of 62
Neuronal Activation of the Gastrointestinal Tract
Shapes the Gut Environment in Mice
Bryan B. Yoo
1*
, Jessica A. Griffiths
1
, Peter Thuy-Boun
2+
, Victor Cantu
3+
, Kelly
Weldon
4,5,+
, Collin Challis
1
, Michael J. Sweredoski
1
, Ken Y. Chan
1#
, Taren M.
Thron
1
, Gil Sharon
1
, Annie Moradian
1%
, Gregory Humphrey
3
, Qiyun Zhu
3
, Justin
Shaffer
3
, Dennis W. Wolan
2
, Pieter C. Dorrestein
3,4,5
, Rob Knight
3,5,6,7
, Viviana
Gradinaru
1
, and Sarkis K. Mazmanian
1*
1
Division of Biology & Biological Engineering, California Institute of Technology, Pasadena, California,
USA, 91125
2
Departments of Molecular Medicine and Integrative Structural and Computational Biology, The Scripps
Research Institute, 10550 North Torrey Pines Road, La Jolla, CA 92037, USA
3
Department of Pediatrics, University of California San Diego, San Diego, CA, USA
4
Collaborative Mass Spectrometry Innovation Center, Skaggs School of Pharmacy and Pharmaceutical
Sciences, University of California San Diego, San Diego, CA, USA
5
UCSD Center for Microbiome Innovation, University of California San Diego, San Diego, CA, USA
6
Department of Computer Science and Engineering, University of California San Diego, San Diego, CA,
USA
7
Department of Engineering, University of California San Diego, San Diego, CA, USA
+
These authors contributed equally
#
Present Address: Stanley Center for Psychiatric Research, Broad Institute, Massachusetts Institute of
Technology, Cambridge, MA, 02142, USA
%
Present Address: Precision Biomarker Labs, Cedars-Sinai, 8700 Beverly Blvd. Davis 2904, Los
Angeles, CA 90048
*
Corresponding Author. Email: byoo@caltech.edu
(B.B.Y); sarkis@caltech.edu
(S.K.M)
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 15, 2021.
;
https://doi.org/10.1101/2021.04.12.439539
doi:
bioRxiv preprint
2
Abstract
The enteric nervous system (ENS) coordinates responses that shape the intestinal
milieu, and participates in bi-directional communication with various organs including the brain.
Tools to study, and knowledge of, ENS development and function lag behind brain research.
Herein, we deploy recombinant adeno-associat
ed viral (rAAV) vectors with enhanced tropism
for the gut to map and activate gut associated neurons in mice with spatial and temporal
resolution. rAAV-mediated fluorescent labelling coupled with whole-tissue clearing methods in
the small intestine and colon reveal novel neuronal architectures in the gut of mice. Viral
delivery of genetically-encoded calcium indicators enable
in vivo
measurement of ENS activity.
Further, we employed chemogentics to specifically activate gut neurons that express choline
acetyltransferase (ChAT
+
) or tyrosine hydroxylase (TH
+
). Targeted activation of ChAT
+
or TH
+
neuronal populations associated with the gastrointestinal (GI) tract alters the intestinal
transcriptome, mouse and microbial proteomes, metagenome, and fecal metabolome. We
reveal previously unknown and broad roles for enteric neurons in modulating GI physiology,
mucosal immunity, and microbiome structure, and provide evidence for novel functions by the
ENS such as regulating fungal colonization and shaping of bile acid profiles in the gut. These
advanced experimental tools and rich datasets with multi-parameter characterization of the gut
ecosystem may facilitate expanded under
standing and study of the ENS.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 15, 2021.
;
https://doi.org/10.1101/2021.04.12.439539
doi:
bioRxiv preprint
3
Introduction
The gastrointestinal (GI) tract consists of diverse cells (e.g., immune, epithelial, endothelial,
endocrine, muscle, neuronal, and even microbial) that coordinate physiology within the gut
(Furness, 2006) and throughout the body (Rao and Gershon, 2016). The mammalian gut
receives and transmits neuronal signals from over 100,000 extrinsic nerve fibers that originate
from the sympathetic and parasympathetic nervous systems (Grundy and Brookes, 2011). The
GI tract is also innervated by an expansive network of intrinsic neurons, collectively termed the
enteric nervous system (ENS), that has over 100 million neurons organized into distinct
compartments within the GI tract, namely the myenteric plexus and submucosal plexus
(Furness et al., 2014). In recent years, extrinsic and intrinsic innervation of the GI tract has been
implicated in various host responses, from digestion (Schneider et al., 2019), to immunity
(Gabanyi et al., 2016; Muller et al., 2014), to complex behaviors (Bravo et al., 2011) in mice.
Specific interactions between neurons of the GI tract and various other cell types highlight the
diverse and expansive roles of the ENS. For example, neural pathways in the gut regulate
nutrient sensation through intestinal enteroendocrine cells (Kaelberer et al., 2018), affect
thesepithelial barrier and mucosal immunity (Jarret et al., 2020; Seillet et al., 2020; Talbot et al.,
2020), and dynamically interface with the microbiome (Lai et al., 2019; Matheis et al., 2020).
While neurons in the central nervous system (CNS) interact with multiple cell types, exposure of
the ENS to changing external inputs such as from diet, microbiome, xenobiotics,
etc
. create
interactions distinct from those in the brain.
With the goal of exploring complex interactions in the GI tract, we employed and optimized
powerful experimental tools, originally developed for brain research, to map the architecture of
the ENS in mice and the functional consequence of neuronal activation of the GI tract. Adeno-
associated viruses (AAVs) have been characterized for their ability to efficiently deliver various
genes to eukaryotic cells (Kotterman and Schaffer, 2014), and have been used in research,
biotechnology, and medical settings (Samulski and Muzyczka, 2014). Employing an engineered
AAV capsid variant, AAV-PHP.S, that was selected for having enhanced tropism within the ENS
and other peripheral ganglia of mice (Chan et al., 2017), we achieved broad expression of
genetically encoded fluorescent reporters and neural activators in the gut. Further, we delivered
a fluorescent calcium reporter (GCaMP6F) into enteric neurons to measure ENS activity in live
animals. Adding to current knowledge gained through immunohistochemical characterization of
the ENS (Furness, 2012), we coupled tissue clearing technologies with AAV-mediated labelling
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 15, 2021.
;
https://doi.org/10.1101/2021.04.12.439539
doi:
bioRxiv preprint
4
of cells to present methodological improvements to traditional approaches that allowed broad
surveyance of ENS architectures within intact GI tissue and the ability to directly measure
in
vivo
neuronal activity in the gut.
Beyond characterization, we sought to investigate the effects of activating specific neuronal
subsets on the gut environment. Choline acetyltransferase (ChAT) and tyrosine hydroxylase
(TH) are the rate limiting enzymes in acetylcholine and catecholamine biosynthesis,
respectively, and are key chemical mediators of neurotransmission in the brain and the
periphery. However, the extent to which these neuronal populations impact complex GI
physiologies beyond motility and secretory functi
on remain less studied. Accordingly, we
transduced GI neurons of mice with chemogenetic modulators (Wess et al., 2013) in discrete
cell populations to achieve cell-specific control of ChAT and TH neurons. We discovered that
stimulating neural activity in the GI tract alters the transcriptional and proteomic landscape of
the intestines, with complementing alterations to the gut metagenome and metabolome. We
report diverse and expansive effects of directed neuronal activation on epithelial biology,
mucosal immunity, angiogenesis, and muscle physiology, among other outcomes. We observe
expansions and contractions of fungal and bacterial taxa. Of the many molecular changes, we
highlight that ENS activation has dramatic effects on the conversion of primary to secondary bile
acids, products of mouse and bacterial co-metabolism that have been widely implicated in
health and disease (de Aguiar Vallim et al., 2013; Fiorucci et al., 2018; Hofmann and Eckmann,
2006; Long et al., 2017; Sakai et al., 1980). The enhanced experimental tools and approaches,
and multi-parameter datasets, introduced in this Tools and Resource may help advance
technologies and understanding of neurogastroenterology and the complex environment of the
gut in mice.
Results
Targeted neuronal labeling enables extensive, high-resolution imaging of the mouse ENS
Various morphologic features of the ENS have been characterized (Furness, 2006). However,
resolving the broad architecture of neurons in the gut with current imaging methods, primarily
using immunohistochemistry (IHC), are limited by several physical obstacles of the GI tract:
large area (~40cm long in a mouse), gut motility / tissue pliability (i.e., with or without luminal
contents, relaxed vs. contracted smooth muscle), and 3-dimensional biogeography (i.e.,
longitudinal and cross-sectional analysis). Furthermore, the ENS is localized within two distinct
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 15, 2021.
;
https://doi.org/10.1101/2021.04.12.439539
doi:
bioRxiv preprint
5
plexuses of the GI tract and studies have often characterized myenteric neurons by
microdissection of the longitudinal muscle / myenteric plexus tissue layer (Fried and
Gulbransen, 2015), a manipulation of the tissue that disassociates axons from their destination
in 3-dimensional space and overlooks the organization of submucosal neurons.
To overcome these limitations, we labelled enteric neurons in living mice with recombinant
adeno-associated viruses (rAAVs) expressing fluorescent proteins, and used tissue clearing
techniques to enhance visualization of intact GI tissue
(https://github.com/mazmanianlab/Yoo_et_al_2021/tr
ee/main/ENS%20quantification). The AAV
capsid variant AAV-PHP.S was used since it has been optimized for systemic delivery in mice
(Deverman et al., 2016) and characterized for its increased tropism to the peripheral nervous
system, including the ENS (Chan et al., 2017). We first looked to compare the general infectivity
of AAV-PHP.S to the traditionally-used AAV9, wh
ich has been empolyed in the past for viral
delivery to the ENS (Benskey et al., 2015). Simila
r to Chan et al., the cargo for both AAVs was
the gene encoding a nuclear localized green fluorescent protein expressed under the strong,
ubiquitous CAG promoter. Following a single intravenous (
i.v
.) systemic injection of the AAVs,
AAV-PHP.S efficiently transduced cells throughout t
he GI tract and exhibited greater infectivity
of cells throughout layers of the GI tract, compared to AAV9, as seen by increased cellular
expression of nuclear localized CAG-GFP (Figure 1A). Gene products driven by the human
Synapsin 1 (hSYN1) promoter have previously been shown to restrict expression to neurons
(Jackson et al.,
2016; Kügler et al., 2003) and
minimize expression in per
ipheral targets such as
the dorsal root ganglia (Haenraets et al., 2017). We packaged fluorescent proteins genes under
the hSYN1 promoter into AAV-PHP.S, delivered them systemically, and found that ~90%
(±1.5%) of cells in the ENS labelled with antibodies against
P
rotein
G
ene
P
roduct 9.5 (PGP9.5),
a pan-neuronal protein, co-localized with virally labelled neurons in the small intestines (SI) and
colon (Figure 1B).
Efforts to study the circuitry and neurochemistry of the gut nervous system have been
hampered by the physical length of the GI tract, making broad characterization of the ENS by
IHC both technically c
hallenging and costly. With AAV-medi
ated labeling, a single systemic
injection of AAV-PHP.S:hSYN1-mNeonGreen at 10
12
viral genomes (vg) was sufficient to label
expansive regions of ENS as seen in the SI and colon Figures 1C, Figure 1 - Figure
Supplement 1A and 1B). This minimizes the technical variability inherent to IHC staining of
numerous tissue sections, and enables uniform imaging of millimeters of tissue at micron
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 15, 2021.
;
https://doi.org/10.1101/2021.04.12.439539
doi:
bioRxiv preprint
6
resolution across many samples. Pan tissue labeling revealed spatial diversity of the ENS, such
as architectural differences of ganglia proximal and distal to the mesentery (insets in Figure 1C),
helping orient images relative to each other and thus facilitating 3-dimensional tissue
reconstruction. Indeed, coupling viral-labeling with tissue clearing methods, such as a sorbitol-
based clearing method (ScaleS) (Hama et al., 2015; Treweek et al., 2015; Yang et al., 2014),
we were able to preserve the structure of intestinal tissue. Imaging intact tissue enabled a global
view of the ENS including extrinsic nerve processes in the gut (Figure 1 - Figure Supplement 1A
and 1B), how ENS architecture changes in association with Peyer’s patches (small intestinal
lymphoid tissues) (Figure 1 - Figure Supplement 1C and 1D), and revealed different patterns of
ENS ganglionic morphology when tissue is fixed with or without luminal contents (Figure 1 -
Figure Supplement 1E-1G).
Multicolor labeling to trace axons across vast tissue distances
When three separate viruses, each encoding a distinct fluorescent protein (PHP.S:hSYN1-
mTurquoise2, PHP.S:hSYN1-mNeonGreen,
or PHP.S:hSYN1-mRuby2 [together AAV-
PHP.S:hSYN1-XFP]), are delivered together at a combined dose of 10
12
vg, stochastic and
combinatorial viral transduction of cells resulted in multi-color labelling of neurons. Labeling of
neighboring cells with different colors demarcated cellular boundaries (Figure 1D) and
morphologies (Figure 1E) much more clearly than IHC, and allowed unambiguous assignment
of axons to specific cell bodies from which they emanate. This approach is also tunable by the
design of genetic promoters for different cell types, for example labeling of enteric glial cell
bodies with virally-expressed tdTomato driven by the glial fibrillary acidic protein (GFAP)
promoter (Figure 1F).
Lastly, passive CLARITY (PACT) tissue clearing (Hama et al., 2015; Treweek et al., 2015; Yang
et al., 2014) facilitated imaging of both myenteric and submucosal neurons within the same
image stack captured by confocal microscopy (Figure 1G). This method allows direct imaging of
the ENS in intact GI tissue via whole mount microscopy, eliminating the need to microdissect
distinct layers of the GI tract and prepare thin sections that full tissue reconstruction challenging,
and also reduces mechanically-introduced artifacts to tissue. Together, this cell labelling
approach provides an effective and economical method to label the ENS at high resolution, with
cell specificity, with spatial positioning throughout the tissue, and is widely deployable without
the need for genetically-engineered mouse strains.
While we demonstrate the cellular versatility
of this method, the remainder of this study focuses on investigating neurons in the gut.
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 15, 2021.
;
https://doi.org/10.1101/2021.04.12.439539
doi:
bioRxiv preprint
7
Characterizing and quantifying ENS architecture
The small intestine and colon display similar, divergent, and synergistic functions (Hansen,
2003). However, it remains poorly understood how the cellular architecture of the ENS
contributes to GI function, and broad, simultaneous characterization of the myenteric and
submucosal plexuses has remained elusive (Foong et al., 2014; Sang and Young, 1998; Steele
et al., 1991). Wild-type mice were transduc
ed with AAV-PHP.S:hSYN1-XFP via a single
i.v.
injection to broadly label enteric neurons in the gut (Figure 2A). Viral infection was uniform
throughout the SI and colon (Figure 2 - Figure Supplement 2A), outside of a small ~1.5cm
section of the medial colon that, for unknown reasons, was consistently not well transduced and
was therefore excluded from further analysis. Seven regions of the SI and two regions of the
colon were sampled proximally to distally. Tissue regions were approximately 1cm in length and
sampled every 2 to 3 cm. General ENS architecture was explored by quantifying the number of
neurons and ganglia, as well as the size of ganglia (i.e., the number of neurons in each
ganglion) in the myenteric and submucosal plexuses (Figures 2B-2G), normalized to the
number of crypts to control for inherent variabilities within tissues. Importantly, no differences
were found in crypt diameter along the GI tract (Figure 2 - Figure Supplement 2B). We saw that
in the SI, the number of neurons and ganglia generally increased toward the distal portions of
the myenteric plexus compared to proximal regions, while the converse was true for neurons
and ganglia in the submucosal plexus (i.e., lower numbers in distal relative to proximal regions)
(Figures 2B and 2C). Additionally, the size of the ganglia (i.e., the number of neurons per
ganglia) increased in the distal regions of the SI myenteric plexus, a feature not observed in the
submucosal plexus (Figure 2D). While neuronal numbers were similar in the proximal and distal
regions of the colonic neural plexuses (Figure 2E), the number of myenteric ganglia increased
(Figure 2F) while the size of each ganglion decreased in the distal colon (Figure 2G).
Interestingly, submucosal neurons in the proximal colon were localized to the natural folds the
tissue (Figure 2A, dashed lines in lower right panel), and these ganglia appear smaller in the
distal colon. We conclude that combining viral labelling with tissue clearing enables
exceptionally broad analysis of both myenteric and submucosal neurons throughout the gut,
uncovering biogeographic features of ENS architecture in mice.
Measuring neuronal activity in the ENS
Existing approaches to measure intracellular calcium flux, a surrogate for neuronal firing, in the
ENS include
ex vivo
preparations of intestinal tissue (Hennig et al., 2015) or surgically
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 15, 2021.
;
https://doi.org/10.1101/2021.04.12.439539
doi:
bioRxiv preprint
8
implanting an abdominal window to obtain
in vivo
ENS calcium recordings (Rakhilin et al., 2016;
2019). These techniques involve extensive manipulation of tissue to obtain access to the
longitudinal muscle and myenteric plexus layer, require physical separation of the ENS from its
extrinsic neural connectivity, and are technically challenging. We sought to develop a more
tractable method to directly assess the activity of specific neurons in the ENS of live mice.
A genetically-encoded calcium indicator, GCaMP6F (Chen et al., 2013), driven by the CAG
promoter was targeted to the ENS via systemic
viral delivery with AAV-PHP.S. Mice were
anesthetized, intestines removed from the peritoneum, and unstimulated calcium transients in
the proximal colon were recorded in individual neurons by
in vivo
confocal microscopy (Figure
2H and Video Supplement 1). Neurons in myenteric ganglia were determined by their size and
those that displayed robust changes in fluorescence intensity were chosen at random for
analysis. Measuring neuronal calcium flux over time revealed an average spike rate of 29.1
spikes/minute with an interquartile range of 19.0 - 37.7 spikes/minute (min.=7.0 spikes/min.,
max.=63.0 spikes/min) (Figure 2I). Time between calcium spikes was calculated for each cell,
and the average peak interval was 2.6 seconds with an interquartile range of 1.6 - 3.2 seconds
(min.=1.0 seconds, max.=8.6 seconds) (Figure 2J). Heatmaps (Figure 2K) and traces (Figure
2L) constructed from GCaMP6F fluorescence intensities further display the heterogeneity of
neuronal firing among and within cells (i.e., tonic vs. phasic, small vs. large fluorescence delta,
duration of spike). These data highlight the heterogeneity of ENS activity and underscores the
resolution of our method in achieving cell specific measurements of neuronal activity
(https://github.com/mazmanianlab/Yoo_et_al_2021/tree/main/gcamp6f).
Mapping neuronal subsets in the ENS
Subpopulations of neurons can be defined by the neurotransmitters they produce, and are often
demarcated by the rate limiting enzymes required for neurotransmitter production. Choline
acetyltransferase (ChAT) and tyrosine hydroxylase (TH) are the rate limiting enzymes in
acetylcholine and catecholamine biosynthesis, respectively. In the GI tract, acetylcholine is a
major excitatory neurotransmitter mainly studied for its role in motility and secretomotor function
(Furness, 2012; Mittal et al., 2017). Although ChAT
+
and TH
+
neurons are intrinsic to the GI
tract, they have yet to be systematically characterized and interrogated for their role(s) in GI
physiology. We employed ChAT-Cre and TH-Cre animals, where the Cre-recombinase (Cre) is
expressed under the control of each respective gene promoter. Technically, genetic constructs
can be engineered to have the payload transgene in a double-floxed inverse orientation (DIO),
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 15, 2021.
;
https://doi.org/10.1101/2021.04.12.439539
doi:
bioRxiv preprint
9
such that the transgene is flipped and expressed (and protein translated) in a Cre-dependent
manner; in other words, AAV’s del
iver transgenes that are expre
ssed in ChAT+ or TH+ neurons
only.
Accordingly, we transduced ChAT-Cre or TH-Cre mice with AAV-PHP.S:hSYN1-DIO-XFP and
cleared tissues using ScaleS to visualize these neuronal subpopulations. Remarkably, each
neuronal population occupied spatially distinct layers of the GI tract, with ChAT
+
neurons
primarily localized to the myenteric plexus whereas TH
+
neurons were more abundant in the
submucosal plexus (Figure 3A). Upon quantification, there are more ChAT
+
than TH
+
neurons in
all assayed regions of the SI (Figure 3 - Figure Supplement 3A), an increase largely due to
higher cell numbers in the myenteric plexus (Figure 3B). However, there were more TH
+
submucosal neurons in the small intestine compared to ChAT
+
cells (Figure 3C), and this spatial
disparity was accentuated when calculating the ratio of myenteric-to-submucosal neurons
(Figure 3 - Figure Supplement 3B). Furthermore, densities of myenteric neurons increased
distally in the SI of TH-Cre mice (Figure 3B, SI-1 vs SI-10/13/15, p < 0.01; SI-4 vs SI-10/13/15 p
< 0.05), though still lower than neuronal numbers in ChAT-Cre mice. In the colon, the ratio of
myenteric and submucosal neurons did not differ between ChAT
+
and TH
+
cells (Figure 3 -
Figure Supplement 3B). These data reveal that spatial localization of ChAT
+
and TH
+
neurons is
distinct. Additional subsets of neurons in the gut can be mapped by this approach.
Activation of ChAT
+
and TH
+
neurons alters the intestinal transcriptome
The unique spatial organization of ChAT
+
and TH
+
neurons suggests potentially distinct
functions that can be investigated through cell-specific activation of each neuronal
subpopulation. ChAT-Cre or TH-Cre mice were infected with Cre-dependent genetic constructs
encoding activating ‘
D
esigner
R
eceptors
E
xclusively
A
ctivated by
D
esigner
D
rugs’ or DREADDs
(AAV-PHP.S:hSYN1-DIO-hM3Dq-mRuby2). hM3Dq is
a modified neurotransmi
tter receptor that
is designed to induce neuronal activation when exposed to Compound 21 (C21), a “designer
drug” that activates only this receptor (Thompson et al., 2018). Accordingly, C21 induces a
calcium transient in a Cre-expressing neuron activated by DREADDs, as shown by the gradual
increase in GCaMP6F fluorescence in an intestinal explant (Figure 3D, Video Supplement 2),
validating the approach.
ChAT-Cre or TH-Cre mice were infected with activating DREADDs or a control virus expressing
a reporter protein instead (AAV-PHP.S:hSYN
1-DIO-mRuby2). C21 was administered
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 15, 2021.
;
https://doi.org/10.1101/2021.04.12.439539
doi:
bioRxiv preprint
10
intraperitoneally (
i.p
.), once daily for 10 consecutive days to all mice (Figure 3E). One hour
following the last C21 injection, 1cm of tissue from the distal SI and proximal colon was
harvested, and gene expression profiled by QuantSeq, a quantitative 3’ mRNA-sequencing
technology (https://github.com/mazmanianlab/Yoo_et_al_2021/tree/main/RNAseq). In neurons,
the rapid and transient expression of immediate early genes (IEGs) is widely used as a measure
of increased neural activity (Wu et al., 2017). Accordingly, the IEGs
Fos
,
Egr1
,
Jun
,
Klf2
(Figures 3F-3I)
were among the most significantly upregulated transcripts in the SI and colon of
both ChAT-Cre and TH-Cre mice harboring the activating DREADD (Figure 3J). Moreover, IEGs
identified here are known to be upregulated during growth and differentiation of highly active cell
types such as immune cells (Bahrami and Drabløs, 2016; Ramirez-Carrozzi et al., 2009),
smooth muscle cells (Miano et al., 1993), and intestinal epithelial cells (Flandez et al., 2008).
In the distal SI, DREADD activation produced similar numbers of differentially expressed genes
(DEGs; p
adj.
<0.05) in ChAT-Cre mice (162 DEGs) (Figure 3F) and TH-Cre mice (165 DEGs)
(Figure 3H). Interestingly however, ~73% of the DEGs were upregulated upon ChAT-Cre
activation (118 up, 44 down) and only ~42% of the DEGs were upregulated upon TH-Cre
activation (69 up, 96 down). This parallels the number of upregulated IEGs in the distal SI of
ChAT-Cre mice (29 IEGs) compared to those in TH-Cre mice (2 IEGs) (Figure 3J), with a few
IEGs (i.e.,
Hbegf, Soca3, Mcl1
) being downregulated in TH-Cre mice (Figure 3H). Similar
proportions of transcripts were upregulated in the proximal colon of both ChAT-Cre (253 DEGs-
169 up, 84 down) (Figure 3G) and TH-Cre mice (192 DEGs – 130 up, 62 down) (Figure 3I).
Gene Set Enrichment Analysis (GSEA) was empl
oyed to identify differ
ences in gene ontology
(GO) annotations between treatment groups (Figures 3K-N and Figure 3 - Table Supplement 1),
providing insight into cellular function. Notably, the most highly enriched GO term in the distal SI
of ChAT-Cre mice was “regulation of smooth muscle cell proliferation” (Figure 3K), whereas in
TH-Cre mice it was the “response to bacteria” (Figure 3M). In the proximal colon, we observed
more similarity in GO pathways between ChAT-Cre and TH-Cre mice following DREADD-
mediated activation (Figures 3L and 3N). These gene expression profiles support the implication
that spatial localization may drive distinct transcriptomic signatures depending on the neurons
activated (see Figures 3A-3C), i.e., ChAT
+
neurons in the myenteric plexus of the SI are
adjacent to muscle cells, while TH
+
neurons in the distal SI are enriched in the submucosal
plexus that neighbor epithelial and immune cells which respond to bacteria. Both neuronal
subsets are abundant in the myenteric plexus of the colon (see Figure 3B). Also, the pathways
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for this preprint
this version posted October 15, 2021.
;
https://doi.org/10.1101/2021.04.12.439539
doi:
bioRxiv preprint