Stc2a inhibits IGF-stimulated somatic growth in favor of organismal survival under hypoxic stress

TL;DR

This study shows how a protein called Stc2a helps animals survive low-oxygen conditions by reducing growth and redirecting energy to essential functions.

Contribution

The study identifies Stc2a as a novel factor that regulates the trade-off between growth and survival under hypoxia.

Findings

Stc2a inhibits IGF signaling by blocking pappalysin metalloproteinases.

Hypoxia-induced Stc2a expression is reduced in Hif2-deficient zebrafish.

Stc2a deletion increases growth but reduces survival under hypoxia.

Abstract

In response to hypoxia, animals reduce somatic growth to shift energy resources toward the maintenance of vital functions and organismal survival. Although this phenomenon is widespread, the systemic factors and mechanisms involved remain poorly understood. Here we report that hypoxia causes major changes in zebrafish transcriptomic landscapes with hormonal activity or hormonal signaling identified as most prominently up-regulated GO term and KEGG pathway. Among the top in this group is Stanniocalcin 2a (Stc2a), a secreted glycoprotein that inhibits insulin-like growth factor (IGF) signaling by binding to pappalysin metalloproteinases and inhibiting their activities. The hypoxic induction of stc2a expression is attenuated in Hif2-deficient fish. Genetic deletion of Stc2a increased the developmental speed and growth rate, resulting in enlarged adult organ and body size. Under hypoxia, stc2a-/- fish grew faster than wild-type fish but showed reduced survival rate. These phenotypes were reversed by inhibiting pappalysin metalloproteinase activity and by blocking IGF signaling. These findings suggest that Stc2a limits IGF-mediated somatic growth in favor of survival and that the induction of Stc2a is part of a conserved mechanism regulating the trade-off between somatic growth and organismal survival under hypoxic stress.

Figures

RNA-seq analysis of hypoxia-treated and normoxia control zebrafish. (A) Volcano plot of differentially expressed genes (DEGs) between hypoxia and normoxia groups. Significance on y-axis as -log10 (p-value) and effect size on x-axis as log2 (fold change). Dotted lines represent cutoffs of adjusted p-value < 0.05 and |log2(fold change)| > 0.6. Up-regulated genes under hypoxia are shown in red, down-regulated in blue, and non-significant change genes in gray. (B) Hierarchical clustering of the DEGs in the normoxia (N) and hypoxia (H) groups. Rows represent individual genes, and columns represent biological replicates from hypoxia and normoxia groups. Gene expression values were normalized and scaled by row (z-scores), with red indicating higher expression and blue indicating lower expression relative to the mean. (C) qRT-PCR confirmation of RNA-seq data. Changes (log2) in the mRNA levels of the 20 genes measured by RNA-seq were plotted against those detected by qPCR. The line indicates the linear correlation between the results of RNA-seq and qPCR. (D, F) KEGG pathway enrichment for up-regulated (D) and down-regulated (F) DEGs. (E, G) GO molecular function enrichment analysis for up-regulated (E) and down-regulated (G) DEGs. Dot size indicates the number of genes mapped to each term, and dot color reflects the statistical significance (adjusted p-value).

Identification of stc2a as a hypoxia-inducible gene. (A) List of genes enriched in the GO term hormone activity. mRNA abundance is ranked by mean transcripts per million (TPM) in the hypoxia group. (B) qRT-PCR analysis results of stc2a mRNA expression at the indicated developmental stages. hpf, hour post fertilization. Data are shown as mean ± SEM. n = 10-15. (C) Relative stc2a mRNA expression extracted from Zebrahub scRNA-seq database. (D) Levels of stc2a mRNA in the indicated adult tissues. RNA was extracted from the indicated tissues and analyzed by RT-qPCR analysis and normalized by 18s rRNA. Data are shown as mean ± SEM. n = 3. (E) qPCR analysis result of stc2a mRNA expression. Fish were subjected to hypoxia at the indicated air O2 levels and periods. RNA was isolated and stc2a mRNA levels examined by RT-qPCR and normalized by β-actin mRNA levels. Note: Because 3% O2 is lethal for advanced larvae, 7% O2 was used in the 96–120 hpf group. Data are shown as mean ± SEM. *p < 0.05, p < 0.01, *p < 0.001, t-test. n = 3-4. (F) Hypoxia-induced stc2a mRNA expression in Hif2a-deficient and wild-type (WT) fish. RNA was isolated and stc2a mRNA levels examined by RT-qPCR, normalized by β-actin mRNA levels.

Loss of Stc2a increases developmental speed and growth rate, resulting in enlarged adult organ and body size. (A) Gross morphology of fish of the indicated genotypes at the indicated stages. Lateral views with anterior to the left and dorsal up. Scale bar = 0.2mm. dpf, day post fertilization. (B, C). Body length (B) and head-trunk angle (HTA) (C) of the indicated mutant fish and wild-type (WT) zebrafish at 51 hpf. WT, n = 76, stc2a(Δ2 + 4)-/-, n = 73, and stc2a(Δ5)-/-, n = 10. p < 0.0001 by one-way ANOVA followed by Tukey’s multiple comparisons test. (D) The somite number of the indicated fish line at 26 hpf. WT, n = 40, stc2a(Δ2 + 4)-/-, n = 25, and stc2a(Δ5)-/-, n = 14. p < 0.0001 by one-way ANOVA followed by Tukey’s multiple comparisons test. (E, F) Body length (E) and body weight (F) of the indicated genotype of 6 month-old. WT, n = 16, stc2a(Δ2 + 4)+/-, n = 14, and stc2a(Δ5)-/-, n = 15. p < 0.01, p < 0.0001 by one-way ANOVA followed by Tukey’s multiple comparisons test. (G–H) Dorsal spine length (G) and pelvic spine length (H) of the indicated genotypes of 1 year-old. Fish were stained by alizarin red and a representative image is shown on the left and quantitative results on the right. *p < 0.05 by unpaired two-tailed t-test. stc2a(Δ2 + 4)+/- n = 6, and stc2a(Δ5)-/-, n = 8. In all above panels, data shown are mean ± SEM.

Stc2a regulates somatic growth in an IGF signaling-dependent manner. (A) Whole body phospho-Akt levels. Lysates of pooled wild-type (WT) and stc2a(Δ2 + 4)-/- larvae (3 dpf) were analyzed by Western blotting using antibodies against phospho-Akt, total Akt, and GAPDH. Representative images are shown in the left panel; quantified results are shown in the middle and right panels. n = 7, ns, no significant, t -test. (B) The same samples described in (A) were analyzed using phospho-ERK, total ERK levels, and GAPDH. n = 7, ns, no significant, t-test. (C) Fish of the indicated genotypes were treated with 1.5 μM BMS-754807 or vehicle from 6–51 hpf. Body length was measured individually. Percent decrease caused by BMS treatment was calculated and shown. n = 58~68. , P < 0.001, t-test. (D) Fish of the indicated genotypes were treated with 8 μM ZnCl2 from 6–51 hpf. Body length was measured individually. Percent decrease caused by ZnCl2 was calculated and shown. n = 36~43. , P <0.01, t-test. (E, F) Fish of the indicated genotypes were treated with 0.3 μM wortmannin (E) or 8 μM U0126 or vehicle from 6–51 hpf. Body length was measured individually. Percent decrease caused by each drug was calculated and shown. n = 37-72. p < 0.0001, t-test. In all above panels, data shown are mean ± SEM.

Stc2a regulates the growth-survival trade-off under hypoxia. (A, B) Loss of Stc2a attenuates hypoxia-induced growth retardation and developmental delay. Fish of the indicated genotypes were subjected to hypoxia (filled bar) or normoxia (open bar) from 25–51 hpf. Body length (A) and head-trunk angle (HTA) (B) were measured and shown as mean ± SEM. n = 10~20. *p < 0.05, p < 0.01, *p < 0.001, p < 0.0001, t-test. (C, D) Fish of the indicated genotypes were subjected to hypoxia treatment from 25–51 hpf in the presence or absence of 8 μM ZnCl2. Body length (C) and HTA (D) was measured and shown as mean ± SEM. n = 58~73. *p < 0.001, p < 0.0001, One-way ANOVA with multiple comparisons. (E) Fish (5 dpf) of the indicated genotypes were exposed to hypoxia (3% O2) and survival rates were assessed and shown. p < 0.0001, Mantel-Cox log rank test. n = 42-100. (F) Fish (5 dpf) of the indicated genotypes were exposed to hypoxia (3% O2) in the presence or absence of 8 μM ZnCl2. The number of survival fish were assessed at the indicated time. p < 0.01, ns, not significant. Mantel-Cox log rank test, n = 45-90.

Proposed model. Hypoxia activates Hif-dependent gene expression, leading to metabolic shift towards glycolysis, protein breakdown, gluconeogenesis, and increases in hormonal signaling. An important hypoxia-induced hormonal factor is Stc2a. The increased Stc2a restricts growth and redirects toward critical survival processes by modulation of pappalysin metalloproteinase activity and IGF signaling. Created with BioRender.com.

Full text

Introduction

Hypoxia poses a fundamental bioenergetic challenge to animals and their cells. Hypoxia has been linked to human diseases, such as ischemia, inflammation, tumorigenesis, and intrauterine growth restriction. In response to low oxygen levels, cells shift their metabolism toward glycolysis and reduce overall metabolic demands by altering gene expression, primarily through the actions of hypoxia-inducible factors (HIFs) (1). HIFs, including HIF1, HIF2, and HIF3, are dimeric proteins composed of an oxygen-regulated α subunit and a constitutive β subunit (1, 2). Under normoxic conditions, HIF-α subunits are hydroxylated by prolyl hydroxylase domain proteins (PHDs) and targeted for degradation by the von Hippel–Lindau (VHL) E3 ubiquitin ligase complex (3). During hypoxia, hydroxylation is inhibited, resulting in the stabilization and accumulation of HIF-α subunits, which then translocate to the nucleus, dimerize with HIF-β, and bind to hypoxia response elements (HREs) in target genes, thereby enhancing their transcription (4–6). At the whole-organism level, animals further prioritize energy expenditure by reallocating resources to sustain essential functions, such as those of the brain and heart, while reducing investment in non-essential activities like growth and reproduction. This phenomenon, observed across diverse taxa, suggests the existence of evolutionarily conserved regulatory mechanisms (7). Despite this, the systemic factors and molecular mechanisms that orchestrate organism-wide responses to hypoxia remain poorly understood.

Zebrafish (Danio rerio) is a valuable vertebrate model for dissecting systemic factors and mechanisms underlying somatic growth and body size regulation, owing to its small size, rapid development, and short generation time. Previous studies in zebrafish have demonstrated that hypoxia retards developmental and growth rates by attenuating insulin-like growth factor (IGF) signaling (8–12). IGFs are evolutionarily ancient polypeptides structurally related to insulin. The actions of IGFs are mediated through the IGF-1 receptor (IGF-1R), a receptor tyrosine kinase. Ligand binding induces IGF-1R tyrosine phosphorylation, activating major intracellular signaling cascades such as PI3K-AKT-mTOR and RAS-MEK-ERK pathways, which drive cell proliferation, growth, differentiation, and survival (13, 14). The IGF system further comprises six types of high-affinity IGF binding proteins (IGFBP1–6), which sequester IGFs and inhibit their interaction with IGF-1R, thereby modulating IGF signaling (15) (16). Some IGFBPs also possess IGF-independent biological activities (17–19). The bioavailability of IGFs is tightly regulated by IGFBP proteases such as the pappalysin family members, including pregnancy-associated plasma protein-A (PAPP-A) and PAPP-A2. These enzymes cleave IGFBPs and release IGFs from the IGFBP complexes and make them available for receptor binding (20, 21). Recent findings have shown that PAPP-A-mediated IGFBP proteolysis is further modulated by stanniocalcin-1 (STC1) and stanniocalcin-2 (STC2) (16, 22, 23), both of which are potent PAPP-A and PAPP-A2 inhibitors (20, 21). While STC1 binds PAPP-A non-covalently, STC2 interacts covalently and irreversibly to PAPP-A/A2, leading to decreased IGF bioavailability and reduced IGF-1R signaling (22). Notably, human loss-of-function mutations in STC2 have been associated with increased adult height and enhanced local IGF signaling (22–24).

The objective of this study is to investigate hypoxia-induced transcriptomic changes using zebrafish and to identify novel systemic factors and mechanisms regulating the trade-off between somatic growth and organismal survival under hypoxic stress. We found that hypoxia caused major changes in the transcriptome landscapes and the most prominently up-regulated pathway is hormonal activity, with stc2a among the top. Our genetic, physiological, and pharmacological evidence indicates that hypoxic induction of Stc2a represents an adaptive mechanism that restricts IGF-stimulated somatic growth, redirecting energy toward critical survival processes in hypoxia.

Results

Hypoxia results in major changes in hormonal signaling and metabolism

RNA-seq analysis of hypoxia-treated and control zebrafish larvae detected a total of 2411 differentially expressed genes (DEGs), including 703 up-regulated and 1708 down-regulated DEGs (Figure 1A). The two groups are distinctly clustered in hierarchical clustering (Figure 1B) and principal component analysis (Supplementary Figure S1A). The RNA-seq results were confirmed by qRT-PCR assays using a different set of biological samples (Figure 1C) and similar changes were detected in 18 out of 20 genes tested (Supplementary Figure S1B).

RNA-seq analysis of hypoxia-treated and normoxia control zebrafish. (A) Volcano plot of differentially expressed genes (DEGs) between hypoxia and normoxia groups. Significance on y-axis as -log10 (p-value) and effect size on x-axis as log2 (fold change). Dotted lines represent cutoffs of adjusted p-value < 0.05 and |log2(fold change)| > 0.6. Up-regulated genes under hypoxia are shown in red, down-regulated in blue, and non-significant change genes in gray. (B) Hierarchical clustering of the DEGs in the normoxia (N) and hypoxia (H) groups. Rows represent individual genes, and columns represent biological replicates from hypoxia and normoxia groups. Gene expression values were normalized and scaled by row (z-scores), with red indicating higher expression and blue indicating lower expression relative to the mean. (C) qRT-PCR confirmation of RNA-seq data. Changes (log2) in the mRNA levels of the 20 genes measured by RNA-seq were plotted against those detected by qPCR. The line indicates the linear correlation between the results of RNA-seq and qPCR. (D, F) KEGG pathway enrichment for up-regulated (D) and down-regulated (F) DEGs. (E, G) GO molecular function enrichment analysis for up-regulated (E) and down-regulated (G) DEGs. Dot size indicates the number of genes mapped to each term, and dot color reflects the statistical significance (adjusted p-value).

KEGG analysis indicated that the up-regulated DEGs are highly enriched in the glycolysis/gluconeogenesis pathway, followed by carbon metabolism, amino acid synthesis etc. (Figure 1D). The topmost enriched pathways, however, are “neuroactive ligand signaling” and “hormonal signaling” (Figure 1D). In addition, “calcium signaling”, “FoxO signaling”, and “insulin signaling” pathways are also enriched (Figure 1D). In agreement, GO (Molecular Function) analysis identified “signaling receptor regulator activity” as the top enriched term in the up-regulated DEGs (Figure 1E). “Hormone activity”, largely overlapping with “signaling receptor regulator activity “, is also enriched (Figure 1E). Other enriched GO terms are oxidoreductase activity, iron ion binding, dioxygenase activity, FAD binding etc. (Figure 1E). The enriched down-regulated KEGG pathways include biosynthesis of co-factor, cornified envelope formation, cytoskeleton in muscle cells, focal adhesion etc. (Figure 1F). The top enriched down-regulated GO terms are endopeptidase activity, glycosyltransferase activity, peptidase regulator activity, iron ion binding, enzyme inhibitor activity etc. (Figure 1G). GO CNET plot of the up-regulated DEGs revealed the top five most significantly enriched molecular function terms are dioxygenase activity, 2-oxoglutarate-dependent dioxygenase activity, hormone activity, oxidoreductase activity acting on paired donors with incorporation or reduction of oxygen, and FAD binding. Correspondingly, the top five enriched terms among down-regulated DEGs were endopeptidase activity, peptidase regulator activity, endopeptidase regulator activity, peptidase inhibitor activity, and monooxygenase activity (Supplementary Figures S1C, D).

Hypoxic induction of stc2a expression

We further analyzed the RNA-seq dataset by ranking the up-regulated DEGs in hormone activity according to their mRNA abundance (Figure 2A). Among the top on the list is stc2a, which has been implicated in human body height regulation and IGF signaling (22–24) During early development, stc2a mRNA levels increased gradually and reached the plateau at 4 days post fertilization (dpf) (Figure 2B). Data mining from the Zebrahub scRNA-seq database (25), indicated that stc2a mRNA is detected in multiple tissues from 0 somite stages to 10 dpf, and is particularly abundant in the central nervous system, periderm, neural crest, paraxial mesoderm (Figure 2C). In the adult stage, the highest stc2a mRNA levels were detected in the brain, followed by kidney, eyes, and fin. (Figure 2D). Importantly, hypoxia increased stc2a mRNA levels at multiple developmental stages (Figure 2E). The hypoxia-induced stc2a mRNA expression was attenuated in hif2a^-/-^ deficient fish (Figure 2E), suggesting that stc2a expression is induced by hypoxia, likely through the action of Hif2.

*Identification of stc2a as a hypoxia-inducible gene. (A) List of genes enriched in the GO term hormone activity. mRNA abundance is ranked by mean transcripts per million (TPM) in the hypoxia group. (B) qRT-PCR analysis results of stc2a mRNA expression at the indicated developmental stages. hpf, hour post fertilization. Data are shown as mean ± SEM. n = 10-15. (C) Relative stc2a mRNA expression extracted from Zebrahub scRNA-seq database. (D) Levels of stc2a mRNA in the indicated adult tissues. RNA was extracted from the indicated tissues and analyzed by RT-qPCR analysis and normalized by 18s rRNA. Data are shown as mean ± SEM. n = 3. (E) qPCR analysis result of stc2a mRNA expression. Fish were subjected to hypoxia at the indicated air O2 levels and periods. RNA was isolated and stc2a mRNA levels examined by RT-qPCR and normalized by β-actin mRNA levels. Note: Because 3% O2 is lethal for advanced larvae, 7% O2 was used in the 96–120 hpf group. Data are shown as mean ± SEM. *p < 0.05, **p < 0.01, **p < 0.001, t-test. n = 3-4. (F) Hypoxia-induced stc2a mRNA expression in Hif2a-deficient and wild-type (WT) fish. RNA was isolated and stc2a mRNA levels examined by RT-qPCR, normalized by β-actin mRNA levels.

Loss of Stc2a increases growth of adult organ and body size

Using CRISPR-Cas9, two mutant zebrafish lines, stc2a(Δ2 + 4)^-/-^ and stc2a(Δ5)^-/-^, were generated for functional analysis. Both are predicted to be null mutations. In both mutant lines, there was a significant reduction in the stc2a mRNA levels, whereas no such changes were found with the mRNA levels of the closely related stc2b (Supplementary Figure S2B). Both mutant fish lines survived to adulthood and reproduced well under standard conditions. The gross morphology of these stc2a^-/-^ fish was indistinguishable from their siblings (Figure 3A). Compared with wild-type fish, however, the body length of stc2a(Δ2 + 4)^-/-^ and stc2a(Δ5)^-/-^ larvae was significantly greater (Figure 3B). Likewise, the head-trunk angle (HTA) and somite number, two parameters of zebrafish developmental speed (73), were significantly greater (Figures 3C, D), suggesting these mutant fish develop more rapidly and grow faster. Next, stc2a(Δ2 + 4)^+/-^ fish were intercrossed and the offsprings were grew under the same conditions for 6 months. Compared with their wild-type and heterozygous siblings, stc2a(Δ2 + 4)^-/-^ fish had greater body length and body weight (Figures 3E, F). Likewise, the lengths of the pelvic and dorsal spines in 1 year-old stc2a(Δ2 + 4)^-/-^ fish were greater than those of the siblings (Figures 3G, H). These results suggest that Stc2a negatively regulates somatic growth, organ size, and body size in adult fish.

*Loss of Stc2a increases developmental speed and growth rate, resulting in enlarged adult organ and body size. (A) Gross morphology of fish of the indicated genotypes at the indicated stages. Lateral views with anterior to the left and dorsal up. Scale bar = 0.2mm. dpf, day post fertilization. (B, C). Body length (B) and head-trunk angle (HTA) (C) of the indicated mutant fish and wild-type (WT) zebrafish at 51 hpf. WT, n = 76, stc2a(Δ2 + 4)-/-, n = 73, and stc2a(Δ5)-/-, n = 10. ****p < 0.0001 by one-way ANOVA followed by Tukey’s multiple comparisons test. (D) The somite number of the indicated fish line at 26 hpf. WT, n = 40, stc2a(Δ2 + 4)-/-, n = 25, and stc2a(Δ5)-/-, n = 14. ****p < 0.0001 by one-way ANOVA followed by Tukey’s multiple comparisons test. (E, F) Body length (E) and body weight (F) of the indicated genotype of 6 month-old. WT, n = 16, stc2a(Δ2 + 4)+/-, n = 14, and stc2a(Δ5)-/-, n = 15. **p < 0.01, ****p < 0.0001 by one-way ANOVA followed by Tukey’s multiple comparisons test. (G–H) Dorsal spine length (G) and pelvic spine length (H) of the indicated genotypes of 1 year-old. Fish were stained by alizarin red and a representative image is shown on the left and quantitative results on the right. p < 0.05 by unpaired two-tailed t-test. stc2a(Δ2 + 4)+/- n = 6, and stc2a(Δ5)-/-, n = 8. In all above panels, data shown are mean ± SEM.

Since genetic deletion of stc1a, a structurally related gene, resulted in increased ionocyte cell number, ectopic calcium deposits, kidney stone-like calcium deposits, and reduced bone mineralization (26, 27), we crossed stc2a(Δ5)^-/-^ fish with Tg (igfbp5a:GFP) fish, a stable transgenic line expressing EGFP in calcium transporting ionocytes or NaR cells (28), and measured the NaR cell number. There were no differences among the different genotypes (Supplementary Figures S3A, B). Alizarin red staining of the juvenile and adult stc2a(Δ2 + 4)^-/-^ fish did not detect notable differences in calcified tissues (Supplementary Figures S3C, D), suggesting that Stc2a does not play an indispensable role in regulating ionocyte proliferation, calcium balance, bone mineralization nor kidney development or function.

Stc2a regulates somatic growth via the Pappalysin-Igfbp-Igf signaling axis

We postulated that Stc2a may regulate somatic growth by inhibiting the pappalysin family metalloproteinase (including PAPP-A and PAPP-A2) and IGF signaling activity. In zebrafish, there are 3 pappalysin family members, including Papp-aa, Papp-ab, and Papp-a2 (29, 30). Currently, there is no specific antibodies against any of these proteins. qRT-PCT analysis showed the levels of papp-aa and papp-a2 mRNA levels were similar between wild-type fish and stc2a^-/-^ fish (Supplementary Figure S4A). Although there was a markedly increase in papp-ab mRNA levels, this change was not statistically different (Supplementary Figure S4). Hypoxia did not alter their mRNA levels of papp-aa and papp-ab, while it caused a modest but statistically significant decrease in papp-a2 mRNA levels. (Supplementary Figure S4). Commercially antibodies against mammalian phospho-IGF1 receptor did not yield specific signal in zebrafish larvae. Western blotting of the whole body lysates did not detect major differences in phospho-Akt and phospho-Erk levels between stc2a(Δ2 + 4)^-/-^ and wild-type fish (Figures 4A, B). The low sensitivity of the assay and the lack of tissue/cell resolution may obscure the results.

*Stc2a regulates somatic growth in an IGF signaling-dependent manner. (A) Whole body phospho-Akt levels. Lysates of pooled wild-type (WT) and stc2a(Δ2 + 4)-/- larvae (3 dpf) were analyzed by Western blotting using antibodies against phospho-Akt, total Akt, and GAPDH. Representative images are shown in the left panel; quantified results are shown in the middle and right panels. n = 7, ns, no significant, t -test. (B) The same samples described in (A) were analyzed using phospho-ERK, total ERK levels, and GAPDH. n = 7, ns, no significant, t-test. (C) Fish of the indicated genotypes were treated with 1.5 μM BMS-754807 or vehicle from 6–51 hpf. Body length was measured individually. Percent decrease caused by BMS treatment was calculated and shown. n = 5868. ****, P < 0.001, t-test. (D) Fish of the indicated genotypes were treated with 8 μM ZnCl2 from 6–51 hpf. Body length was measured individually. Percent decrease caused by ZnCl2 was calculated and shown. n = 3643. **, P <0.01, t-test. (E, F) Fish of the indicated genotypes were treated with 0.3 μM wortmannin (E) or 8 μM U0126 or vehicle from 6–51 hpf. Body length was measured individually. Percent decrease caused by each drug was calculated and shown. n = 37-72. ***p < 0.0001, t-test. In all above panels, data shown are mean ± SEM.

We therefore tested the possible role of IGF signaling using BMS-754807, a potent IGF1R inhibitor that has been previously tested in zebrafish (11, 31–33). If Stc2a regulates somatic growth by inhibiting IGF signaling, then inhibition of IGF signaling should reverse these phenotypes. Treatment with BMS-754807 decreased body lengths in both stc2a(Δ2 + 4)^-/-^ fish and wild-type control fish. The magnitude of decrease, however, was significantly greater in the stc2a((Δ2 + 4)^-/-^ fish (Figure 4C). Likewise, ZnCl_2_, which inhibits pappalysin-mediated Igfbp degradation in human cells and in zebrafish (34), caused a greater percent decrease in body length in the stc2a((Δ2 + 4)^-/-^ larvae compared to wild-type siblings (Figure 4D). To further test the role of IGF signaling, the mutant fish were treated with wortmannin, a PI3K inhibitor (35), and U0126, a MAPK inhibitor (36). Both inhibitors decreased body length in the wild-type as well as stc2a(Δ2 + 4)^-/-^ fish. Again, the magnitude of decrease was significantly more pronounced in the stc2a((Δ2 + 4)^-/-^ fish (Figures 4E, F).

stc2a-/- fish grew faster but had low survival rate under hypoxia

Hypoxia caused growth retardation and developmental delays in wild-type zebrafish embryos (Figure 5A) (8, 11). Although hypoxia reduced the body length and HTA in stc2a(Δ2 + 4)^-/-^ (Figure 5A), the mutant fish still grew bigger compared to wild-type siblings under hypoxia (Figure 5A). Likewise, while hypoxia lowered HTA values in both genotypes (Figure 6B), the mutant fish developed faster (Figure 5B). Treatment of mutant fish with ZnCl_2_ reduced the body length and HTA value to the levels of wild-type fish (Figures 5C, D). We noted that mutant fish were prone to die under hypoxia and quantified their survival rate under hypoxia. They were significantly lower compared to the wild-type fish (Figure 5E, Supplementary Table S1). Addition of ZnCl_2_ restored the survival rate to the levels of wild-type control groups (Figure 5F, Supplementary Table S1), suggesting that pappalysin enzyme activity is required in the elevated growth and mortality in stc2a^-/-^ mutant fish under hypoxic stress.

*Stc2a regulates the growth-survival trade-off under hypoxia. (A, B) Loss of Stc2a attenuates hypoxia-induced growth retardation and developmental delay. Fish of the indicated genotypes were subjected to hypoxia (filled bar) or normoxia (open bar) from 25–51 hpf. Body length (A) and head-trunk angle (HTA) (B) were measured and shown as mean ± SEM. n = 1020. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, t-test. (C, D) Fish of the indicated genotypes were subjected to hypoxia treatment from 25–51 hpf in the presence or absence of 8 μM ZnCl2. Body length (C) and HTA (D) was measured and shown as mean ± SEM. n = 5873. ***p < 0.001, ****p < 0.0001, One-way ANOVA with multiple comparisons. (E) Fish (5 dpf) of the indicated genotypes were exposed to hypoxia (3% O2) and survival rates were assessed and shown. ****p < 0.0001, Mantel-Cox log rank test. n = 42-100. (F) Fish (5 dpf) of the indicated genotypes were exposed to hypoxia (3% O2) in the presence or absence of 8 μM ZnCl2. The number of survival fish were assessed at the indicated time. *p < 0.01, ns, not significant. Mantel-Cox log rank test, n = 45-90.

Proposed model. Hypoxia activates Hif-dependent gene expression, leading to metabolic shift towards glycolysis, protein breakdown, gluconeogenesis, and increases in hormonal signaling. An important hypoxia-induced hormonal factor is Stc2a. The increased Stc2a restricts growth and redirects toward critical survival processes by modulation of pappalysin metalloproteinase activity and IGF signaling. Created with BioRender.com.

Discussion

The findings made in this study suggest that Stc2a functions as an important regulator of the somatic growth and organismal survival trade-off under hypoxic stress. We provided in vivo data showing that stc2a expression is induced by hypoxia. Loss of Stc2a increases the growth rate and developmental speed, leading to enlarged adult body size. When exposed to hypoxia, stc2a^-/-^ null animals exhibit accelerated growth but reduced survival. These phenotypes were reversed by inhibiting pappalysin metalloproteinase activity and by blocking IGF signaling, suggesting that these functions are mediated by pappalysin metalloproteinase and IGF signaling (Figure 6).

Stc2a is a member of the STC/Stc glycoprotein family. The first Stc protein was discovered from the Corpuscles of Stannius (CS), an endocrine organ in bony fish (37–39). Subsequent studies indicate that STC proteins are found in a wide range of species ranging from humans to eukaryotes such as Fungi, cnidarians, sponges, nematodes, and that most species have multiple STC genes (40). Humans, for example, have STC1 and STC2, while many teleost fish including zebrafish have 4, including stc1a, stc1b, stc2a, and stc2b (40). Previous studies have shown that loss of Stc1a, while had no effect on zebrafish somatic growth, increased the proliferation of calcium transporting ionocytes, resulting in abnormal calcium uptake, kidney stone formation, cardiac and body edema, and premature death (26, 27). The effect of Stc1a in ionocytes is mediated through its action on Papp-aa-mediated Igfbp5a proteolysis (26, 34). The functions of Stc2a and other Stc proteins, however, have not been previously reported. In this study, we show that genetic deletion of Stc2a increased growth rate and adult body and organ size. This finding agrees with published genetic studies in mice and humans. Stc2 knockout mice exhibit increased body growth (41), while mice overexpressing human STC2 are smaller (42). Human carriers of STC2 loss-of-functional mutation are taller than non-carriers (24). In a genome wide association study, STC2 and its binding partners (i.e., PAPP-A and PAPP-A2) are found in loci associated with human heights (43). Likewise, freshwater stickleback with different alleles with either increased or decreased stc2a expression are associated with decreasing or increasing spine length (44). Together, these findings suggest that STC2/Stc2a functions as a negative growth regulator in a wide range of species.

It has been suggested that STC2 inhibits somatic growth by inhibiting PAPP-A/A2-mediated IGFBP proteolysis and by reducing local IGF signaling (23). While this notion is well supported by in vitro biochemical evidence and clinical observation, there is little direct evidence in vivo. Taking advantage of zebrafish larvae and the availability of stc2a^-/-^ fish lines, we tested the importance of Stc2a-Papp-a-IGF signaling axis on somatic growth in vivo. Our results indicate that the elevated growth observed in Stc2a deficient fish was abolished by ZnCl_2_ treatment and by pharmacological blockade of the Igf1 receptor-mediated signaling. Likewise, inhibition of PI3 kinase and MAP kinase signaling reduced the growth rate of stc2a^-/-^ fish to the wild-type levels. These data have provided strong in vivo evidence that Stc2a inhibits somatic growth by negatively inhibiting pappalysin family metalloprotease activity and IGF signaling under normoxic conditions.

A new and important finding made in this study is that while loss of Stc2a increases somatic growth, it decreases organismal survival under hypoxic stress. Both actions are mediated by IGF signaling. This conclusion is supported by several lines of evidence. First, zebrafish stc2a is transcriptionally upregulated by hypoxia in vivo. Second, while hypoxia slows down the growth rate, the mutant fish still grew faster than the wild-type fish, suggesting that Stc2a deficient fish have an advantage in somatic growth under low oxygen conditions. Meanwhile, stc2a^-/-^ mutants exhibited increased mortality under hypoxia. Both of these phenotypes were reversed by inhibiting pappalysin metalloproteinase activity. Hypoxia strongly induces the expression of igfbp1a and igfbp1b in zebrafish (8, 10, 45). Morpholino-based knockdown of Igfbp1a partially alleviates hypoxia-induced growth retardation by binding to IGF ligands and inhibiting their interaction with the Igf1 receptor (8). Likewise, the loss of irs2b, but not its paralog irs2a, blunts MAPK-activation and catch-up growth in hypoxia-treated and reoxygenated zebrafish embryos (46). The findings made in this study reveal yet another layer of the hypoxia adaptive response, showing that a specific physiological mechanism (Stc2a-Papp-a-Igfbp) is engaged in lowering IGF signaling to prioritize energy for organismal survival over somatic growth under hypoxic stress (Figure 6). Previous studies have shown that hypoxia causes developmental delay in rapidly developing zebrafish embryos (8). In the present study, larval zebrafish were used and the hypoxia treatment period was shorter and less severe. Therefore, developmental delay is unlikely to have a significant impact in the role of Stc2a found in the present study.

The notion that hypoxic induction of Stc2a inhibits IGF-stimulated somatic growth to divert resources away from growth for survival is in line with our transcriptomic analysis results. Among the top up-regulated KEGG pathways in hypoxia-treated fish is “neuroactive ligand signaling”. This finding, together with the high expression of Stc2a in the adult brain, is consistent with the idea of prioritization of brain function. Hypoxia also caused a metabolic shift towards glycolysis and gluconeogenesis, as indicated by KEGG analysis results. Among the enriched genes are pfkpb3, hk1, pkma, and gpia. Pfkfb3 is a key regulator of glycolysis and plays a crucial role for the metabolic changes seen in rapidly proliferating cancer cells, a phenomenon known as the Warburg effect (47, 48). The hk1 gene encodes hexokinase 1, which catalyzes the first step of glycolysis. There was also a significant upregulation of ucp3 (uncoupling protein 3), which is involved in uncoupling substrate oxidation from the ATP synthesis and reducing oxygen-dependent ATP production (49). Pkma or pyruvate kinase enzyme catalyzes the last step of glycolysis. The gpia gene encodes glucose phosphate isomerase a, which is critical in gluconeogenesis. Hypoxia treatment also resulted in a significant downregulation of gck (glucokinase) and ppp1r3c2a (protein phosphatase 1 regulatory subunit 3C), indicating a shift away from glycogen synthesis (50) One of the most enriched groups of down-regulated DEGs are peptidases-endopeptidases and their regulators/inhibitors. This together with the up-regulation of gluconeogenesis and amino acid synthesis is indicative of change in protein breakdown in these animals to meet the energy demand to a level that can be met by the limited oxygen supply. We speculate these changes help to divert energy from growth to vital function and survival. Future functional studies are needed to test whether these metabolic changes are causal to the observed increase in mortality in stc2a^-/-^ mutant fish.

In conclusion, our results suggest that Stc2a limits IGF-mediated growth in favor of survival under hypoxic stress. STC2 gene has been reported to be up-regulated by hypoxia in culture human cells and mouse retina (43, 51, 52). A meta-analysis of 128 hypoxia-related human RNA-seq datasets found human STC2 upregulated in 73 datasets across varying tissue types and hypoxia severities (53). Therefore, the hypoxic induction of STC2/Stc2a is likely conserved across species. Future studies will be needed to elucidate the functional role of STC2 in the hypoxia response in mammals and humans. There may also be species differences in the HIF isoform(s) involved in the hypoxic regulation of STC2/Stc2 expression. While previous studies in human cell culture systems suggest a possible role of HIF1 in regulating STC2, we found that the hypoxic induction of stc2a expression is impaired in Hif2a-deficient zebrafish, suggesting that the hypoxic induction of stc2a expression may be mediated by Hif2 in zebrafish.

Limitations of the study

Because zebrafish larvae are tiny, several dozen were pooled for Western blotting analysis of phospho-Akt and phospho-Erk levels. No difference was detected between stc2a^-/-^ and wild-type fish. This was likely due to low sensitivity of Western blotting and the lack of resolution. Future studies are needed to develop more sensitive and quantitative assays to detect local IGF signaling activities. In this study, ZnCl_2_ was used as a generic pappalysin metalloproteinase inhibitor. The specific pappalysin metalloproteinase isoform(s) involved with the reported Stc2a action is unclear. qRT-PCR analysis did not detect major changes in papp-aa and papp-a2 mRNA levels in stc2a-/- mutant fish. While there was a trend of increase in papp-ab mRNA levels, it was not statistically significant. The actual protein levels and their activities are currently unclear. Among the 3 zebrafish pappalysin members, Papp-aa could cleave human IGFBP5 and IGFBP4 in vitro, whereas it did not cleave other 4 human IGFBPs (34). Zebrafish Papp-ab has been shown to cleave human IGFBP4 and IGFBP5 but not the other four IGFBPs (29). Zebrafish Papp-a2 can cleave human IGFBP3 and IGFBP5, but not other IGFBPs (30). Zebrafish has a total of 9 igfbp genes, including igfbp1a, 1b, 2a, 2b, 3, 5a, 5b, 6a, and 6b (8, 10, 18, 19, 54–56). The specific Igfbp(s) involved in the Stc2a action reported in this study remains unknown. Future studies will be needed to develop reagents and tools to measure these pappalysin metalloproteinase activities, identify their substrates, and develop conditional knockout and isoform-specific inhibitors to dissect their role(s) in somatic growth and survival trade-off under hypoxia stress.

Materials and methods

Unless noted otherwise, chemical and molecular reagents were purchased from Fisher Scientific (Pittsburgh, PA, United States). Restriction enzymes were purchased from New England Biolabs (Ipswich, MA, United States) or Promega (Madison, WI, United States). TRIzol were purchased from Life Technologies (Carlsbad, CA, United States). Oligo primers were ordered from Integrated DNA Technologies (Coralville, IA, United States). PureLink RNA Mini Kit, PureLink DNase Set, DTT, RNaseOUT Recombinant Ribonuclease Inhibitor, and M-MLV reverse transcriptase were purchased from Invitrogen (Waltham, MA, United States). Anti-GAPDH primary antibody purchased from Proteintech (Rosemont, Illinois, United States). All other primary antibodies purchased from Cell Signaling Technology (Danvers, MA, United States). Secondary antibodies purchased from LI-COR Biosciences (Lincoln, Nebraska, United States). BMS-754807 was purchased from JiHe Pharmaceutica (Beijing, China). Alizarin Red and ZnCl_2_ were purchased from Sigma (St. Louis, MO, USA). Wortmannin was purchased from Calbiochem (Gibbstown, NJ). U0126 was purchased from Selleck Chemicals (Houston, Texas).

Experimental animals

Zebrafish were maintained following standard zebrafish husbandry guidelines (57). All experiments using zebrafish were conducted in line with guidelines approved by the Institutional Animal Care & Use Committee, University of Michigan. Embryos and larvae were raised in standard E3 medium as reported previously (58). 0.003% (w/v) N-phenylthiourea (PTU) was added to the E3 medium to prevent pigmentation when required. Modified low calcium media was prepared following a previously reported protocol (32). In addition to wild-type (WT) fish, Tg(igfbp5a:GFP) fish (28), hif2αbΔ10^-/-^ (e.g., epas1b.2, (59)), stc1a^-/-^ (26), stc2a(Δ2 + 4)^-/-^ and stc2a(Δ5)^-/-^ (this study) were used.

RNA sequencing and differential expression analysis

Zebrafish larvae were subjected to hypoxia (6% O_2_) or normoxia (20.9% O_2_, atmospheric level) from 81 to 96 hours post-fertilization (hpf). Each group comprised three to four biological replicates, with 30 larvae pooled per replicate. Total RNA was extracted using the PureLink™ RNA Mini Kit (Invitrogen) and treated with DNase using the PureLink™ DNase Set (Invitrogen). Independent RNA sample sets (n = 3~4) were submitted for library preparation and sequencing at the Advanced Genomics Core, University of Michigan. Poly(A) RNA libraries were constructed using the NEBNext Poly(A) mRNA Magnetic Isolation Module and NEBNext UltraExpress RNA Library Prep Kit (New England Biolabs). Sequencing was performed on the Illumina NovaSeq X Plus platform, generating 89.5–104.6 million paired-end reads (151 bp) per sample. Raw sequencing reads were processed to remove low-quality bases and adapter sequences using Trimmomatic (v0.39) (60), and read quality was evaluated before and after trimming using FastQC (v0.12.1) (61). Trimmed reads were then aligned to the zebrafish reference genome (GRCz11, Ensembl release 113) using STAR (v2.7.10) (62), with gene annotations obtained from Ensembl (63). Read counts were obtained using featureCounts (v2.0.7) within the Subread package (64). DEseq2 (v1.46.0) was used to analyze the differential expression between the groups (65). Genes with adjusted p-value (padj) < 0.05 and |log_2_ fold change| > 0.6 were considered differentially expressed genes (DEGs). Gene Ontology (GO) and KEGG pathway enrichment analyses of DEGs were performed using the org.Dr.eg.db annotation database (v3.20.0) (Carlson) and the clusterProfiler R package (v4.14.4) (66). Enriched terms with p-value and q-value < 0.05 were considered statistically significant. Top GO terms and KEGG pathways were visualized using the dot plot function, and gene-concept networks (CNETs) were generated using the cnetplot function, both implemented in the clusterProfiler R package. All statistical analyses were conducted using R software (v4.4.1).

Gene expression analysis by Real-time RT-qPCR and data mining

RNA was isolated from a pool of 15~30 zebrafish larvae or from adult zebrafish tissue as reported (67). RNA was reverse-transcribed to cDNA using oligo-dT primers and M-MLV reverse transcriptase (Invitrogen). qPCR was performed using SYBR Green (Bio-Rad) on a StepONE PLUS real-time thermocycler (Applied Biosystems) as previously reported (68). The expression level of a target gene transcript was normalized by 18s rRNA or β-actin mRNA levels. PCR primers were designed based on sequences as described in previous studies, the NCBI Gene database, and by NCBI Primer Blast (27, 69) and are listed in Supplementary Table S2. The spatial expression information of stc2a mRNA was extracted from the Single-cell RNA-seq dataset from ZebraHub (https://zebrahub.org) using Scanpy (v1.11.1) (25). Expression of stc2a mRNA was extracted and averaged across tissue categories defined by annotation. Mean expression values were computed per tissue, and relative expression was calculated by normalizing to the sum of expression across all tissues. Visualization and downstream analysis were performed using Seaborn and Matplotlib.

Generation of stc2a-/- lines by CRISPR/Cas9

Two sgRNAs targeting the stc2a gene were designed using CHOPCHOP (http://chopchop.cbu.uib.no/). Their sequences are: stc2a-gRNA1-oligo1: 5’- GCTGCTGCTCTCCGTATTGG-3’ and stc2a-gRNA3-oligo3: 5’-GGGTGACTCTCGTGCACATC-3’. sgRNA(30–40 ng/ul) mixed with Cas9 mRNA (200–400 ng/ul) were injected into Tg(*igfbp5a:*GFP) at the 1-cell stage (70). The injected F0 fish were raised to adulthood and crossed with Tg(igfbp5a:GFP) fish (28). The DNA of the F1 fish were extracted by fin clipping and analyzed by sanger sequencing. Heterozygous F1 male and female fish were crossed to generate the F2 fish.

Genotyping

Genomic DNA, isolated from adult fin or whole larval lysate, were digested with proteinase K (60 μg/mL) in SZL buffer (50 mM KCl, 2.5 mM MgCl_2_, 10 mM Tris-HCl (pH 8.3), 0.45% NP-40, 0.45% Tween 20, 0.01% gelatine). Samples were digested at 60°C for 2 hours and at 95°C for 15 minutes. The stc2a^-/-^ fish genotyping was performed by PCR and by direct DNA sequencing as previously reported (71).

Morphology and developmental tracking

The bright-field images of larvae zebrafish were acquired using a stereomicroscope (Leica MZ16F, Leica, Wetzlar, Germany) equipped with a QImaging QICAM camera (QImaging, Surrey, BC, Canada). Head-trunk-angle, adult zebrafish body length, body weight, and brain weight was measured following published protocols (8, 11). Larval body length measured from inner ear stone to tail end. Image J was used for image analysis and data quantification.

Alizarin red staining

Alizarin red staining was performed as described previously (72). The bright-field images were acquired as described above. The whole-body images were joined together with Adobe Photoshop. Dorsal and pelvic spine lengths measured by ImageJ.

Drug treatment

All drugs used in this study, except ZnCl_2_, were dissolved in DMSO and further diluted to desired concentrations. ZnCl_2_ was dissolved in distilled water. Zebrafish larvae were treated with drugs and vehicle as described previously (28). Drug solutions were changed daily.

Hypoxia treatment and survival curves

Hypoxia treatments were performed using the Invivo2–300 Hypoxia Workstation with an I-CO_2_N_2_IC advanced gas mixing system (Baker Ruskinn, Sanford, ME). Gas tanks were purchased from Cryogenic Gases (Detroit, MI). CO_2_ levels were kept constant at 1.4%. In addition to monitoring the gas values shown by this machine, an oxygen meter was utilized for a secondary reading (Sper Scientific, Scottsdale, AZ). The temperature was maintained at ~ 28°C. Zebrafish larvae were set up in 6-well plates at a density of 15 larvae per well with 3–5 mL of E3 or drug medium. The number of dead larvae was counted hourly based on appearance and motor response to stimulus.

Western blotting

Zebrafish larvae (30–40 per group) were homogenized in RIPA buffer (50 mM Tris-HCl, 150 mM NaCl, 1 mM EGTA, 0.1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, pH 7.4) containing a cocktail of protease inhibitors and phosphatase inhibitors on ice for 30 seconds. Zebrafish homogenates were centrifuged at 13,000 rpm, 4°C, for 20 minutes. The protein concentration of the supernatant was measured by Bradford assay (Bio-rad) and normalized. 2X urea loading buffer (150 mM Tris pH 6.8, 6 M Urea, 6% SDS, 40% glycerol, 100 mM DTT, 0.1% Bromophenol blue). Samples were heated at 95°C for 5 minutes and subjected to 12% SDS-PAGE gels and transferred to a nitrocellulose membrane for western blot analysis. After incubation with primary and secondary antibodies, membranes were scanned using the Odyssey CLx imaging system (LI-COR). The ratios of phosphorylated to total protein were calculated after the images had been analyzed. The primary antibodies were mouse anti-GAPDH (1:4000, Proteintech), rabbit anti-Akt (1:1000, Cell Signaling Technology), rabbit anti-P-Akt (S473) (D9E) XP^®^ (1:2000, Cell Signaling Technology), rabbit anti-p44/42 MAPK (1:2000, Cell Signaling Technology), rabbit anti-Phospho-p44/42 MAPK (D13.14.4E) XP^®^ (1:2000, CellSignaling Technology). The secondary antibodies, goat anti-mouse IRDye 680LT and goat anti-rabbit IRDye 800CW were purchased from LI-COR Biosciences and used at 1:10,000 dilution.

Statistical analysis

Statistical analysis was performed using GraphPad Prism 9 software (GraphPad Software, Inc., San Diego, CA). Values are shown as means ± SEM. Statistical significance between experimental groups was performed using unpaired two-tailed t-test, one-way ANOVA followed by Tukey’s multiple comparison test, or two-way ANOVA with multiple comparisons. Differences between the experimental groups in survival curves were analyzed using the Mantel-Cox log-rank test. Statistical significances were accepted at *p < 0.05, **p<0.01, ***p < 0.001, ****p < 0.0001.