PNAS
2024 Vol. 121 No. 25 e2322588121
https://doi.org/10.1073/pnas.2322588121
1 of 11
RESEARCH ARTICLE
|
Significance
We show that tissue
-
specific
proteomic analysis is achievable
and can be done efficiently at an
individual level in a small
nematode, with resolution
sufficient for genetic analysis on
a single
-
animal basis. With data
collected from single animals, we
produced high
-
quality sets of
proteins that described the
proteomes of the gonad and the
intestine. Comparison of these
proteomes with the organs’
transcriptomes improved our
understanding of interorgan
protein transport. We applied
single
-
tissue proteomic to
describe the proteome of the gut
granules in the nematode
intestine, a specialized lysosome
-
related organelle (LRO) with
important functions but which
is not well characterized; we
identified additional putative LRO
proteins and verified two by
subcellular localization.
Preprint server: Pre
-
print on bioRxiv: doi:
https://doi.
org/10.1101/2023.12.29.573592
.
Author contributions: C.
-
H.T., T.
-
Y.W., B.L., T.
-
F.C., and
P.W.S. designed research; C.
-
H.T., T.
-
Y.W., H.P., and B.L.
performed research; H.P. contributed new reagents/
analytic tools; C.
-
H.T., T.
-
Y.W., and P.W.S. analyzed data;
and C.
-
H.T., T.
-
Y.W., T.
-
F.C., and P.W.S. wrote the paper.
Reviewers: M.d.B., Institute of Science and Technology
Austria; and A.J.W., University of Massachusetts Medical
School.
The authors declare no competing interest.
Copyright © 2024 the Author(s). Published by PNAS.
This article is distributed under
Creative Commons
Attribution
-
NonCommercial
-
NoDerivatives License 4.0
(CC BY
-
NC
-
ND)
.
1
C.
-
H.T. and T.
-
Y.W. contributed equally to this work.
2
To whom correspondence may be addressed. Email:
tfchou@caltech.edu or pws@caltech.edu.
This article contains supporting information online at
https://www.pnas.org/lookup/suppl/doi:10.1073/pnas.
2322588121/-
/DCSupplemental
.
Published June 11, 2024.
CELL BIOLOGY
Single
-
tissue proteomics in
Caenorhabditis elegans
reveals
proteins resident in intestinal lysosome
-
related organelles
Chieh
-
Hsiang Tan
a,1
, Ting
-
Yu Wang
b,1
, Heenam Park
a
, Brett Lomenick
b
, Tsui
-
Fen Chou
a,b,2
, and Paul W. Sternberg
a,2
Contributed by Paul W. Sternberg; received December 25, 2023; accepted May 6, 2024; reviewed by Mario de Bono and Albertha J. Walhout
The nematode intestine is the primary site for nutrient uptake and storage as well as the
synthesis of biomolecules; lysosome
-
related organelles known as gut granules are impor
-
tant for many of these functions. Aspects of intestine biology are not well understood,
including the export of the nutrients it imports and the molecules it synthesizes, as well
as the complete functions and protein content of the gut granules. Here, we report a
mass spectrometry (MS)–based proteomic analysis of the intestine of the
Caenorhabditis
elegans
and of its gut granules. Overall, we identified approximately 5,000 proteins each
in the intestine and the gonad and showed that most of these proteins can be detected
in samples extracted from a single worm, suggesting the feasibility of individual
-
level
genetic analysis using proteomes. Comparing proteomes and published transcriptomes
of the intestine and the gonad, we identified proteins that appear to be synthesized in
the intestine and then transferred to the gonad. To identify gut granule proteins, we
compared the proteome of individual intestines deficient in gut granules to the wild
type. The identified gut granule proteome includes proteins known to be exclusively
localized to the granules and additional putative gut granule proteins. We selected two
of these putative gut granule proteins for validation via immunohistochemistry, and our
successful confirmation of both suggests that our strategy was effective in identifying
the gut granule proteome. Our results demonstrate the practicability of single
-
tissue
MS–based proteomic analysis in small organisms and in its future utility.
tissue
-
specific mass
-
spectrometry | lysosome
-
related organelle | microproteomics | yolk protein
Comprising just 20 cells (1, 2), the intestine of
Caenorhabditis elegans
is not only the
primary organ for nutrient uptake and storage but also the site for metal detoxification
(3–7) and the synthesis and export of various molecules (8–10). The intestine also plays
a key role in the aging process, as well as in the responses to stress and pathogens (11–13).
Some of these roles depend on a distinct type of organelle commonly known as gut gran-
ules, a type of lysosome
-
related organelle (LRO) (14–17). Likely analogous to the rhabditin
granules observed in other nematodes in the order Rhabditida (18–23), gut granules are
robustly present in the intestinal cells (16, 23). Studies have linked them to many intestinal
functions, including metabolism and storage of nutrients and trace metals, biogenesis of
ascarosides, and immunity and stress response (6, 7, 10, 24–27). However, many functions
of the intestine, such as how the absorbed nutrients are passed on to other tissues and the
characteristics of the gut granules, remain poorly understood.
While proteomic analyses have been applied extensively in
C. elegans
research, especially
in the fields of aging (28–33) and stress response (34–36), tissue
-
level analyses have been
hindered by the small body size of the animal. A variety of strategies have been adopted to
overcome this limitation, including tissue enrichment with mutants (37, 38) and labeling
proteins in specific tissues with noncanonical amino acids (39, 40) or biotin (41, 42).
Analyzing tissue isolates could still offer the cleanest proteome with a much
-
reduced back-
ground, but it has been mostly prohibitive for somatic tissues, as they have to be manually
extracted. Advancements in mass spectrometry (MS)–based proteomics have made analyses
of increasingly small amounts of samples possible (43), including analyses of the
C. elegans
proteome using single worms (44–46). Individual animals are the basic units of genetic
and physiological research, and the ability to perform tissue
-
specific proteomics at this level
would be a valuable analytical tool.
We performed MS–based proteomic analysis of manually extracted intestines and
gonads of
C. elegans
and showed that most of the detected proteins can be detected using
samples extracted from a single animal. We produced high
-
quality sets of proteins that
described the proteomes of the gonad and the intestine. In a comparison of these pro-
teomes with the organs’ transcriptomes, we identified proteins that appear to be synthe-
sized in the intestine and then transferred to the gonad. We applied single
-
tissue
proteomics to analyze the proteome of the gut granules by comparing the proteome of
Downloaded from https://www.pnas.org by George Porter on June 12, 2024 from IP address 131.215.225.155.
2 of 11
https://doi.org/10.1073/pnas.2322588121
pnas.org
individual intestines deficient in gut granules to the wild type.
Our method identified proteins known to be exclusively localized
to the granules and additional putative gut granule proteins. We
validated two by immunohistochemistry: MRP
-
3 and W05H9.1,
which we named LRO
-
1 (LROs protein).
Our results show that tissue
-
level MS–based proteomic analyses
in small organisms, such as nematodes, can be done effectively at
an individual animal level, allowing its utilization in phenotypic
and genetic analyses.
Results
MS–Based Proteomic Analysis of Isolated Nematode Tissues.
We first established a method for MS–based proteomic analysis of
the
C. elegans
intestine and gonad using manually isolated tissues.
We extracted and isolated individual intestines (Fig. 1
A
and
C
)
and gonads (Fig. 1
A
and
B
) through dissection. Specifically, the
intestinal tissues were severed at both ends and the gonads severed
at the spermatheca to disconnect them from the rest of the worm
and enable their removal. Samples were then snap
-
frozen and
prepared “in
-
tube” using a “single
-
pot” strategy similar to those
used in single
-
cell analysis (47). The samples were heated, sonicated,
and digested with Lys
-
C and trypsin before MS analysis (Fig. 1
D
,
See
Materials and Methods
). Prior to adopting the current method,
we tested other methods commonly used for preparing samples of
larger quantities, all of which resulted in a very significant loss of
material. The need to treat these tissues similarly to individual cells
was not unexpected due to the small size of the worm and especially
its organs. A single worm’s intestinal tissue contains only 20 cells,
and to avoid contamination with the connected components of the
alimentary system, the intestines were not extracted in their entirety.
The Intestinal and Gonadal Proteomes of
C. elegans
.
Using MS,
we detected a total of 5,962 proteins from a total of 30 wild
-
type
gonad and intestine samples. Of these, 5,243 and 4,848 were
identified in the gonad and the intestine, respectively (Fig. 2
A
and
Dataset S1
). From a single experiment with four independent
samples of each organ, each consisting of tissues collected from
five worms, we were able to detect a total of 5,464 proteins. Of
these proteins, 4,720 and 4,267 were identified in the gonad
and the intestine, respectively (Fig. 2
B
and
Dataset S1
). Overall,
samples of the same tissue type show a high degree of similarity
in both the identified proteins and their abundances while being
clearly distinctive from samples of the other tissue type, albeit
with notable heterogeneity across individual samples (Fig. 2
C
and
D
and
Dataset S1
).
Of the 5,962 total proteins detected, 5,297 were repeatedly
detected in the same tissue type. From these, 3,501 were shared
in both the intestine and the gonad, 1,060 were detected only in
the gonad, and 726 were detected only in the intestine (Fig.
2
E
and
Dataset S1
). We subject these protein groupings to Gene
Ontology enrichment analysis (48, 49), and found that, as
expected, proteins involved in housekeeping processes such as
ribosome assembly were enriched biological processes of the shared
protein group. By contrast, proteins involved in processes expected
to be prominent in the proliferating germ cells, such as cell cycle,
were enriched in the group only found in the gonad. Proteins
involved in processes expected to be prominent in the intestinal
cells, such as lipid metabolism and immune response, were
enriched in the group only found in the intestine (Fig.
2
E
and
Dataset S1
).
Single
-
Tissue Proteomic Analysis.
To test whether our method
allows suitable coverage of the proteome, we performed MS analysis
of samples each consisting of one tissue type gathered from a single
animal. We found that large portions of the tissue proteome could
indeed be identified from the tissue of a single animal. Specifically,
we identified a total of 4,012 proteins from six independent
dissected single gonads, with up to 3,316 (82.7%) identifiable
from a single gonad (Fig. 3
A
and
Dataset S2
). Similarly, a total
of 3,732 proteins were identified from six independent dissected
intestines, with up to 2,889 (77.4%) identifiable from a single
intestine (Fig. 3
B
and
Dataset S2
). In both cases, the proteins
identifiable from tissues gathered from a single worm covered
over half of all the proteins we identified in our research (Figs. 2
A
and 3
A
and
B
). As with the multitissue samples, the tissues of the
same type were much more similar to one another than they were
to the other tissue both in terms of proteins identified and their
abundances (Fig. 3
C
and
D
and
Dataset S2
). Similar results were
also obtained from an independent trial (
SI Appendix
, Fig. S1
and
Dataset S3
). The proteins identified from each tissue type are also
largely consistent, with nearly half of all the identified proteins
identifiable in every animal, and with most of the identifiable
proteins quantifiable suggesting that this is a viable method
for tissue proteome analysis (Fig. 3
E
,
SI Appendix
, Fig. S2
, and
Datasets S2
and
S4
).
Taking advantage of the significant quantities of data we col-
lected from our single
-
tissue MS analysis, we compiled a “high
confidence” list of proteins identified in either the gonad or the
intestine (Fig.
3
F
and
Dataset S2
). The listed proteins were those
that were consistently found in the associated tissue type (found
in at least half of datasets from each of the two trials in which each
sample was a single
-
tissue type collected from a single animal). In
this way, we identified 2,368 proteins frequently found in gonad
samples and 1,713 frequently found in intestinal samples. From
these proteins, we then identified proteins that were consistently
enriched in the gonad or the intestine (Fig.
3
G
and
Dataset S2
),
based on the differential abundance in the two tissues (>twofold
difference from the other tissue type in both trials). 450 and 418
Fig. 1.
MS–based proteomic analysis of extracted nematode tissues. (
A
) A
young adult stage hermaphrodite
C. elegans
with the intestine (blue shade)
and gonad (red shade) colored for identification. (
B
) An extracted gonad arm
severed at the spermatheca. (
C
) Extracted intestinal tissue. (
D
) Schematic
representation of the workflow. Intestinal tissues and gonads were extracted
from young adult worms, snap
-
frozen, heated, sonicated, and digested prior
to MS and data analysis. The figure was created with
BioRender.com
.
Downloaded from https://www.pnas.org by George Porter on June 12, 2024 from IP address 131.215.225.155.
PNAS
2024 Vol. 121 No. 25 e2322588121
https://doi.org/10.1073/pnas.2322588121
3 of 11
proteins were found to be consistently enriched in the gonad and
intestine.
Proteome–Transcriptome Comparison Reveals Gonadal Proteins
with an Intestinal Origin.
We compared our proteomic analysis
to the transcriptomic data of the same tissue types by Han et al.
(50). Similar to our study, their sample material was also collected
through dissection and physical isolation, and gene enrichment
based on cross
-
comparison of the gonad and the intestine.
Comparing the proteins that were enriched in the gonad or the
intestine to the mRNAs enriched in the respective organs, we
found that most of the enriched proteins (82.7% of the gonadal
proteins and 62.2% of the intestinal proteins) were associated
with a higher expression level of their corresponding genes in the
same organ (Fig. 4
A
and
B
and
Dataset S5
). Although part of this
difference is likely due to differences in methods, it also genuinely
reflects the differences between the abundance of the proteins
and the mRNA expression levels. For example, a few proteins
are known to be synthesized in the intestine of
C. elegans
and
then exported to other tissues. A well
-
established example is the
vitellogenins (yolk protein precursors), encoded by six genes (
vit
-
1
~
6
) that are expressed exclusively in the intestine (51, 52), but
after synthesis in the intestine, they are transported to the maturing
oocyte in the gonad (8). To further investigate intestine
-
to
-
gonad
Fig. 2.
The intestinal and gonadal proteomes of
C. elegans
. (
A
) A total of 5,962 proteins were detected in our wild
-
type gonad and intestine samples. 5,243 were
identified in the gonad, and 4,848 were identified in the intestine. 4,129 proteins were identified in both the intestine and the gonad, while 1,114 were identified
only in the gonad, and 719 were identified only in the intestine. (
B–D
) Overview of proteins identified in the pooled tissue samples. Tissues were collected from
five animals in each of the four samples used per tissue type. (
B
) The data from the four pooled samples for each tissue were analyzed together. From these
pooled tissue samples, 5,464 proteins were detected. 3,532 of those were shared in both tissues, while 1,188 were identified only in the gonad and 744 only in
the intestine. (
C
and
D
) The tissue samples of the same type show a high degree of similarity in both the identified proteins and their abundances while being
clearly distinctive from those of the other tissue type. (
C
) PCA of the pooled tissue samples. Samples were divided into two distinct groups separated on the PC1.
(
D
) Heatmap displaying the proteins of the pooled
-
tissue samples. The Z
-
score, representing the distance in SDs from the mean within each row, was computed
by subtracting the mean protein abundance from each individual abundance and then dividing by the SD across all samples in that row. (
E
) Gene Ontology
biological processes enrichment analysis of proteins identified at least twice in the same tissue types. 5,297 proteins were detected more than once in the same
tissue type. From these, 3,501 were shared in both tissue types, with an overrepresentation of proteins involved in essential housekeeping processes such as
ribosome assembly. 1,060 were detected only in the gonad, with an overrepresentation of proteins involved in processes expected to be predominant in the
germ cells, such as cell cycle. 726 were detected only in the intestine, with an overrepresentation of proteins involved in processes expected to be predominant
in the intestinal cells, such as immune response, metal transport, and lipid metabolism. The Euler diagrams are area
-
proportional. Lists of proteins described
in this figure can be found in
Dataset S1
.
Downloaded from https://www.pnas.org by George Porter on June 12, 2024 from IP address 131.215.225.155.
4 of 11
https://doi.org/10.1073/pnas.2322588121
pnas.org
Fig. 3.
Single
-
tissue proteomic analysis. (
A
) A total of 4,012 proteins were identified from six samples of gonadal tissue, each isolated from a single animal, with
up to 3,316 (82.7%) identified in a single gonad sample. (
B
) A total of 3,732 proteins were identified from six samples of intestinal tissue each extracted from a
single animal, with up to 2,889 (77.4%) identified in a single intestinal tissue sample. (
C
and
D
) The single
-
tissue samples of the same type show a high degree
of similarity and were clearly distinctive from that of the other tissue type. (
C
) PCA of the single
-
tissue samples. Samples were divided into two distinct groups
separated on the PC1. (
D
) Heatmap of the single
-
tissue samples. The Z
-
score, representing the distance in SDs from the mean within each row, was computed
by subtracting the mean protein abundance from each individual abundance and then dividing by the SD across all samples in that row. (
E
) The vast majority
of the identified proteins were repeatedly identified in tissues isolated from different animals, with nearly half of them identified in every animal. (
F
) A high
-
confidence set of proteins present in each tissue, 2,368 in the gonad and 1,713 in the intestine, were identified based on consistent identification of the proteins
across different animals and experimental sets. Of these proteins, 450 and 418 were consistently enriched in the gonad and intestine, respectively. (
G
) Center:
A volcano plot showing the relative abundance of the proteins between the gonad and the intestine. Points representing proteins in the high
-
confidence sets
found to be enriched in the gonad (68) and intestine (143) are highlighted with the corresponding colors (green and purple, respectively), and some of the most
abundant proteins are labeled. The vitellogenins, which are highly abundant in both tissues, are also labeled in yellow. Flanking the volcano plot on either side
are lists of highly abundant proteins selected from those highly enriched (maxed out fold
-
change) in either the gonad (382) or the intestine (275) and cannot
be placed in the plot. Two intestinal
-
enriched proteins, LRO
-
1 (plot) and MRP
-
3 (side), are also labeled and will be discussed more in later sections. The values
used in this panel were based on one of the two experimental sets, but the highlighted (colored) proteins were those that were consistently enriched in the
respective tissue. Green: gonad; purple: intestine; Yellow: vitellogenins. The Euler diagrams are area
-
proportional. Lists of proteins described in this figure can
be found in
Dataset S2
.
Downloaded from https://www.pnas.org by George Porter on June 12, 2024 from IP address 131.215.225.155.