17
Xenopus Egg Extracts
and Ferrell, 2007; Pomerening et al., 2005; Sha et al., 2003;
Solomon et al., 1990). The experiments ultimately confrmed the model that cell cycle transitions in extract, which
occur independently of nuclear activity, are bistable: only
interphase and M phase are stable states, and both positive
and negative feedback loops ensure that transitions between
the two are rapid and concerted.
Having worked out the nature of the transitions between
S phase and M phase, another important systems-level question that was addressed using egg extracts is: How are cell
cycle transitions propagated within the cytoplasm? This is of
particular interest in large cells such as the amphibian egg.
The surface contraction waves in the egg had been shown to
represent a wave of cell cycle progression from the top (animal pole) of the egg to the bottom (Pérez-Mongiovi et al.,
1998; Rankin and Kirschner, 1997 ). This wave model was
subsequently explored in vitro using egg extract to elucidate
the nature of the cell cycle wave (Chang and Jr, 2013). To
do this, egg extract was supplemented with a recombinant
fuorescent nuclear protein and loaded into capillary tubes.
Cell cycle progression within the tubes was then analyzed by
time-lapse analysis of nuclear accumulation of the f uorescent protein, which occurs only during interphase, when the
nuclear envelope is intact. In this system, the accumulation
and loss of nuclear signal serves as a surrogate for cell cycle
progression. The results indicate that the cell cycle progresses as a trigger wave, expanding from sites of initiation
and radiating out. Similar approaches were subsequently
used to analyze propagation of apoptosis in egg extract, and
recently a previously unknown impact of nuclei on cell cycle
progression was analyzed using a related approach (Afanzar
et al., 2020; Cheng and Ferrell, 2018).
The rapid oscillatory nature of the cell cycle in the early
frog embryo has clearly enabled important studies on the
nature of the cell cycle, but it is also a limitation. Because
the cell cycle proceeds independently of transcription, early
Xenopus models are generally not appropriate for the study
of cell cycle entry or exit, which are transcription dependent
and emerge later in development. In addition, the absence
of Gap/Growth phases in the early embryo has somewhat
limited the utility of egg extracts in the study of events that
unfold during G1 and G2 in somatic cells.
2.5. DNA REPLICATION CONTROL
The ability to control cell cycle transitions in vitro using
egg extracts opened the door for the study of cellular events
downstream of the cell cycle. Perhaps the feld that most benefted from in vitro control of the cell cycle was the study of
DNA replication. Low-speed Xenopus egg extracts have several distinct advantages for DNA replication research. First,
because eggs are stockpiled with material to support the f rst
12 divisions without new RNA synthesis, they contain all
the components to assemble and replicate ~4000 nuclei per
egg. Second, the low-speed extracts are rich in the membranous component that can dock and fuse to form nuclear
envelopes at exit from M phase. Additionally, the extracts
are essentially devoid of nuclei, allowing the addition of
model replication substrates, typically Xenopus sperm nuclei
that have been isolated and demembranated by detergent
treatment. Finally, the absence of plasma membrane makes
trivial the addition of labeled nucleotides, drugs, and other
interventions important to understand DNA replication. As
in many systems, the early embryonic divisions in Xenopus
embryos are much more rapid than those that occur later in
development. In the cleavage-stage Xenopus embryo, cleavages occurred in about 30 minutes, with complete genome
duplication between each cleavage. Although replication
occurs somewhat more slowly in egg extracts, it is largely
complete within 45–60 minutes and very synchronous. This
makes Xenopus egg extract a particularly powerful and tractable tool for studying DNA replication in vitro. What can
take 6–20 hours in somatic cells takes only 2 hours from
start to fnish in the egg extract. Here we describe the contributions that were made to the understanding of DNA replication control through the use of Xenopus egg extracts.
The method of extract preparation would prove to be critical to the study of DNA replication. Using extract prepared
by generating a clarifed supernatant following homogenization of eggs, Méchali and Harland demonstrated eff cient
and complete replication of single-stranded DNA (1982).
However, they were unable to detect replication of doublestranded DNA, limiting the usefulness of this approach. Blow
and Laskey generated a low-speed extract from Xenopus eggs
and showed that it was capable of initiating DNA replication
of both sperm nuclei and plasmid DNA. The extract, following the technique established by Lohka and Masui, was
prepared by centrifugation of packed eggs, which led to stratifcation of the egg components. Critically, the membrane-rich
cytosolic layer could be collected without clarif cation. They
showed that replication of sperm nuclei in this extract was
quite effcient, while replication of plasmid DNA occurred but
was less effcient. They also made the critical observation that
in both cases nuclear envelopes would form around the substrates during incubation in the extract. This suggested that
the nuclear envelope played a critical role in DNA replication
and helped to explain why double-stranded DNA replication
failed in the clarifed extract of Méchali and Harland.
One critical observation made using the turbid cytosolic
extracts was that replication seemed to be well controlled: it
happened once and only once in nuclei added to the extract.
The concept of replication licensing was developed to explain
this phenomenon (Blow, 1993; Blow and Laskey, 1988;
Coverley et al., 1993). The licensing concept suggested that
there were two essential phases of DNA replication control
that were mutually exclusive. Licensing of replication origins
was initially proposed to be dependent on exposure of chromatin to cytosolic components in the extract (“licensing factors”) that were then excluded by the nuclear envelope after it
had formed (Blow and Laskey, 1988). Thus, replication origins, which were licensed by progression through M phase
when the nuclear envelope was broken down, would subsequently initiate DNA replication in interphase but be unable
to replicate again until progression through the next M phase.
Xenopus Egg Extracts
and Ferrell, 2007; Pomerening et al., 2005; Sha et al., 2003;
Solomon et al., 1990). The experiments ultimately confrmed the model that cell cycle transitions in extract, which
occur independently of nuclear activity, are bistable: only
interphase and M phase are stable states, and both positive
and negative feedback loops ensure that transitions between
the two are rapid and concerted.
Having worked out the nature of the transitions between
S phase and M phase, another important systems-level question that was addressed using egg extracts is: How are cell
cycle transitions propagated within the cytoplasm? This is of
particular interest in large cells such as the amphibian egg.
The surface contraction waves in the egg had been shown to
represent a wave of cell cycle progression from the top (animal pole) of the egg to the bottom (Pérez-Mongiovi et al.,
1998; Rankin and Kirschner, 1997 ). This wave model was
subsequently explored in vitro using egg extract to elucidate
the nature of the cell cycle wave (Chang and Jr, 2013). To
do this, egg extract was supplemented with a recombinant
fuorescent nuclear protein and loaded into capillary tubes.
Cell cycle progression within the tubes was then analyzed by
time-lapse analysis of nuclear accumulation of the f uorescent protein, which occurs only during interphase, when the
nuclear envelope is intact. In this system, the accumulation
and loss of nuclear signal serves as a surrogate for cell cycle
progression. The results indicate that the cell cycle progresses as a trigger wave, expanding from sites of initiation
and radiating out. Similar approaches were subsequently
used to analyze propagation of apoptosis in egg extract, and
recently a previously unknown impact of nuclei on cell cycle
progression was analyzed using a related approach (Afanzar
et al., 2020; Cheng and Ferrell, 2018).
The rapid oscillatory nature of the cell cycle in the early
frog embryo has clearly enabled important studies on the
nature of the cell cycle, but it is also a limitation. Because
the cell cycle proceeds independently of transcription, early
Xenopus models are generally not appropriate for the study
of cell cycle entry or exit, which are transcription dependent
and emerge later in development. In addition, the absence
of Gap/Growth phases in the early embryo has somewhat
limited the utility of egg extracts in the study of events that
unfold during G1 and G2 in somatic cells.
2.5. DNA REPLICATION CONTROL
The ability to control cell cycle transitions in vitro using
egg extracts opened the door for the study of cellular events
downstream of the cell cycle. Perhaps the feld that most benefted from in vitro control of the cell cycle was the study of
DNA replication. Low-speed Xenopus egg extracts have several distinct advantages for DNA replication research. First,
because eggs are stockpiled with material to support the f rst
12 divisions without new RNA synthesis, they contain all
the components to assemble and replicate ~4000 nuclei per
egg. Second, the low-speed extracts are rich in the membranous component that can dock and fuse to form nuclear
envelopes at exit from M phase. Additionally, the extracts
are essentially devoid of nuclei, allowing the addition of
model replication substrates, typically Xenopus sperm nuclei
that have been isolated and demembranated by detergent
treatment. Finally, the absence of plasma membrane makes
trivial the addition of labeled nucleotides, drugs, and other
interventions important to understand DNA replication. As
in many systems, the early embryonic divisions in Xenopus
embryos are much more rapid than those that occur later in
development. In the cleavage-stage Xenopus embryo, cleavages occurred in about 30 minutes, with complete genome
duplication between each cleavage. Although replication
occurs somewhat more slowly in egg extracts, it is largely
complete within 45–60 minutes and very synchronous. This
makes Xenopus egg extract a particularly powerful and tractable tool for studying DNA replication in vitro. What can
take 6–20 hours in somatic cells takes only 2 hours from
start to fnish in the egg extract. Here we describe the contributions that were made to the understanding of DNA replication control through the use of Xenopus egg extracts.
The method of extract preparation would prove to be critical to the study of DNA replication. Using extract prepared
by generating a clarifed supernatant following homogenization of eggs, Méchali and Harland demonstrated eff cient
and complete replication of single-stranded DNA (1982).
However, they were unable to detect replication of doublestranded DNA, limiting the usefulness of this approach. Blow
and Laskey generated a low-speed extract from Xenopus eggs
and showed that it was capable of initiating DNA replication
of both sperm nuclei and plasmid DNA. The extract, following the technique established by Lohka and Masui, was
prepared by centrifugation of packed eggs, which led to stratifcation of the egg components. Critically, the membrane-rich
cytosolic layer could be collected without clarif cation. They
showed that replication of sperm nuclei in this extract was
quite effcient, while replication of plasmid DNA occurred but
was less effcient. They also made the critical observation that
in both cases nuclear envelopes would form around the substrates during incubation in the extract. This suggested that
the nuclear envelope played a critical role in DNA replication
and helped to explain why double-stranded DNA replication
failed in the clarifed extract of Méchali and Harland.
One critical observation made using the turbid cytosolic
extracts was that replication seemed to be well controlled: it
happened once and only once in nuclei added to the extract.
The concept of replication licensing was developed to explain
this phenomenon (Blow, 1993; Blow and Laskey, 1988;
Coverley et al., 1993). The licensing concept suggested that
there were two essential phases of DNA replication control
that were mutually exclusive. Licensing of replication origins
was initially proposed to be dependent on exposure of chromatin to cytosolic components in the extract (“licensing factors”) that were then excluded by the nuclear envelope after it
had formed (Blow and Laskey, 1988). Thus, replication origins, which were licensed by progression through M phase
when the nuclear envelope was broken down, would subsequently initiate DNA replication in interphase but be unable
to replicate again until progression through the next M phase.