present. I found that re-replication induced by overexpression of ScCDCô was com prom ised in a cdcl8-K46 m u tan t at 36°C (Fig. 3.7). H ow ever, the experim ental system s used w ere different. I used low er expression levels of ScCDCS b u t maximised the ability to induce re-replication by co-expressing cdtl. It is possible that ScCdc6p is less effective at interacting with a fission yeast protein or perform ing an essential function and therefore w hen it is expressed at relatively low levels, as in m y experim ents, c d c l8p is required to carry o u t th at function. H ow ever, increased expression of ScCdc6p could
Chapter 3 A nalysis of cdc18 h o m o lo g u es
compensate for the fact that it functioned less effectively and c d c l8p w ould not be required. A possible explanation for this could be related to the fact that cdcISp interacts w ith c d tlp (Nishitani et al., 2000). There is no cdtl gene in budding yeast so ScCdc6p m ay interact more weakly w ith cd tlp . CdcISp could therefore prom ote the interaction to allow re-replication w hen ScCdc6p is expressed at a lower level. As w ith the ability to complement, it m ay be interesting to overexpress the cdcl8 hom ologue w ith the cdtl hom ologue from a particular organism and see w hether the proteins are then able to prom ote re-replication.
It w ould appear that none of the cdcl8 homologues exam ined here can functionally substitute for cdcl8. Despite the fact that the cdcl8 hom ologues behave similarly in different organisms, there could be changes in both function and regulation. Some of these differences will be discussed later (Chapter 5).
One of the m ajor differences betw een the organism s w hich could be directly relevant to the function of cdcISp is origin specification. It is know n that the definition of an origin varies greatly betw een different organism s. This suggests that cdcl8 hom ologues could be unable to recognise Schizosaccharomyces pombe origins. Alternatively, higher eukaryotes m ay have a m ore complex pre-replicative complex, possibly involving more proteins. These could be required for the cdcl8 hom ologue to interact at origins. W hen these are absent, as in fission yeast, the hom ologue m ay be unable to bind. As progression is m ade in this direction, we m ay be able to discover if either of these hypotheses are correct.
H s C d c 6 Z l C d c 6 c d c 1 8 S c C d c 6 c o n s e n s u s . . 1 0 . . . S Q A Q A T I S F P I s r s q s s i q f pI i g c h t p r r c nI • 2 0 • • • « 3 0 • tKL^AjgHKAKNSSDA (KTgoTgAKEVSgAKS 'id^ aJidctnItnq . . m S a i p i t p t k I . • • • . 4 0 • • E P T N V Q T V T C E I C S S V S L P L H S P . a . . * . . I R R N ...7 I 7 F D D A P A T . 6 0 • • • . 7 0 SggggLGDDNLCN S^^QLGDDNRCN R K g r L A S S H F g T P Ri^iaOKKLQFTDV * * *---- — — —I — — **** — —___ . . . .8 0 . 7 1 : H L P P C S P P K g G K g 7 1 : P T L S C S P P K ^ r [ 6 4 : R I K T E L G E L a E E n . . . 9 0 : N G P g I S H T L K S T G Q r a r . . T P K [?DLtBn. . . F P 1 0 0 . . . 1 1 0 . . . 1 2 0 F D N Q L T I F T K H P S K R E L A K B H Q M R I I F D E N Q A A A A F l T P L S g L B K l g D P Y Q l l k e n k k p k l p t t p q t p k t^ I t|q i v t p ï | t a e