Abstract

   Angiotensin-converting enzyme 2 (Ace2) is a well-studied enzyme with
   important physiological functions in mammals. However, its roles in
   non-mammalian animals remain largely unexplored. This study
   investigates the function of Ace2 in zebrafish using
   CRISPR/Cas9-generated ace2 mutants (−/−). The ace2^−/− mutants
   exhibited significant growth retardation with reduced body length and
   weight. High ace2 expression was found in the intestine, indicating its
   importance in gut function. Ace2 deficiency led to marked decreases in
   specific plasma and muscle amino acids, including cationic and neutral
   amino acids such as L-(+)-arginine, L-(+)-lysine, leucine, and valine,
   suggesting impaired amino acid absorption. The gut microbiota of
   ace2^−/− mutants showed reduced diversity and altered bacterial
   communities, with marked increases in expression of genes involved in
   bacterial carbohydrate metabolism and marked decreases in amino acid
   biosynthesis pathways. These findings are consistent with enhanced
   carbohydrate metabolism in the liver and reduced amino acids in muscle.
   Our study demonstrates the essential role of Ace2 in growth, amino acid
   metabolism, and gut microbiota balance in zebrafish, offering novel
   insights into its functions across vertebrate species.

   Subject terms: Development, Metabolism, Zebrafish
     __________________________________________________________________

   Ace2^−/− zebrafish show growth delay, smaller size, reduced intestinal
   amino acid absorption, altered plasma/muscle amino acid profiles, and
   gut microbiota dysbiosis, underscoring Ace2’s pivotal role in fish
   growth and metabolism.

Introduction

   Ace2 is an enzyme with diverse biological functions beyond its
   classical roles in mammals^[50]1,[51]2. It has been implicated in lipid
   metabolism, inflammation, and regulation of gut microbiota. These
   functions indicate that Ace2 plays important roles in various
   physiological and pathological processes, including metabolic disorders
   and immune regulation^[52]3–[53]6. However, its roles in non-mammalian
   animals remain largely unexplored.

   Recent studies have highlighted the importance of Ace2 in
   gastrointestinal health, including nutrient absorption, gut barrier
   integrity, and microbiota regulation, which are critical for
   maintaining metabolic and inflammatory balance^[54]7. Ace2 also plays a
   pivotal role in controlling gut dysbiosis and vascular permeability,
   factors associated with conditions such as diabetes and cardiovascular
   diseases^[55]8,[56]9. Furthermore, Ace2 deficiency has been linked to
   impaired intestinal barrier integrity and gastrointestinal dysfunction
   in animal models^[57]7,[58]10, as well as changes in gut microbiota and
   intestinal health in certain pathological conditions like COVID-19.

   Despite extensive research on Ace2 in mammalian systems, its roles in
   non-mammalian vertebrates, such as teleosts, remain largely unexplored.
   To assess the universality function of Ace2 across vertebrates, we
   conducted this study by using zebrafish as a model, which exists a
   homolog of Ace2. Zebrafish are widely used in biological research due
   to their genetic manipulability and the applicability of findings to
   other vertebrates^[59]11–[60]16. Fish exhibit higher protein
   utilization and lower carbohydrate utilization, efficiently using
   proteins as an energy source while limiting
   carbohydrates^[61]17,[62]18. This trait is evident in the gut
   microbiota, which regulates the host’s energy metabolism. The gut
   microbiota composition in fish is closely related to protein-optimized
   metabolic pathways, essential for nutritional balance and energy
   efficiency^[63]19,[64]20. Our study aims to understand the function of
   Ace2 and how Ace2 influences metabolic processes and intestinal
   function in zebrafish, providing insights into its role beyond
   mammalian systems.

   In our study, we analyzed the spatiotemporal expression patterns of the
   ace2 gene across various tissues in zebrafish. We then created a mutant
   zebrafish line for ace2 using CRISPR/Cas9 technology to explore the
   roles of Ace2 in growth, gut microbiota, and metabolism regulation. The
   study findings revealed that compared to heterozygous controls (+/−),
   ace2^−/− mutants exhibited significant growth delays, including
   reductions in body length and weight. Furthermore, ace2 gene knockout
   significantly impacted gut microbiota diversity in the mutants, with
   marked reductions in the diversity indices and significant changes in
   the composition of the microbiota. These results suggest that Ace2
   plays a crucial role in the growth and development of zebrafish and may
   influence the host’s overall metabolic state by regulating the
   diversity and composition of gut microbiota. Therefore, this research
   sheds light on the functional roles of Ace2/ace2 in zebrafish and
   offers new perspectives on its role in vertebrate physiology and the
   interactions among host growth, gut microbiota, and metabolism.

Results

Ace2 deficiency led to growth retardation in zebrafish

   To explore the roles of Ace2 in zebrafish, we first analyzed the tissue
   distribution of ace2. The expression levels of ace2 were relatively
   high in the intestine and liver compared to other tissues
   (Fig. [65]1a), suggesting its important role in gut and liver. To
   further verify the functionality of Ace2 in zebrafish, we generated an
   ace2 null mutant fish line with CRISPR/Cas9 technology. An sgRNA
   targeting exon 1 of ace2 was used to generate mutations; we selected a
   7-base pair deletion [ace2(-7)] as it is predicted to induce frameshift
   mutation that disrupts Ace2 protein synthesis (Fig. [66]1b and
   Supplementary Fig. [67]1a). In order to determine whether this allele
   affected gene expression, qPCR was performed using three different
   primer pairs targeting separate regions of ace2 mRNA. The results
   showed that ace2 expression was clearly reduced in the mutants compared
   to controls (Supplementary Fig. [68]1b), supporting the successful
   knockout.

Fig. 1. Disruption of Ace2 in zebrafish lead to growth retardation.

   [69]Fig. 1
   [70]Open in a new tab

   a Quantitative real‑time PCR analysis of ace2 mRNA expression in
   different zebrafish tissue (n = 3/each group), normalized to ef1a as
   the internal control. b Schematic representation of the genomic
   structure of zebrafish ace2 and the target site of CRISPR. Gross
   morphology (c), body length (d) and body weight (e) of zebrafish ace2
   homozygous mutants (ace2^−/−, n = 5) and sibling heterozygous controls
   (ace2^+/−, n = 6) at 50 dpf. Gross morphology (f), body length (g) and
   body weight (h) of zebrafish ace2 homozygous mutants (ace2^−/−, n = 21)
   and sibling heterozygous controls (ace2^+/−, n = 14) at 110 dpf. i The
   relative body weight was normalized the aged-match controls at 50, 80,
   and 110 dpf (50 dpf: ace2^+/−, n = 6; ace2^−/−, n = 5; 80 dpf:
   ace2^+/−, n = 19; ace2^−/−, n = 13; 80 dpf: ace2^+/−, n = 14; ace2^−/−,
   n = 21). j The growth hormones (GH) encoded gene gh mRNA expression
   levels in zebrafish pituitary at 80 dpf assessed by quantitative
   real‑time PCR analysis (ace2^+/−, n = 5; ace2^−/−, n = 6). The
   significance was analyzed by Student’s t-test. (**P < 0.01;
   ***P < 0.001).

   Notably, the growth of ace2 homozygous mutants (−/−) was significantly
   reduced compared to the heterozygous controls (+/−) between 16-50 days
   post fertilization (dpf) in co-housed feeding (Supplementary
   Fig. [71]2a, b and Fig. [72]1c–e). No significant difference in body
   length was observed at 12 dpf (Supplementary Fig. [73]2a), but a clear
   reduction became detectable by 16 dpf in body length. This growth
   reduced in ace2^−/− mutants were consistently observed at 20, 30, and
   50 dpf (Supplementary Fig. [74]2a, b and Fig. [75]1c–e). To exclude
   potential growth differences caused by co-housed feeding, we raised the
   fish separately under identical conditions. Both body length and weight
   were significantly lower in ace2^−/− mutants compared to controls at 40
   dpf under separate feeding conditions (Supplementary Fig. [76]2c, d),
   indicating a clear reduction in growth rate. The above results
   underscored the critical role of Ace2 in growth.

   Also, to investigate the molecular mechanisms underlying the growth
   retardation observed in Ace2 knockout zebrafish and to uncover the
   associated transcriptional changes, we performed transcriptomic
   analysis on 16 dpf larvae. The results revealed significant alterations
   in the transcriptomic profiles between the ace2^−/− mutants and
   controls. Notably, several immune-related pathways, such as the
   intestinal immune network for IgA production, phagosome, and Toll-like
   receptor signaling pathways, were markedly downregulated (Supplementary
   Fig. [77]3). These findings suggest that the loss of ace2 triggers
   widespread changes in gene expression within the zebrafish intestine.

   To further assess the growth performance of ace2^−/− mutants, we
   compared them with their sibling heterozygous controls at 110 dpf, the
   adult stage (Fig. [78]1f). Although both body length and body weight
   were significantly lower in ace2^−/−mutants compared to the controls at
   110 dpf, the difference was less pronounced than at 50 dpf
   (Fig. [79]1i). These results suggest that Ace2 disruption slowed down
   but does not completely block growth. We also assessed the mRNA
   expression levels of growth hormone (gh) in the pituitary. Gh, which is
   essential for normal physical growth, was significantly lower in
   ace2^−/− mutants compared to that in the controls (Fig. [80]1j). The
   reduced expression of gh in ace2^−/− mutants indicated a potential link
   between Ace2 and the regulation of gh, possibly mediated through
   nutritional status. The expression of gh is closely related to the
   intake and absorption of nutrients^[81]21,[82]22. In summary, our
   results suggest that Ace2 is vital for the normal growth of zebrafish,
   and its effect on gh may be through the regulation of nutrient
   absorption and metabolism in intestine.

Ace2 deficiency affected the amino acids metabolism in zebrafish

   The expression of amino acid transport protein B0AT1 (a
   sodium-dependent neutral amino acid transporter) in the intestinal
   tract of Ace2-deficient mice was significantly decreased. It has been
   previously reported that Ace2 heterodimerizes with B0AT1 in small
   intestinal mucosal epithelial cells to form a transport complex,
   thereby regulating the transport of corresponding amino acids, which
   revealed the role of Ace2 in the intestinal amino acid transport
   process of mice^[83]23. Also, single-cell sequencing databases from two
   studies have both shown that zebrafish ace2 is specifically expressed
   in intestinal epithelial cells, which are the primary sites for amino
   acid transport and absorption in the intestine, suggesting an important
   role for ace2 in the transport and absorption of amino acids in the
   zebrafish^[84]24,[85]25. In this study, to assess the physiological
   consequences of Ace2 loss, we detected the content of total amino acids
   in ace2^−/− mutants and their sibling controls at 110 dpf. Both plasma
   and muscle showed significant reductions in total amino acids in the
   absence of Ace2, underscoring its critical role in amino acid
   metabolism (Fig. [86]2a, b). Histological analysis revealed that muscle
   density was significantly lower in ace2^−/− mutants than in controls,
   indicating that muscle content was lower in the in ace2^−/− mutants
   than in the control (Fig. [87]2c, d). Furthermore, amino acid specific
   metabolome analysis showed that disruption of Ace2 caused significant
   changes in muscle amino acid metabolism (Fig. [88]2e–g). In zebrafish
   ace2^−/− mutants, the levels of cationic amino acids, such as
   L-(+)-Arginine and L-(+)-Lysine, were markedly reduced, with a more
   than three-fold decrease compared to controls (Fig. [89]2h).
   Additionally, neutral amino acids, including L-Leucine and valine, also
   showed notable reductions (Fig. [90]2h). These results suggest that
   Ace2 deficiency primarily affects the transport of cationic and neutral
   amino acids, which are crucial for fish growth, highlighting Ace2’s
   important role in regulating fish growth through modulation of amino
   acid metabolism.

Fig. 2. Disruption of Ace2 in zebrafish reduced the amino acids.

   [91]Fig. 2
   [92]Open in a new tab

   Quantification of total amino acids in zebrafish a plasma and b muscle
   at 110 dpf. The significance was analyzed by Student’s t-test.
   (**P < 0.01; ***P < 0.001; ace2^+/−, n = 6; ace2^−/−, n = 7). c The
   histological analysis in zebrafish muscle at 110 dpf. d The muscle
   density analysis in zebrafish at 110 dpf. Muscle density is calculated
   (software: Image J) by dividing the area of muscle fibers (the red
   part) in a tissue slice by the total area of the muscle. The
   significance was analyzed by Student’s t-test. (***P < 0.001; ace2^+/−,
   n = 6; ace2^−/−, n = 9). e PLS-DA (Partial Least Squares Discriminant
   Analysis) within zebrafish muscle between the ace2 homozygous mutants
   and sibling heterozygous controls at 110 dpf (ace2^+/−, n = 6;
   ace2^−/−, n = 7). f Clustering heatmap of different amino acid patterns
   for all samples in zebrafish muscle at 110 dpf (ace2^+/−, n = 6;
   ace2^−/−, n = 7). g The distribution of amino acid metabolites in each
   sample revealed by Z-score analysis (ace2^+/−, n = 6; ace2^−/−, n = 7).
   h The contents of different amino acids in zebrafish muscle at 110 dpf.
   The significance was analyzed by Student’s t-test. (*P < 0.05;
   **P < 0.01; ***P < 0.001; ace2^+/−, n = 6; ace2^−/−, n = 7).

Differences in bacterial composition between the mutant and control groups

   Zebrafish Ace2 is primarily expressed in the intestine and liver, with
   the highest expression in the intestine (Fig. [93]1a). Growth depends
   on food utilization and intestinal function, with the intestine being
   crucial for nutrient absorption and energy acquisition^[94]26. The
   intestine also hosts microbes that regulate metabolism and support
   growth^[95]27. Therefore, we performed third-generation full-length 16S
   rRNA amplicon sequencing on the intestines of both zebrafish Ace2
   mutants and control groups. Venn analysis results showed that there
   were 1,883 operational taxonomic units (OTUs) in all zebrafish guts,
   among which were 1,742 and 537 OTUs of the control and mutant groups,
   respectively (Fig. [96]3a). OTUs represent clusters of closely related
   microbial sequences, serving as a fundamental unit for assessing
   microbial diversity and community composition^[97]28. A reduction in
   OTU count in the mutant group suggests a loss of microbial diversity,
   which may have implications for gut function and host metabolism. Alpha
   diversity, as reflected by the Chao1 and Shannon indices, which showed
   microbial richness and diversity respectively, was significantly lower
   in ace2^−/− mutants (268 ± 30 and 3.49 ± 0.47) compared to ace2^+/−
   controls (598 ± 59 and 5.74 ± 0.85) (Fig. [98]3b), indicating a
   reduction in gut microbial diversity and potential community imbalance
   in the mutants. Beta-diversity, which assesses the differences in
   microbial community composition between groups, was analyzed using
   non-metric multidimensional scaling (NMDS). The results (stress =
   0.085) revealed a substantial variation in the gut microbial community
   structures between these two groups, further supported by analysis of
   similarity (Anosim) (Fig. [99]3c). More importantly, the structure of
   zebrafish intestinal bacterial community in the gut of ace2^−/− mutants
   (−/−) was more similar than that of controls (Fig. [100]3c, d). Hence,
   the results indicated that there was a significant reduction in the
   diversity of the gut microbiome in the Ace2 mutants, with significant
   changes in the intestinal flora. This suggests that Ace2 plays an
   important regulatory role in gut microbiome and may be a crucial factor
   affecting host growth.

Fig. 3. Disruption of Ace2 in zebrafish significantly decreased intestinal
microbial diversity and altered intestinal community composition.

   [101]Fig. 3
   [102]Open in a new tab

   a Bacterial OTU numbers in the ace2 homozygous mutants and sibling
   heterozygous controls at 110 dpf, respectively (n = 6/each group). b
   Difference in the gut bacterial α-diversity of zebrafish between the
   ace2 homozygous mutants and sibling heterozygous controls at 110 dpf
   (n = 6/each group). c Non-metric multidimensional scaling (NMDS) and d
   analysis of similarity (Anosim) results showed that there were
   significantly (P = 0.001) differences in gut community structure
   between the ace2 homozygous mutants and sibling heterozygous controls
   at 110 dpf (n = 6/each group). e Differences in the bacterial community
   composition within zebrafish gut between the ace2 homozygous mutants
   and sibling heterozygous controls at 110 dpf (n = 6/each group). f
   LEfSe results of bacterial indicator species in guts of the ace2
   homozygous mutants and sibling heterozygous controls at 110 dpf
   (n = 6/each group).

   The bacterial composition in the gut of ace2^−/− mutants exhibited
   significant changes compared to controls (Fig. [103]3e). In gut of
   ace2^+/− controls, the ten most relatively abundant species were the
   unclassified Desulfovibrionaceae, Staphylococcus cohnii, Aeromonas
   enteropelogenes, Cetobacterium somerae, Chryseobacterium formosense,
   alpha proteobacterium LS96, Sinirhodobacter hungdaonensis, unclassified
   Alphaproteobacteria, Corynebacterium tuberculostearicum, and
   Leuconostoc pseudomesenteroides, which accounted for 18.36%, 4.68%,
   3.29%, 1.85%, 0.71%, 0.58%, 0.52%, 0.51%, 0.48% and 0.42%,
   respectively. While in the gut of ace2^−/− mutants, the ten most
   relatively abundant species were the Cetobacterium somerae, Aeromonas
   enteropelogenes, Shewanella putrefaciens, Cetobacterium sp. ZOR0034,
   unclassified Sutterellaceae, unclassified Desulfovibrionaceae,
   Staphylococcus cohnii, Rubrobacter bracarensis, Rhodovulum sp. JC2237
   and Coraliomargarita sp.,which accounted for 37.31%, 16.96%, 4.11%,
   1.54%, 1.26%, 0.99%, 0.78%, 0.61%, 0.29%, and 0.22%, respectively.
   Thus, knockout of Ace2 altered the distribution of dominant bacterial
   species in ace2^−/− mutants. We further observed notable differences in
   the relative abundances of some bacterial species in the zebrafish gut
   between these two groups. Specifically, there was a significant
   (P < 0.05 in all cases) increase in the relative abundance of 12
   bacterial species, such as the Cetobacterium somerae and Aeromonas
   enteropelogenes of ace2^−/− mutants (Supplementary Data [104]1).
   Meanwhile, there was a significant (P < 0.05 in all cases) decrease in
   the relative abundance of 46 bacterial species, such as the
   unclassified Desulfovibrionaceae and Staphylococcus cohnii of ace2^−/−
   mutants (Supplementary Data [105]1). Linear Discriminant Analysis
   Effect Size (LEfSe) analysis, which used to identify statistically
   significant biomarkers that explain differences in microbial
   communities, further revealed that C. somerae, A. enteropelogenes and
   S. putrefaciens were the bacterial biomarkers in the gut of ace2^−/−
   mutants (Fig. [106]3f). Therefore, knockout of ace2 exerted a stronger
   selective pressure on the bacterial taxa in zebrafish guts.
   Furthermore, knockout of ace2 significantly changed the bacterial
   composition of zebrafish and so likely affects gut bacterial function.

   Similar alterations were observed at earlier developmental stages. At
   16 dpf, ace2^−/− mutants showed reduced α-diversity and distinct
   microbial structures, accompanied by growth defects. At 30 dpf,
   although α-diversity differences were no longer significant,
   β-diversity remained altered, indicating persistent microbiome
   remodeling (Supplementary Fig. [107]4). Together, these results
   highlight the role of Ace2 in shaping gut microbial communities
   throughout development, with potential implications for nutrient
   metabolism and growth.

Differences in potential bacterial functions between the mutant and control
groups

   The principal component analysis (PCA) results, which was based on
   microbial content from 16S rRNA sequencing, indicated that the
   predicted functional contents of the bacterial community in zebrafish
   gut of ace2^−/− group were clearly separated from those of control
   group (Fig. [108]4a). Several pathways were significantly (P < 0.05)
   enriched in the gut of ace2^−/− group, especially the genes associated
   with starch and sucrose metabolism, purine metabolism, homologous
   recombination, pyrimidine metabolism, pentose phosphate pathway,
   beta-Lactam resistance, phosphotransferase system (PTS), lysine
   biosynthesis, amino sugar and nucleotide sugar metabolism, ribosome,
   quorum sensing, porphyrin and chlorophyll metabolism (Fig. [109]4b). In
   contrast, the genes associated with arginine biosynthesis, glyoxylate
   and dicarboxylate metabolism, “phenylalanine, tyrosine and tryptophan
   biosynthesis”, 2-Oxocarboxylic acid metabolism, “glycine, serine and
   threonine metabolism”, biosynthesis of amino acids, microbial
   metabolism in diverse environments, biosynthesis of antibiotics and
   oxidative phosphorylation were significantly (P < 0.05 in all cases)
   less abundant in ace2^−/− group (Fig. [110]4b). Overall, the potential
   functions of the gut bacterial community were markedly different
   between the control and ace2^−/− zebrafish. For instance, bacterial
   pathways involved in protein metabolism and biosynthesis were markedly
   decreased in ace2^−/− mutants (Fig. [111]2), consistent with their
   reduced amino acid profiles while pathways related to carbohydrate
   metabolism showed enhanced (Fig. [112]4b), indicating potentially
   compromised energy utilization and storage in ace2^−/− mutants might be
   changed.

Fig. 4. Disruption of Ace2 in zebrafish significantly altered the bacterial
functional in gut.

   [113]Fig. 4
   [114]Open in a new tab

   a PCA results of predicted functional content of bacterial community in
   zebrafish gut between the ace2 homozygous mutants and sibling
   heterozygous controls at 110 dpf. b Predicted bacterial functional gene
   comparisons in zebrafish gut between these two groups. *P < 0.05,
   ***P < 0.001.

   We further analyzed the predicted functional composition of the gut
   microbiota in juvenile zebrafish (at 16 and 30 dpf). Similar to the
   adult stage, pathways related to protein metabolism and biosynthesis
   were significantly downregulated in ace2^−/− mutants, likely due to
   impaired amino acid transport caused by Ace2 deficiency. However,
   unlike in adults, carbohydrate metabolism pathways did not show a
   marked upregulation at this stage (Supplementary Fig. [115]5).

Glucose metabolism was enhanced in the absence of Ace2

   In the context of Ace2 deficiency, there was a noted decrease in
   zebrafish growth (Fig. [116]1), accompany by a reduction in total amino
   acid levels in both the blood and muscle (Fig. [117]2). This
   insufficiency in amino acid supply likely prompted a shift in metabolic
   pathways related to nutrient metabolism. Also, Ace2 knockout resulted
   in a diminished capacity for amino acid metabolism within the gut
   microbiota, while concurrently enhancing bacterial pathways related to
   starch and sucrose metabolism, indicating increased expression of genes
   associated with carbohydrate metabolism (Fig. [118]4b).

   By 110 dpf, the body weight difference between Ace2 mutants and
   controls decreased, indicating compensatory growth (Fig. [119]1i). To
   investigate the underlying mechanism, we examined key genes in the
   hepatic insulin-like growth factor (IGF) signaling, a critical
   downstream effector of the hypothalamic-pituitary-somatotropic (HPS)
   axis involved in somatic growth and cell proliferation^[120]29,[121]30.
   Our results showed elevated expressions of the igf signaling genes
   igf1, igf2a, and igf2b in Ace2-deficient zebrafish. Furthermore, there
   was significant upregulation in the expressions of igf receptors
   igf1ra, igf1rb, and igf2r (Fig. [122]5a–f), indicating activation of
   the IGF signaling pathway that may contribute to the compensatory
   growth^[123]31.

Fig. 5. Gene expression levels of glucose metabolism pathways assessed by
quantitative real‑time PCR analysis.

   [124]Fig. 5
   [125]Open in a new tab

   a–f Expression of different genes of Igf signaling in zebrafish liver
   between the ace2 homozygous mutants and sibling heterozygous controls
   at 110 dpf. g Expression of glucose transporter gene glut2 in zebrafish
   liver between the ace2 homozygous mutants and sibling heterozygous
   controls at 110 dpf. h Expression of glycogen synthesis gene ugp2a in
   zebrafish liver between the ace2 homozygous mutants and sibling
   heterozygous controls at 110 dpf. i–l Expression of glycolysis gene
   hk1, hk2, pfkla and pklr in zebrafish liver between the ace2 homozygous
   mutants and sibling heterozygous controls at 110 dpf. Data shown are
   mean ± SEM, and statistical significance was analyzed by Student’s
   t-test. (*P < 0.05; **P < 0.01; n = 5/each group). m PCA (Principal
   component analysis) of non-targeted metabolomics within zebrafish
   muscle between the ace2 homozygous mutants and sibling heterozygous
   controls at 110 dpf (ace2^+/−, n = 4; ace2^−/−, n = 4). n Volcano plot
   of differential metabolites in zebrafish muscle between the ace2
   homozygous mutants and sibling heterozygous controls at 110 dpf
   (ace2^+/−, n = 4; ace2^−/−, n = 4). Each dot represents a metabolite.
   The x-axis shows the log[2] fold change (ace2^−/− vs ace2^+/−), and the
   y-axis indicates the -log[10] (P-value). Red dots represent
   significantly increased metabolites in ace2^−/− (log[2]FC > 1 and P <
   0.05), blue dots indicate significantly decreased metabolites
   (log[2]FC < -1 and P < 0.05), and grey dots represent metabolites with
   no significant difference. o KEGG pathway-based Differential Abundance
   Score analysis of differential metabolites in zebrafish muscle between
   the ace2 homozygous mutants and sibling heterozygous controls at 110
   dpf (*P < 0.05; **P < 0.01; ***P < 0.001; ace2^+/−, n = 4; ace2^−/−,
   n = 4). The plot shows the distribution of significantly altered
   metabolites in KEGG pathways between ace2^−/− and ace2^+/− muscle. The
   x-axis represents the differential abundance score, calculated as
   (number of increased metabolites/ number of decreased metabolites)
   divided by the square root of the total number of detected metabolites
   in each pathway. A higher score indicates a tendency toward
   upregulation in the pathway, while a lower score suggests
   downregulation.

   Activation of IGF signaling and compensatory growth suggest that Ace2
   mutants undergo metabolic remodeling in response to impaired amino acid
   metabolism. Consistent with this, elevated expression of glucose
   transporter (glut2) and glycogen synthesis-related (ugp2a) genes
   (Fig. [126]5g, h), along with glycolytic enzymes (hk1, hk2, pklr;
   Fig. [127]5i–l), might indicate a potential shift toward carbohydrate
   catabolism. To further substantiate this observation, we performed
   non-targeted metabolomic profiling of muscle in addition to the
   previous targeted amino acid analyses (Fig. [128]5m–o). PCA revealed a
   clear metabolic distinction between ace2^−/− mutants and controls
   (Fig. [129]5m), with 155 metabolites significantly increased and 139
   decreased (Fig. [130]5n). KEGG-based differential abundance score
   analysis revealed a coordinated downregulation of amino acid metabolism
   pathways, including arginine biosynthesis, arginine and proline
   metabolism, lysine degradation, D-amino acid metabolism, and
   glutathione metabolism. Conversely, the citrate cycle (TCA cycle), a
   core component of carbohydrate metabolism, was significantly increased
   (Fig. [131]5o). Together, these findings suggest that Ace2 deficiency
   induces metabolic reprogramming characterized by impaired amino acid
   metabolism and enhanced glucose utilization, which may compensate for
   growth retardation through IGF signaling.

Discussion

   In this study, we generated an ace2 zebrafish mutant line using
   CRISPR/Cas9 technology to provide compelling insights into the
   biological roles of Ace2 in vertebrates, particularly in fish models.
   The ace2^−/− exhibited not only growth retardation but also significant
   alterations in gut microbial diversity and function, which correlated
   with changes in nutrient metabolism. The observed reductions in amino
   acids in ace2^−/− mutants highlight the role of Ace2 in amino acid
   transport and metabolism, aligning with findings in other vertebrates
   and indicating a conserved function across species. Moreover, the
   alterations in the gut microbiota and their functional pathways suggest
   a link between Ace2, gut microbial, and host metabolism. These findings
   support the hypothesis that Ace2 is crucial for amino acid absorption
   and growth. Additionally, the microbial shifts in the absence of Ace2
   help the fish adapt to amino acid deficiencies, providing an
   alternative metabolic strategy to support growth. This indicates that
   by reducing or interfering with amino acid transport, the host can
   enhance glucose utilization mediated by the gut microbiota, thereby
   regulating growth and development and partially compensating for growth
   demands (Fig. [132]6).

Fig. 6. A working model on the role of Ace2 in zebrafish growth and metabolic
remodeling.

   [133]Fig. 6
   [134]Open in a new tab

   Ace2 deficiency leads to reduced amino acid absorption in the
   intestine, resulting in growth retardation. Ace2 deficiency also
   significantly alters the composition and function of the gut
   microbiota. The host’s energy metabolism shifts from being amino
   acid-dominant to carbohydrate-dominant. The altered gut microbiota and
   the shift to carbohydrate metabolism partially compensate for the
   growth inhibition caused by Ace2 deficiency, even though growth
   differences remained significant.

   Ace2 is crucial for growth, as demonstrated in various studies and
   supported by our findings on zebrafish ace2 mutants. Ace2 mutants
   (ace2^−/−) showed significant growth retardation, including reduced
   body length and weight compared to controls. This is consistent with
   the previous studies in mice, who reported that Ace2 deficiency impairs
   intestinal barrier integrity and leads to gastrointestinal dysfunction,
   subsequently affecting overall growth^[135]7,[136]10. Ace2 knockout
   mice exhibited a body weight reduction of 4%-20% compared to the
   control group at various testing ages; however, in our study, we found
   that Ace2 deficiency also resulted in a severe weight decrease in
   zebrafish, with body weight in mutant groups being 82% and 51% lower
   than that of the age-matched control group at 50 and 110 dpf,
   respectively (Fig. [137]1), suggesting the significant role of Ace2 on
   fish growth.

   Growth is the result of nutrient utilization, and the intestine plays a
   key role in this process as the primary site for the absorption of
   nutrients such as proteins, fats, carbohydrates, and vitamins^[138]32.
   The high expression of Ace2 in the intestine, as observed in our study,
   further underscores its role in nutrient absorption for growth, which
   is agreed with the previous studies on Ace2 involvement in intestine
   function^[139]1,[140]4.

   In Ace2-deficient mice, the transport of neutral amino acids like
   tryptophan is impaired due to the disruption of the Ace2-B0AT1
   transport complex in the small intestine^[141]23. When wild-type mice
   and Ace2 mutant mice were fed a high-protein diet, the relative weight
   gain of Ace2 mutant mice was similar to that of the wild-type controls,
   indicating that a high-protein diet can compensate for the effects of
   insufficient amino acid transport due to the Ace2 mutation. However,
   when wild-type mice and Ace2 mutant mice were fed a low-protein diet,
   the wild-type mice continued to grow effectively, while the growth of
   Ace2 mutant mice completely stalled, indicating that under
   low-nutrition conditions, Ace2-mediated amino acid transport plays a
   major role in growth^[142]10. Therefore, Ace2-mediated amino acid
   transport is crucial for efficient protein utilization in mice, and
   this role is particularly important in protein depleted diets.

   The growth of ace2^−/− mutants was significantly inhibited, even under
   conditions of overfeeding to the point of visible satiety, indicating
   that the Ace2 mutation may impair nutrient absorption in zebrafish.
   Unlike other animals, fish have a lower ability to utilize
   carbohydrates; proteins and their breakdown products, amino acids, are
   the primary nutrients for fish growth^[143]26. Therefore, sufficient
   protein/amino acid levels in the intestine are crucial for fish growth
   and development^[144]17,[145]33. In zebrafish, the knockout of Ace2
   significantly impacts amino acid absorption, and given the
   protein-dependent metabolic mode of fish, the deficiency of amino acids
   severely affects zebrafish growth (Fig. [146]2a–g).

   In addition, apart from tryptophan, there is a significant
   downregulation of a wide range of amino acids in zebrafish muscle,
   including those highly related to growth such as L-(+)-Arginine,
   L-(+)-Lysine, L-Isoleucine, L-Leucine, and L-Methionine (Fig. [147]2h).
   A high-protein diet in fish can promote the expression of growth
   hormone-releasing peptide transcripts in the intestines of gilthead
   seabream, thereby increasing Gh levels, indicating the regulatory role
   of nutrients on Gh^[148]34. Adding arginine to the diet of silver
   crucian carp can activate the IGF signaling pathway, thereby promoting
   their growth, demonstrating the important role of arginine and IGF
   signaling pathway in fish growth^[149]35.

   Notably, unlike in mice where Ace2 deficiency mainly affects neutral
   amino acids, Ace2 knockout in zebrafish led to significant reductions
   not only in neutral but also in cationic amino acids such as arginine
   and lysine. Our results collectively suggested that ace2 plays a
   broader role in amino acid metabolism in fish than in mammals. Two
   possible mechanisms may explain this: (1) the Ace2-B0AT1 complex in
   fish may have acquired a novel function in transporting cationic amino
   acids; (2) Ace2 may form alternative transporter complexes with other
   B0AT1 homologs specifically mediating cationic amino acid uptake.
   Further investigation is needed to elucidate the underlying mechanisms.

   Research indicates that Ace1 (angiotensin-converting enzyme) plays a
   critical role in maintaining gut homeostasis. In zebrafish, knockout of
   ace1-a molecule in the renin-angiotensin system (RAS) with functions
   opposing those of ace2-leads to significant gut inflammation,
   characterized by elevated inflammatory markers and altered mucosal
   defense, further confirming the importance of Ace1 in gut
   health^[150]36. Additionally, the high expression levels of ace1 and
   ace2 in the gut suggest their vital roles in regulating gut
   physiological functions.

   Recent studies, including our findings, demonstrate that knocking out
   either ace1 or ace2 significantly impacts gut function. Knockout of
   ace2 primarily results in disrupted amino acid metabolism, alterations
   in the gut microbiota, and growth inhibition, while ace1 knockout
   mainly manifests as heightened inflammatory responses and increased
   susceptibility to enteritis. The complex interplay between Ace1 and
   Ace2 in regulating gut functions warrants further investigation.

   We performed transcriptome and 16S amplicon sequencing on Ace2 mutants
   and found that Ace2 deficiency led to a significant downregulation of
   immune-related pathways at 16 dpf. Since inflammatory responses are
   typically associated with upregulation of immune markers, this immune
   suppression is unlikely to result from inflammation^[151]37,[152]38.
   Instead, it may be due to impaired growth in Ace2 mutants, which
   hinders the proper development of the immune system. Instead, impaired
   growth in Ace2 mutants may hinder proper immune system development.
   Given the considerable ace2 expression in the spleen (Fig. [153]1a)—a
   tissue rich in immune cells^[154]39—Ace2 knockout may affect immune
   cell development and function. Additionally, Ace2 loss in the intestine
   could directly impair the local immune environment and barrier
   function. This systemic immune dysfunction, involving both spleen and
   intestine, may disrupt intestinal homeostasis, as evidenced by reduced
   microbial diversity, altered community structure, and increased
   abundance of potential pathogens (Supplementary Fig. [155]4c, d). These
   observations differ from Ace1 mutants, which primarily exhibit
   inflammation-driven gut abnormalities, suggesting that Ace2 plays a
   distinct and critical role in coordinating growth, immune maturation,
   and gut homeostasis.

   Future studies should focus on the molecular mechanisms of Ace1 and
   Ace2, particularly their synergistic or antagonistic roles in gut
   inflammation and metabolic regulation. In our research, ace2 knockout
   caused significant growth inhibition and a decrease in amino acid
   levels, highlighting an important role in amino acid transport and/or
   metabolism. However, whether ace1 knockout results in similar
   nutritional deficiencies and growth issues remains unclear and merits
   further exploration. Unraveling the interactions between Ace1 and Ace2
   in gut regulation could provide new insights into gut health research.
   Additionally, future work will also investigate the relevant
   ligand-receptor interactions and signaling mechanisms that may mediate
   the effects observed in Ace2 mutants, which could further elucidate the
   complex regulation of gut homeostasis.

   Ace2 deficiency significantly influenced amino acid metabolism and
   consequently the composition and function of the gut microbiota. Our
   study demonstrated substantial reductions in total amino acid levels in
   both plasma and muscle of ace2^−/− mutants, indicating impaired amino
   acid metabolism. The host nutrition absorption mode changes can lead to
   intestinal microbial populations^[156]40,[157]41. The composition of
   gut microbiota in fish varies according to their dietary
   habits^[158]42,[159]43. In our study, despite consistent daily feeding
   routines, the impairment of Ace2 function in mutants led to disrupted
   intestinal amino acid absorption, significantly altering the microbial
   composition in ace2^−/− mutants compared to controls (Fig. [160]3). The
   reduced amino acid levels directly affected the gut microbiota, as
   these nutrients are crucial for microbial growth and function.

   Gut is a complex ecosystem where a vast array of microorganisms
   coexists symbiotically with the host, jointly regulating the host
   metabolism and growth. Previous research has found that the diversity
   of gut microbiota in fish is positively correlated with their
   developmental stages, suggesting the critical role of gut microbiota
   diversity in fish growth and development^[161]44,[162]45. The gut
   microbiota supports and promotes host growth by participating in
   nutrient absorption, energy acquisition, and metabolic regulation. In
   this study, we performed full-length 16S rRNA amplicon sequencing on
   the intestines of ace2 mutant zebrafish and control groups. The results
   showed that the gut microbiota diversity in ace2 mutants was
   significantly reduced (Fig. [163]3), and there were notable changes in
   the gut microbial community (Fig. [164]4), suggesting that Ace2 plays
   crucial roles in regulating the gut microbiota and affecting host
   growth.

   The PCA results and pathway enrichment analysis reveal that the Ace2
   mutation significantly alters the metabolic functions of the gut
   microbiota in zebrafish (Fig. [165]5). Notably, an increase in
   carbohydrate metabolism pathways, such as starch and sucrose
   metabolism, suggests a compensatory response to impaired protein
   metabolism due to Ace2 deficiency^[166]10,[167]33. Conversely, the
   decrease in amino acid biosynthesis pathways and energy production
   cycles like the TCA cycle underscores the metabolic disruptions that
   hinder normal growth and physiological functions^[168]17,[169]32.
   Additionally, reduced pathways related to host defense mechanisms
   suggest a potential increase in vulnerability to pathogens^[170]7.
   Also, these microbial functional changes likely arise from shifts in
   community composition rather than gene expression alterations within
   individual species. In ace2^−/− mutants, diminished amino acid
   absorption coupled with suppressed intestinal immunity creates a
   nutrient environment favoring the proliferation of
   carbohydrate-utilizing and immune-sensitive bacteria. This suggests
   that microbial restructuring may be driven by selective pressures
   stemming from host metabolic and immune modifications. These findings
   highlight the critical role of Ace2 in maintaining balanced gut
   microbiota and overall host metabolism.

   Due to technical limitations, we were unable to collect gut microbiota
   samples before 16 dpf, making it unclear whether dysbiosis is the cause
   or consequence of the growth defects observed in ace2^−/− mutants. One
   possibility is that early loss of Ace2 disrupts mucosal immunity,
   thereby affecting microbiota composition. Alternatively, amino acid
   deficiency may trigger an energy trade-off in the host, impairing
   immune function and growth. These mechanisms may coexist. Future
   studies using early-stage sampling, germ-free models, or microbiota
   transplantation may help clarify their causal relationships and
   underlying mechanisms.

   The distinct differences in bacterial community and function between
   the mutant and control groups suggested that the Ace2 mutation has
   profound effects on the gut microbiota metabolic landscape. In our
   study, the most significant change in the gut microbiota due to the
   knockout of Ace2 was the marked upregulation of Cetobacterium somerae
   (Fig. [171]3f). In zebrafish, a study demonstrated that the gut
   bacterium Cetobacterium somerae plays a key role in regulating fish
   growth. The Cetobacterium somerae promotes host insulin expression and
   glucose utilization by activating the parasympathetic nervous system
   through its metabolic product, acetate^[172]19. In Ace2 mutant
   zebrafish, the enrichment of carbohydrate metabolism pathways suggests
   a compensatory mechanism for the decrease of amino acid and protein
   metabolism, potentially driven by the increased presence of
   Cetobacterium somerae, reflecting an adaptation to protein-derived
   energy deficits^[173]33.

   Fish, which predominantly rely on proteins for energy metabolism, have
   a higher protein utilization rate and lower carbohydrate utilization
   rate compared to other vertebrates, making them particularly sensitive
   to disruptions in amino acid absorption^[174]17,[175]18. In
   Ace2-deficient zebrafish, the significant reduction in amino acids
   severely impacts growth due to their dependence on proteins as a
   primary energy source. However, the gut microbiota appears to
   adaptively respond to this nutritional challenge. Our findings showed
   that the Ace2 knockout led to a reorganization of the gut microbiota,
   which helped the host adapt to the amino acid deficiency. This
   microbial shift supports an alternative metabolic strategy, enhancing
   glucose metabolism pathways to compensate for the reduced protein
   utilization, as evidenced by the upregulation of genes involved in
   glucose transport and glycolysis in Ace2 mutants. This adaptive
   response is consistent with the studies, which highlighted the gut
   microbiota’s role in regulating host energy metabolism and nutrient
   absorption.

   The metabolic remodeling in adaption of amino acid deficiency for
   growth was not only supported by gut microbiota changes but also the
   gene expression profile in liver. Increased expression of genes
   involved in glucose transport (glut2) and glycogen synthesis (ugp2a)
   suggests an adaptive mechanism for improved glucose storage and
   utilization (Fig. [176]5g, h). Additionally, the upregulation of
   glycolysis pathway genes hk1, hk2, and pklr highlights an increased
   capacity for glucose breakdown and energy production, compensating for
   amino acid metabolism deficiencies (Fig. [177]5i–l). The concurrent
   enhancement of glycolysis and glycogen synthesis pathways reflects an
   overall strengthening of glucose metabolism. This conclusion is further
   supported by untargeted metabolomics analysis of muscle, which revealed
   a reduction in amino acid metabolism and a corresponding increase in
   glucose metabolism.

   Meanwhile, as developmental age progresses, the growth differences
   between ace2^−/− mutants and controls decrease, indicating that this
   gut microbiota-driven adaptive metabolic remodeling plays a role in the
   growth process. However, despite this adaptation, significant growth
   differences remain between ace2^−/− mutants and controls
   (Fig. [178]1I), indicating that enhanced glucose metabolism alone
   cannot fully offset the loss of amino acid metabolism.

   This metabolic plasticity underscores the critical role of Ace2 in
   balancing amino acid and glucose metabolism for optimal growth, as
   amino acids are crucial for protein synthesis while glucose provides
   essential energy. The disruption of this balance due to Ace2 deficiency
   emphasizes the enzyme’s importance in comprehensive nutrient
   metabolism. Ace2 deficiency leads to impaired amino acid absorption,
   causing a metabolic reshaping of the gut microbiota, which compensates
   by enhancing carbohydrate metabolism. However, fish have low efficiency
   in utilizing carbohydrates, resulting in restricted growth despite
   adaptive changes in gut microbiota and liver metabolism. This
   highlights the critical role of Ace2 in balancing amino acid and
   glucose metabolism to support growth.

   In Ace2-deficient zebrafish, early growth is suppressed, but body
   weight differences decrease with age, indicating compensatory growth.
   This phenomenon resembles the “catch-up growth” observed under nutrient
   restriction or hypoxia, which typically involves activation of IGF
   signaling pathways^[179]31,[180]46. For example, Kamei et al. reported
   that zebrafish embryos rely on IGF1R/Akt/Erk signaling to achieve
   accelerated growth following re-oxygenation^[181]31. Our study revealed
   that Ace2 deficiency impairs amino acid metabolism, and compensatory
   growth may depend on altered nutrient utilization. Similarly, Ace2
   knockout mice exhibit a metabolic shift toward increased glucose
   utilization, characterized by elevated respiratory exchange ratio,
   reduced palmitate oxidation, and upregulated expression of glucose
   transporters in muscle^[182]47, consistent with the glucose metabolic
   adaptations observed in our zebrafish model.

   IGF-1 is downregulated during resource scarcity and upregulated upon
   recovery^[183]46. IGF-1 administration in mice fed a low-protein diet
   nearly restored their body weight to normal levels, indicating that
   exogenous IGF-1 supplementation can directly counteract growth
   inhibition caused by nutrient deficiency^[184]48. Fang et al.
   demonstrated that IGF-1 supplementation during starvation significantly
   increased gluconeogenesis, restored blood glucose levels to normal, and
   enhanced lipid mobilization, as evidenced by elevated plasma glycerol
   and free fatty acids^[185]49. Together, these findings support the idea
   that activation of the IGF signaling pathway alleviates metabolic
   stress and promotes compensatory growth. We propose that growth delay
   caused by Ace2 loss is partially compensated through IGF-dependent
   metabolic regulation. However, it remains unclear whether metabolic
   reprogramming triggers IGF signaling to drive compensatory growth or if
   IGF signaling induces metabolic adaptations that promote compensatory
   growth; this causal relationship requires further investigation.

   Our study demonstrated the significant role of Ace2 in zebrafish. In
   conclusion, this study provides genetic evidence for the function and
   interaction between Ace2, gut microbiota, and metabolism in zebrafish.
   The key discoveries can be summarized as follows: (1) Ace2 is crucial
   for normal growth and development, with Ace2 knockout mutants
   exhibiting developmental delays and significant changes in gut
   microbiota diversity; (2) Ace2 regulates a broader range of amino acids
   in fish compared to mammals, highlighting its important role in amino
   acid absorption; (3) The interplay between Ace2 mediated amino acid
   absorption and gut microbiota is vital for metabolic and physiological
   process in zebrafish, underscoring the enzyme’s multifaceted role in
   nutrient metabolism.

Materials and methods

Animal

   The zebrafish were reared at 28 ± 1°C with a photoperiod of 14-h light
   and 10-h dark in a flow-through aquarium system (Haixing, China). The
   larvae were reared in nursery tanks with paramecia and artemia before
   transfer to the aquarium system, and the adults were fed with artemia
   and commercial dry food. All experiments were carried out according to
   the protocols approved by the Institutional Animal Care and Use
   Committee of Sun Yat-sen University (Approval No.
   SYSU-IACUC-2024-B0723). We have complied with all relevant ethical
   regulations for animal use.

Zebrafish mutant line generation and genotyping

   The CRISPR/Cas9 technique was employed to generate the ace2 mutant in
   zebrafish, following established protocols and our previous study,
   which is briefly descripted as follows. An sgRNA targeting exon I of
   ace2 was designed using online tools ([186]https://www.crisprscan.org)
   to reduce the potential off-target effects, with the oligo sequence
   detailed in Supplementary Table [187]1. The gRNA was synthesized
   through in vitro transcription from DraI-digested pDR274 (Addgene
   Plasmid #42250) using the mMACHINE T7 system, adhering to the
   manufacturer’s guidelines. Subsequently, 4.6 nl of a mixture containing
   gRNA (60 ng/μl) and Cas9 protein (New England Biolabs, Ipswich, MA) was
   injected into one- or two-cell-stage embryos using the IM 300
   Microinjector (NARISHIGE, Japan). Screening of F0 mutants was performed
   using high-resolution melting analysis (HRMA) and heteroduplex mobility
   assay (HMA), with sequencing for confirmation^[188]50,[189]51. F0
   founders with mosaic mutations were crossed with wild-type fish to
   produce heterozygous F1 (+/−) offspring. Male and female F1 siblings
   with the same mutation were then crossed to obtain homozygous F2 (−/−)
   individuals.

   In addition, to minimize potential off-target effects introduced by
   CRISPR/Cas9 genome editing, we performed multiple generations of
   outcrossing and selected heterozygous carriers based on genotype rather
   than phenotype. As a result, the growth data analyzed in this study
   were collected from later generations: data in Fig. [190]1 were
   obtained from the F5 generation, and data in Supplemental Fig. [191]2
   were obtained from the F7 generation. This approach supports the
   interpretation that the observed slow growth phenotype is primarily
   attributable to the ace2 mutation.

Sampling and histological examination

   The entire process can be referenced from our previously published
   article and is summarized as follows^[192]12,[193]15. Fish were
   collected at various time points for phenotype assessment. Following
   anesthetization with MS222 (Sigma, St. Louis, MO), the fish were
   euthanized. Each fish’s gross morphology was documented with a digital
   camera (Canon EOS 700D). For histological examination, the fish were
   fixed in PFA fixative for a minimum of 24 hours. Dehydration and
   infiltration were conducted using the ASP6025S Automatic Vacuum Tissue
   Processor (Leica, Wetzlar, Germany). Samples were embedded in paraffin,
   and serial sections of 5 µm thickness were prepared. The sections were
   stained with hematoxylin and eosin (H&E) and examined using the Nikon
   ECLIPSE Ni-U microscope (Nikon). Images were captured with the Aperio
   VERSA 8 scanning system (Leica, Germany). Sibling wild-type (+/+)
   and/or heterozygous (+/−) fish served as controls for phenotype
   analysis. This entire procedure was performed as described in our
   previous publication.

RNA extraction and quantitative real-time PCR

   The entire procedure, as detailed in our previous publication, is
   summarized as follows^[194]12,[195]15. Total RNA was isolated from the
   respective tissues using TRIzol (Invitrogen, Waltham, MA) as the
   manufacturer’s instructions. M-MLV reverse transcriptase (Invitrogen)
   was used for reverse transcription. Real-time PCR was carried out on
   the CFX384 Real-Time System (Bio-Rad) with primers listed in
   Table [196]S1. Primer specificity was verified through melting curve
   analysis. Each PCR assay included a standard curve for quantification.
   Gene expression levels were normalized to ef1a and expressed as fold
   changes relative to the control group.

The 16S rRNA, transcriptome sequencings and bioinformatic analysis

   The entire procedure, as detailed in our previous publication, is
   summarized as follows^[197]52. Genomic DNA from the gut contents of
   zebrafish was extracted using the PowerFecal DNA Isolation Kit (Mobio,
   Carlsbad, CA, USA). The bacterial full-length 16S rRNA gene was
   amplified using the 27 F and 1492 R primers, following the previously
   described amplification program and amplicon sequencing library
   construction^[198]53. The prepared library was sequenced on a PacBio
   Sequel II platform (Pacific Biosciences, Menlo Park, CA, USA) by
   Biomarker Technologies Co. Ltd. (Beijing, China), producing single-end
   reads. Raw 16S rRNA sequencing data have been deposited in the SRA
   database of NCBI under accession numbers PRJNA1125140. The raw
   sequences were processed and demultiplexed using SMRT (version 8.0) to
   obtain circular consensus sequencing (CCS) reads. CCS assignment for
   each sample was performed with Lima (version 1.7.0). Quality control
   using cutadapt (version 2.7) discarded CCS reads without primers and
   those outside the 1,200-1,650 bp range. Chimera sequences were
   identified and removed with the UCHIME algorithm (V8.1) as described
   previously, resulting in clean reads. Sequences with ≥ 97% similarity
   were clustered into operational taxonomic units (OTUs) using USEARCH
   (V10.0). Taxonomy was annotated using the RDP classifier
   ([199]http://sourceforge.net/projects/rdpclassifier/), diversity
   analysis followed the methods described in, and the phylogenetic tree
   was constructed as outlined in previous studies^[200]54. The functional
   prediction of the bacterial community was conducted using
   Tax4Fun^[201]55.

Estimation of ecological processes governing the bacterial community assembly
in zebrafish gut

   In accordance with our prior publication, we utilized a phylogenetic
   bin-based null model (iCAMP) to assess the predominant ecological
   processes influencing community assembly in the zebrafish gut across
   both control and mutant groups. Initially, the observed taxa were
   categorized into 12 distinct ‘bins’ based on their phylogenetic
   relationships. Each bin’s assembly process was determined through a
   null model analysis incorporating beta net relatedness index (βNRI) for
   phylogenetic diversity and modified Raup-Crick metric (RC) for
   taxonomic β-diversities. For bins, percentages of homogeneous and
   heterogeneous selection were calculated based on pairwise comparisons
   with βNRI values falling below -1.96 and above +1.96, respectively.
   Subsequently, RC was employed to categorize remaining pairwise
   comparisons with |βNRI | ≤ 1.96: percentages of homogenizing dispersal
   and dispersal limitation were derived from comparisons with RC values
   below -0.95 and above +0.95, while those with |βNRI | ≤ 1.96 and
   |RC | ≤ 0.95 indicated percentages attributed to drift. This analysis
   was iteratively applied across all bins, with individual process
   fractions weighted by the relative abundance of each bin. These results
   were then synthesized to estimate the relative significance of each
   process at the community-wide level.

Zebrafish plasma extraction and total amino acids determination

   The zebrafish were raised until 110 dpf to ensure sufficient body size
   for blood collection. For this study, total amino acids in the plasma
   were assessed. Each fish was anesthetized using MS-222 (Sigma-Aldrich)
   prior to sacrifice. Once fully anesthetized, blood samples were
   collected from the heart using either a glass capillary or a 10 µl
   pipette tip. The collected blood was immediately transferred into tubes
   pre-rinsed with heparin-sodium to prevent coagulation. The samples were
   then centrifuged at 3000 rpm (850 g) for 30 minutes at 4 °C to separate
   plasma. The supernatant (plasma) was carefully collected and stored for
   analysis. The total amino acid concentration in the plasma was measured
   using the Total Amino Acid Assay Kit (Jiancheng, Nanjing, China),
   following the manufacturer’s instructions.

UHPLC-MS/MS analysis of amino acids in muscle

   Each of the 21 amino acid standards (10 mg) was dissolved in 0.1 mol/L
   HCl and diluted to 1 mL to obtain individual stock solutions. These
   were combined and serially diluted with 25 mM TCA to prepare a series
   of working solutions. Each working solution was then mixed in equal
   volume with a 400 ng/mL internal standard mixture to generate
   calibration standards, which were transferred into 1.5 mL EP tubes for
   LC-MS analysis.

   Accurately weigh 20 mg of tissue, add 141 μL of water and 100 μL of
   0.15% DOC (deoxycholic acid), and mix thoroughly. Then add 4 μL of an
   internal standard solution (100 μg/mL Lys-d4/Try-d5/Gln-d4), mix well,
   and ultrasonicate for 10 minutes (5 °C, 40 KHz). Add 5 μL of 10 M
   trichloroacetic acid (TCA), mix, and freeze the sample to precipitate
   proteins for 10 minutes. Centrifuge at 14,000 rcf for 10 minutes at
   4°C, collect 25 μL of the supernatant, and add 375 μL of water. Vortex
   to mix thoroughly, and filter through a 0.2 μm PTFE membrane (Biotage,
   Sweden).

   The LC-MS/MS analysis of the samples was conducted on an ExionLC AD
   system coupled with a QTRAP® 6500+ mass spectrometer (Sciex, USA) at
   Majorbio Bio-Pharm Technology Co. Ltd. (Shanghai, China). In summary,
   samples were separated using an AdvanceBio MS Spent Media column (2.1 ×
   50 mm, 2.7 µm) at 40°C. Metabolite separation was performed at a flow
   rate of 0.5 mL/min using a gradient mobile phase composed of 95%
   acetonitrile in water with 0.1% formic acid and 10 mM ammonium formate
   (solvent A) and 95% acetonitrile in water with 0.1% formic acid and
   10 mM ammonium formate (solvent B), over a 6-minute run time. The
   gradient was as follows: 0-3 min, from 90% to 60% B; 3-4 min,
   maintained at 60% B; 4-4.1 min, from 60% to 90% B; 4.1-6.0 min,
   maintained at 90% B. Samples were kept at 4 °C during analysis.

   The UHPLC system, paired with the QTRAP® 6500+ mass spectrometer
   equipped with an electrospray ionization (ESI) source, collected the
   mass spectrometric data in both positive and negative ionization modes.
   The operational parameters were as follows: source temperature at
   550 °C; collision-activated dissociation (CAD) gas at medium pressure;
   Ion Source Gas 1 and Gas 2 both at 50 psi; and ion-spray voltage
   floating (ISVF) at +5500 V for positive mode and -4500 V for negative
   mode.

Untargeted metabolomics profiling of muscle

   50 mg of muscle tissue was precisely weighed into a 2 mL
   microcentrifuge tube with a 6 mm stainless steel bead. Then, 400 μL of
   methanol:water (4:1, v/v) containing internal standards (e.g.,
   L-2-chlorophenylalanine, 0.02 mg/mL) was added. Samples were
   homogenized at –10 °C, 50 Hz for 6 minutes, followed by ultrasonic
   extraction at 5 °C, 40 kHz for 30 minutes. After incubation at -20 °C
   for 30 minutes, the samples were centrifuged at 13,000 × g, 4 °C for
   15 minutes. The supernatant was transferred to autosampler vials with
   inserts for LC-MS analysis. A pooled QC sample was prepared by
   combining 20 μL of supernatant from each sample. LC-MS analysis was
   performed on a UHPLC-Q Exactive HF-X system (Thermo Fisher Scientific,
   USA) using an ACQUITY UPLC HSS T3 column (100 × 2.1 mm, 1.8 μm; Waters,
   USA) at 40°C. The mobile phases were: A (95% water + 5% acetonitrile,
   0.1% formic acid) and B (47.5% acetonitrile + 47.5% isopropanol + 5%
   water, 0.1% formic acid). Injection volume was 3 μL. Mass spectra were
   acquired in both positive and negative ion modes with electrospray
   ionization. Key parameters: spray voltage ±3500 V, capillary temp
   325 °C, heater temp 425 °C, sheath gas 50 arb, aux gas 13 arb, S-lens
   RF 50, scan range 70–1050 m/z, resolution 60,000 (full MS) and 7,500
   (MS/MS), collision energy 20, 40, and 60.

RNA-seq analysis

   RNA was extracted from both control and ace2^−/− larvae at 16 days
   post-fertilization (dpf). Total RNA was extracted using MagPure
   Universal RNA Kit (Magen, China). RNA purity and integrity were
   monitored using a NanoPhotometer spectrophotometer (Implen, USA) and
   Agilent 2100 (Agilent, USA), respectively. The mRNA was purified from
   total RNA using poly-T oligo-attached magnetic beads. Paired-end
   sequencing libraries were constructed using the VAHTS Universal V6
   RNA-seq Library Kit for Illumina (Vazyme, China) with unique index
   codes and sequenced with 2×150 bp chemistry on the Illumina NovaSeq X
   Plus platform (Illumina, USA).

   Low-quality and sequencing adapter-contaminated RNA-seq reads were
   filtered using fastp (v0.23.4)^[202]56 with parameters “-n 0 -q
   20”.RNA-seq reads were aligned to the reference genome of
   D.rerio(GenBank: GCA_000002035.4) using HISAT2 (v2.1.0)^[203]57. The
   featureCounts(v2.0.2)^[204]58 software was employed to enumerate the
   number of reads aligned to each gene. The gene abundance was calculated
   and normalized using FPKM (fragments per kilobase per million fragments
   mapped). DEGs (differentially expressed genes) between groups were
   determined using the Deseq2 package^[205]59. Differential gene
   expression analysis was conducted with a fold change criterion of ≥2 or
   ≤ 0.5, and P-value < 0.05 was considered significant. Subsequently, the
   KEGG pathways (Kyoto Encyclopedia of Genes and Genomes) were analyzed
   for enrichment. Raw reads are accessible at the NCBI database under
   BioProjectID: PRJNA1248424.

Statistics and reproducibility

   Venn analysis was used to compare the OTU numbers in zebrafish gut
   between the Control and Mut groups. The comparison of α-diversity
   indices between these two groups was conducted using the Student’s
   t-test, along with an analysis of taxonomic composition at the species
   level. LEfSe was used to determine the bacterial indicators in the gut
   of the Mut zebrafish at the species level. To assess differences in the
   community structure between these two groups based on the Bray Curtis
   distance, the NMDS and Anosim were employed. The PCA was employed to
   analyze the similarity of predicted function content (level 3) between
   these two groups, and then the Student’s t-test was used to identify
   differentially functional genes.

   Sample numbers in the figure legends were selected to satisfy
   statistical power, ensure experimental robustness, and reflect
   available resources. Each experimental group comprised at least three
   independent phenotypic replicates. All values are presented as the
   mean ± SEM, p values and statistical significance (p < 0.05) were
   analyzed by Student’s t-test or one-way ANOVA using Prism 10 (GraphPad
   Prism, San Diego, CA). The significance levels are indicated in the
   figures by exact p values or asterisk (*p < 0.05; **p < 0.01;
   ***p < 0.001; p > 0.05 or ns, not significant). Sample sizes correspond
   to independent biological replicates, and all experiments were
   conducted at least twice. All individual data values plotted in the
   figures were contained in Supplementary Date (Supplementary
   Date [206]1).

Reporting summary

   Further information on research design is available in the [207]Nature
   Portfolio Reporting Summary linked to this article.

Supplementary information

   [208]Supplementary information^ (3.4MB, pdf)
   [209]42003_2025_8626_MOESM2_ESM.docx^ (13KB, docx)

   Description of Additional Supplementary Files Blank
   [210]Supplementary Data 1^ (119.6KB, xlsx)
   [211]Reporting Summary^ (1.3MB, pdf)

Acknowledgements