The activin receptor signaling pathway, a member of the transforming growth factor-β superfamily, is highly conserved across vertebrate species but WGD events in the teleost lineage have resulted in significant expansion of the signaling pathway in these species (Huminiecki et al. 2009; Phelps et al. 2013). Despite relative conservation of many signaling pathway components at the protein level, significant differential regulation has been identified between fish and mammals (Rescan et al. 2001; Garikipati et al. 2006; Gabillard et al. 2013). We examined the function of the activin receptor signaling pathway across RBT development and in the skeletal muscle of fry, parr, and juvenile fish to gain insight into the function of the pathway in teleost fish. Only a limited number of pathway components were expressed at 80dd of development but bambia1, exhibited very high levels of expression in the early embryo which differed significantly from the pattern of expression exhibited by bambia2. Bambi is a membrane-bound inhibitor of the activin receptor signaling pathway with a known function in vertebrate development (Onichtchouk et al. 1999; Grotewold et al. 2001). It has been found to be co-expressed with Bmp4 in several vertebrate species and is believed to have a role in regulating Bmp embryonic patterning (Reichert et al. 2013), however, loss of function studies have not been able to establish a clear developmental phenotype (Chen et al. 2007).
While most activin ligands and receptors increased expression throughout development the activin type IIb receptors (acvr2b) were downregulated over time (Fig. 1). Activation of the activin receptor signaling pathway requires the recruitment of both type II and type I serine-threonine kinase receptors after ligand binding (Massagué 2012). Loss of acvr2b function in mice leads to reduced follicle-stimulating hormone levels and some skeletal and facial deformities, but the mice are viable (Matzuk et al. 1995a). The observation of opposing expression patterns between acvr2a and acvr2b receptors suggests the potential for key functional differences between these receptors during embryonic development in trout. In mammals, the activin type II receptors (A and B) have overlapping functions and can compensate for the loss of a single receptor in skeletal muscle (Lee et al. 2020). Whether these receptors have redundant functions in embryonic development in mammals is unknown.
The developmental timing of the expression of activin receptor ligands corresponded to the timing of myod1 expression (i.e., 170 dd; Fig. 1c and 1d). In fact, except for moderate expression of inhbaa1 observed at 80 dd, there was minimal expression of the activin ligands prior to 170 dd (Fig. 1). Both myostatin and activin A are regulators of skeletal muscle growth in mammals (Latres et al. 2017). While myostatin contributes to the regulation of teleost skeletal muscle growth (Ohama et al. 2020), the role of activin A in fish is largely unknown. The activin ligands were expressed at a significantly higher level in the skeletal muscle tissue of fry, parr, and juvenile fish than in embryos. This is especially representative of inhbab2 and mstnb genes which exhibited little to no expression during embryonic development but were significantly up-regulated in post-hatch skeletal muscle (Fig. 1). Activin A has been challenging to study in mammals given its multi-functional role in embryonic development, muscle growth, and reproductive development (Latres et al. 2017; Lee et al. 2020). Duplication of activin A in RBT may have facilitated diversification (i.e., subfunctionalization) of the activin A ohnologs such that the functions that are carried out by a single gene in mammals are now regulated by independent genes in fish. Interestingly, activin B (inhbb) exhibited double or triple the level of expression throughout RBT development than activin A (Fig. 1c). All four of the activin B ohnologs also increased their expression level during RBT development while only the activin Aa ohnologs were expressed in RBT embryos (Fig. 1c). This may highlight a greater role for activin B in RBT development than activin A. In mice, inhbb knockout animals are viable but have eyelid fusion defects at birth and are therefore prone to eye lesions and females have significantly impaired reproductive ability (Vassalli et al. 1994). Since activin signaling molecules are dimers they are able to function as either homo or heterodimers between activin A and B proteins (e.g., AA, BB or AB; (Appiah Adu-Gyamfi et al. 2020). The signaling dynamics of activin dimers during development is complex and poorly understood but there is strong evidence that activin A can compensate for the loss of activin B (Vassalli et al. 1994).
Diversification of gene function was observed in the differential expression pattern of activin Cb (inhbcb) during RBT muscle development. Activin C was not expressed in RBT embryos but activin Cb1 (inhbcb1) and inhbcb2 were down-regulated and up-regulated in juvenile skeletal muscle, respectively (Fig. 2c). Activin C is expressed in the liver and adipose tissue of mammals with no recorded function in skeletal muscle (Goebel et al. 2022). It was originally suggested that activin C may be an inhibitor of activin A by forming heterodimers with reduced function, but recent studies have shown that activin C is able to stimulate the activin receptor signaling pathway but that it has lower affinity for acvr2 receptors and is resistant to follistatin inhibition (Goebel et al. 2022). More research is needed to understand how activin C functions in fish and what its role is in skeletal muscle growth, given its high expression in this tissue.
The activin Aa2 (inhbaa2) and Ab2 (inhbab2) genes targeted in CRISPR/Cas9 gene knockout RBT represented one ohnolog from each of the salmonid specific ohnolog pairs (i.e., inhbaa and inhbab, Ss4R; (Berthelot et al. 2014). The expression of inhbaa2 exhibited a consistent increase in expression throughout development while inhbab2 was not expressed during development but was significantly up-regulated in skeletal muscle after hatch (Figs. 1 and 2). Given this expression pattern, it is hypothesized that only inhbaa2 has a role during embryonic development and that inhbab2 may have specialized for other physiological functions, such as regulating skeletal muscle growth. Activin A knockout mice die shortly after birth from cranial facial deformities (Matzuk et al. 1995b). We did not observe any developmental defects in the inhbaa2 or inhbab2 knockout fish. However, while all of the InDel mutations identified in inhbab2 knockout fish resulted in frameshift mutations, only one of the inhbaa2 targeted fish was identified with multiple null mutations. The remaining inhbaa2 targeted fish either had in-frame or wildtype genotypes in addition to having one null allele. While embryonic viability is difficult to quantify in gene-edited founders given the high mortality that naturally occurs after microinjection, the unexpectedly high percentage of in-frame mutant alleles in inhbaa2 targeted fish compared to inhbab2 targeted fish may indicate a role for activin Aa2 in RBT embryonic development. We did not observe any significant difference in weight, length or condition factor in the gene-edited models that were developed (Fig. 3b). This is not unexpected as the activin pathway is known to exhibit functionally redundant genes (Lee et al. 2020) and duplication of the pathway in RBT has only increased the potential for redundant relationships between the activin A ohnologs. Myostatin knockout double muscling phenotypes in fish are often limited and may not manifest as an increase in overall size but instead as an increase in muscle mass per body length (i.e., condition factor; Phelps et al. 2013; Kim et al. 2019)). This results in fish with a significantly bulky appearance but may actually exhibit reduced body length compared to controls (Phelps et al. 2013; Kim et al. 2019). Given the extensive duplication of the activin receptor signaling pathway in salmonids and the resulting regulatory complexity, more functional studies are needed to determine the role of activin A in embryonic development in trout. The detailed expression profile of the activin pathway as well as the relative lack of full biallelic inhbaa2 knockout fish identified in this study provide clues to the potential importance of activin A for RBT development and that duplication of activin A has led to unique specializations of the gene family in salmonids.