ArticlePDF Available

Regulation of host gene expression by gastrointestinal tract microbiota in Chinook Salmon (Oncorhynchus tshawytscha)

June 2023
Molecular Ecology 32(15)

License
CC BY-NC 4.0

Authors:

University of British Columbia

Differences in gut microbiome composition are linked with health, disease and ultimately host fitness; however, the molecular mechanisms underlying that relationship are not well characterized. Here, we modified the fish gut microbiota using antibiotic and probiotic feed treatments to address the effect of host microbiome on gene expression patterns. Chinook salmon (Oncorhynchus tshawytscha) gut gene expression was evaluated using whole transcriptome sequencing (RNA-Seq) on hindgut mucosa samples from individuals treated with antibiotic, probiotic and control diets to determine differentially expressed (DE) host genes. Fifty DE host genes were selected for further characterization using nanofluidic qPCR chips. We used 16S rRNA gene metabarcoding to characterize the rearing water and host gut microbiome (bacterial) communities. Daily administration of antibiotics and probiotics resulted in significant changes in fish gut and aquatic microbiota as well as more than 100 DE genes in the antibiotic and probiotic treatment fish, relative to healthy controls. Normal microbiota depletion by antibiotics mostly led to downregulation of different aspects of immunity and upregulation of apoptotic process. In the probiotic treatment, genes related to post-translation modification and inflammatory responses were up-regulated relative to controls. Our qPCR results revealed significant effects of treatment (antibiotic and probiotic) on rabep2, aifm3, manf, prmt3 gene transcription. Moreover, we found significant associations between members of Lactobacillaceae and Bifidobacteriaceae with host gene expression patterns. Overall, our analysis showed that the microbiota had significant impacts on many host signalling pathways, specifically targeting immune, developmental and metabolic processes. Our characterization of some of the molecular mechanisms involved in microbiome-host interactions will help develop new strategies for preventing/ treating microbiome disruption-related diseases.

Content uploaded by Javad Sadeghi

Content may be subject to copyright.

Molecular Ecology. 2023;00:1–20.

1wileyonlinelibrary.com/journal/mec

Received: 1 Septem ber 2022

Revised: 3 May 2023

Accepted: 25 May 2023

DOI : 10.1111/me c.17039

ORIGINAL ARTICLE

Regulation of host gene expression by gastrointestinal tract

microbiota in Chinook Salmon (Oncorhynchus tshawytscha)

Javad Sadeghi1 | Subba Rao Chaganti2 | Daniel D. Heath1,3

This is an op en access arti cle under the ter ms of the Creative Commons Attribution-NonCommercial License , which permits use, dis tribu tion and reprod uction

in any medium, provided the original work is properl y cited an d is not use d for comm ercial purposes.

1Great La kes Inst itute for Environmental

Research, Universit y of Winds or, Windso r,

Ontario, Canada

2Cooperative In stitu te for Gre at Lakes

Research, Universit y of Michig an, Ann

Arbor, Michigan, USA

3Depar tment of Integrative Biology,

University of Windsor, Win dsor, Ontario,

Canada

Correspondence

Daniel D. Heath, Great Lakes Institute for

Environmental Research, Unive rsit y of

Windso r, Windsor, ON N9B 3P4, Cana da.

Email: dheath@uwindsor.ca

Funding information

Natural Sciences and Engineering

Research Council of Cana da; Ontario

Trillium Scholarship

Handling Editor: Nick Fountain- Jones

Abstract

Differences in gut microbiome composition are linked with health, disease and ulti-

mately host fitness; however, the molecular mechanisms underlying that relationship

are not well characterized. Here, we modified the fish gut microbiota using antibiotic

and probiotic feed treatments to address the effect of host microbiome on gene ex-

pression patterns. Chinook salmon (Oncorhynchus tshawytscha) gut gene expression

was evaluated using whole transcriptome sequencing (RNA- Seq) on hindgut mucosa

samples from individuals treated with antibiotic, probiotic and control diets to de-

termine differentially expressed (DE) host genes. Fifty DE host genes were selected

for further characterization using nanofluidic qPCR chips. We used 16S rRNA gene

metabarcoding to characterize the rearing water and host gut microbiome (bacterial)

communities. Daily administration of antibiotics and probiotics resulted in significant

changes in fish gut and aquatic microbiota as well as more than 100 DE genes in the

antibiotic and probiotic treatment fish, relative to healthy controls. Normal microbiota

depletion by antibiotics mostly led to downregulation of different aspects of immu-

nity and upregulation of apoptotic process. In the probiotic treatment, genes related

to post- translation modification and inflammatory responses were up- regulated rela-

tive to controls. Our qPCR results revealed significant effects of treatment (antibiotic

and probiotic) on rabep2, aifm3, manf, prmt3 gene transcription. Moreover, we found

significant associations between members of Lactobacillaceae and Bifidobacteriaceae

with host gene expression patterns. Overall, our analysis showed that the microbiota

had significant impacts on many host signalling pathways, specifically targeting im-

mune, developmental and metabolic processes. Our characterization of some of the

molecular mechanisms involved in microbiome- host interactions will help develop

new strategies for preventing/ treating microbiome disruption- related diseases.

KEYWORDS

antibiotics, fish microbiome, host– microbe interactions, probiotics, transcriptome

SADEGHI et al.

1 | INTRODUCTION

Nearly all animals examined to date show complex interac tions with

their associated microbial communities. It is evident that there are bi-

directional interactions between the gut microbiome and the host in

humans (Davison et al., 2017; Dayama et al., 2020; Meisel et al., 2018)

and non- human animals (Fuess et al., 2021; Muehlbauer et al., 2021;

Naya- Catala et al., 2021). These interactions affect a wide range of

host phenotypes including metabolism, immunity and physiology

(McFall- Ngai et al., 2013). Recent studies have shown that host ge-

netics can also shape their gut microbiome (Lopera- Maya et al., 2022;

Piazzon et al., 2020). The evidence for benefits provided by the gut

microbiota is growing, for example gut microbiota can improve nutri-

tion absorption (Krajmalnik- Brown et al., 2012), facilitate resistance

against pathogens (Ducarmon et al., 20 19), train the immune system

and even modif y behaviour and mental state (Surana & Kasper, 2017 ).

Moreover, the gut microbiota gain substantial benefits from their host

(e.g. available nutrients and suitable habitat) resulting in a mutualis-

tic relationship with the host. This provides the context for a unique

coevolved process in which host and their gut microbiome inter-

act in a mutualistic adaptive scenario (Escalas et al., 2021; Groussin

et al., 2020; Minich et al., 2022). Coevolution is defined as th e recipro-

cal adaptation process experienced by two organisms as the result of

their reciprocal selection pressures; it is possible for the microbiome

to evolve at the individual species level, as well as a community re-

sponse to host- mediated selection (Koskella & Bergelson, 2020).

Many studies have shown the importance of the gut microbiome

in healthy and diseased host states, which ultimately affects host fit-

ness (Bozzi et al., 2021; Manor et al., 2020; Yao et al., 2018). The gut

microbiome has been shown to alter host gene expression (Davison

et al., 2017; Nichols & Davenport, 2021), perha ps a mechanism for the

effect of the microbiome on the host. However, the mechanisms and

direction of these effects is still not clear since the evidence is largely

correlational. Does a change in microbiome composition cause changes

in hos t gene expression , and if so, which genes will be most impacted?

It is clearly important to characterize the mechanisms through which

the microbiome can cause changes in host gene expression.

Fish live in diverse aquatic environments, but they all harbour

complex and diverse microbiomes, and those microbial communi-

ties start developing when the eggs are laid (Llewellyn et al., 2014).

The bidirectional interaction bet ween the host gut and its associ-

ated microbes may arguably be better established in fish, relative

to terrestrial animals, as fish are in constant direct contact with the

aquatic environmental microbiome through their gut, gills and skin.

Moreover, given the long evolutionary history of fish as a group,

studying host– microbe co- evolution in fish may provide unique in-

sights into the host– microbe relationships in general (Montalban-

Arques et al., 2015). Charac terizing the mechanisms of how the gut

microbiot a and gene expression processes of the host interact in a

symbiotic manner will help explain the physiological processes that

maintain the balance among these intricate cross- kingdom interac-

tions and ultimately, help attempts to prevent dysbiosis (Nichols &

Davenport, 2021).

Most studies on host– microbiome interactions are correlative or

associative analys es wi thout clearly defi ned cause and effec t (Surana

& Kasper, 2017). To move beyond such studies, we must more di-

rectly address causation through perturbation experimental analyses

(Xia & Sun, 2017). Using probiotics and antibiotics to alter gut micro-

biome (bacterial communities) in healthy hosts can provide valuable

experimental insight into the mechanisms of host- microbiome inter-

actions. Antibiotics can be used for antibiotic- induced microbiome

depletion (AIMD): this leads to changes in the structure and function

of the gut microbial communities (Ferrer et al., 2017 ). Furthermore,

probioti cs ca n al so be us ed to al ter the gut mic ro biome in a controll ed

manner, as well as stimulate the host intes tinal immun e sy stem (Le e &

Bak, 2011). Experimental perturbations of the gut microbial commu-

nity with probiotic strains in human and animal disease treatment is

well documented (Azad et al., 2018). However, the effect of probiot-

ics in healthy individuals is not as well charac terized.

The direction and nature of host- gut microbiome interactions is still

an open question in the study of the microbiome, although it is likely bi-

directional and experimental analyses of the mechanisms behind these

interactions are needed. Here, our goal was to explore a broad range of

host gut tissue responses induced by the experimental manipulation of

the gut microbiome. We chose Chinook salmon (Oncorhynchus tshawyts-

cha) as our study organism as they are reared for commercial and con-

servation purposes and provide logistical advantages for a study such

as ours. Specifically, we used antibiotic, probiotic and control diet treat-

ments to manipulate the gut microbiome in families of Chinook salmon.

We used 16S rRNA metabarcoding of the gut bacterial community, cou-

pled with host gut tissue transcriptomics to (i) quantify treatment effects

on the host gut and the fish rearing water bacterial community composi-

tions, (ii) determine the response of the host gut tissue transcriptome to

the treatments and (iii) use gene transcriptional profiling Taqman™ qPCR

to characterize the host response to the treatment- altered gut micro-

biome. Given the long evolutionary history of the relationship between

fish and their microbiomes, we expect strong bidirectional effects, but

predicted that the effects of the microbiome on the host are more pro-

nounced. We specifically hypothesized that the host transcriptional

responses to each treatment could be attributed to the abundance of

specific bacterial taxa. The results obtained provide insight into the

co- evolved symbiotic relationship between host and its associated mi-

crobiome that may inform future studies exploring host- microbiome in-

teractions and evolution. Additionally, our work will help in better using

microbiome manipulation (probiotics, antibiotics) to improve health in

fishes and potentially in other animals, including humans.

2 | MATERIALS AND METHODS

2.1 | Study design

We domesticated Chinook salmon from Yellow Island Aquaculture

Ltd, an organic salmon farm on Quadra Island, BC, Canada to create

a nested breeding design with two sires crossed with one dam (2 × 1)

replicated six times. Eggs were fertilized in October 2019 and the

1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17039 by Cochrane Canada Provision, Wiley Online Library on [08/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

SADEGHI et al .

eggs were incubated in replicate cells of vertical stack incubation

trays. When the offspring reached the first feeding (March 2020)

offspring from replicated incubation tray cells were transferred to

200 L tanks with well water flow of 2 L per minute with continuous

aeration with a 16:8 h light– dark cycle. Fish were fed ~3% of their

body weight three times per day until October 24th, 2020. At that

time, 5 fish per family were moved to new 200 L tanks for a total

of 72 tanks (12 (families) * 2 (replicates) * 3 (treatments— see below)).

2.2 | Microbiome manipulation

We manipulated the gut microbiome of the fish in the tanks using con-

trol (untreated) feed, antibiotic treated feed and probiotic treated feed:

2.2.1 | Antibiotic treatment

Oxytetracycline (OTC) and chloramphenicol (CAP), t wo broad spec-

trum antibiotics, were selected for the trial. Twenty- four t anks (for

12 (families) * 2 (replicates)) were labelled as antibiotic and were

treated with OTC (83 mg/kg/day concentration) (Kokou et al., 2020;

Leal et al., 2019; Rosado et al., 2019) for 6 days. After 6 days, the

fish were switched to a combination of chloramphenicol (CAP)

(42 mg/kg/day) (Bilandzic et al., 2012) plus the OTC for four more

days, for a total of 10 days of antibiotic treatment. Fish were fed

three times a day at approximately 3% of their body weight .

2.2.2 | Probiotic treatment

Twenty- four tanks (for 12 (families) * 2 (replicates)) were labelled

as probiotic treatment and fed commercially available Jamieson

Probiotic Complex with 60 billion colony forming units (CFU)

(Jamieson Laboratories, Canada; Table S1). Specifically, the probiotic-

treated feed (three capsules per 100 gram of feed) was coated with

10 mL of sodium alginate (1%) and 10 mL of 0.5% calcium chloride

prior to mixing with the probiotic powder. Fish were fed three times

a day at approximately 3% of their body weight.

2.2.3 | Control

Twenty- four tanks (for 12 (families) * 2 (replicates)) were labelled as

control group and fish were fed with regular feed without probiotic

or antibiotic for 10 days. Fish were fed three times a day at approxi-

mately 3% of their body weight (Figure 1).

2.3 | Sampling

All fish were terminally sampled after the 10- day trial (November

3, 2020). The fish were not fed on the day of sampling. The final

mean mass of the fish was 23.3 g (±7.2 SE) across all families and

treatments (no treatment effect on fish body weight was detected).

Three fish were dip net ted from each tank and humanely euthanized

immediately in an overdose solution of clove oil (Toews et al., 2019).

Of the 72 tanks, four tanks (control) had 100% mor tality and those

replicates were excluded from the study, bringing the total num-

ber of samples to 204 fish (72 probiotic treated fish, 72 antibiotic

treated fish and 60 control fish) and 68 water samples (one per tank).

The sampled fish were immediately weighed and dissected, with the

entire GI tract placed in a 50 mL tube with 35 mL of a highly con-

centrated salt buffer (ammonium sulfate, 1 M sodium citrate, 0.5 M

EDTA, H2SO4 to bring the pH to 5.2) for preservation for later RNA

and DNA extraction. Additionally, 500 mL water samples were col-

lected from each of the tanks (N = 68) before sampling the fish and

filtered immediately using 0.22- micron pore size, 47 mm diameter

polycarbonate filters (Isopore™, Millipore, MA). All samples (tissue

in preservative and the filters) were stored at −20°C, until used for

DNA or RNA extraction.

2.4 | Lab analyses

The lab analyses consisted of three related but separate protocols

(Figure 1). The first was to assess the bacterial composition of the

fish gut and rearing water microbiomes using 16S rRNA metabar-

coding. The second was to determine the whole transcriptome re-

sponse to treatment by RNA- Seq of gut tissue from offspring from a

single family. The third analysis was designed to better characterize

the transc riptional profi le resp onse to the treatment s using nanoflu-

idic array qPCR analysis of 50 gene loci selected using the RNA- Seq

analysis.

2.5 | Bacterial DNA extraction and 16S rRNA gene

library preparation

DNA was extracted from fish hindgut content using a sucrose lysis

buffer solution method previously described (Shahraki et al., 2019)

and extracted DNA was subsequently stored at −20°C, until fur-

ther analysis. Additionally, the PCR conditions and 16S rRNA primer

sets (first and second PCR) were the same as those used in previ-

ously described methods (Sadeghi et al., 2021). Briefly, the V5

( 7 8 7 F - a c c t g c c t g c c g - A T T A G A T A C C C N G G T A G ) a n d V 6 ( 1 0 4 6 R -

a c g c c a c c g a g c - C G A C A G C C A T G C A N C A C C T ) v a r i a b l e r e g i o n s o f t h e

16S rR NA were selected fo r PC R amplificati on with a PCR cyc le pro-

gram of 95°C for 3 min followed by 28 cycles of 95°C for 30 s, 55°C

for 30 s and 72°C for 1 m, and a final step at 72°C for 7 m. A second

short- cycle PCR (7 cycles) using purified first PCR products ligated

the adaptor and barcode (10– 12 bp) sequences to the amplicons as

required for sample identification and sequencing. During the first

and second PCR, nine samples failed amplification and 263 samples

(195 gut samples, and 68 water samples) remained for the gel ex-

traction. For each 96 well PCR plate, one negative control consisting

SADEGHI et al.

of PCR mix (of first and second PCR) with ultra- pure water instead

of DNA template was included. The pooled purified PCR amplicon

mix (i.e. sequencing library) was sequenced on an ION S5 next-

generation sequencing system.

2.6 | 16S Metabarcode sequence data processing

The resulting FASTQ file was analysed using the Quantitative

Insights Into Microbial Ecolog y (QIIME2- 2020.11) platform (Bolyen

et al., 2019). The FASTQ sequence file was demultiplexed and the

DADA2 pipeline was used to denoise single- end sequences, derep-

licate and filter chimeras. This was followed by amplicon sequence

variant (ASV) picking using the removeBimeraDenovo function with

the “consensus” method, while default values were used for the

other parameters (Callahan et al., 2016). Taxonomic classification

was done through the feature- classifier plugin (Bokulich et al., 2018)

using the SILVA 138- 99 reference database (Quast et al., 2013). This

plugin supports taxonomic classification of features using the Naive

Bayes method. All ASVs were aligned with maff t (Katoh et al., 2002)

and used to construct a phylogeny with fasttree (Price et al., 2010).

A total of 8,820,568 sequences with 19,776 ASVs were obtained for

the 267 samples (195 gut samples, 68 water samples and four nega-

tive controls). The four negative controls had 1– 7 reads and were

excluded from the rest of the study. Using a taxon filter- table, ASVs

related to eukaryotes, mitochondria, chloroplasts (combined ~1%)

and unassigned (1%) were removed, resulting in a total of 8,655,659

(98%) sequences remaining. Furthermore, samples with low se-

quence depth (<3000 reads), low abundance taxa (<10 ASVs) and

ASVs that showed up in only one sample were removed. This de-

creased the total number of samples to 255 samples (189 gut sam-

ples, 66 water samples) with 8,217,478 sequences and 2888 ASVs.

The eight deleted samples were not related to specific treatment

type or family (antibiotic treatment (one water sample), probiotic

treatment (four gut samples, and one water sample) and control (two

gut samples)). Alpha diversity indices (Chao1) (a metric for species

richness) and Faith's phylogenetic diversit y (PD) (a metric that incor-

porates both species richness and species evenness while correcting

for phylogenetic distance) of bacterial communities were calculated

using the QIIME2 alpha diversity plugin. The ASV table was rarefied

to 30 0 0 reads per sam ple for the alpha di versity est imati on (rarefac-

tion curves plateaued at 3000 reads). Bray– Curtis and Jaccard dis-

similarity distance matrixes were calculated to estimate β- diversity.

2.7 | RNA extraction

RNA was extracted from host hindgut tissue using TRIzol® reagent

(Life Technologies, Mississauga, ON, CAT = 15,596,018) following

the manufacturer's protocol. RNA was dissolved in sterile water and

FIGURE 1 Experimental design and the number of samples collected for each experiment.

SADEGHI et al .

treated with TURBO™ DNase (Life Technologies, Mississauga, ON)

to remove genomic DNA contamination and preser ved at −80°C

until RNA sequencing or cDNA synthesis and qPCR were per formed

(see below).

2.8 | RNA sequencing and transcriptome assembly

A total of 18 gut tissue samples from one family, but from all three

treatments (6 fish per treatment), were used for transcriptome anal-

yses by RNAseq. Fish from one family were used to minimize dif-

ferences due to genetic variability among individuals. RNA quality

was assessed using the Eukaryotic RNA 600 0 Nano assay on a 2100

Bioanalyzer (A gilent , Mississauga, ON). All samples had an RIN >7

and a 28S:18S rRNA ratio >1.0. RNAseq libraries were prepared and

sequenced at the McGill University and Genome Quebec Innovation

Centre using the Illumina NovaSeq 6000 S4 PE100 protocol and

100- bp paired- end sequencing. To remove potentially contaminat-

ing rRNA sequences, raw sequences were filtered against eight de-

fault rRNA databases using SortMeRNA v2.1 (Kopylova et al., 2012).

The sequences were then quality- filtered using Trimmomatic v0.38

(Bolger et al., 2014). The non- rRNA sequences were aligned to the

Chinook salmon (GCF_002872995.1_Otsh_v1.0; https://www.ncbi.

nlm.nih.gov/assem bly/GCF_00287 2995.1/) reference genome using

the splicing aligner HISAT2 (Kim et al., 2015). FeatureCounts (Liao

et al., 2014), was used to calculate the number of transcript se-

quence fragments assigned to each gene.

2.9 | Differential expression gene analysis

The output from FeatureCounts was impor ted into DESeq2 (version

‘1.32.0’) (Love et al., 2 014) in R (R version 4.1.1) (R Core Team, 2013)

for normalization and differentially expressed genes analysis.

2.10 | qPCR primer/probe optimization and

cDNA synthesis

2.10.1 | Primer and probe optimization

Fifty transcripts (genes) that were significantly DE between antibi-

otic and probiotic treatments vs. the control treatment in the DESeq2

analysis were selected for printing on OpenArray Taqman qPCR

chips (Table S2). Four endogenous control genes (β- 2- microglobulin,

β- actin, ribosomal protein L13, and glyceraldehyde- 3- phosphate

dehydrogenase (GAPDH)) were selected from previous stud-

ies (Geffroy et al., 2021; Limbu et al., 2018; Toews et al., 2019) to

normalize the transcription profiles of the candidate transcripts.

Primers for the candidate transcripts were designed using Geneious

Software v7.1.5 (http://www.genei ous.com) and optimized on DNA

from Chinook salmon fry. After PCR optimization, the primers were

tested on a subset of our cDNA samples with SyBr® Green Dye I

(Thermo Fisher Scientific) following the manufacturer's protocol on

the QuantStudio 12 K Flex Real- Time PCR System (Thermo Fisher

Scientific). After testing positive for amplification of the expected

sized fragment using SyBr® Green assays, new qPCR primers and

Taqman® probes were developed using Primer Express® Sof tware

v3.0.1 ( Thermo Fisher Scientific) for all 54 genes (50 candidate and

4 control genes; Table S1). The qPCR primers spanned intron- exon

boundaries with a short amplicon size (50– 100 bp). The Taqman®

probe was designed for a melting temperature between 57 and

60°C.

2.10.2 | cDNA synthesis

RNA was quality tested on a random subset of the samples both

on a 2100 Bioanalyzer and on 2% agarose gels. RNA integrit y num-

ber (RIN) values were consistent among samples, ranging between

7 and 9.8, while gel images showed the expected rRNA bands. The

RNA concentration for each sample was est imated by Spa rk® multi-

mode microplate reader and NanoQuant Plate™ (Tecan, Morrisville,

NC, USA). All total RNA preparations had purity values of 1.8– 2.1

(A260/A280) with concentrations ranging from 200 0 to 5000 ng/μL.

TURBO DNA- free™ Kits (Thermo Fisher Scientific, cat. no. AM1907)

were used to remove genomic DNA contamination. Total RNA

was converted to cDNA using High Capacity cDNA Kits (Applied

Biosystems, Ontario, Canada), following the manufacturer's proto-

col. Reverse transcriptase reactions contained 10 μL of total RNA

at a concentration of 200 ng/μL, 2 μL of 10× RT random primers

(Applied Biosystems), 0.8 μL of dNTP (100 mM), 50 U of MultiScribe

RT (Applied Biosystems) and 40 U of RNase Inhibitor (Applied

Biosystems) in a 2 μL of 10× RT buffer at a final volume of 20 μL. RT

reactions were incubated at 25°C for 10 min followed by 37°C for

120 min and were stopped by incubating at 85°C for 5 min. cDNA

samples were stored at −20°C until further analysis.

2.11 | OpenArray high- throughput qPCR

TaqMan® OpenArray® chips from Applied Biosystems (Burlington,

ON, Canada) were used to quantify transcription at the 54 genes (50

candidate and 4 endogenous control genes) on a QuantStudio 12K

Flex Real- Time PCR System following the manufacturer's protocol.

Forty- eight cDNA samples were run (two chips for 48 samples) for

each of the 54 genes on each chip. A 5 μL reaction volume, which in-

cluded 1.2 μL of cDNA (100 ng/μL/per sample), 1.3 μL of ddH2O and

2.5 μL of TaqMan® OpenArray® Real- Time PCR Master Mix (Applied

Biosystems, Burlington, ON, Canada) was used, aliquoted across a

384- well plate and then loaded onto the TaqMan® OpenArray®

chips using the OpenArray® AccuFill System. A total of 10 chips were

used for 213 cDNA samples. The samples were randomly distributed

among the chips. ExpressionSuite Soft ware (Applied Biosystems,

Thermo Fisher Scientific, Carlsbad, CA , USA) was used to analyse

the endogenous control genes. Of four endogenous control genes,

SADEGHI et al.

β- actin was selected for normalization due to lower among- sample

variation compared to the three other endogenous control genes.

Sub se quently, all 10 chips we re normalized with the selec ted endog-

enous control gene (β- Actin) together in ExpressionSuite Software

v1.0.3 (Applied Biosystems, Burlington, Ontario, Canada). Moreover,

ExpressionSuite Software was used to calculate raw critical thresh-

old (CT) values and the rel at iv e cr it ic al threshold values (ΔCT). Values

produced by this platform are already corrected for the efficiency of

the amplification (Molina- Lopez et al., 2020). We tested for replicate

effect using paired sample T test in SPSS (IBM SPSS Statistics for

Windows, Version 27.0. Armonk, NY: IBM Corp). As we found no

ev id e nc e for a repl ica te ef f ect (p value >.0 5) , CT and ΔCT values were

averaged bet ween the replicate and only one CT or ΔCT value was

used for each gene.

2.12 | Statistical analysis

2.12.1 | Treatment effects on bacterial community

composition

Aquatic bacterial community composition

To test for the effect of treatment on the bacterial community com-

position in the hold tank water, taxonomical compositions of the

bacterial communities were visualized using stacked barplots and

pie charts of the relative abundance of the bac teria at the phylum

and family level using the online tool MicrobiomeAnalyst (Chong

et al., 2020) as well as R packages (“microbiome” and “phyloseq”).

Moreover, differences in alpha diversity indices (Chao1 and PD)

among the treatments (antibiotic, probiotic, control) for the tank

water bacterial communities were tested using a Kruskal– Wallis

(KW) rank test. In the case of a significant association, a post hoc

Dunn tests with Bonferroni corrected p values was done. To visu-

alize among- treatment divergence in the tank water bacterial com-

munities, a Principal- coordinate analysis (PCoA) using two measures

of community dissimilarity (Bray– Curtis, and Jaccard) were created.

Thus, the significance of the observed clusters was assessed using

permutational multivariate analysis of variance (PERMANOVA)

and permutational analysis of multivariate dispersions (PERMDISP)

in Primer 6 (v6.1.15) as well as QIIME2 (qiime diversity beta- group-

significance). Pair wise comparisons were performed in cases of sig-

nificant PERMANOVA among treatment groups.

Fish gut bacterial community composition

The effect of treatment on taxonomic composition of the gut sample

bacterial communities was visualized using pie chart s and stacked

barplot s of the relative abundance of the bacterial taxa at the fam-

ily and phylum level (Chong et al., 2020). To identify treatment and

parental effects on gut microbial community, alpha (Chao1 and

PD) diversity indices for gut samples were compared using the KW

rank test in SPSS (IBM SPSS Statistics for Windows, Version 27.0.

Armonk, NY: IBM Corp). To visualize treatment ef fects on bac terial

communit y structure, a PCoA using the Bray- Curtis distance matrix

was used to generate scatterplot of the first two PCoA axes in R

(version 4.1.1). Moreover, PERMDISP and PERMANOVA analyses

were performed in R (version 4.1.1) to test for treatment and pa-

rental (dams, sires) effects on bacterial community composition.

Additionally, QIIME2 was used for creating a PCoA plot as well as

PERMANOVA analysis using the Jaccard distance matrix. Pairwise

comparisons were performed when significant differences among

the treatment groups were detected to identify specific treatment

effects.

Comparison between fish gut and aquatic bacterial community

composition

Fish gut bacterial community composition was compared against the

rearing water bacterial communit y at both the alpha and beta diver-

sity level. Alpha diversity measures (Chao1 and PD) of gut and water

samples were compared using Mann– Whitney U test in SPSS (IBM

SPSS Statistics for Windows, Version 27.0. Armonk, NY: IBM Corp).

PCoA first and second axes were used to visualize clustering of the

samples based on sample type (gut or water) based on both Jaccard

and Bray- Curtis distance matrixes. Subsequently, PERMDISP and

PERMANOVA analyses were per formed in R (version 4.1.1) to test

sample t ype effect on bacterial community composition.

2.12.2 | Gut transcriptome response to treatment

The DESeq2 (version ‘1.32.0’) package in R (version 4.1.1) was used

to identify dif ferentially expressed transcripts in the host gut tran-

scriptome between any of the treatment groups in three pairwise

comparisons (antibiotic vs. control, probiotic vs. control, antibiotic

vs. probiotic). The package uses a Wald test to test the significance

of gene transcription differences. To identify differentially ex-

pressed transcripts, Benjamini– Hochberg corrections for multiple

testing was used (false discover y rate (FDA) < 0.05). We identified

differentially expressed transcripts as those genes with thresholds

of FDR < 0.05 and |log2 FC| > 1. Volcano plots of differentially ex-

pressed genes between the treatments were generated by using the

FC and the log- scaled adjusted p value using the EnhancedVolcano

package (Blighe et al., 2021) in R.

2.12.3 | Transcriptional profile (qPCR) response

to treatment

The 50 selected candidate transcripts (hereafter “genes”) were

tested to determine which genes showed a transcription response

to ei t her of th e tre atm e nts. Two ge n e s (cfap58, ubr4) were dropped

from the analysis due to failure of PCR amplification for most of

the samples, thus 48 candidate genes were included for the rest

of the study. To reduce the number of independent variables and

to avoid over fitting the models, we used principal component

analyses (PC A) on the qPCR data for the 48 selected genes using

“prcomp” (which is a part of the R statistical analysis package) and

SADEGHI et al .

factoextra package (1.0.7) (Kassambara & Mundt, 2017) in R (ver-

sion 4.1.1). Based on a threshold of Eigenvalue >1, and % variance

explained >2%, the first nine PC axes were selected. We used lin-

ear mixed models (LMM) (lmerTest package (v3.1.3)) (Kuznetsova

et al., 2017) in R with the selected PC axes to test for the effect of

treatment (fixed effect), and the random effects of dam, sire, fish

bod y weight, tank ID and chip ef fect, with all interaction te rms for

fixed and random factors on gene transcription patterns. Chip ID,

body weight, dam, treatment × dam, treatment × sire effects were

nonsignificant before FDR correction and were removed from the

model. When any of the nine PC s were found to exhibit sign if ic ant

effects with any of the independent variables (treatments, dam,

sire, bo dy we ig ht , t ank I D or chip ef fe ct), we examin ed the in divid-

ual gene transcription loading values. We used fviz_contrib within

the fa c to ex t ra pack ag e (1.0 .7 ) to identi fy ge ne s wit h co nt rib ut io ns

to the PC greater than expected (Kassambara & Mundt, 2017). The

id en tif ied gene s wer e incl ud e d in a sec on d anal ysis that use d LMM

with the ΔCT values for the selected genes and the same inde-

pendent variables (treatment, dam, sire, body weight, tank ID and

chip effect), including all interaction terms for fixed and random

factors. Nonsignificant factors (Chip ID, body weight, dam and all

interactions) were removed from the model and the analysis was

re- run. Lastly, a sequential Bonferroni p value correction was ap-

plied for multiple testing correction (Rice, 19 89).

2.12.4 | Correlation between gut bacterial

community and host transcriptional profile

To investigate the direct effect of variation in the gut microbi-

ome composition on host gene expression patterns, Spearman's

ra nk co r re lat ion co ef ficie nt (S pea r man's rh o) was per forme d u si ng

the function co r.tes t in R (R version 4.2.3). We selected common

bacterial taxa (bacteria families) with more than 5% contribution

to total sequence reads counts within each treatment; (7 t axo-

nomic families) and individual genes with evidence for possible

treatment effects (p value <.1 (nine genes)) from the gene- level

analysis described above. Moreover, a Holm- Bonferroni (sequen-

tial) p value correc tion was applied for multiple testing correction

(Rice, 1989). We visualized the pattern of correlation across all

genes and bacterial taxa using a heatmap generated in the pheat-

map function in R.

3 | RESULTS

3.1 | Impact of antibiotics and probiotics on

aquatic and fish microbiome

3.1.1 | Microbial community associated with water

We characterized the rearing tank water bacterial commu-

nities at two taxonomic levels; the phylum and family. Tank

bacterial community diversity diverged among the treatments,

with the top 10 most abundant families making up the majority of

reads. Proteobacteria were the most common phylum among all

treatments (control (70%), antibiotic (68%) and probiotic (51%)).

Bacteroidota (13%) and Actinobacteriota (17%) were also common

phylum in the control treatment water. Moreover, in the antibiotic

treated water, Firmicutes (24%) and Bacteroidota (12%) were com-

mon phyla after Proteobacteria. On the other hand, in the probiotic

tr e a ted wa ter, Bac ter o i dot a (12%) an d Firmi c ute s (8% ) wer e th e com-

mon phyla after Proteobacteria (Figure S1). At the family level, the

most common aquatic associated bacterial taxa were members of

Comamonadaceae, a family of the Betaproteobacteria accounting for

28%, 30% and 35% bacterial taxa in control, probiotic and antibiotic

waters, respectively. Mycoplasmataceae were found in all samples,

but at re latively hi gh er abu nd an ce in antib iotic challenge water com-

pared to probiotic and control waters. Members of Oxalobacteraceae

were also found in all sampled tanks but at higher abundance in the

probiotic and control tanks relative to the antibiotic tanks. Other

notable freshwater- associated bacterial taxa at the family level

were Flavobacteriaceae, Pseudomonadaceae, Sporichthyaceae and

Aeromonadaceae (Figure 2a).

To quantify treatment effects on the aquatic bacterial com-

munities, alpha and beta diversity indices for water samples were

compared for the three treatment groups (antibiotic, probiotic and

control). Alpha diversit y analysis (Chao1, PD) showed no signif-

icant differences among the groups (Chao1: KW 5, p > .05; PD:

KW 3, p > .05). However, our PCoA plot showed clear separation

of antibiotic treatment group from the other groups (Figure 2b).

PERMDISP (p value <.05) and PERMANOVA (F- value: 8.9; R-

squared: 0.22; p value <.001) results confirmed that the overall

community structures were significantly different among the

three groups. Pairwise comparison also showed that the three

groups are different from each other, but with the probiotic treat-

ment group compared to antibiotic treatment group showing the

highest dissimilarity (probiotic- control F: 2.17, p < .001; probiotic-

antibiotic F: 2.86, p < .001; control- antibiotic F: 2.77, p < .001).

Moreover, the average dissimilarity within treatments was higher

for the control tanks (73.2%) compared to our probiotic (65.4%)

and antibiotic treatment tanks (61.8%). Jaccard distance matrix

analyses showed similar results with significant differences among

the groups (Figure S2a, Tables S3 and S4).

3.1.2 | Microbial community associated with gut

Firmicutes were the most common phylum for the control (46%) and

probiotic (49%) gro up fish (Figure S3a). On the other hand, members

of Desulfobacterota were the most common bacteria in the antibi-

otic treated fish gut microbiomes (Figure S3a). We also compared

members of Firmicutes phylum among the treatments at the fam-

ily level. Within the Firmicutes phylum, Mycoplasmataceae was the

most common gut associated bacterial taxa across all treatments, in

addition to other important t axa (Figure S3b). For example, control

SADEGHI et al.

and probiotic treated fish had Mycoplasmataceae (control (65%), pro-

biotic (50%)), Streptococcaceae (control (30%), probiotic (28%)) and

Lactobacillaceae (control (2%), probiotic (17%)) present. However,

in the antibiotic group, different families were present within

Firmicutes phylum (Mycoplasmataceae (68%), Streptococcaceae (14%)

and Leuconostocaceae (5%)) (Figure S3b). At the family level, the most

common gut associated bacterial taxa across all treatment groups

were members of Desulfovibrionaceae (related to Desulfobacterota

phylum) and Mycoplasmataceae (Figure 3a). While Streptococcaceae

had high relative abundances in control group, samples in probiotic

groups had high relative abundances Lactobacillaceae. Moreover,

members of Pseudomonadaceae had high relative abundances in

antibiotic group (Figure 3a). Unlike in the tank water microbiome,

Mycoplasmataceae was higher in the control and probiotic groups

compared to the antibiotic group. At the genus level, we also found

two potential fish associated pathogen groups, Enterovibrio and

Photobacterium (from Vibrionaceae family), in the fish gut micro-

biome; however, they were at low abundance based on their read

count.

To id e nti f y th e tre a tme nt an d par e ntal (dam s and si res) ef fec t s on

the gut bacterial community, alpha diversity indices for gut samples

were compared. Alpha diversity analysis (Chao1, PD) for the gut mi-

crobiome showed no significant differences among the treatments

(Chao1: KW 2.8, p > .05; PD: KW 3.2, p > .05), sires (Chao1: KW 6.9,

FIGURE 2 (a) Relative abundance

(top 10) of water bacterial community

composition presented at the family level.

The ‘other’ taxa category includes the sum

of all bacterial families. (b) Scatterplot of

the first two axes from the PCoA of the

tank water bacterial community where

the treated fish were held. Treatment is

shown by colour with the 95% ellipses.

SADEGHI et al .

p > .05; PD: KW 6.8, p > .05) and dams (Chao1: KW 5.3, p > .05; PD:

KW 8.9, p > .05). Beta diversity variation was also explored using

Bray- Curtis distance matrices and a PCoA plot. The PCoA plot

showed weak separation among the samples based on treatments

(Figure 3b). PERMDISP and PERMANOVA results confirmed that the

overall bacterial community structures were significantly different

FIGURE 3 (a) Relative abundance

(top 10) of gut bacterial community

composition presented at the family level.

The ‘other’ taxa category includes the sum

of all bacterial families. (b) Scatterplot of

the first two axes from the PCoA of the

Chinook salmon gut bacterial community.

Treatment is shown by colour with the

95% ellipses.

Source df SS MS Pseudo- F p(perm)

Treatment 241,246 20,623 6.1 .001

Dams 523,505 4700 1.1 .22

Sires (dams) 6 24,2 59 4043 1.3 .06

Res 151 4.6 3 107. 6 – –

Tot a l 186 6.5 – – –

Note: The value under “p(perm)” is the significance probabilit y.

TABLE 1 Multivariate statistical testing

(PERMANOVA) of effects of treatment,

dams and sires (nested within dams) on

microbial communit y beta diversit y (Bray–

Curtis dissimilarity matrix).

SADEGHI et al.

among the treatments ( Table 1). Treatment alon e had the highest in-

fluence on the gut microbial communit y (PERMDISP: p value <.005;

PERMANOVA: Pseudo- F:6.1 , p value <.05). Pairwise comparisons

also showed that the three treatment groups exhibit significant dif-

ference in beta- diversit y, with the probiotic vs. control treatment

samples showing the highest dissimilarity (probiotic– control F: 3.01,

p < .001; probiotic– antibiotic F: 2.85, p < .001; control– antibiotic F:

1.52 , p < .05). Moreover, the average within treatment group bacte-

rial community dissimilarit y was higher for the control (82.2%) than

the probiotic (77%) and antibiotic treatments (80.5%), indicating that

the control group had higher diversity than the other two groups

in the fish hindgut. Dams alone did not have significant effects.

However, sires had marginal significant effect effects on bacterial

community structures (Table 1). Jaccard distance matrix analyses

also showed significant differences among the treatment groups for

the fish gut microbiome (Figure S2b, Tables S3 and S4).

3.1.3 | Association between gut and aquatic

microbial community

We evaluated the relationship between the tank water microbiome

and the fish gut microbiome. Chao1 and PD (diversity measures)

showed significant differences in the species richness of the two

sample types; overall, diversity was significantly higher in the water

samples than gut samples (p < .001, Mann– Whitney U test: 2191.5).

The PCoA plot (Figure 4) showed clear separation between the gut

and water samples. Moreover, PERMDISP and PERMANOVA test

also revealed that the clusters showed in PCoA plot were signifi-

cantly different (PERMDISP: p value <.01; PERMANOVA: Pseudo- F:

39. 6 , p value <.05). Additionally, Jaccard distance matrix analyses

revealed similar results with significant differences among the

treatment groups for fish gut microbiome composition (Figure S2b,

Tables S3 and S4).

3.2 | Treatment effects on the host gut

transcriptome

To determine if antibiotic and probiotic- induced changes in the mi-

crobio me led change s in the hos t gut transc riptome, RNA- Seq was

used to determine host transcript levels in the hindgut. Pairwise

treatment comparisons resulted in 96 (control vs. antibiotic; 35

control upregulated and 61 control downregulated), 105 (control

vs. probiotic; 61 control upregulated and 44 control downregu-

lated), 120 (antibiotic vs. probiotic; 84 antibiotic upregulated and

36 antibiotic downregulated) transcripts that were differentially

expressed among treatments (Benjamini- Hochberg false- discovery

rate (BH FDR) 0.1, |log2 FC| > 0.25). However, for selecting candi-

date genes fo r th e OpenA rr ay high- thro ug hp ut qPC R an al ys es , we

took a conservative approach and we only selected genes with

transcripts that were significantly expressed at |log2 FC| > 1 and

FDR p value <.05 (Figure S4). This decreased the differentially ex-

pressed transcripts to 29 (control vs. antibiotic), 29 (control vs.

probiotic) and 27 transcripts (antibiotic vs. probiotic) (Table 2). For

the control vs. antibiotic group comparisons, the selected genes

related to cellular process (e.g . cell ac ti vation , cell commu nic ation,

cell cycle and cell death) were upregulated and genes related to

metabolism and response to stimuli and stress were downregu-

lated in antibiotic group (Table 2). While in the control vs. probi-

otic group comparisons, genes related to regulation of a variety

of functions (regulation of meiosis, intracellular protein transport,

angiogenesis, transmembrane transporter, cell adhesion, negative

regulation of apoptotic process) were downregulated and genes

related to post- translation modifications were upregulated in the

probiotic treated fish (Table 2). Moreover, when we compared

antibiotic against probiotic group transcription, genes related to

cellular process (mostly apoptotic process) were up- regulated in

antibiotic group whil e gen es related to cell a dhe sion, regulatio n of

transcription were upregulated in probiotic group (Table 2).

FIGURE 4 Scatterplot of the first

two axes from the PCoA of the Chinook

salmon gut as well as water bacterial

communit y. Sample type is shown by

colour with the 95% ellipses.

SADEGHI et al .

3.3 | OpenArray high- throughput qRT- PCR

The LMM analysis showed PCs 4, 5, 6, 7 and 9 were significantly

affected by treatment (Table 3). We identified only those genes

whose contributions to the significantly affected principal com-

ponent axes were import ant (Figure S5) and selected them for

analyses. In our analysis, we also included tank, body weight and

OpenAr ray chip ID as random ef fects to correct fo r possible tech-

nical, environmental and body size effec ts. Chip and body weight

were not significant for any of the genes and were dropped from

our analyses. Sire effects (nested within dam) were not significant

after FDR correction. Moreover, a significant tank effect was ob-

served for only one gene (anxa1, p < .05) before FDR correction.

We found no significant effects for dam- by- treatment or sire- by-

treatment interactions. After including FDR correction into our

model, aifm3, manf and prmt3 still showed a significant treatment

effect (Table 4).

3.4 | Correlation between gut bacterial

community and host transcriptional profile

Spearman's rank correlation analysis was carried out to evaluate

the potential link between bacterial taxon abundance (at the fam-

ily level) for taxa common to the gut and differentially transcribed

ge ne s, wh il e cont rol lin g for tr eat men t and fam ily ef fec t. Th e abun -

dance of Lactobacillaceae, Bifidobacteriaceae and Aeromonadaceae

were negatively and positively correlated with several gene

transcription levels (Figure 5). However, after incorporating

Holm- Bonferroni p value correction, only Lactobacillaceae and

Bifidobacteriaceae was negatively correlated with manf and prmt3

genes (Figure 5, and Table S5).

4 | DISCUSSION

Interactions between fish hosts and their microbiomes have been

an under- studied area of research, perhaps due to the complexity

of the host- microbiome relationship making the detection of spe-

cific microbial features that impact the host phenotype challenging.

We approached this problem by manipulating gut microbiomes and

measured the impact on key candidate gene regulation— such ef-

fects are likely mechanisms for microbes to affect host phenotype

and health. We found that our treatments resulted in changes in host

gene expression patterns, and those changes were mostly related

to immune function and cell motility/integrity. By correcting for the

direct effects of the treatment, as well as the quantitative genetic

effects of family, we showed that changes in microbial communi-

ties do lead to changes in host physiology. Given the putative func-

tion of the responding genes, our work indicates a likely effect on

host fitness as well. Indeed, many recent studies have shown that

microbial symbionts are critical biological components for host traits

closely associated with fitness, such as immune system development

and function (Fuess et al., 2021; Langlois et al., 2021; Rosshart

et al., 2017).

Thi s is the first study to consid er and comp are the impac t of pro-

biotics and antibiotics administered to captive fish on the rearing

water microbial communities and we found that the aquatic micro-

bial communities in the rearing tanks were significantly influenced

by th e fe ed treat ment s. This was not exp ected as the fish foo d tr eat-

ment itsel f repre sented a sma ll prop or tion of the t ank volume , espe-

cially given the low flow through water effect. One possible factor

is that up 90% of administered antibiotics are excreted in the urine

and faeces of the fish, still in the active form (Polianciuc et al., 2020).

The common bacterial phyla we report in the tank water were also

reported in other studies that showed Proteobacteria, Bacteroidota

and Firmicutes are the dominant taxa in water where fish are held

(Chiarello et al., 2015; He et al., 2018; Stevick et al., 2019; Uren

Webster et al., 2018; Zhang et al., 2019). Nevertheless, we observed

significant treatment effects on the rearing water bacterial com-

munities, one possible explanation would be antibiotic- associated

diarrhoea leading to more fish gut- associated microbial excretion.

Another reason could be antibiotic- susceptible taxa being replaced

by taxa resistant to antimicrobial agents (e.g. Mycoplasmataceae

(Firmicutes)) (antibiotic (15%), control (1%), probiotic (3%)). Since

the aquatic microbiome itself plays a role in maintaining fish health

(Blancheton et al., 2013) quantifying the unexpected effects of feed-

based treatment on the rearing water is unexpected and important

as it may contribute to dysbiosis and poor health outcomes in the

fish. Although the negative effects of antibiotics on healthy fish have

been reported before, few studies have considered the effect of an-

tibiotic treatment on the rearing water microbiome. Furthermore,

our study showed that probiotic feed treatment also affec ted the

water microbiome. Previous studies showed that treating water

with probiotics can improve water quality (Elsabagh et al., 2018;

Tabassum et al., 2021).

The microbial communities present in fish rearing water are

thought to affect the initial colonization of the fish microbiota during

development (Llewellyn et al., 2014; Talwar et al., 2018). However,

similar to other studies (Uren Webster et al., 2018; Wu et al., 2018),

our fish gut microbiomes were distinct from the water sample mi-

crobiomes. This indicates that the fish host gut microbiome is likely

largely independent of the water microbial community and that

other factors such as diet and host genome may be contributing dis-

propor tionally (Talwar et al., 2018).

Our principal goal was to use probiotic and antibiotic treat-

ments to alter the Chinook salmon gut microbiome to determine

the potential role of gut microbiota composition variation in host-

microbiome interactions. However, we also assessed how the gut

microbial community reacted to the treatments. We found that,

while fish gut bacterial communit y alpha diversity was not af-

fect ed by the tr eat me nts , bet a divers it y was signi fic an tly dif fere nt

among all three treatments. Similar results were reported in other

studies, indicating community richness (alpha diversity) did not re-

spo nd to treatme nt w ith probi otics and ant ibiot ic s, but bet a di ver-

sity did (Hernandez- Perez et al., 2022; Kokou et al., 2020; Laursen

SADEGHI et al.

TABLE 2 Comparison of gene expression levels and differentially expressed gene distributions between the treatment groups (|log2 Fold

Change| > 1 and FDR p value <.05).

Genes Gene name Gene abbreviation Base mean log2 FD p value p (adj)

Control vs. antibiotic

LOC11 223 4113 Transmembrane protein 220- like tmem220 40.0 4.68 1.09E−0 8 .000

ubr4 Ubiquitin protein ligase E3 component

n- recognin 4

ubr4 14 9.1 −2 .75 5. 55E−07 .005

LOC112248188 Annexin A1- like anx a1 123.0 −5.56 6.34E−07 .005

LOC11224182 1 Kinase D- interacting substrate of 220 kDa

B - l i k e

kidins220 558.8 −1.1 2 1. 35E− 06 .0 07

LOC11224182 0 NTPase KAP family P- loop domain-

containing protein 1- like

nkpd1 422.9 −1 . 36 1.2 0E− 06 .007

LOC112255867 Neuronal acet ylcholine receptor subunit

alpha- 3- like

chrna3 88.5 1.10 1.73 E−0 6 .008

LOC112220969 Macrophage- stimulating protein

r e c e p t o r - l i k e

mst1r 4 7. 2 −3.38 2.10E−06 .008

LOC112248679 Golgin subfamily A member 7- like gol ga7 6.7 −6.86 1.97E−0 6 .008

LOC11 2245911 Vacuolar protein sorting- associated protein

33A

vps33a 5 7. 2 5.09 3.59E−06 .009

LOC112 21610 0 Cyclin- dependent kinase 11B cd k11b 32 .7 −4.1 3 .3 8E−05 .03

LOC112231356 Free fat ty acid receptor 2- like ffar2 21.3 −3.66 4.05E−06 .0 09

LOC112259970 Vacuolar protein- sorting- associated

protein 25- like

vps25 12.4 −3.52 3.82E−06 .009

LOC112258258 Septin- 8- A SEPTIN8 42.1 3.84 4.24E−06 .008

xrr a1 X- ray radiation resistance associated 1 xrra1 2201.2 1.07 6.86E−06 .011

LOC11 224 6816 HCLS1 binding protein 3 hs1bp3 160 .1 −4.37 6 .6 6E−06 .0 11

nsrp1 Nuclear speckle splicing regulatory

protein 1

ns rp1 21.1 −3.91 6 .33E−06 .0 11

LOC112239977aApoptosis inducing factor mitochondria

associated 3

aifm3 14.9 −5.0 0 6. 54E−06 .011

LOC112232613 Peroxisomal biogenesis factor 14 pe x14 1 7. 2 −4. 31 7. 21E − 0 6 .011

cfap58 Cilia and flagella associated protein 58 cfap58 302.4 1.73 1.04E− 05 . 014

LOC112218626 WAS protein family homologue 1 was h1 58.3 1.22 1 .11E− 05 .014

LOC11 2215983 Natriuretic peptide B nppb 34.2 1. 61 1. 06E−05 . 014

LOC112 261371 Echinoidin echinoidin 58.3 1.2 1 .18E− 05 .014

LOC11 2245791 Multidrug resistance- associated protein

7 - l i k e

mrp7 28.4 −4.01 9.57 E− 0 6 .0 14

LOC11 223 4 485 Uncharacterized uncharacterized 192.0 −4.76 1. 23E−05 .015

LOC112243336 SET domain containing 2, his tone lysine

methyltransferase

setd2 86.5 −1 . 2 9 1.37E−05 .016

LOC112251319 Phosphatase and actin regulator 1 phactr1 23.1 −3.97 2 .82E−0 5 .030

LOC112265673 Serine protease 16 prss16 41.2 3.69 4.34E−05 .043

LOC112249106 Piezo- type mechanosensitive ion channel

component 1

pi ezo1 41. 2 3.69 4.34E−05 .043

LOC112232636 ER degradation enhancer, mannosidase

alpha- like 2

edem2 2 9.8 4.52 4 .83E−05 .0 47

Control vs. probiotics

LOC11 2242158 Sorting nexin- 10A snx10b 7818 .8 1. 24 9.94 E−10 .000

LOC112232343 Heat shock factor- binding protein 1- like hs bp1 131.3 −6.57 1.91E− 08 .000

LOC11 2245911 Vacuolar protein sorting- associated

protein 33A

vps33a 5 7. 2 5.40 3.37E−07 .002

SADEGHI et al .

Genes Gene name Gene abbreviation Base mean log2 FD p value p (adj)

LOC112246837 COMM domain containing 10 com md10 31.4 −5. 88 3.12E−07 .002

LOC11224182 0 NTPase KAP family P- loop domain-

containing protein 1- like

nkpd1 422.9 −1 . 35 5 .06E−07 .002

LOC112220855 COP9 signalosome complex subunit 6 cops6 138.4 −5.74 5.03E−07 .002

LOC112218768 WEE2 oocyte meiosis inhibiting kinase wee2 353.9 1 .15 9.1 9E−07 .003

LOC112253929 Homeobox protein PKNOX1 pknox1 18.1 −5.02 8. 06E−07 .003

LOC11 223 4113 Transmembrane protein 220- like tmem220 40.0 3.60 1.73E− 06 .004

prmt3aProtein arginine methyltrans ferase 3 prmt3 138.8 −5.25 3.34E−06 .007

LOC1122 55055 L SM3 homologue, U6 small nuclear RNA

and mRNA degradation associated

lsm3 114. 8 −7. 2 7 .00001 .017

LOC11 226 4620 THO complex subunit 4 alyref 138.8 −5.25 .000003 .007

LOC112248608 B- cell linker protein- like blnk 37. 6 −4.01 3.57E−06 .007

LOC11 224 670 0 Cytochrome b- c1 complex subunit 6,

mitochondrial- like

uqcrh 97. 9 4.75 9. 3 5E −06 .015

LOC11 221960 6 Transmembrane protein 38B tmem38b 114.8 −7. 2 7 1.11E−05 .0 17

LOC112249934 S y n d e c a n - 1 sdc1 71.6 −3 .41 1.27E−05 .017

LOC112214559 Lysophospholipid acyltransferase l pcat4 42.1 3.46 1. 27E− 05 .0 17

LOC112217687 R - s p o n d i n - 3 - l i k e rspo3 90.9 1.38 2.33E− 05 .028

LOC11 226 4739 Proteasome subunit beta type- 4- like psmb4 97.9 −4.25 3 .2 9E−0 5 .034

LOC112 2615 03 Sodium- dependent multivitamin

transporter

slc5a6 291.4 1.80 3.4 4E− 05 .035

LOC11 226 34 81 CD151 antigen cd 151 18.5 4.07 3.89E−05 .037

LOC112232610 Phosphogluconate dehydrogenase pgd 155.3 −3 .97 4.62E−05 .040

rabep2aRab G TPase- binding effector protein 2 rabep2 46.8 −3 .05 4 .56E−05 .040

sidt2 SID1 transmembrane family, memb er 2 sidt2 19.7 4.66 4.79E−05 .0 40

LOC11 224 820 4 Low- density lipoprotein receptor- related

protein 2

lrp2 12.9 5.16 5.71E−05 .045

LOC11 2242147 Uncharacterized uncharacterized 291.3 1.8 3.44E−0 5 .034

LOC11 22 54714 Ras- related protein R ab- 34- like rab34b 558.7 −1. 0 3 2.59E− 06 .0 05

LOC11 223 4 485 Uncharacterized uncharacterized 191.9 −5. 59 7.1 1E− 0 8 .0007

LOC11224182 1 Kinase D- interacting substrate of 220 kDa

B - l i k e

kidins220 558.7 −1 . 0 3 2. 59E−06 .005

Antibiotic vs. probiotic

LOC112 21614 6 FERM domain cont aining 4Bb frmd4bb 24.1 3.93 3.97E−0 8 .001

LOC112232343 Heat shock factor- binding protein 1- like hs bp1 131.3 −6 .47 1. 22E−07 .0 02

LOC112220855 COP9 signalosome complex subunit 6 cops6 138.4 −6.27 2.0 5E−07 .002

LOC112248188 Annexin A1- like anx a1 123.0 5 .74 2 .91E−07 .002

LOC11 2265459 Arsenite methyltransferase- like as3mt 57. 5 −4.33 2.39E−07 .002

LOC112232636 ER degradation enhancer, mannosidase

alpha- like 2

edem2 2 9.8 −5.67 3 .23E−07 .002

LOC112218768 WEE2 oocyte meiosis inhibiting kinase wee2 353.9 1 .19 1.16 E−06 .004

LOC112232613 Peroxisomal biogenesis factor 14 pe x14 1 7. 2 4.82 1 .14E− 06 .004

LOC11 225288 3 Transient receptor potential cation

channel subfamily V member 5

trpv5 160 .4 −1 . 01 1 .43E−06 .005

LOC112222087aMesencephalic astrocyte- derived

neurotrophic factor

manf 43.9 −2. 96 1.98E−06 .006

LOC112246837 COMM domain containing 10 com md10 31.4 −5. 39 4.10 E−0 6 .010

TABLE 2 (Continued)

(Continues)

SADEGHI et al.

et al., 2017; Rasmussen et al., 2022). One possible reason for this

is that using antibiotics does not necessarily mean a reduced di-

versity of bacterial taxa. Indeed, a review showed that individuals

with dysbiosis (potentially caused by treatment) can have even

more diverse microbial community than healthy individuals (Berg

et al., 2020). For example, Rosado et al. (2 019) showed that treat-

ment of farmed seabass (Dicentrarchus labrax) with OTC caused a

decrease in core bacterial community diversity in the gill and an

increase in the skin. One reason that our probiotic treatment did

not change bacterial community alpha diversity may be that we

treated healthy fish. Previous studies in humans have shown that

probiotics in healthy patients (healthy state) does not greatly im-

pact the resident microbial populations (Eloe- Fadrosh et al., 2015;

Lahti et al., 2013). In general, external stimuli that af fect the

intestinal environment can drive a hierarchical series of microbi-

ome responses; resistance, resilience, redundancy or finally dys-

biosis−depending on if the disturbance overcomes the intestinal

microbial ecosystem (Lozupone et al., 2012; Moya & Ferrer, 2016;

Sommer et al., 2017). It appears that the microbial responses to

probiotics in our study is either resistance or resilience, as previ-

ous studies have shown that the bacterial communities tended to

be more resil ient to ex te rnal sti muli. On the oth er hand, treat me nt

with antibiotics tends to result in either of resilience, redundancy

or dysbiosis. Moreover, apart from the treatment effect, overall,

the f ish gut microbiome in our study was divided into two cluste rs

and none of our measured factors could explain the two clusters

(Figure 3a). This could be due to other unmeasured factors (e.g.

sex) that may be contributing to variation in the fish microbiome.

Genes Gene name Gene abbreviation Base mean log2 FD p value p (adj)

nsrp1 Nuclear speckle splicing regulatory

protein 1

ns rp1 21.1 4.00 4.35E−06 . 010

LOC112239977 Apoptosis inducing fac tor mitochondria

associated 3

aifm3 14.9 5.17 4.30E−06 .010

LOC112249580 Transcription factor 12 tcf12 40.7 −1. 1 5.9675 4E−06 .01

LOC11 2220311 Occludin a ocln 2096 .9 1.05 7.9 4 E−06 .016

LOC112231356 Free fatty acid receptor 2- like ffar2 21.3 3 .47 1.10 E−05 .020

LOC11 226 4739 Proteasome subunit beta type- 4- like psmb4 97.9 −4 .59 2 .06E−0 5 .033

LOC112237710 Dispanin subfamily A member 2b dspa2b 780.5 1.89 2 .69E−05 .035

LOC112262831 Polyubiquitin ub 687. 1 2.32 2 .65E−05 .035

LOC11 2245 658 Trafficking kinesin- binding protein 1 trak1 36.8 2.52 2.38 E−05 .035

LOC11 2215983 Natriuretic peptide B nppb 34.2 −1 . 5 4 2.60E−05 .035

LOC112218750 Protein mono- ADP- ribosyltransferase p ar p12 39. 3 3.13 2.94E−05 .036

LOC112225266 Interferon alpha/beta receptor 2 ifnar2 158.6 1.07 3.60E−05 .043

LOC1122 174 07 Protein PML pml 442.7 1.36 4.4 0E− 05 .048

an o7 Anoctamin 7 an o7 62.4 −3.31 4.93E− 05 .0 49

LOC112245 441 Uncharacterized uncharacterized 20.3 2.37 4.7 7E−05 .049

LOC112225425 T cell differentiation protein 2 mal2 155.6 1.05 6.50E−05 .049

aIndicate that these genes were also signific ant in our OpenA rray high- throughput qRT- PCR analysis.

TABLE 2 (Continued)

PCA axes

Type III sum of

squares df

Mean

square FSig.

PC1 64.65 232.32 1.75 0.17

PC2 5.19 22. 59 0.44 0.64

PC3 7.9 1 23.95 1.42 0.285

PC4 27.45 213.72 9.5 9 0.0002***a

PC5 31.93 215.96 14.4 4 1.038e−05***

PC6 12.64 26.32 5.99 0.0097**

PC7 12.19 26.09 4.44 0.01*

PC8 3.16 21.58 1.64 0.23

PC9 10.62 25.31 5.93 0.003**

aSignificant codes: .01 < p ≤ .05*, .001 < p ≤ .01**, p ≤ .001***.

TABLE 3 LMM model of PC1- 9

(Eigenvalue >1 and % variance explained

>2%) on the qPCR data for the 48

selected genes test for the effect of

treatment.

SADEGHI et al .

We predicted that the gut microbial community would respond

to the treatments through an increase in beneficial gut bacteria

(probiotic treatment) or through a decrease in the beneficial mi-

crobes with a related increase in the number of potential patho-

gens (antibiotic treatment). This was based on the expectation

that antibiotics can cause dysbiosis in the gut, resulting in elevated

levels of opportunistic pathogens (Dethlefsen & Relman, 2011;

Francino, 2015), while prebiotics and probiotics are expected to

increase the frequency of gut barrier- protecting bacteria such as

Lactobacillaceae and Bifidobacteriaceae (Xiao et al., 2014). In this

study, bacteria with potential probiotic properties (Lactobacillaceae,

Bifidobacteriaceae, Streptococcaceae) were higher in the probiotic

group compared to other treatment groups, as expected. On the

other hand, Pseudomonadaceae and Aeromonadaceae had higher

relative abundances in the antibiotic treated fish. Similar patterns

of response to probiotics and antibiotics in bacterial community

structure and composition have been reported by others (Falcinelli

et al., 2016; Kokou et al., 2020; Navarrete et al., 2008; Rutten

et al., 2015). For example, Kokou et al. (2020) showed that after

7 days of antibiotic treatment, the European seabass (Dicentrarchus

labrax) microbiome increased in Staphylococcus, Pseudomonas gen-

era (Proteobacteria). OTC treatment was reported to reduce gut

microbial diversity in Atlantic salmon, while enhancing possible

opportunistic pathogens belonging to Aeromonas spp. likely due

to eliminating competing microorganisms (Navarrete et al., 2008).

Moreover, Falcin elli et al. (2016) showed that Firmicutes, specifically

Lactobacillus genus, were significantly higher in probiotic treated

Zebrafish (Danio rerio) larvae relative to controls.

Studies in humans (Qin et al., 2010) and fishes (Boutin et al., 2014)

have reported that the gut microbiome varies subst antially at the in-

dividual and population level, and the transcriptome of the fish gut

appears to correlate with this variation (Franzosa et al., 2014; Qin

et al., 2010 ). Moreover, Thaiss et al. (2016) showed that treatment

with antibiotics will change the mouse gut microbiome and that the

microbiome in turn regulates fluctuations in the host transcriptome

and epigenome. In our study, we showed that our treatment altered

the gut microbiota, then we tested if these changes were associated

with changes in host gene expression. Specifically, we showed that

several genes related to cellular processes such as cell activation,

cell communication and cell death were upregulated after treat-

ment with antibiotics in the feed. Although previous studies have

shown a direct effect of antibiotic treatment on gene transcription

in humans (Ryu et al., 2017), antibiotic treatment had a limited ef-

fect on gene expression in germ- free mice (Morgun et al., 2015; Ruiz

et al., 2017 ), providing evidence that the microbiome mediates the

effects of orally administered antibiotics on the host. In this study,

we found that our antibiotic treatment resulted in the upregu-

lation of genes related to cell death. Moreover, bacteria from the

Genes

Probiotics

vs. control

Antibiotics

vs. control Treatment

Sire (nested

within dam)

Tank

(sire(dam))

uqcrh 0.9 0.14 0 .11 0.036* 0.32

sidt2 0.07 0.17 0.08 0.09 0.43

rabep2 0.05*a0.60 0.015* 0.78 0.45

pi ezo1 0.06 0.25 0.18 1.00 1.00

ffar2 0.71 0.14 0.09 0.83 0.89

trpv5 0. 52 0.88 0.62 0.33 0.98

aifm3 0.04* 0.89 0.002**b0.009** 0.99

ub 0.78 0.14 0.05* 0.62 1.00

dspa2b 0.4 0.20 0.06 0.07 1.00

pml 0.58 0.39 0.60 0.01* 0.99

nkpd1 0.27 0.33 0.47 0.04* 1.00

tmem38b 0.41 0.02* 0.07 0.13 0.98

pknox1 0.63 0.44 0.43 1.0 0 1.00

manf 0.0001*** 0.87 5.6e−06*** 1.00 0.14

ifitm3 0.44 0.35 0.25 0.17 0.49

ifnar2 0.4 0.80 0.7 7 0.02* 0.99

anx a1 0.57 0.31 0.19 0.13 0.02*

prmt3 0.0027** 0.16 0.001*** 0.37 1.00

Note: Body weight, dam, treatment × dam, treatment × sire effects were nonsignificant before FDR

correction and were removed from the model. Treatment was considered as fixed effects, with

body weight, dam and sire effect s as random effe cts. The dependent variable was log transformed

ΔCT.

aSignificant codes: .01 < p ≤ .05*, .001 < p ≤ .01**, p ≤ .001***.

bSignificant bold p value indicates significant af ter p value correction.

TABLE 4 Result s of the LMM analysis

for significance levels for treatment, dam,

sire (nested in dam) and tank (nested in

sire nested in dam) effects for each.

SADEGHI et al.

Firmicutes and Bacteroidetes phyla were reduced while members of

the Proteobacteria phylum increa sed. Za rrinpar et al. (2018) showed

a similar shift the bacterial community in the mouse cecal; however,

a cecal tr anscriptome ana ly si s sh ow ed that the cha nges in th e bac te -

rial community resulted in changes in the expression of genes related

to cellular growth and proliferation, as well as cell death and survival

pathways. This suggest s that colonic remodelling after treatment

with antibiotics is directly driving changes in the host transcriptome.

Additionally, in our antibiotic treatment group, we showed increased

transcription of the m rp7 (multidrug resistance- associated protein 7-

like) gene. Moreover, our qPCR analyses showed the upregulation of

aifm3 gene in antibiotic group. A study by Stoddard et al. (2019) in

zebrafish showed that after introducing antibiotics to fish, inflam-

matory gene transcription was downregulated and apoptotic genes

such as aifm3 were upregulated within 24 h.

Antibiotics are designed to pass the gut barrier and become

systemic; however, probiotics are live microorganisms that are not

able to pass the lumen barrier. Probiotics can directly modulate

host physiology by interacting with host cells (mostly immune cells),

and through indirect changes in microbiome composition (Langlois

et al., 2021). We showed tha t ge nes rel ated to pos t- tra ns lation mo di-

fications were over- expressed in the probiotic treatment group, rela-

tive to the control and antibiotic treatment groups. Previous studies

showed that probiotic diet supplements elicit a proinflammatory

response in fish (Nayak, 2010) and honeybees (Daisley et al., 2020),

which promotes more effective pathogen clearance and improved

disease resistance. In this study, we found that our treatment with

probiotics indeed changes the bacterial community composition with

increased numbers of potential probiotics taxa (Lactobacillaceae and

Bifidobacteriaceae). Moreover, our treatment with probiotics showed

fewer genes related to apoptosis process responding, relative to the

antibiotics group. However, this was not the case for the control

treatment, which was expected as the fish in control group were

healthy. Finally, we noticed that our probiotic treatment did change

the expression of several genes related to immune function as re-

ported in other studies (Petrof et al., 2004; Tomosada et al., 2013).

For example, Tomosada et al. (2013), showed that Bifidobacteria

strains can have immunoregulatory effect in the intestinal epithelial

cells by modulation the ubiquitin- editing enzyme. Moreover, similar

to this study, Willms et al. (2022) also showed that beneficial bac-

teria can promote intestinal angiogenesis in Zebrafish. The precise

mechanism of action of probiotics remains to be elucidated, espe-

cially in healthy states.

One approach to characterize the bidirectional interactions be-

tween the host and the microbiome is to perturb the gut and mea-

sure the response of the host (such as in AIMD studies). In this study,

we used antibiotics and probiotics to modify the microbial communi-

ties within the gut and measured host gene transcription responses

to those modifications. We explored this effect using correlation

between multiple common bacterial taxa and host gene transcrip-

tion. The results of that analysis were consistent with a microbiome-

mediated effect on the host. We found that specific microbial taxa

are af fecting the regulation of several host genes; for example, the

abundance of Lactobacillaceae and Bifidobacteriaceae were negatively

associated with the transcription of the prmt3 and manf host genes.

Previous work has shown that prmt3 gene as a post- translational

modification is involved in a number of cellular processes, such as

protein trafficking, signal transduction and transcriptional regulation

(Bedford & Richard, 2005; Choi et al., 2008). Moreover, the upreg-

ulation of manf gene can active innate immune cells and repairing

damaged tissue (Neves et al., 2016; Sereno et al., 2017). However,

further studies will be required to determine the specific association

of Lactobacillaceae with manf host gene.

The direction of interaction between fish gut and microbiome is

not clear, yet it is the basis of the co- evolution of the host with its

associated microbiomes. In this study, we experimentally modified

the fish gut microbiome and evaluated host gut tissue responses

to those perturbation using transcriptome analysis and transcrip-

tional profiling coupled with a controlled breeding design to control

for host genome variation. Short term (10 days) perturbation of the

juvenile Chinook salmon gut microbiome with antibiotics and pro-

biotics affected the microbiome composition and host gene expres-

sion patterns. This study achieved a number of important goals: (1)

characterized the effects of antibiotics and probiotics on the aquatic

bacterial community (2) characterized juvenile Chinook salmon

gut microbiome response to antibiotic and probiotic treatment (3)

FIGURE 5 Hierarchical clustering of the 7 core bacterial taxa

and association with gene expression. Columns correspond to the

7 core bacterial taxa; rows correspond to 9 selected differentially

expressed genes. Red and blue denote positive and negative

associations, respectively. The intensity of the colours represents

the degree of association between the genus abundance and

bacterial taxa are based on Spearman's Rank Correlation coefficient

rho. Stars in each square represent significant p values (adjusted).

SADEGHI et al .

characterized the host gut tissue transcriptional response to antibi-

otic and probiotic treatments. We showed that our treatments with

antibiotics and probiotics not only changed the Chinook salmon mi-

crobiome (composition), but we also observed significant changes at

the gene expre ssion level in the gut tis su e of the fis h. This study pro -

vides insight into a long- standing co- evolved symbiotic relationship

between fish gut tissue and its associated microbiome. Moreover,

understanding factors influencing the fish gut microbiome and its

influence on host health and fitness will help in better sustainable

growth for the aquaculture.

AUTHOR CONTRIBUTIONS

J.S, S.R.C and D.D.H. conceived and planned the experiments. J.S

carried out field work. J.S. carried out the wet laboratory sample

preparations and experiments. All authors contributed to selecting

the models and the computational framework for the data analysis.

J.S., S.R.C and D.D.H. contributed to the interpretation of the re-

sults. J.S. took the lead in performing the research, analysing data

and writing the manuscript. S.R .C and D.D.H. provided critical feed-

back and helped shape the research, analysis and manuscript.

ACKNOWLEDGMENTS

We thank the staff at Yellow Island Aquaculture (h t t p : / / y e l l o w i s l a

ndaqu acult ure.ca/), especially Dr. John Heath, Dr. Ann Heath, Jane

Drown and Earl Heath for their help during the breeding and sam-

pling. We especially thank Shelby Mackie and Jonathon Leblanc

(Environmental Genomics Facilit y, GLIER) for helping with DNA

robotic extraction, high throughput metbarcoding library sequenc-

ing and Open Array nanofluidic chip analyses. We thank Zahra S.

Taboun and Keta Patel for their help with RNA sequencing analy-

sis. This research received financial support from Canada's Natural

Sciences and Engineering Research Council (NSERC) to DDH and JS

received an Ontario Trillium Scholarship (OTS).

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest .

DATA AVAILAB ILITY STATE MEN T

The raw 16S rRNA gene sequencing data are available at the Sequence

Read Archive (SR A) of NCBI with PRJNA872508 BioProject acces-

sion number. The RNAseq data have been deposited in NCBI's Gene

Expression Omnibus (Edgar et al., 2002) and are accessible through

GEO Series accession number GSE211372 (https://www.ncbi.nlm.

nih.gov/geo/query/ acc.cgi?acc=GSE21 1372). For a complete list

of packages and code for microbiome analyses, see https://github.

c o m / j a v a d 3 0 / R e g u l a t i o n - o f - H o s t - G e n e - E x p r e s s i o n - b y - G a s t r o i n t e

s t i n a l - T r a c t - M i c r o b i o t a - i n - C h i n o o k - S a l m o n - O n c o r h y n c .

ORCID

Javad Sadeghi https://orcid.org/0000-0002-2002-462X

Subba Rao Chaganti https://orcid.org/0000-0001-9796-3986

Daniel D. Heath https://orcid.org/0000-0001-5762-3653

REFERENCES

Azad, M., Sarker, M., Li, T., & Yin, J. (2018). Probiotic species in the

modulation of gut microbiota: An over view. BioMed Research

International, 2018, 9478630.

Bedford, M. T., & Richard, S. (2005). Arginine methylation: An emerging

regulator of protein function. Molecular Cell, 18(3), 263– 272.

Berg, G., Rybakova, D., Fischer, D., Cernava, T., Verges, M. C., Charles, T.,

Chen, X., Cocolin, L., Eversole, K., Corral, G. H., Kazou, M., Kinkel,

L., Lange, L., Lima, N., Loy, A., Macklin, J. A., Maguin, E., Mauchline,

T., McClure, R., … Singh, B. K. (2020). Microbiome definition re-

visited: Old concepts and new challenges. Microbiome, 8, 103.

Bilandzic, N., Tankovic, S., Varenina, I., Kolanovic, B. S., & Smajlovic, M.

(2012). Chloramphenicol residues in muscle of rainbow trout fol-

lowing t wo different dose treatments. Bulletin of Environmental

Contamination and Toxicology, 89, 4 61– 466.

Blancheton, J., Attramadal, K., Michaud, L., d'Orbcastel, E. R., & Vadstein,

O. (2013). Insight into bac terial population in aquaculture systems

and its implication. Aquacultural Engineering, 53, 30– 39.

Blighe, K., Rana, S., & Lewis, M. (2021). EnhancedVolcano: Publication-

ready volcano plots with enhanced colouring and labeling. In:

1.12.0 Rpv, editor.

Bokulich, N. A., Kaehler, B. D., Rideout, J. R., Dillon, M., Bolyen, E.,

Knight, R., Huttley, G. A., & Gregory Cap oraso, J. (2018). Optimizing

taxonomic classification of marker- gene amplicon sequences with

QIIME 2's q2- feature- classifier plugin. Microbiome, 6, 90.

Bolger, A. M., Lohse, M., & Usadel, B. (2014). Trimmomatic: A flexible

trimmer for Illumina sequence data. Bioinformatics, 30, 2114– 2120 .

Bolyen, E., Rideout, J. R., Dillon, M. R., Bokulich, N. A., Abnet, C. C., Al-

Ghalith, G. A., Alexander, H., Alm, E. J., Arumugam, M., Asnicar,

F., Bai, Y., Bisanz, J. E., Bittinger, K ., Brejnrod, A., Brislawn, C. J.,

Brown, C. T., Callahan, B. J., Caraballo- Rodríguez, A. M., Chase,

J., … Caporaso, J. G. (2019). Reproducible, interactive, scalable

and extensible microbiome data science using QIIME 2. Nature

Biotechnology, 37, 852– 857.

Boutin, S., Sauvage, C., Bernatchez, L., Audet, C., & Derome, N. (2014).

Inter individual variations of the fish skin microbiota: Host genetics

basis of mutualism? PLoS One, 9, e102649.

Bozzi, D., Rasmussen, J. A., Caroe, C ., Sveier, H., Nordoy, K ., Gilber t, M .

T. P., & Limborg, M. T. (2021). Salmon gut microbiota correlates with

disease infection status: Potential for monitoring health in far med

animals. Animal Microbiome, 3, 30.

Callahan, B. J., McMurdie, P. J., Rosen, M. J., Han, A. W., Johnson, A. J.,

& Holmes, S. P. (2016). DADA2: High- resolution sample inference

from Illumina amplicon dat a. Nature Methods, 13, 581– 583.

Chiarello, M., Villeger, S., Bouvier, C ., Bettarel, Y., & Bouvier, T. (2015).

High diversity of skin- associated bacterial communities of marine

fishes is promoted by their high variability among body parts, indi-

viduals and species. FEMS Microbiology Ecology, 91, fiv061.

Choi, S., Jung, C. R., Kim, J. Y., & Im, D. S. (20 08). PRMT3 inhibits ubiq-

uitinat ion of rib osomal protein S2 and together forms an active en-

zyme complex. Biochimica et Biophysica Acta (BBA)- General Subjects,

1780 (9), 1062– 10 69.

Chong, J., Liu, P., Zhou, G., & Xia, J. (2020). Using MicrobiomeAnalyst for

comprehensive statistical, functional, and meta- analysis of microbi-

ome data. Nature Protocols, 15, 799– 821.

Daisley, B. A ., Pitek, A. P., Chmiel, J. A., Gibbons, S., Chernyshova, A . M.,

Al, K . F., Faragalla , K. M., Bur ton, J. P., Thompson, G. J., & Reid, G.

(2020). Lac tobacillus spp. attenuate antibiotic- induced immune and

microbiota dysregulation in honey bees. Communications Biology, 3,

534.

Davison, J. M., Lickwar, C. R., Song, L., Breton, G., Crawford, G. E., &

Rawls, J. F. (2017). Microbiota regulate intestinal epithelial gene

expression by suppressing the transcription fac tor hepatoc yte nu-

clear factor 4 alpha. Genome Research, 27, 1195– 1206 .

SADEGHI et al.

Dayama, G., Priya, S., Niccum, D. E., Khoruts, A., & Blekhman, R. (2020).

Interactions between the gut microbiome and host gene regulation

in cystic fibrosis. Genome Medicine, 12, 12.

Dethlefsen, L., & Relman, D. A. (2011). Incomplete recovery and individ-

ualized responses of the human distal gut microbiota to repeated

antibiotic perturbation. Proceedings of the National Academy of

Sciences of the United States of America, 108 (Suppl 1), 4554– 4561.

Ducarmon, Q. R., Zwittink, R. D., Hornung, B. V. H., van Schaik, W.,

Young, V. B., & Kuijper, E. J. (2019). Gut microbiota and coloniza-

tion resistance against bacterial enteric infection. Microbiolog y and

Molecular Biology Reviews, 83, e00007- 19.

Edgar, R ., Domr achev, M., & Lash , A . E. (2002). Gene expression omni-

bus: NCBI gene expression and hybridization array data repositor y.

Nucleic Acids Research, 30(1), 207– 210.

Eloe- Fadrosh, E. A., Brady, A., Crabtree, J., Drabek, E. F., Ma, B.,

Mahurkar, A., Ravel, J., Haverkamp, M., Fiorino, A. M., Botelho,

C., Andreyeva, I., Hibberd, P. L., & Fraser, C. M . (2015). Functional

dynamics of the gut microbiome in elderly people during probiotic

consumption. mBio, 6, e00231- 15.

Elsabagh, M., Mohamed, R., Moustafa, E. M., Hamza, A., Farrag, F.,

Decamp, O., Dawood, M. A. O., & Eltholt h, M. (2018). Assessing the

impact of bacillus strains mixture probiotic on water quality, growth

performance, blood profile and intestinal morpholog y of Nile tila-

pia, Oreochromis niloticus. Aquaculture Nutrition, 24, 1613– 1622.

Escalas, A ., Auguet, J.- C., Avouac, A., Seguin, R., Gradel, A ., Borrossi, L.,

& Villéger, S. (2021). Ecological specialization within a carnivorous

fish family is supported by a herbivorous microbiome shaped by

a combination of gut traits and specific diet. Frontiers in Marine

Science, 8, 91.

Falcinelli, S., Rodiles, A., Unniappan, S., Picchietti, S., Gioacchini, G.,

Merrif ield, D. L., & Carnevali, O. (2016). Probiotic treatment re-

duces appetite and glucose level in the zebrafish model. Scientific

Reports, 6, 18061.

Ferrer, M., Mendez- Garcia, C., Rojo, D., Barbas, C., & Moya, A . (2017).

Antibiotic use and microbiome function. Biochemical Pharmacology,

134, 114– 126.

Francino, M. P. (2015). Antibiotics and the human gut microbiome:

Dysbioses and accumulation of resist ances. Frontiers in Microbiology,

6, 1543.

Franzosa, E. A ., Morgan, X. C ., Segata, N., Waldron, L ., Reyes, J., Earl, A.

M., Giannoukos, G., Boylan, M. R., Ciulla, D., Gevers, D., Izard, J.,

Garret t, W. S., Chan, A. T., & Huttenhower, C . (2014). Relating the

metatranscriptome and metagenome of the human gut. Proceedings

of the National Academy of Sciences of the United States of America,

111, E2329– E2338.

Fuess, L. E., den Haan, S., Ling, F., Weber, J. N., Steinel, N. C., & Bolnick,

D. I. (2021). Immune gene expression covaries with gut microbiome

composition in stickleback. mBio, 12, e00145- 21.

Geffroy, B., Gesto, M., Clota, F., Aerts, J., Darias, M. J., Blanc, M. O.,

Ruelle, F., Allal, F., & Vandeputte, M. (2021). Parental selection for

growth and early- life low stocking densit y increase the female- to-

male ratio in European sea bass. Scientific Reports, 11, 13620.

Groussin, M., Mazel, F., & Alm, E. J. (2020). Co- evolution and co-

speciation of host- gut bacteria systems. Cell Host & Microbe, 28,

12– 2 2 .

He, X., Chaganti, S. R., & Heath, D. D. (2018). Population- specific re-

sponses to interspecific competition in the gut microbiota of two

Atlantic Salmon (Salmo salar) populations. Microbial Ecology, 75,

1 4 0 – 1 5 1 .

Hernandez- Perez, A., Zamora- Briseno, J. A., Soderhall, K., & Soderhall, I.

(2022). Gut microbiome alterations in the crustacean Pacifastacus

leniusculus exposed to environmental concentrations of antibiotics

and ef fect s on susceptibilit y to bacteria challenges. Developmental

and Comparative Immunology, 126, 104181.

Kassambara, A., & Mundt, F. (2017). Factoextra: Extract and visualize the

results of multivariate data analyses. R Package Version, 1, 3 3 7 – 3 5 4 .

Katoh, K., Misawa, K., Kuma, K., & Miyata, T. (2002). MAFF T: A novel

method for rapid multiple sequence alignment based on fast

Fourier transform. Nucleic Acids Research, 30, 3059– 3066.

Kim, D., Langmead, B., & Salzberg, S. L. (2015). HISAT: A fast spliced

aligner with low memory requirements. Nature Methods, 12,

3 5 7 – 3 6 0 .

Kokou, F., Sasson, G., Mizrahi, I., & Cnaani, A. (2020). Antibiotic effec t

and microbiome persistence vary along the European seabass gut.

Scientific Reports, 10, 10003.

Kopylova, E., Noe, L., & Touzet, H. (2012). SortMeRNA : Fast and ac-

curate filtering of ribosomal RNAs in metatranscriptomic dat a.

Bioinformatics, 28, 3211– 3217.

Koskella , B., & Berge lson, J. (2020). The study of host- microbiome (co)

evolution across levels of selection. Philosophical Transactions

of the Royal Society of London. Series B, Biological Sciences, 375,

20190604.

Krajmalnik- Brown, R., Ilhan, Z . E., Kang, D. W., & DiBaise, J. K. (2012).

Effects of gut microbe s on nutrient absorption and energy regula-

tion. Nutrition in Clinical Practice, 27, 201– 214.

Kuznetsova, A., Brockhoff, P. B., & Christensen, R. H. (2017). lmerTest

package: Tests in linear mixed effects models. Journal of Statistical

Software, 82, 1– 26.

Lahti, L., Salonen, A., Kekkonen, R. A., Salojarvi, J., Jalanka- Tuovinen, J.,

Palva, A ., Orešič, M., & de Vos, W. M. (2013). Associations between

the human intestinal microbiota, Lactobacillus rhamnosus GG and

serum lipids indicated by integrated analysis of high- throughput

profiling data. Pee rJ, 1, e32.

Langlois, L., Akhtar, N., Tam, K. C., Dixon , B., & Reid, G. (2021). Fishing

for the right probiotic: Host- microbe interactions at the interface

of effective aquaculture strategies. FEMS Microbiology Reviews, 45,

fuab030.

Laursen, M. F., Laursen, R. P., Larnkjaer, A ., Michaelsen, K. F., Bahl, M. I.,

& Licht, T. R. (2017). Administration of t wo probiotic strains during

earl y ch il dh ood doe s no t af fe ct the end ogenous gut mic ro bi ot a com-

position despite probiotic proliferation. BMC Microbiology, 17, 175.

Leal, J. F., Santos, E. B., & Esteves, V. I. (2019). Oxytetracycline in inten-

sive aquaculture: Water quality during and after its administration,

environmental fate, toxicity and bacterial resistance. Reviews in

Aquaculture, 11(4), 1176– 1194.

Lee, B. J., & Bak, Y. T. (2011). Irritable bowel syndrome, gut microbiota and

probiotics. Journal of Neurogastroenterology and Motility, 17, 252– 266.

Liao, Y., Smyth, G. K., & Shi, W. (2014). featureCounts: An efficient gen-

era l purpo se program for as sig ning seque nce rea ds to geno mic fea-

tures. Bioinformatics, 30, 923– 93 0.

Limbu, S. M., Zhou, L., Sun, S. X., Zhang, M. L., & Du, Z. Y. (2018). Chronic

exposure to low environmental concentrations and legal aquacul-

ture doses of antibiotics cause systemic adverse effect s in Nile

tilapia and provoke differential human health risk. Environment

International, 115, 205– 219.

Llewellyn, M. S., Boutin, S., Hoseinifar, S. H., & Derome, N. (2014).

Teleost microbiomes: The state of the art in their characterization,

manipulation and importance in aquaculture and fisheries. Frontiers

in Microbiology, 5, 207.

Lopera- Maya, E. A., Kurilshikov, A., van der Graaf, A., Hu, S., Andreu-

Sanchez, S., Chen, L., Vila, A. V., Gacesa, R., Sinha, T., Collij, V.,

Klaassen, M. A . Y., Bolte, L. A., Gois, M. F. B., Neerincx , P. B. T.,

Swertz, M. A., LifeLines Cohort Study, Harmsen, H. J. M., Wijmenga,

C., Fu, J., … Sanna, S. (2022). Effect of host genetics on the gut mi-

crobiome in 7,738 participants of the Dutch microbiome project.

Nature Genetics, 54, 143– 151.

Love, M. I., Huber, W., & Anders, S. (2014). Moderated estimation of fold

change and dispersion for RNA- seq data with DESeq2. Genome

Biology, 15, 550.

Lozupone, C. A., Stombaugh, J. I., Gordon, J. I., Jansson, J. K., & Knight, R.

(2012). Diversity, stability and resilience of the human gut microbi-

ota. Nature, 489, 220– 230.

SADEGHI et al .

Manor, O., Dai, C . L., Kornilov, S. A ., Smith, B., Price, N. D., Lovejoy, J. C.,

Gibbons, S . M., & Magis, A. T. (2020). Health and disease markers

correlate with gut microbiome composition across thousands of

people. Nature Communications, 11, 5206.

McFall- Ngai, M., Hadfield, M. G ., Bosch, T. C., Carey, H. V., Domazet-

Loso, T., Douglas, A. E., Dubilier, N ., Eberl, G., Fukami, T., Gilbert,

S. F., Hent schel, U., King, N., K jelleb erg, S., Knoll, A. H., Kremer,

N., Mazmanian, S. K., Metcalf, J. L., Nealson, K., Pierce, N. E., …

Werneg reen, J. J. (2013). Animals in a bacterial world, a new imper-

ative for the life sciences. Proceedings of the National Academy of

Sciences of the United States of America, 110, 3229– 3236.

Meisel, J. S., Sfy roera, G., Bar tow- McKenney, C., Gimblet, C., Bugayev,

J., Horwinski, J., Kim, B., Brestoff, J. R., Tyldsley, A. S., Zheng, Q.,

Hodkinson, B. P., Artis, D., & Grice, E. A. (2018). Commensal mi-

crobiota modulate gene expression in the skin. Microbiome, 6, 20.

Minich, J. J., Härer, A ., Vechinski, J., Frable, B. W., Skelton, Z. R .,

Kunselman, E., Shane, M. A ., Perry, D. S., Gonzalez, A., McDonald,

D., Kn ig ht, R. , Mi cha el , T. P., & Alle n, E. E. (2 022). Hos t biol ogy, ecol -

ogy and the environment influence microbial biomass and diversity

in 101 marine fish species. Nature Communications, 13(1), 6978.

Molina- Lopez, J., Ricalde, M. A. Q., Hernandez, B. V., Planells, A., Otero,

R., & Planells, E. (2020). Effect of 8- week of dietary micronutrient

supplementation on gene expression in elite handball athletes. PLoS

One, 15, e0232237.

Montalban- Arques, A., De Schry ver, P., Bossier, P., Gorkiewicz, G.,

Mulero, V., Gatlin, D. M., 3rd, & Galindo- Villegas, J. (2015). Selective

manipulation of the gut microbiota improves immune status in ver-

tebrates. Frontiers in Immunology, 6, 512.

Morgun, A., Dzutsev, A., Dong, X., Greer, R . L., Sexton, D. J., Ravel, J.,

Schuster, M., Hsiao, W., Matzinger, P., & Shulzhenko, N. (2015).

Uncovering effects of antibiotic s on the host and microbiota using

transkingdom gene networks. Gut, 64, 1732– 1743.

Moya, A., & Ferrer, M. (2016). Functional redundancy- induced stability

of gut microbiota subjected to disturbance. Trends in Microbiology,

24, 402– 413.

Muehlbauer, A. L., Richards, A. L., Alazizi, A., Burns, M. B., Gomez, A.,

Clayton, J. B., Petrzelkova, K., C ascardo, C., Resztak, J., Wen, X.,

Pique- Regi, R., Luca, F., & Blekhman, R. (2021). Interspecies vari-

ation in hominid gut microbiota controls host gene regulation. Cell

Reports, 37, 110057.

Navarrete, P., Mardones, P., Opazo, R., Espejo, R., & Romero, J. (2008).

Oxytetracycline treatment reduces bacterial diversity of intestinal micro-

biota of Atlantic salmon. Journal of Aquatic Animal Health, 20, 177– 183.

Naya- Catala, F., do Vale Pereira, G., Piazzon, M. C., Fernandes, A. M .,

Calduch- Giner, J. A., Sitja- Bobadilla, A., Conceição, L. E. C., &

Pérez- Sánchez, J. (2021). Cross- talk between intestinal microbiota

and host gene expression in Gilthead Sea bream (Sparus aurata) ju-

veniles: Insight s in fish feeds for increased circularity and resource

utilization. Frontiers in Physiology, 12, 748265.

Nayak, S. K. (2010). Probiotics and immunity: A f ish perspective. Fish &

Shellfish Immunology, 29, 2– 14.

Neves, J., Zhu, J., Sous a- Victor, P., Konjikusic, M., Riley, R., Chew, S., Qi,

Y., Jasper, H., & Lamba, D. A. (2016). Immune modulation by MANF

promotes tissue repair and regenerative success in the retina.

Science, 353, aaf3646.

Nichols, R. G., & Davenport, E. R. (2021). The relationship between

the gut microbiome and host gene expression: A review. Human

Genetics, 14 0, 747– 760.

Petrof, E. O., Kojima, K., Ropeleski, M. J., Musch , M. W., Tao, Y., De

Simone, C., & Chang, E. B. (20 04). Probiotic s inhibit nuclear fac tor-

kappaB and induce heat shock proteins in colonic epithelial cells

through proteasome inhibition. Gastroenterology, 127, 1474– 1487.

Piazzon, M. C., Naya- Catala, F., Perera, E., Palenzuela, O., Sitja- Bobadilla,

A., & Perez- Sanchez, J. (2020). Genetic selection for growth drives

differences in intestinal microbiota composition and parasite dis-

ease resistance in gilthead sea bream. Microbiome, 8, 168 .

Polianciuc, S. I ., Gur zau, A . E., K iss, B., Stefan, M. G., & Loghin, F. (2020).

Antibiotics in the environment: Causes and consequences. Medicine

& Pharmacy Reports, 93, 231– 240.

Price, M. N., Dehal, P. S., & Arkin, A. P. (2010). FastTree 2– approximately

maximum- likelihood trees for large alignments. PLoS One, 5, e9490.

Qin, J., Li, R., Raes, J., Arumugam, M., Burgdorf, K. S., Manichanh, C.,

Nielsen , T., Pons, N., Levenez, F., Yamada, T., Mende, D. R., Li, J.,

Xu, J., Li, S ., Li, D., Cao, J., Wang, B., Liang , H., Zheng, H., … Wang,

J. (2010). A human gut microbial gene catalogue established by

metagenomic sequencing. Nature, 464, 59– 65.

Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T., Yarza, P.,

Peplies, J., & Glöckner, F. O. (2013). The SILVA ribosomal RNA gene

database project: Improved data processing and web- based tools.

Nucleic Acids Research, 41, D590– D596.

R Core Team. (2013). R: A language and environment for statistical comput-

ing. R Foundation for Statistical Computing.

Rasmussen, J. A., Villumsen, K. R ., Ernst, M., Hansen, M., Forberg, T.,

Gopalakrishnan, S., Gilbert, M. T. P., Bojesen, A. M., Kristiansen,

K., & Limborg, M. T. (2022). A multi- omics approach unravels

metagenomic and metabolic alterations of a probiotic and synbi-

otic additive in rainbow trout (Oncorhynchus mykiss). Microbiome,

10(1), 1– 19.

Rice, W. R. (1989). Analyzing tables of statistical tests. Evolution, 43,

223– 225.

Rosado, D., Xavier, R., Severino, R., Tavares, F., Cable, J., & Perez- Losada,

M. (2019). Ef fect s of dis ea se , an ti biotic treatme nt and recove r y tr a-

jectory on the microbiome of farmed seabass (Dicentrarchus l abrax).

Scientific Reports, 9, 18946.

Rosshar t, S. P., Vassallo, B. G ., A ngelet ti, D., Hutchinson, D. S., Morgan,

A. P., Takeda, K., Hickman, H. D., McCulloch , J. A., Badger, J. H.,

Ajami, N. J., Trinchieri, G ., Pardo- Manuel de Villena, F., Yewdell, J.

W., & Rehermann, B. (2017). Wild mouse gut microbiota promotes

host fitness and improves disease resist ance. Cell, 171, 1015– 1028

e13.

Ruiz, V. E., Battaglia, T., Kurt z, Z. D., Bijnens, L., Ou, A., Engstran d, I.,

Zheng, X., Iizumi, T., Mullins, B. J., Müller, C . L., Cadwell, K.,

Bonneau, R., Perez- Pere z, G. I., & Blaser, M. J. (2017). A single early-

in- life macrolide course has lasting effects on murine microbial net-

work topology and immunity. Nature Communications, 8, 518.

Rutten, N. B., Gorissen, D. M., Eck, A ., Niers, L . E., Vlieger, A. M .,

Besseling- van der Vaart , I., Budding, A . E., Savelkoul, P. H., van der

Ent, C. K ., & Rijke rs , G. T. (2015). Lon g ter m development of gut mi-

crobiota composition in atopic children: Impact of probiotics. PLoS

One, 10, e0137681.

Ryu, A. H., Eckalbar, W. L., Kreimer, A., Yosef, N., & Ahituv, N. (2017). Use

antibiotics in cell culture with caution: Genome- wide identification

of antibiotic- induced changes in gene expression and regulation.

Scientific Reports, 7, 7533.

Sadeghi, J., Chaganti, S. R., Shahraki, A. H., & Heath, D. D. (2021).

Microbial community and abiotic effects on aquatic bacterial

communities in north temper ate lakes. The Science of the Total

Environment, 781, 146771.

Sereno, D., Muller, W. E. G., Bausen, M., Elkhooly, T. A., Markl, J. S., &

Wiens, M. (2017). An evolutionary p erspective on the role of mes-

encephalic astrocyte- derived neurotrophic factor (MANF ): At the

crossroads of poriferan innate immune and apoptotic pathways.

Biochemistry and Biophysics Reports, 11, 161– 173.

Shahraki, A. H., Chaganti, S. R., & Heath, D. (2019). Assessing high-

throughput environmental DNA extraction methods for meta-

barcode characterization of aquatic microbial communities. Journal

of Water and Health, 17, 37– 49.

Sommer, F., Anderson, J. M., Bharti, R ., Raes, J., & Rosenstiel, P. (2017).

The resilience of the intestinal microbiota influences health and dis-

ease. Nature Reviews. Microbiology, 15, 630– 638.

Stevick, R . J., Sohn, S., Modak, T. H., Nelson, D. R ., Rowley, D. C., Tammi,

K., Smolowit z, R., Markey Lundgren, K., Post, A. F., & Gómez- Chiarri,

SADEGHI et al.

M. (2019). Bacterial community dynamics in an oyster hatcher y in

response to probiotic treatment. Frontiers in Microbiology, 10, 1060.

Stoddard, M., Huang, C., Enyedi, B., & Niethammer, P. (2019). Live imag-

ing of leukocyte recruitment in a zebrafish model of chemical liver

inj ur y. Scientific Reports, 9, 28.

Surana, N. K., & Kasper, D. L . (2017). Moving beyond microbiome- wide

associations to causal microbe identification. Nature, 552, 244– 247.

Tabassum, T., Mahamud, A. S. U., Acharjee, T. K., Hassan, R., Snigdha, T.

A., Islam, T., Alam, R., Khoiam, U., Akter, F., Azad, R., Al Mahamud,

A., Uddin Ahmed, G., & Rahman, T. (2021). Probiotic supplemen-

tations improve growth, water quality, hematolog y, gut microbiota

and intes tinal morpholog y of Nile tilapia. Aquaculture Reports, 21,

100972.

Talwar, C., Nagar, S., Lal, R ., & Negi, R. K. (2018). Fish gut microbiome:

Current approaches and future perspectives. Indian Journal of

Microbiology, 58, 397– 414.

Thaiss, C. A., Levy, M., Korem, T., Dohnalova, L ., Shapiro, H., Jaitin, D. A.,

David, E., Winter, D. R., Gury- BenAri, M., Tatirovsky, E., Tuganbaev,

T., Federici, S., Zmora, N., Zeevi, D., Dori- Bachash, M., Pevsner-

Fischer, M., Kartvelishvily, E., Brandis, A., Harmelin, A., … Elinav,

E. (2016). Microbiota diurnal rhythmicity programs host transcrip-

tome oscillations. Cell, 167, 1495– 1510 e12.

Toews, S. D., Wellband, K. W., Dixon, B., & Heat h, D. D. (2019). Variation

in juvenile Chinook salmon (Oncorhynchus tshawytscha) transcrip-

tion profiles among and within eight population crosses from

British Columbia, Canada. Molecular Ecology, 28, 1890– 1903.

Tomosada, Y., Villena, J., Murata , K., Chiba, E., Shimazu, T., Aso,

H., Iwabuchi, N., Xiao, J. Z., Saito, T., & Kitazawa, H . (2013).

Immunoregulatory effect of bifidobacteria strains in porcine in-

testinal epithelial cells through modulation of ubiquitin- editing en-

zyme A 20 expression. PLoS One, 8, e59259.

Uren Webster, T. M., Consuegra, S., Hitchings, M., & Garcia de Leaniz,

C. (2018). Interpopulation variation in the Atlantic Salmon micro-

biome reflects environmental and genetic diversity. Applied and

Environmental Microbiology, 84, e00691- 18.

Willms, R. J., Jones, L. O., Hocking, J. C ., & Foley, E. (2022). A cell atlas

of microbe- responsive processes in the zebrafish intestine. Cell

Reports, 38(5), 110311 .

Wu, Z. B., Gatesoupe, F. J., Li, T. T., Wang, X. H., Zhang, Q. Q., Feng, D.

Y., Feng, Y. Q., Chen, H., & Li, A. H. (2018). Significant improvement

of intestinal microbiota of gibel carp (Carassius auratus gibelio)

after traditional Chinese medicine feeding. Journal of Applied

Microbiology, 124, 829– 841.

Xia, Y., & Sun, J. (2017). Hypothesis testing and statistical analysis of mi-

crobiome. Genes Diseases, 4, 138– 148.

Xiao, S., Fei, N., Pang, X., Shen, J., Wang, L., Zhang, B., Zhang, M.,

Zhang, X., Zhang, C., Li, M., Sun, L., Xue, Z., Wang, J., Feng, J., Yan,

F., Zhao, N., Liu, J., Long, W., & Zhao, L. (2014). A gut microbiota-

targeted dietary intervention for amelioration of chronic in-

flammation underlying metabolic syndrome. FEMS Microbiology

Ecology, 87, 357– 367.

Yao, Z., Yang, K., Huang, L., Huang, X ., Qiuqian, L., Wang, K., & Zhang,

D. (2018). Disease outbreak accompanies the dispersive structure

of shrimp gut bacterial community with a simple core microbiota.

AMB Express, 8, 120.

Zarrinpar, A., Chaix, A., Xu, Z. Z., Chang, M. W., Marotz, C. A., Saghatelian,

A., Knight, R., & Panda , S. (2018). A ntibiotic- induced microbiome

depletion alters metabolic homeostasis by affecting gut signaling

and colonic metabolism. Nature Communications, 9, 2872.

Zhang, Z., Li, D., Xu, W., Tang, R., & Li, L. (2019). Microbiome of Co-

cultured fish exhibit s host selection and niche differentiation at the

organ sc ale. Frontiers in Microbiology, 10, 2576.

SUPPORTING INFORMATION

Additional supporting information can be found online in the

Suppor ting Information section at the end of this article.

How to cite this article: Sadeghi, J., Chaganti, S. R., &

Heath, D. D. (2023). Regulation of host gene expression by

gastrointestinal tract microbiota in Chinook Salmon

(Oncorhynchus tshawytscha). Molecular Ecology, 00, 1–20.

https://doi .org /10.1111/me c.1703 9

Investigating the role of genetic variation in vgll3 and six6 in the domestication of gilthead seabream ( Sparus aurata Linnaeus) and European seabass ( Dicentrarchus labrax Linnaeus)

Article

Full-text available

Nov 2023

Gene function conservation is crucial in molecular ecology, especially for key traits like growth and maturation in teleost fish. The vgll3 and six6 genes are known to influence age‐at‐maturity in Atlantic salmon, but their impact on other fish species is poorly understood. Here, we investigated the association of vgll3 and six6 in the domestication of gilthead seabream and European seabass, both undergoing selective breeding for growth‐related traits in the Mediterranean. We analysed two different sets of samples using two different genotyping approaches. The first dataset comprised farmed and wild populations from Greece, genotyped for SNPs within the two genes (‘gene‐level genotyping’). The second dataset examined 300–600 k SNPs located in the chromosomes of the two genes, derived from a meta‐analysis of a Pool‐Seq experiment involving farmed and wild populations distributed widely across the Mediterranean (‘chromosome‐level genotyping’). The gene‐level analysis revealed a statistically significant allele frequency differences between farmed and wild populations on both genes in each species. This finding was partially supported by the chromosome‐level analysis, identifying highly differentiated regions may be involved in the domestication process at varying distances from the candidate genes. Noteworthy genomic features were found, such as a CpG island in gilthead seabream and novel candidate genes in European seabass, warranting further investigation. These findings support a putative role of vgll3 and six6 in the maturation and growth of gilthead seabream and European seabass, emphasizing the need for further research on their conserved function.

Host biology, ecology and the environment influence microbial biomass and diversity in 101 marine fish species

Article

Full-text available

Nov 2022

Fish are the most diverse and widely distributed vertebrates, yet little is known about the microbial ecology of fishes nor the biological and environmental factors that influence fish microbiota. To identify factors that explain microbial diversity patterns in a geographical subset of marine fish, we analyzed the microbiota (gill tissue, skin mucus, midgut digesta and hindgut digesta) from 101 species of Southern California marine fishes, spanning 22 orders, 55 families and 83 genera, representing ~25% of local marine fish diversity. We compare alpha, beta and gamma diversity while establishing a method to estimate microbial biomass associated with these host surfaces. We show that body site is the strongest driver of microbial diversity while microbial biomass and diversity is lowest in the gill of larger, pelagic fishes. Patterns of phylosymbiosis are observed across the gill, skin and hindgut. In a quantitative synthesis of vertebrate hindguts (569 species), we also show that mammals have the highest gamma diversity when controlling for host species number while fishes have the highest percent of unique microbial taxa. The composite dataset will be useful to vertebrate microbiota researchers and fish biologists interested in microbial ecology, with applications in aquaculture and fisheries management.

Effect of host genetics on the gut microbiome in 7,738 participants of the Dutch Microbiome Project

Article

Full-text available

Feb 2022
Nat Genet