203
High-Resolution Proteomic Analyses
2015). This platform has been widely adopted, enabling
fundamental discoveries and new insight across many disciplines in biomedicine. We later applied this platform to
profling ten early developmental stages that provided a new
view of development and differentiation of over 200 cell
types in early Xenopus embryo (Briggs et al., 2018)—work
which was recognized along with zebraf sh counterparts
as the Science Magazine Breakthrough of 2018 (Harland,
2018; Pennisi, 2018). Yet, as revealing as mRNA expression
profles might be, they do not fully capture and sometimes
are strongly discordant with the protein expression, as we
(Peshkin et al., 2015) and others (Smits et al., 2014) have
shown. It is particularly relevant in the early embryo in
which a maternal protein dowry has been synthesized in the
mother and endocytosed by oocytes; as a result, many proteins might have no trace of the respective mRNA expression in cells resulting from progressive oocyte cleavages
(Peshkin et al., 2015; Sonnett et al., 2018b). Therefore, it
would be very attractive to profle protein expression at the
single-cell level. A critical barrier on this path is the absence
of reverse translation biochemistry. While a molecule of
mRNA can be reverse-transcribed into cDNA and then
amplifed exponentially for detection and quantitation, there
has so far been no discovery of a counterpart reaction for
proteins, despite some intriguing proposed methods (Cook
et al., 1977; Martin, 2006; Nashimoto, 2001). Attempts are
being made to work directly with miniscule material quantities available from a single cell by reducing the loss of protein during sample preparation steps, but these show very
few reproducibly measurable proteins and a limited dynamic
range (Cheung et al., 2020). Xenopus offers a way out of this
dilemma by providing large single cells. Technically speaking, single-cell proteomics in Xenopus begins with an egg
(Smits et al., 2014; Lindeboom et al., 2019), since a 1.43-mmdiameter oocyte of X. laevis (Leibovich et al., 2020) is in
fact a single cell. After a few initial divisions, blastomeres
are still large—recently an effort has been made to examine individual blastomeres in early embryos and characterize lineage-specifc protein deposit. The Nemes and Moody
labs profled single blastomeres with mass spectrometry
and were able to quantify ~400 proteins of individual cells
in the 16-cell stage (Onjiko et al., 2015; Lombard-Banek
et al., 2016). First, they were able to minimize derivatization steps to enhance analytical sensitivity and use label-free
quantif cation (Lombard-Banek et al., 2016) to obtain a low
limit of detection and quantifcation for proteins in complex cell digests. Separately, both the Nemes and Moody
(Onjiko et al., 2015) and Klein and Garcia (Saha-Shah et al.,
2019) groups used pulled glass pipettes to capture material
from specifc blastomere cells, and animal-vegetal differences have been reproducibly measured. The Dovichi lab
has also compared the protein expression among different
blastomeres and observed that the blastomere-to-blastomere
heterogeneity in 8-, 16-, 32-, and 50-cell embryos increases
with development stage (Sun et al., 2016), ref ecting progressive cellular differentiation.
13.9. DISCUSSION AND FUTURE DIRECTIONS
In this chapter, we discussed the successes of mass spectrometrybased proteomics in Xenopus, including methodology—sample
preparation and bioinformatics analysis—and biological
fndings. It has been recognized for decades that Xenopus
complements the more commonly used mouse and zebraf sh
animal models in the study of human biology and disease.
Xenopus is particularly attractive for proteomics studies due
to the ease of obtaining large amounts of protein without
artifcial perturbations needed, for example, synchronization
of cells (with regard to cell cycle) or embryos (with regard
to fertilization). While even a single egg or an early embryo
in X. laevis contains suffcient amounts of non-yolk protein
for some measurements of relative and absolute ( Smits et al.,
2014 ) protein abundance, ten embryos provide a protein
amount suffcient for most protein measurements. Over 1000
sibling embryos can be obtained from a single clutch, and an
in-vitro fertilization can be done with near-perfect (within
one minute) synchrony.
This brief chapter could not exhaustively cover every
aspect of Xenopus proteomics, and we encourage the reader
to follow through other recent method reviews (Gupta et al.,
2018; Sonnett et al., 2018b; Gilchrist et al., 2020). The f eld of
Xenopus proteomics is in its infancy, and we have many reasons to expect great advances in developmental and cellular
biology from bringing together the strengths of Xenopus as
a system with advances in mass spectrometric instrumentation, biochemistry, machine learning, and analytics. The key
component that will enable these advances is bioinformatics.
This is true both for making existing methods more accessible to scientists without advanced computational skills and
for creating new approaches to analysis of the data. Hosting a
proteomic processing pipeline at Xenbase could go a long way
to achieving this by maintaining the most complete and current set of reference proteins, using the optimal set of parameters for the spectra searches, and facilitating the archival and
interactive access to the resulting datasets; it would also be
benefcial for ongoing refnement of gene models.
We can anticipate some of the developments in the f eld.
The push for single-cell proteomics will inevitably come
full circle—having studied the repertoire of cell types and
distilled abundant and specifc cell type markers, we will
develop ways to enrich single cells of a given type from
dissociated embryos to extract enough material for celltype-focused deep proteomics. In addition, we can expect
the single-cell mRNA sequencing data for both Xenopus
species to improve, which will allow for a more thorough
classifcation of spatial expression across development that
can then be used to deconvolve bulk protein data. It will
be useful to not only increase our catalogue of cell type
specifc mRNAs but also have more certainty about which
mRNAs—and potentially proteins—are expressed in all
embryonic cell types.
As CRISPR knock-ins become commonplace in Xenopus,
constructs for lineage tracing and controllable gene regulatory
High-Resolution Proteomic Analyses
2015). This platform has been widely adopted, enabling
fundamental discoveries and new insight across many disciplines in biomedicine. We later applied this platform to
profling ten early developmental stages that provided a new
view of development and differentiation of over 200 cell
types in early Xenopus embryo (Briggs et al., 2018)—work
which was recognized along with zebraf sh counterparts
as the Science Magazine Breakthrough of 2018 (Harland,
2018; Pennisi, 2018). Yet, as revealing as mRNA expression
profles might be, they do not fully capture and sometimes
are strongly discordant with the protein expression, as we
(Peshkin et al., 2015) and others (Smits et al., 2014) have
shown. It is particularly relevant in the early embryo in
which a maternal protein dowry has been synthesized in the
mother and endocytosed by oocytes; as a result, many proteins might have no trace of the respective mRNA expression in cells resulting from progressive oocyte cleavages
(Peshkin et al., 2015; Sonnett et al., 2018b). Therefore, it
would be very attractive to profle protein expression at the
single-cell level. A critical barrier on this path is the absence
of reverse translation biochemistry. While a molecule of
mRNA can be reverse-transcribed into cDNA and then
amplifed exponentially for detection and quantitation, there
has so far been no discovery of a counterpart reaction for
proteins, despite some intriguing proposed methods (Cook
et al., 1977; Martin, 2006; Nashimoto, 2001). Attempts are
being made to work directly with miniscule material quantities available from a single cell by reducing the loss of protein during sample preparation steps, but these show very
few reproducibly measurable proteins and a limited dynamic
range (Cheung et al., 2020). Xenopus offers a way out of this
dilemma by providing large single cells. Technically speaking, single-cell proteomics in Xenopus begins with an egg
(Smits et al., 2014; Lindeboom et al., 2019), since a 1.43-mmdiameter oocyte of X. laevis (Leibovich et al., 2020) is in
fact a single cell. After a few initial divisions, blastomeres
are still large—recently an effort has been made to examine individual blastomeres in early embryos and characterize lineage-specifc protein deposit. The Nemes and Moody
labs profled single blastomeres with mass spectrometry
and were able to quantify ~400 proteins of individual cells
in the 16-cell stage (Onjiko et al., 2015; Lombard-Banek
et al., 2016). First, they were able to minimize derivatization steps to enhance analytical sensitivity and use label-free
quantif cation (Lombard-Banek et al., 2016) to obtain a low
limit of detection and quantifcation for proteins in complex cell digests. Separately, both the Nemes and Moody
(Onjiko et al., 2015) and Klein and Garcia (Saha-Shah et al.,
2019) groups used pulled glass pipettes to capture material
from specifc blastomere cells, and animal-vegetal differences have been reproducibly measured. The Dovichi lab
has also compared the protein expression among different
blastomeres and observed that the blastomere-to-blastomere
heterogeneity in 8-, 16-, 32-, and 50-cell embryos increases
with development stage (Sun et al., 2016), ref ecting progressive cellular differentiation.
13.9. DISCUSSION AND FUTURE DIRECTIONS
In this chapter, we discussed the successes of mass spectrometrybased proteomics in Xenopus, including methodology—sample
preparation and bioinformatics analysis—and biological
fndings. It has been recognized for decades that Xenopus
complements the more commonly used mouse and zebraf sh
animal models in the study of human biology and disease.
Xenopus is particularly attractive for proteomics studies due
to the ease of obtaining large amounts of protein without
artifcial perturbations needed, for example, synchronization
of cells (with regard to cell cycle) or embryos (with regard
to fertilization). While even a single egg or an early embryo
in X. laevis contains suffcient amounts of non-yolk protein
for some measurements of relative and absolute ( Smits et al.,
2014 ) protein abundance, ten embryos provide a protein
amount suffcient for most protein measurements. Over 1000
sibling embryos can be obtained from a single clutch, and an
in-vitro fertilization can be done with near-perfect (within
one minute) synchrony.
This brief chapter could not exhaustively cover every
aspect of Xenopus proteomics, and we encourage the reader
to follow through other recent method reviews (Gupta et al.,
2018; Sonnett et al., 2018b; Gilchrist et al., 2020). The f eld of
Xenopus proteomics is in its infancy, and we have many reasons to expect great advances in developmental and cellular
biology from bringing together the strengths of Xenopus as
a system with advances in mass spectrometric instrumentation, biochemistry, machine learning, and analytics. The key
component that will enable these advances is bioinformatics.
This is true both for making existing methods more accessible to scientists without advanced computational skills and
for creating new approaches to analysis of the data. Hosting a
proteomic processing pipeline at Xenbase could go a long way
to achieving this by maintaining the most complete and current set of reference proteins, using the optimal set of parameters for the spectra searches, and facilitating the archival and
interactive access to the resulting datasets; it would also be
benefcial for ongoing refnement of gene models.
We can anticipate some of the developments in the f eld.
The push for single-cell proteomics will inevitably come
full circle—having studied the repertoire of cell types and
distilled abundant and specifc cell type markers, we will
develop ways to enrich single cells of a given type from
dissociated embryos to extract enough material for celltype-focused deep proteomics. In addition, we can expect
the single-cell mRNA sequencing data for both Xenopus
species to improve, which will allow for a more thorough
classifcation of spatial expression across development that
can then be used to deconvolve bulk protein data. It will
be useful to not only increase our catalogue of cell type
specifc mRNAs but also have more certainty about which
mRNAs—and potentially proteins—are expressed in all
embryonic cell types.
As CRISPR knock-ins become commonplace in Xenopus,
constructs for lineage tracing and controllable gene regulatory