15
Xenopus Egg Extracts
that cyclin mRNA (in this case from sea urchins) was necessary and suffcient to drive oscillations in MPF activity
( Murray and Kirschner, 1989 ). To do this, they destroyed all
endogenous RNA in the extract by nuclease treatment and
then added back synthetic cyclin mRNA (in the presence of
RNase inhibitor). This resulted in multiple cell cycles in vitro.
In this cycling extract, the Cyclin protein accumulated during interphase and was degraded at exit from M phase. This
elegant experiment showed that everything required to drive
the oscillation between interphase and mitosis was present
in the egg cytosol and, importantly, that the level of Cyclin
protein oscillated with MPF activity. The highly conserved
kinase activated by Cyclin had previously been identif ed as
one of the cell cycle genes mutated in Hartwell’s screen in
budding yeast ( cdc2) and a similar screen done in f ssion
yeast (cdc28) ( Hartwell et al., 1974 ; Lörincz and Reed,
1984 ). This kinase was indeed so conserved that the human
gene was able to complement the temperature-sensitive
allele in f ssion yeast, confrming the universal nature of cell
cycle control in eukaryotes ( Lee and Nurse, 1987 ).
The observation that cytoplasmic factors drive cell cycle
progression was also consistent with a striking observation
made previously by Hara and Kirschner (Hara et al., 1980).
They had found that enucleated Xenopus eggs that were
stimulated to exit meiotic arrest underwent a series of coordinated movements called surface contraction waves. When
the eggs were examined from the side, these movements
caused the eggs to appear to “bounce” up and down with the
same periodicity as the cell cycle. The fact that these events
occurred in the absence of nuclei was consistent with the
notion that the cell cycle machinery represented an autonomous, cytoplasmic oscillator to which cytoskeletal components would respond. In the following years, we learned that
cycling extracts and “bouncing eggs” are detectable in frog
eggs because these cells are insensitive to external inputs:
checkpoints that become active later in development are
not yet fully active, so the cytoplasmic oscillatory machine
trundles along, unaffected by feedback from DNA replication, a mitotic spindle, or cell cleavage. In the context of the
Xenopus embryo, the ability to undergo more than a dozen
cleavage divisions without new gene expression forms the
basis for its remarkable usefulness as an in vitro model for
studying cell division.
2.3. PROTEIN DEGRADATION DRIVES
THE CELL CYCLE
In their extract experiments, Murray and Kirschner established that a particular deletion mutant of Cyclin (Δ90)
was able to drive entry into mitotic (M) phase but was not
destroyed like wild-type Cyclin, preventing M phase exit
(Murray et al., 1989). These results suggested that regulated
proteolysis of Cyclin was necessary for exit from M phase.
To better understand how Cyclin destruction was regulated,
the Kirschner group purifed the activity leading to Cyclin
degradation from Xenopus egg extract. Prior to degradation,
Cyclin had been seen to accumulate high molecular weight
derivatives, which were ultimately shown to be ubiquitin
conjugates (Glotzer et al., 1991). The activity eventually
purifed was later identifed as an E3 ubiquitin ligase, which
was named the anaphase promoting complex (APC) (King
et al., 1995). Satisfyingly, this large (20S) protein complex
was shown to contain subunits homologous to yeast proteins
that had previously been shown to control Cyclin degradation (Irniger et al., 1995). While the APC was purif ed from
frog egg extract, the analogous complex was purif ed from
clam embryos by Hershko and colleagues and named the
Cyclosome (Sudakin et al., 1995). Thus, again, yeast genetics
and extract biochemistry converged on a conserved mechanism of cell cycle control in all eukaryotes. The cell cycle
resulted from the accumulation of mitotic Cyclin protein,
which in turn led to activation of MPF, which in turn led to
activation of the anaphase promoting complex/Cyclosome
(APC/C), which in turn led to destruction of Cyclin, resetting the cycle.
It soon became evident that the APC/C had other substrates in addition to Cyclin. As mentioned previously,
Murray and Kirschner had found that deletion of the N
terminus of mitotic Cyclin (Cyclin Δ90) resulted in mitotic
arrest. The nature of the arrest was interesting and unique:
although MPF levels remained high due to the persistence
of Cyclin, the APC/C remained active, creating an artif cial
“early anaphase”-like state in vitro. This discovery provided
an ideal system with which to identify additional APC/C
substrates. By incubating small pools of in vitro translated
and radiolabeled proteins in egg extracts with active APC, it
was possible to identify additional substrates based on their
instability in Cyclin Δ90-treated extract, compared to interphase controls. In this manner, several key effectors of cell
cycle control were identifed (King et al., 1997; McGarry
and Kirschner, 1998; Stukenberg et al., 1997; Zou et al.,
1999). In time it would be shown that the APC/C in somatic
cells was activated by two different substrate specif city factors: Cdc20 protein upon mitotic exit and Cdh1 in the Gap1
or Growth1 (G1) phase of the cell cycle (Fang et al., 1998a).
Although Cdh1 activity is undetectable in Xenopus egg
extracts, the addition of recombinant Cdh1 protein to interphase extract results in an artifcial G1-like state in which
Cdh1-dependent substrates of the APC/C were degraded.
Based on these observations, additional small pool screens
were performed in egg extracts to identify G1 substrates of
the APC/C (Ayad et al., 2005, 2003; Rankin et al., 2005).
Similar screens were performed to identify mitotic phosphoproteins, based on their shift in electrophoretic mobility in mitotic versus interphase extracts (Lustig et al., 1997;
Stukenberg et al., 1997 )
Xenopus egg extracts were also used to investigate the
mechanisms of APC-dependent degradation. Because degradation of radiolabeled APC substrates is readily detectable
in egg extracts, it was relatively straightforward to identify
specifc sequences in substrate proteins that promote their
recognition and degradation by screening for mutations
that disrupted degradation. Cdc20 substrates were shown
Xenopus Egg Extracts
that cyclin mRNA (in this case from sea urchins) was necessary and suffcient to drive oscillations in MPF activity
( Murray and Kirschner, 1989 ). To do this, they destroyed all
endogenous RNA in the extract by nuclease treatment and
then added back synthetic cyclin mRNA (in the presence of
RNase inhibitor). This resulted in multiple cell cycles in vitro.
In this cycling extract, the Cyclin protein accumulated during interphase and was degraded at exit from M phase. This
elegant experiment showed that everything required to drive
the oscillation between interphase and mitosis was present
in the egg cytosol and, importantly, that the level of Cyclin
protein oscillated with MPF activity. The highly conserved
kinase activated by Cyclin had previously been identif ed as
one of the cell cycle genes mutated in Hartwell’s screen in
budding yeast ( cdc2) and a similar screen done in f ssion
yeast (cdc28) ( Hartwell et al., 1974 ; Lörincz and Reed,
1984 ). This kinase was indeed so conserved that the human
gene was able to complement the temperature-sensitive
allele in f ssion yeast, confrming the universal nature of cell
cycle control in eukaryotes ( Lee and Nurse, 1987 ).
The observation that cytoplasmic factors drive cell cycle
progression was also consistent with a striking observation
made previously by Hara and Kirschner (Hara et al., 1980).
They had found that enucleated Xenopus eggs that were
stimulated to exit meiotic arrest underwent a series of coordinated movements called surface contraction waves. When
the eggs were examined from the side, these movements
caused the eggs to appear to “bounce” up and down with the
same periodicity as the cell cycle. The fact that these events
occurred in the absence of nuclei was consistent with the
notion that the cell cycle machinery represented an autonomous, cytoplasmic oscillator to which cytoskeletal components would respond. In the following years, we learned that
cycling extracts and “bouncing eggs” are detectable in frog
eggs because these cells are insensitive to external inputs:
checkpoints that become active later in development are
not yet fully active, so the cytoplasmic oscillatory machine
trundles along, unaffected by feedback from DNA replication, a mitotic spindle, or cell cleavage. In the context of the
Xenopus embryo, the ability to undergo more than a dozen
cleavage divisions without new gene expression forms the
basis for its remarkable usefulness as an in vitro model for
studying cell division.
2.3. PROTEIN DEGRADATION DRIVES
THE CELL CYCLE
In their extract experiments, Murray and Kirschner established that a particular deletion mutant of Cyclin (Δ90)
was able to drive entry into mitotic (M) phase but was not
destroyed like wild-type Cyclin, preventing M phase exit
(Murray et al., 1989). These results suggested that regulated
proteolysis of Cyclin was necessary for exit from M phase.
To better understand how Cyclin destruction was regulated,
the Kirschner group purifed the activity leading to Cyclin
degradation from Xenopus egg extract. Prior to degradation,
Cyclin had been seen to accumulate high molecular weight
derivatives, which were ultimately shown to be ubiquitin
conjugates (Glotzer et al., 1991). The activity eventually
purifed was later identifed as an E3 ubiquitin ligase, which
was named the anaphase promoting complex (APC) (King
et al., 1995). Satisfyingly, this large (20S) protein complex
was shown to contain subunits homologous to yeast proteins
that had previously been shown to control Cyclin degradation (Irniger et al., 1995). While the APC was purif ed from
frog egg extract, the analogous complex was purif ed from
clam embryos by Hershko and colleagues and named the
Cyclosome (Sudakin et al., 1995). Thus, again, yeast genetics
and extract biochemistry converged on a conserved mechanism of cell cycle control in all eukaryotes. The cell cycle
resulted from the accumulation of mitotic Cyclin protein,
which in turn led to activation of MPF, which in turn led to
activation of the anaphase promoting complex/Cyclosome
(APC/C), which in turn led to destruction of Cyclin, resetting the cycle.
It soon became evident that the APC/C had other substrates in addition to Cyclin. As mentioned previously,
Murray and Kirschner had found that deletion of the N
terminus of mitotic Cyclin (Cyclin Δ90) resulted in mitotic
arrest. The nature of the arrest was interesting and unique:
although MPF levels remained high due to the persistence
of Cyclin, the APC/C remained active, creating an artif cial
“early anaphase”-like state in vitro. This discovery provided
an ideal system with which to identify additional APC/C
substrates. By incubating small pools of in vitro translated
and radiolabeled proteins in egg extracts with active APC, it
was possible to identify additional substrates based on their
instability in Cyclin Δ90-treated extract, compared to interphase controls. In this manner, several key effectors of cell
cycle control were identifed (King et al., 1997; McGarry
and Kirschner, 1998; Stukenberg et al., 1997; Zou et al.,
1999). In time it would be shown that the APC/C in somatic
cells was activated by two different substrate specif city factors: Cdc20 protein upon mitotic exit and Cdh1 in the Gap1
or Growth1 (G1) phase of the cell cycle (Fang et al., 1998a).
Although Cdh1 activity is undetectable in Xenopus egg
extracts, the addition of recombinant Cdh1 protein to interphase extract results in an artifcial G1-like state in which
Cdh1-dependent substrates of the APC/C were degraded.
Based on these observations, additional small pool screens
were performed in egg extracts to identify G1 substrates of
the APC/C (Ayad et al., 2005, 2003; Rankin et al., 2005).
Similar screens were performed to identify mitotic phosphoproteins, based on their shift in electrophoretic mobility in mitotic versus interphase extracts (Lustig et al., 1997;
Stukenberg et al., 1997 )
Xenopus egg extracts were also used to investigate the
mechanisms of APC-dependent degradation. Because degradation of radiolabeled APC substrates is readily detectable
in egg extracts, it was relatively straightforward to identify
specifc sequences in substrate proteins that promote their
recognition and degradation by screening for mutations
that disrupted degradation. Cdc20 substrates were shown