Pheromone Perception in Fish: Mechanisms and Modulation by Internal Status

Synopsis

Pheromones are chemical signals that facilitate communication between animals, and most animals use pheromones for reproduction and other forms of social behavior. The identification of key ligands and olfactory receptors used for pheromonal communication provides insight into the sensory processing of these important cues. An individual’s responses to pheromones can be plastic, as physiological status modulates behavioral outputs. In this review, we outline the mechanisms for pheromone sensation and highlight physiological mechanisms that modify pheromone-guided behavior. We focus on hormones, which regulate pheromonal communication across vertebrates including fish, amphibians, and rodents. This regulation may occur in peripheral olfactory organs and the brain, but the mechanisms remain unclear. While this review centers on research in fish, we will discuss other systems to provide insight into how hormonal mechanisms function across taxa.

Introduction

Communication is essential to the social life of animals, whether for finding a mate, signaling social status, or warning of nearby threats. Chemical cues are a crucial medium for social communication. Karlson and Luscher (1959) first defined pheromones as chemicals secreted by an individual (the sender) that are detected by a conspecific (the receiver), evoking a change in physiology and/or behavior. Pheromones may be specific compounds or blends of compounds (Roelofs et al. 1974; Sorensen and Stacey 1999; Stacey 2014). Animals have evolved to use pheromones for many forms of conspecific interactions, such as reproduction, aggression, migration, and signaling danger (Liberles 2014; Stacey 2014). Work across animal models indicates that pheromone-guided behavior is further modulated by multiple factors including overall physiological condition and experiences of the individual (Gomez-Diaz and Benton 2013).

In this review, we discuss the physiological mechanisms that regulate pheromonal perception by the receiver. In particular, we focus on fish as model systems for vertebrate pheromone communication. Fish are the most speciose group of vertebrates, and will yield insights into how pheromone signals have evolved with different ecological conditions and reproductive strategies (Sorensen and Stacey 1999). Fish have evolved to use pheromones for a variety of social behaviors including reproduction, migration, aggression, and predator avoidance (Sorensen and Stacey 1999; Stacey 2003; Chung-Davidson et al. 2011; Keefer and Caudill 2014; Sorensen and Wisenden 2014). Sex pheromones are among the best-studied chemical signals in fish (Stacey 2014) and across the animal kingdom (Gomez-Diaz and Benton 2013). As these molecules robustly elicit behavioral responses, they are excellent candidates for experimentation. While ongoing research endeavors to pair key pheromone ligands with pheromone receptors across diverse taxa, there is an outstanding need to understand how pheromonal communication is regulated in senders and receivers (for reviews in insects, mammals, and fish; see Tirindelli et al. 2009; Sorensen and Wisenden 2014; Fleischer and Krieger 2018). This can lend insights into the mechanisms underlying behavioral plasticity (Dey et al. 2015; Orlova and Amsalem 2019), sex-specific behaviors (Stowers and Logan 2010), and speciation (van Schooten et al. 2020). Sex pheromones also have utility for invasive species control (Sorensen and Stacey 2010; Fissette et al. 2021), wildlife protection (Larsson 2016), and the development of sustainable aquaculture and fisheries management (Pandey 2005).

Pheromone perception is plastic in fish with regard to physiological status (Cardwell et al. 1995; Li et al. 2023). Here, we define pheromone perception as the olfactory sensitivity to pheromone ligands, and the downstream neural processes leading to behavior. Plasticity could emerge from changes at multiple levels of olfactory processing, from pheromone detection in the olfactory organs, signal integration in the olfactory bulbs (OB), and central processing in the brain. We highlight African cichlid fish as an outstanding model for pheromone research in fish because they have evolved a diverse repertoire of social behaviors, which are regulated by complex chemosensory signaling mechanisms (Fernald and Hirata 1979; Maruska and Fernald 2012; Keller-Costa et al. 2014a, 2014b, 2015; Escobar-Camacho and Carleton 2015; Simões et al. 2015; Field and Maruska 2017; Field et al. 2018; Li et al. 2023). The physiological control of social behaviors in cichlids is well studied (O’Connell et al. 2013; Alward et al. 2019; Maruska et al. 2022); and genome editing tools can also be leveraged to understand the genetic control of behavior in cichlid fish (Juntti et al. 2016; Alward et al. 2020; Li et al. 2021). Thus, it is feasible in cichlid fish to gain a mechanistic understanding of pheromonal communication and its plasticity.

How do hormones shape plasticity in the olfactory system, particularly for pheromonal communication? Here, we review the current knowledge of how hormones regulate state-dependent plasticity of pheromone-guided behavior in fish. We first briefly review known fish pheromones and their associated behaviors. We discuss the neural substrates upon which hormones act in the fish olfactory circuit from the peripheral olfactory organ to the central nervous system (CNS). We discuss the potential signaling pathways that hormones activate to modulate pheromone sensitivity and behavior, noting work from other models for insights into how these mechanisms may be conserved across animals. We highlight the need for future research to elucidate the role of hormones in modulating pheromonal communication in different social contexts.

Diversity of pheromonal cues

Pheromones can be released from fish in a variety of secretions, including bile, urine, gonadal fluids, and skin extracts (Smith 1992; Sorensen and Stacey 1999; Stacey et al. 2003; Burnard et al. 2008; Chung-Davidson et al. 2011; Buchinger et al. 2014). Pheromonal ligands of varied chemical forms have been identified in fish, including amines, amino acids, prostaglandins, bile acids, and steroids (Colombo et al. 1980; Sorenson et al. 1988; Li et al. 1995; Murphy et al. 2001; Yambe et al. 2006; Huertas et al. 2010; Buchinger et al. 2014; Stacey 2014; Giaquinto et al. 2015; Scott et al. 2019; Zhu et al. 2023). Peptide pheromones have been identified in terrestrial vertebrates (Kiyukama et al. 2002; Haga et al. 2010), and peptide-sensitive sensory neurons have been proposed to detect species-specific cues in swordtail fish, Xiphophorus spp. (Cui et al. 2017). Some well-studied pheromones are derived from hormones, or are hormones themselves (Sorensen 1992). It is proposed that an ancestral pheromone sensation ability derives from an “eavesdropping” receiver that gained the ability to infer the physiological state of a conspecific fish via the detection of released hormones or their metabolites (Sorensen and Stacey 1999). The eavesdropper’s ability to sense and respond to conspecific cues provides a fitness advantage, which in turn leads to selection for the ability to detect these nascent pheromones.

For such pheromones that are derived from hormones, changes in endocrine state associated with reproductive or dominance status may thus lead to alterations in the titers of released pheromones. These pheromones include the progestin 17,20β-dihydroprogesterone (DHP) and its metabolites. First identified as a hormone produced prior to ovulation (Fostier et al. 1973; Nagahama et al. 1983), DHP was later shown to act as a pheromone that promotes mating behavior in male goldfish, Carassius auratus (Dulka et al. 1987; Sorensen et al. 1987; Stacey et al. 1989; DeFraipoint and Sorensen 1993; Sorensen and Scott 1994). Similarly, prostaglandin F_2α (PGF_2α) is produced following ovulation, and its signaling is necessary and sufficient to induce female mating behavior (Stacey and Liley 1974; Stacey and Peter 1979; Juntti et al. 2016). PGF_2α signaling attracts males to fertile females, revealing its dual role as hormone and pheromone (Sorensen et al. 1988; Li et al. 2023). Androstenedione, a precursor to testosterone and estradiol, also acts as a pheromone in goldfish to promote mating and aggression (Poling et al. 2001; Sorensen et al. 2005). For each of these pheromones, their release is modulated in concert with the behavioral state of the sender, often through urination, such that it is released more at particular times or locations. For example, the urination rate of goldfish and cichlids Oreochromis mossambicus and Astatotilapia burtoni is modulated based on their internal status as well as their social context (Appelt and Sorensen, 1999, 2007; Barata et al. 2007; Maruska and Fernald 2012; Field and Maruska 2017), enabling receivers to regulate their behavior accordingly.

Sex hormones can be released in native form via gills, urine, or feces, though other routes exist (Vermeirssen and Scott 1996; Stacey 2014). Hormones may be modified for enhanced excretion by conjugation with sulfate or glucuronate groups, or other modifications that increase solubility. Olfactory sensitivity toward these modified forms may also vary across species and social status. For example, the cichlid A. burtoni is only sensitive to conjugated steroids, but goldfish responds to free steroids (Sorensen et al. 1990; Cole and Stacey 2006; Sato and Sorensen 2018). The potent goldfish pheromone PGF_2α is not attractive in A. burtoni, but its actions within the female is still necessary for pheromonal activity, suggesting it is metabolized into a still-unidentified pheromone (Cole and Stacey 2006; Li et al. 2023).

Many different conjugates can exist for a single sex hormone (Stacey 2014). The diversity of hormone modifications may result from the evolution of species-specific pheromones, as closely related species can show different sensitivity to the same conjugates. In the male black goby Gobius niger, the pheromone etiocholanolone glucuronide (Colombo et al. 1980) attracts females, whereas in the round goby Neogobius melanostomus, 5β-reduced 11-oxo-etiocholanolone drives attraction (Arbuckle et al. 2005). Hormone modification may also function in context-dependent signaling within a species, as attraction to free or conjugated forms can change depending on internal status. In the round goby, 11-oxo-etiocholanolone glucuronide and sulfate conjugates are attractive to reproductively active females, while the free form is attractive to non-reproductive females (Corkum et al. 2008; Katare et al. 2011; Tierney et al. 2013).

Not all intraspecific chemosensory cues are derived from sex hormones. Schreckstoff, or “fright substance” was first identified from damaged skin of the minnow Phoxinus phoxinus. Chemical cues from damaged skin are aversive when presented to conspecifics, and skin damaged by predators or by pathogens elicits fright behaviors specific to the species (Frisch 1938; Smith 1992; Chivers et al. 2007). This cue appears to be comprised of glycosaminoglycan chondroitin and other substances (Mathuru et al. 2012). Its benefits to bystanders are debated, but it has been suggested that it helps fish avoid locations where conspecifics have been recently predated. Other examples include cadaverine, a diamine produced in decaying flesh that causes avoidance in the zebrafish, Danio rerio (Hussain et al. 2013), and 4-hydroxyphenylacetic acid (4-HPAA), which induces ovulation in zebrafish (Behrens et al. 2014). Spermine has been identified as a male polyamine sex pheromone derived from semen in the lamprey Petromyzon marinus (Scott et al. 2019). While it is not hormonally derived, only mature males produce spermine, and it is only attractive to post-ovulatory females. In addition, the amino acid l-kynurenine, produced by ovulating female salmon Oncorhynchus masou, is detected at low concentrations by spermiated male salmon (Yambe et al. 2006). Bile acids represent a group of pheromones that is not obviously dependent on the internal status of the sender, but have important roles as sex pheromones in lamprey, (Li et al. 1995, 2002; Siefkes and Li 2004), and as migratory cues in salmon (Døving et al. 1980; Jones and Hara 1985; Buchinger et al. 2014). These examples illustrate that fish use a diverse repertoire of ligands as sex pheromones. We turn next to how neural circuits sense these molecules and how the internal hormonal state of the fish changes perception.

Detection of pheromones

Pheromone detection begins when molecules bind to olfactory receptors on olfactory sensory neurons (OSNs). In fish, there are three main populations of OSNs in the olfactory epithelium (OE), named for their anatomical features: ciliated, microvillous, and crypt OSNs. Additional minor OSN populations have been reported in zebrafish, namely the Kappe neurons (Ahuja et al. 2014) and the pear-shaped neurons (Wakisaka et al. 2017), but more research is required to understand their function. Like the mammalian olfactory system, individual OSNs in fish appear to express a single olfactory receptor that confers sensitivity to a limited number of ligands (Buck and Axel 1991; Barth et al. 1997; Malnic et al. 1999). Fish olfactory receptors are G protein-coupled receptors (GPCRs), that are derived from one of four multigene receptor families. Classical odorant receptors (ORs) and trace amine-associated receptors (TAARs) are expressed in ciliated OSNs, vomeronasal-like type 1 receptors (V1Rs; aka ORA receptors) are expressed in crypt neurons, and the vomeronasal-like type 2 receptors (V2Rs; aka OlfC receptors) are expressed in microvillous neurons (Kermen et al. 2013). Based on anatomical evidence and a limited number of receptors for which a ligand has been identified, OR-expressing ciliated OSNs are likely responsive to bile acids, prostaglandins, and sex steroids (Nikonov and Caprio 2001; Hansen et al. 2003; Yabuki et al. 2016; Sato and Sorensen 2018). V2R-expressing microvillous OSNs, homologous to mammalian vomeronasal sensory neurons (VSNs), respond to food odors such as amino acids and nucleotides (Friedrich and Korsching 1997; Hansen et al. 2003; Koide et al. 2009; DeMaria et al. 2013; Wakisaka et al. 2017; Sato and Sorensen 2018; Poncelet and Shimeld 2020; Kawamura and Nikaido 2022). The odor specificity of V1R-expressing OSNs (including crypt) is less clear, but they may respond to social odors such as kin odor, bile acids, and sex steroids (Bazáes and Schmachtenberg 2012; Biechl et al. 2016; Cong et al. 2019; Kawamura and Nikaido 2022). Olfactory information is conveyed by OSN axons to discrete sites of the OB called glomeruli where axons from OSNs bearing a specific receptor converge. The axons of ciliated, microvillous, and crypt OSNs terminate into specific, non-overlapping regions of the OB, revealing separation of odor signals in a chemotopic manner (Hansen et al. 2005; Sato et al. 2005; Koide et al. 2009; Ahuja et al. 2013).

Identified pheromone ligand–receptor pairs

Identifying a ligand–receptor pair is a difficult task, and few pheromone receptors have been confirmed in fish (Table 1). Cyprinid fish males sense post-ovulatory females that produce high levels of PGF_2α (Sorensen et al. 1988, 1989; Sveinsson and Hara 2000; Olsén et al. 2001; Appelt and Sorensen 2007). F-prostaglandins activate ciliated OSNs in the lake whitefish Coregonus clupeaformis, suggesting specific sensory mechanisms are dedicated to individual pheromonal signals (Laberge and Hara 2003). This experiment paved the way for identifying two olfactory receptors responsible for detecting PGF_2α, Or114-1 and Or114-2 in the zebrafish (Yabuki et al. 2016). The receptors were identified by double in situ hybridization for individual OR transcripts and neural activity markers. Neural activity screens also identified Taar13c as a receptor for cadaverine in zebrafish (Hussain et al. 2013). Researchers have also screened for a ligand that activates a specific receptor using biochemical purification techniques such as liquid chromatography and mass spectrometry (LC–MS). 4-hydroxyphenylacetic acid (4-HPAA), was identified as a ligand for Ora1 in zebrafish using this approach (Behrens et al. 2014). Following these pioneering experiments, additional ligand–receptor pairs will be identified in fish species in the future.

Table 1.

List of identified pheromone–receptor pairs with demonstrated behavioral functions in fishes

Family | species	Pheromone	Sender	Response	Olfactory receptor	Key references
Petromyzontidae
P. marinus (sea lamprey)	3kPZS, PZS	Reproductive male: larvae	Female attraction	Or320a, Or320b	Li et al. (1995), Siefkes and Li (2004), Buchinger et al. (2020), Zhang et al. (2020)
P. marinus	Spermine	Reproductive male	Female attraction	Taar348	Scott et al. (2019)
Cyprinidae
B. schwanenfeldii (tinfoil barb), C. auratus (goldfish), C. carpio (common carp), D. rerio (zebrafish), E. frenatus (redtail sharkminnow)	PGF2α, 15-keto-PGF2α	Reproductive female	Male attraction	Or114-1, -2	Sorensen et al. (1988, 1989), Cardwell et al. (1995), Belanger et al. (2010), Yabuki et al. (2016)
D. rerio	4-HPAA	Unknown	Oviposition	Ora1	Behrens et al. (2014)
D. rerio	Cadaverine	Decaying fish	Avoidance	Taar13c	Hussain et al. (2013)
D. rerio	LCA	Unknown	Attraction	Ora1	Cong et al. (2019)
Cichlidae
H. chilotes (Victorian biglip hap cichlid)	4-HPAA	Unknown	Unknown	Ora1	Kawamura and Nikaido (2022)
H. chilotes	LCA	Unknown	Unknown	Ora1	Kawamura and Nikaido (2022)

Potential for modulation in the OE

How does olfactory processing change according to an animal’s physiological status? Sex hormones can convey internal status information to peripheral and central neurons, and thereby modulate perception and behavior. Several studies indicate that sex steroid signaling changes olfactory sensitivity to pheromones in fish. Comparison of animals of different sexes, dominance status, or androgen treatment shows that sensitivity to sex pheromones can be enhanced by androgens, while non-sexual odors are unaffected (Cardwell et al. 1995; Belanger et al. 2010; Ghosal and Sorensen 2016). Androgens have also been shown to induce attraction to female sex pheromones in male juvenile fish (Cardwell et al. 1995; Yambe and Yamazaki 2000; Yambe et al. 2003). Indeed, social status regulates androgen titers and olfactory processing of social information in the fish brain (Nikonov and Maruska 2019). Female goldfish treated with androgens perform male-typical olfactory-guided behaviors (Stacey and Kobayashi 1996; Ghosal and Sorensen 2016). Sex hormones appear to modulate pheromone sensitivity in the OE, but the mechanisms are unknown. In various juvenile cyprinid species treated with androgens, the olfactory organs exhibit sensitivity to female pheromones typical of adult males (Cardwell et al. 1995; Belanger et al. 2010; Ghosal and Sorensen 2016). The change in pheromone sensitivity is associated with expression of male reproductive behavior, suggesting that peripheral androgens enhance male-like sensation of pheromones. This does not rule out a central role for androgen signaling; it is likely that hormones modulate pheromone sensitivity in higher brain regions for olfactory processing as well (Fig. 1).

Fig. 1.

Hormone levels fluctuate with physiological states such as the female reproductive cycle. Hormones relay internal status across peripheral and central substrates of olfactory neural circuits, changing the perception of pheromones and the behavioral output.

Beyond androgens, other sex hormones can influence the receiver’s perception of pheromones. In female mice, the behavioral response to male pheromones is differentially regulated by hormones across the estrous cycle. During estrus, ovulated females are attracted to male sex pheromones, while in diestrus, non-ovulated females become indifferent to male pheromones as a result of progestin signaling (Dey et al. 2015). In males, estrogen, a masculinizing factor in mice, modulates pheromone sensitivity in the vomeronasal organ (VNO) (Halem et al. 2001). Further, estrogens act directly on VSNs to modulate their pheromone sensitivity (Cherian et al. 2014). In the African cichlid A. burtoni, PGF_2α levels rise in female circulation following ovulation. PGF_2α acts on receptors in the brain to permit the expression of mating behavior with a courting male (Juntti et al. 2016). Moreover, the peak in PGF_2α in the female also causes the release of a still-unidentified sex pheromone from the female (Li et al. 2023). This pheromone elicits robust attraction in sexually mature males, but it does not attract juvenile males or females. Electrophysiological recordings in cyprinids from the olfactory organs showed that juveniles and females are less sensitive to the female pheromone PGF_2α, while sexually mature males are highly sensitive and robustly attracted to PGF_2α, indicating modulation at the level of the OE (Cardwell et al. 1995; Stacey and Kobayashi 1996; Belanger et al. 2010; Ghosal and Sorensen 2016). Factors beyond reproductive status and sex can also play an important role. Nutritional status affects sensitivity to sex pheromones in Drosophila, with fed female flies demonstrating increased sensitivity to male sex pheromones compared to starved females (Lebreton et al. 2015). These examples together demonstrate the importance of an animal’s internal status impacting their perception of a pheromonal signal.

What mechanism(s) might underlie peripheral sensitivity to pheromones? In fish, hormone receptors are expressed in the OE, suggesting an avenue through which hormones and hormonal pheromones modulate sensitivity in a receiver. Receptors for progestins, androgens, estrogens, and prostaglandins have been detected in the fish OE (Pottinger and Moore 1997; Kolmakov et al. 2008; Strobl-Mazzulla et al. 2008; Tubbs et al. 2010; Zhang et al. 2018; Chen et al. 2022). Hormones may reach these sites through circulation (Hansen and Zeiske 1998) or may be synthesized directly within the OE (Lupo et al. 1986; Cherian et al. 2014). These provide a basis for hormones to modulate OE activity.

Hormone receptors could in principle affect neural dynamics of the OE through several cellular pathways including transcriptional changes and neuromodulation (Fig. 2A). Canonically, sex steroid receptors are nuclear hormone receptors, which function as ligand-activated transcription factors. The transcriptional changes initiated by hormone concentration rises could enhance pheromone sensitivity through a variety of genomic targets. For example, hormone signaling could enhance transcription of a pheromone-sensitive receptor, leading to selectively enhanced sensitivity to a pheromone. Some olfactory receptors exhibit sexually dimorphic expression in mice and zebrafish (Vihani et al. 2020; Wang et al. 2020), consistent with hormonal regulation, but the significance of these receptors or the mechanisms for enhanced transcription have not been tested. Alternatively, subcellular signaling downstream of pheromone receptor activation could be enhanced in a cell-type specific manner. There may be enhanced transcription of pheromone receptor-coupled G-proteins, the enzymes they activate, or ion channels they potentiate. In addition, hormone signaling could alter general neuronal mechanisms of excitability including resting membrane potential, input resistance, or firing rate. However, these explanations would require transcriptional changes to preferentially affect pheromone-sensitive neurons over OSNs tuned to non-reproductive cues. To our knowledge, expression studies of hormone receptors or their cofactors have not been performed with cellular resolution in fish (e.g., in situ hybridization, single-cell sequencing). It is also possible that during the ongoing renewal of OSNs through neurogenesis, hormone signaling may enhance the selection of pheromone sensitive receptors (or survival of OSNs expressing pheromone receptors), thus biasing the OE to be more sensitive to pheromones. Careful studies of the time course of pheromone sensitivity change following hormone signaling will elucidate the actions of these hormone receptors. Furthermore, the specific genomic targets of hormone signaling in neural tissues have been difficult to elucidate. Recent work in mice has identified a trove of hormone receptor targets (Gegenhuber et al. 2022), and future work in teleost olfactory centers may suggest how hormones change neural activity through transcription.

Fig. 2.

Models for direct hormone modulation of sensation in the OE. (A) Hormone receptors may modulate sensation via altered gene transcription or neuromodulation of OSN signals. (B) Hormone receptors may act as receptors for pheromones, leading to direct activation of OSNs. (C) Hormones may act on non-neuronal cells to modify epithelium morphology, cell development, and/or cell–cell interactions. Additional indirect pathways for peripheral modulation of olfactory signals may exist, including feedback via recursive connections to the OE from CNS sites including the terminal nerve, or by other developmental mechanisms.

In contrast to a model in which transcriptional changes mediate hormone signaling, it is also possible that hormone signaling at the membrane may initiate rapid, neuromodulatory responses. The nuclear hormone receptors have been observed to interact with steroid signaling systems at the membrane, regulating reproductive behaviors in multiple vertebrate species (Micevych et al. 2017). In addition, transmembrane steroid receptors have also been observed to transduce hormone signals. One of the first non-nuclear hormone receptors for steroids discovered in any vertebrate was a transmembrane progestin receptor expressed in the teleost ovary, mPRα (Zhu et al. 2003). This membrane progestin receptor belongs to the progestin and adipoQ receptor (PAQR) family (Tang et al. 2005; Thomas 2022). In the female reproductive tract, mPRα (aka, PAQR7) mediates oocyte maturation prior to ovulation (Zhu et al. 2003). Another membrane progestin receptor, Pgrmc1, is also expressed in teleost oocytes (Mourot et al. 2006). In zebrafish, Pgrmc1 knockout decreases PAQR7 expression, reducing sensitivity to progesterone and impairing oocyte maturation (Wu et al. 2018). Pgrmc1 may act as an adaptor protein to facilitate PAQR7 surface expression in vertebrate tissues (Thomas et al. 2014; Thomas 2022). Interestingly, Pgrmc1 is also expressed in the mouse VNO, a sensory epithelium that detects non-volatile pheromones in rodents (Dey et al. 2015). Here, progesterone binding causes silencing of specific pheromone-sensitive neurons. As a result, during the postovulatory phase when females are not fertile, they are also not attracted to male pheromones. PAQRs including PAQR7 have also been identified in fish and rodent olfactory transcriptomes (Kolmakov et al. 2008; Saraiva et al. 2015; Abaffy et al. 2023), but their function is not well known. Taken together, this suggests that Pgrmc1/PAQR-dependent signaling could regulate pheromone sensitivity in the OE.

Other membrane hormone receptors have been identified in fish and mouse OE (Kolmakov et al. 2008; Cherian et al. 2014; Kanageswaran et al. 2016). Their actions can include rapid release of intracellular calcium, modulation of cell signaling via cAMP or phosphatidylinositol signaling, and interaction with second messengers (Thomas 2012). Thus, hormones could modulate OSN activity in the presence of pheromones on a sub-second timescale. An additional GPCR, Gper1, has also been identified that has high estrogen binding affinity (Filardo et al. 2002). Both the GPCR- and membrane-associated hormone receptor systems have been shown to regulate neural activity in the CNSs of vertebrates (Cornil et al. 2006; Srivastava et al. 2013; Krentzel et al. 2018). In mice, Gper1 is expressed throughout the VNO in VSNs, and 17β-estradiol decreases the firing rate of the VNO neurons in response to male mouse urine (Cherian et al. 2014). Though these examples have been shown in mice, the Gper1 gene is conserved across vertebrate species (Pang and Thomas 2010), making it and other non-canonical hormone signaling pathways worth studying in fishes. In addition to acting as neuromodulators of OSN signaling, traditional hormone receptors could act as receptors for pheromones (Fig. 2B). As some pheromones are hormones themselves, or modified versions thereof (Stacey 2014), these receptors are poised in the OE to sense such molecules. In some fish, PGE₂ has been suggested to act as a pheromone, and the goldfish OE is sensitive to PGE₁ and PGE₂ (Sorensen et al. 1988). Sensitivity to PGE₂ has been shown in the OE of goldfish and pufferfish (Kolmakov et al. 2008; Chen et al. 2022). Prostaglandin receptors could either sense pheromones or modulate olfactory circuits. However, while functional genetic studies demonstrate that a canonical OR (Or114) detects the pheromone PGF_2α, the relative contribution of classical hormone receptors versus ORs in olfaction remains untested.

Aside from OSNs, the OE is composed of various non-neuronal cell types that play important roles in regulating and maintaining the function of the sensory epithelium, and these could be important sites of modulation (Fig. 2C). The sustentacular and olfactory ensheathing cells of the mammalian main OE provide structural support for neurons. Their essential role in regulating olfactory sensitivity has been highlighted by cell-nonautonomous effects on OSN transcription following infection of sustentacular cells by SARS-CoV2 virus, leading to anosmia (Zazhytska et al. 2022). Non-OSN cell types also regulate olfactory signaling through interactions with odorants. Sustentacular cells secrete enzymes that degrade odorants in the nasal cavity, and produce carrier molecules that bind odorants (Nagashima and Touhara 2010). These odorant-carrier complexes could either bind OSNs (Laughlin et al. 2008) or cause their degradation. Upon odorant binding to a carrier, sustentacular cells internalize these complexes for trafficking to lysosomes (Strotmann and Breer 2011). Odorant binding proteins of the same lipocalin class as those which play a role in tetrapod pheromone signaling are present in the fish genomes, however, no functional tests have implicated these molecules in fish pheromone signaling. Intriguingly, non-OSN cells of the mouse, rat, and frog OE also express various hormone receptors (Baly et al. 2007; Breunig et al. 2010; Abaffy et al. 2023). This provides a basis for plasticity in olfactory signaling resulting from endocrine changes, but such mechanisms remain to be tested. We also note that neural connections from the CNS back to the OE via the terminal nerve (TN) may modulate olfactory processing (Satou 1990; Zielinski et al. 2000; Edwards et al. 2007 ; Kawai et al. 2009), and we discuss this pathway below.

Modulation of olfactory signals in the CNS

Anatomy of central olfactory circuits

In addition to modulating olfactory sensation, hormone action in the CNS is likely to play an important role in shaping pheromone-guided behavior (Fig. 3). The OB is the primary structure that receives olfactory input through the olfactory nerve. All OSNs expressing the same olfactory receptor gene converge on a single glomerulus in the OB (Imai et al. 2010). As a result, odorant-induced neural activation in the bulb occurs in a spatially defined pattern that reflects the chemical nature of the olfactory stimulus. In zebrafish, ciliated cells mainly project to the dorsal and medial OB; microvillous cells project to the lateral OB (Sato et al. 2005, 2007) (Fig. 3). Glutamatergic mitral and ruffed cells are the principal output cells from the OB in fish (Edwards and Michel 2002). In each glomerulus, dendrites of mitral cells receive synaptic inputs from OSNs and project to higher brain regions via the olfactory tracts (Fuller et al. 2006; Miyasaka et al. 2009). From the OB, information is relayed through the medial and lateral olfactory tracts via mitral cells (Hamdani et al. 2000, 2001). Granule cells are interneurons in the OB, which mediate lateral inhibition to enhance signal-noise ratios in the olfactory system (Nunez-Parra et al. 2013; Parsa et al. 2015). These GABAergic granule cells are located in the inner layer of the OB and make dendro-dendritic synaptic connections with mitral and ruffed cells (Kermen et al. 2013). The medial and lateral olfactory tracts convey different information, shown through studies in cod and goldfish in which electrical stimulation of the medial olfactory tract induces alarm reaction or reproductive behavior, while lateral olfactory tract stimulation induces feeding behaviors (Døving and Selset 1980; Demski and Dulka 1984). Higher brain regions receiving direct innervation from the OB include the ventral portion of the ventral telencephalon (Vv) and the posterior portion of the dorsal telencephalon (Dp), as well as diencephalic regions including the habenula (Hb), posterior tubercle (PT), and the preoptic area (POA) and other regions of the hypothalamus (Miyasaka et al. 2009, 2014; Kermen et al. 2013; Olivares and Schmachtenberg 2019). Receptors for steroid hormones and neuromodulators appear to be broadly expressed in the CNS, providing a substrate for hormonal modulation across the brain (Tensen et al. 1997; Illing et al. 1999; Okubo et al. 2000; Norton et al. 2008; Maruska and Fernald 2010; Munchrath and Hofmann 2010).

Fig. 3.

Internal state modulates pheromone processing in the CNS at multiple levels, altering behavior. The presence and concentration of a given pheromone carries information about the state and proximity of a conspecific animal. Pheromones are detected at the OE, information is transmitted to the OB, then to other central olfactory centers. The receiver’s state, including its circulating hormone milieu and neuromodulatory activity, conveys information about its social and reproductive status. Hormones and neuromodulators act at multiple levels of pheromone processing. For some brain regions (e.g., Vv, POA) hormones modulate pheromone processing and consequent behavior. Other regions are known to be involved in pheromone processing, but hormone modulation has not yet been demonstrated. Central integration of pheromone information with internal state information culminates in updating the receiver’s behavioral state to suit the context. This requires recruitment of central neural circuits responsible for motivational state, as well as peripheral motor systems responsible for locomotion and autonomic motor systems responsible for processes like gamete release. Other mechanisms exist as well, including direct modulation at the OE.

Mechanisms for modulation in the CNS

The regulation of gonadal hormone release relies on activation of the hypothalamic–pituitary–gonadal (HPG) axis. This axis regulates reproductive functions and is also subject to external influences, such as social signals (Francis et al. 1993; Soma et al. 1996; Maruska 2014), which lead to changes in hormone levels. The modulation of olfactory processing in the CNS could be due to fluctuations in hormone levels, hormone receptor expression levels, or both. Hormone signaling in the OB and other brain regions can proceed via either transcriptional (aka “genomic”) or rapid actions that do not require new transcription.

Modulation through steroid hormones and prostaglandins

In teleost fish, the OB expresses receptors for gonadal steroid hormones, including androgens and estrogens, as well as aromatase (Gelinas and Callard 1997; Harbott et al. 2007; Maruska and Fernald 2010; Munchrath and Hofmann 2010; Ouyang et al. 2021). Additionally, levels of steroid hormones change in response to social and reproductive status (Oliveira et al. 2002; Salazar and Stoddard 2009; Kidd et al. 2013), which may permit status-dependent modulation in this region. However, how exactly steroid hormones modulate processing by the teleost OB is still unclear. In the mammalian OB, sex steroid hormones may modulate odor processing by regulating glomerular activation or by changing neural circuit architecture. In female mice, removing circulating sex hormones by ovariectomy reduces the number of activated glomeruli following odorant, while castration of male mice leads to activation of more glomeruli (Kass et al. 2017). In rats, estradiol increases, while testosterone decreases, the number of mitral cells present in the accessory OB, a group of neurons that receive input from the VNO (Pérez-Laso et al. 1997). Androgen and estrogen receptors are expressed in several types of neurons and non-neuronal cells in the OB (Brann et al. 2020). Thus, these receptors could alter sensitivity by changing the number of output cells in the OB, changing their connectivity, or altering firing patterns (Fig. 3). In ferrets, androgens increase the activity of the main OB in response to reproductive female pheromones (Kelliher et al. 1998). Because androgen receptors are expressed in the main OB granule cell layer of male and female ferrets, androgens may act directly on these cells to augment their responsiveness to female reproductive pheromones. Nevertheless, it is still unclear whether this modulation increases overall sensitivity to all odor cues or fine-tunes the olfactory system to detect certain pheromones, such as reproductive odors.

Regions downstream of the OB are also potential sites for modulation by an animal’s physiological state. The fish Dp is homologous to the primary olfactory cortex in mammals, and expresses receptors for androgens, estrogens, and progesterone (Yaksi et al. 2009; Munchrath and Hofmann 2010; Blumhagen et al. 2011). Despite this, no direct evidence has yet implicated the Dp as a site of hormonal modulation of pheromone-driven behavior in fish. In mice, the homologous primary olfactory cortex is essential for olfactory learning (Wilson and Sullivan 2011), and social odor learning is sensitive to modulation by oxytocin (Choe et al. 2015). Neural oxytocin signaling is elevated in various socially salient contexts, including aggression and mating, and likely promotes memory of such encounters. Olfactory learning recruits the Dp in fish (Jacobson et al. 2018), but further research is needed to evaluate modulation of this process by hormones or neuromodulators.

The other major telencephalic target of OB projections, the Vv, does exhibit status-based modulation of pheromone processing. In males of the cichlid fish A. burtoni, responsiveness of the Vv to social odors is modulated in a sex- and dominance-dependent manner (Nikonov and Maruska 2019). Vv cells in dominant males are highly responsive to female odor, while in subordinate males the Vv is relatively unresponsive to female odor. Interestingly, in subordinate males, the Vv is highly responsive to dominant-male odor, suggesting a shift in sensitivity related to stimulus salience. This provides an avenue through which dominant males enhance sensitivity to reproductive cues while subordinate males sense dominant male cues, perhaps to avoid aggression or sense opportunities to ascend in dominance status. In the cichlid A. burtoni, androgen signaling is a key hormonal mediator of dominance-related changes to behavior, and androgen receptors are found in the Vv in high density (Harbott et al. 2007; Munchrath and Hofmann 2010; Alward et al. 2020). In male A. burtoni, androgen levels are positively correlated with social status, and thus dominant status may modulate Vv sensitivity to pheromone cues. In rodents the Vv-homologous septal area is implicated in regulating pheromone-driven sexual and aggressive behavior, both androgen-dependent processes (Xiao et al. 2005; Marie-Luce et al. 2013; Tan and Stowers 2020). A direct link between androgen signaling in the Vv and sensitivity to particular social odors in fish remains to be demonstrated. Moreover, it remains to be seen whether the differential sensitivity to social odors is indeed a mechanism that underlies differences in behavioral response in fishes. It is also worth noting that in another cichlid fish, the tilapia O. mossambicus, castration abolishes courtship behavior while leaving aggressive behavior intact (Almeida et al. 2014). Therefore, the roles of androgen signaling in regulating aggression vary amongst cichlid fishes.

Hypothalamic nuclei are essential and conserved regulators of reproductive and other social behaviors across vertebrates from fish to mammals (Sakuma 2008; Ogawa et al. 2021). Various lines of evidence indicate that hypothalamic nuclei use pheromonal information to guide reproductive behavior in fish. In the cichlid A. burtoni, elevated PGF_2α in females acts on receptors in the POA to gate sexual motivation in the presence of a receptive male (Juntti et al. 2016), and similar processes may also occur in goldfish (Stacey and Peter 1979). In female goldfish, PGF_2α drives attraction to a pheromonal mixture containing the male pheromone androstenedione (Sorensen and Levesque 2021). The mechanisms of PGF_2α potentiation of mating are not known, but signaling through its receptor, Ptgfr, may enhance calcium currents, promoting action potential generation. In male zebrafish, detection of the female sex pheromone PGF_2α activates neurons of the POA, lateral and caudal hypothalamic areas, as well as the Vv (Yabuki et al. 2016). Work in male goldfish has also shown that another female sex pheromone, 17α,20β-dihydroxy-4-pregnen-3-one, activates POA neurons (Kawai et al. 2015). In mammalian models, these same hypothalamic regions are central to hormone modulation of pheromone-driven behavior. Female mice have a steroid-sensitive population of cells in the POA that regulates sexual motivation, and estradiol modulates their sensitivity to male odor (McHenry et al. 2017). In male mice, neurons expressing estrogen receptor alpha in the POA mediate olfactory-guided investigation of conspecifics as well as mounting behavior (Wei et al. 2018). The lateral hypothalamic area in mice contains a group of cells that express orexin (hypocretin) in a hormone- and estrus-cycle-dependent manner (reviewed in Jennings and De Lecea 2020), and orexin signaling originating in the lateral hypothalamus is involved in driving sexual motivation (Muschamp et al. 2007).

While the importance of the ventromedial hypothalamus (VMH) in pheromone-mediated behavior is well-established in mammals, no evidence yet implicates the fish homolog of the mammalian VMH (the anterior tuberal nucleus). The VMH is central for sexual behavior and aggression in rodents, and steroid-sensitive neuron populations mediate these pheromone-driven behaviors (Lin et al. 2011; Yang et al. 2013; Lee et al. 2014; Ishii et al. 2017; Jennings and De Lecea 2020; Itakura et al. 2022). These neurons also exhibit dramatic estrus-cycle-dependent structural and transcriptomic plasticity necessary for proper expression of sexual behavior in synchrony with estrus (Inoue et al. 2019; Knoedler et al. 2022). In male mice, partially overlapping cell populations in the VMH regulate sexual behavior and aggression, and a critical Esr1⁺ population regulates the progression of sequential sexual behaviors as well as the production of aggression in an activity-dependent manner (Lin et al. 2011; Lee et al. 2014). It remains to be seen whether similar cell populations mediate similar behavioral triggers and transitions in the fish anterior tuberal nucleus.

The remaining central targets of OB projections have not been shown to play a role in hormone modulation of pheromone-driven behavior, despite the expression of hormone receptors. The PT has been implicated in lamprey as a critical relay in an olfactory-motor circuit (Derjean et al. 2010), sending projections to a midbrain cell population that in turn projects to hindbrain reticulospinal neurons, which drive a locomotor swimming pattern. Hormone actions in these PT cells could modulate the sensitivity of initiating chemotaxis to a pheromone source, but further work is needed to understand this circuit in other fishes. The Hb, along with Dp, appear to play a role in resolving the behavioral response to competing pheromones. In zebrafish, the alarm cue Schrekstoff triggers an aversive behavioral reaction driven by activation of the Dp and habenula (Krishnan et al. 2014; Diaz-Verdugo et al. 2019; Jesuthasan et al. 2021). Exposure to Schrekstoff in combination with water conditioned by mating fish, however, suppresses the alarm response by suppressing activation of the Dp (Diaz-Verdugo et al. 2019). Similar to the dynamics described above in the Vv, hormone receptors in the Dp and/or habenula may bias sensitivity to one pheromone over another, favoring the recruitment of distinct cell populations responsible for a specific behavior given the chemical context and the organism’s internal state (Blumhagen et al. 2011; Jetti et al. 2014). Little research has addressed the mechanisms by which detection of chemical cues can modulate the behavioral response to other cues in fish, despite the fact that in nature organisms are probably often exposed to multiple chemical cues simultaneously or in succession. For example, a male fish that is currently occupied with food seeking may be alerted to the presence of a nearby fertile female by her pheromones. The male must then decide, based on some integration of this chemical information with information about its internal state (hunger level, reproductive state), which chemical cue to follow.

Modulation through non-steroidal mechanisms

In addition to steroid hormones, the teleost OB also receives several types of neurotransmitters, including cholinergic inputs (e.g., acetylcholine, noradrenaline) and gonadotropin releasing hormone (GnRH) through the TN ganglion, as well as serotonergic inputs from raphe nuclei (Satou 1990; Edwards et al. 2007; Lillesaar et al. 2009; Kawai et al. 2010). Nonapeptides, including vasopressin (gene ortholog of fish arginine–vasotocin) and oxytocin (gene ortholog of fish isotocin) and their receptors, are also found in the teleost OB (Kline et al. 2011; Almeida et al. 2012; Huffman et al. 2012). This suggests potential top-down mechanisms from higher brain regions modulate the function of OB via different neurotransmitters. Here, we focus on the modulatory mechanisms of GnRH, serotonin, and nonapeptides, which have been shown to convey status information to olfactory circuits.

Most teleost fish have three variants of GnRH commonly referred to as GnRH1, 2, and 3. GnRH1 regulates the release of pituitary gonadotropins, which ultimately control gonadal steroid production via the HPG. GnRH3 is present in the TN ganglion which projects to the OB, and is an important neuromodulator of reproductive behavior in cichlids (Ogawa et al. 2006). Notably, in zebrafish, a loss of the gnrh1 gene has resulted in this species’ use of GnRH3 to activate the pituitary gonadotropes (Okubo and Aida 2001). In the cichlid O. mossambicus, GnRH3 neurons in the TN are sexually dimorphic in number and correlated to male nesting behavior (Kuramochi et al. 2011; Narita et al. 2018). In Japanese rice fish Oryzias latipes, TN GnRH3 neurons serve as a switch to activate mating preferences (Okuyama et al. 2014). The baseline levels of TN GnRH3 neuron activity inhibits female receptivity for any males; however, visual familiarity facilitates TN GnRH3 neuron activity, which correlates with female mating preference for the familiarized male (Okuyama et al. 2014).

In the OB, levels of GnRH3 and its receptor fluctuate as a function of reproductive status (Collins et al. 2001; Maruska and Fernald 2010). In cichlids, GnRH receptor expression increases in reproductive female OB, suggesting that reproductive females may increase GnRH sensitivity in the OB, and thereby enhance olfactory sensitivity for mate search (Maruska and Fernald 2010). But how does GnRH3 alter olfactory sensitivity in the OB? In goldfish, GnRH3⁺ fibers from the TN are found in both the mitral and the granule cell layers of the OB, and the release of GnRH3 enhances synaptic transmission from mitral to granule cells (Kawai et al. 2010) (Fig. 3). These two cell types form reciprocal dendro-dendritic synapses, where mitral cells form excitatory glutamatergic synapses with granule cells, while the granule cells form GABAergic synapses with mitral cells (Satou et al. 2006). Studies in mammals have shown these circuits enhance the tuning specificity of odor responses through lateral inhibition (Geramita et al. 2016; Shmuel et al. 2019). GnRH3 modulation could either enhance the overall olfactory sensitivity or selectively modulate the responsiveness to reproductive pheromones to facilitate mate seeking. GnRH3⁺ fibers in the goldfish are abundantly distributed in both the lateral and the medial parts of the OB, which separately receive OSN stimuli from food cues and sex pheromones, respectively (Sorensen et al. 1991; Sato et al. 2005; Lastein et al. 2006; Hamdani and Døving 2007; Koide et al. 2009; Yaksi et al. 2009; Yabuki et al. 2016). Thus, one possible mechanism is that GnRH3 alters the olfactory responsiveness for multiple categories of odorants simultaneously and this effect contributes to the modulation of some aspects of reproductive behaviors. Alternatively, though GnRH3⁺ fibers are distributed in both the lateral and the medial parts of the OB, it is still possible that GnRH3 from the TN is not equally dispersed to these regions, or GnRH3 acts differentially across glomeruli. If that is the case, then GnRH3 could fine-tune the responsiveness to reproductive pheromones by selectively activating the medial parts of the OB to facilitate a mate search. The OB also receives serotonergic inputs from raphe nuclei (Lillesaar et al. 2009). Levels of serotonin and its receptor have been shown to correlate with social status (Edwards and Kravitz 1997; Loveland et al. 2014). In mice, serotonergic fibers from the raphe nuclei innervate OB glomeruli, where serotonin directly binds to 5-HT2A receptors to excite tufted cells and depolarize mitral cells, and then modulate odor information processing in the OB (Brill et al. 2016). However, the teleost OB lacks tufted cells; thus how serotonergic inputs regulate OB function requires further study.

Vasopressin and oxytocin are synthesized by neurons localized in the parvocellular and magnocellular nuclei of the POA, and project to multiple extrahypothalamic regions and the neurohypophysis where they are released in the circulatory system (Goodson and Bass 2001; Goodson et al. 2003; Thompson and Walton 2009). In fish, brain vasopressin and oxytocin fluctuate in response to social and reproductive status (Greenwood et al. 2008; Almeida et al. 2012; Kleszczyńska et al. 2012), and their receptors are expressed in the granule cell layer of the OB (Huffman et al. 2012). These findings suggest a potential role for nonapeptide modulation of the OB during pheromone-guided behavior. Indeed, whole-brain isotocin levels of male zebrafish increase in response to sensing female reproductive pheromone PGF_2α (Altmieme et al. 2019). Studies in mammals suggest that nonapeptides modulate olfactory processing through the principal output cells. In mice, direct infusion of oxytocin in the OB increases olfaction-driven performance in a social interaction task. Oxytocin increases odor-evoked responses but reduces the spontaneous firing rate of mitral and tufted cells, which enhances discrimination of odor cues (Sun et al. 2021). However, how exactly nonapeptides modulate the teleost OB processing requires further investigation.

Pheromone processing across the whole brain

After initial encoding by the OE, pheromonal information is processed in downstream brain regions, driving changes in neuronal activation as well as changes in transcriptomic state that may persist for hours or days (Chung-Davidson et al. 2008; Lado et al. 2013; Simões et al. 2015). Hormone receptors are distributed in these olfactory recipient brain regions, which also sense the circulating hormonal milieu that conveys information regarding the individual’s reproductive and social status. While certain pheromone-driven social behaviors have been experimentally linked to specific cell populations and brain regions as described above, more general models of social behavior regulation emphasize interactions between brain regions in a distributed network called the Social Behavior Network (SBN) (Goodson 2005; O’Connell and Hofmann 2011; Ogawa et al. 2021). Indeed, many of the regions implicated in pheromone processing above (Vv, POA, and VMH) are members of the SBN. The SBN is characterized in part by its sensitivity to hormones and neuromodulators, which act acutely on the SBN to sculpt functional connectivity between network nodes in real time (Remage-Healey 2014; Johnson and Young 2017).

The nonapeptides oxytocin and vasopressin are well-studied neuromodulators of social behavior across vertebrates (Godwin and Thompson 2012; Johnson and Young 2017). While these nonapeptides are well-established regulators of behaviors with known pheromonal components like aggression and mating, there is not yet direct evidence in fish that they modulate pheromone perception. Nonapeptide receptors are expressed in the fish Vv, POA, tuberal hypothalamus (encompassing VMH and LH), and dorsal telencephalon, providing a broad anatomical substrate necessary for modulation of central pheromone processing (Moons et al. 1989; Lema et al. 2015). Nonapeptide receptor expression across the brain is indeed modulated according to social and reproductive status (Almeida et al. 2012; Kleszczyńska et al. 2012; Lema et al. 2015), and vasotocin expression is also regulated by social status (Greenwood et al. 2008). It is, therefore, plausible that nonapeptide signaling is another key molecular mediator of internal status effects on central pheromone processing, although further work is needed to confirm this in fish.

Nonapeptides and other neuromodulators discussed in this review act acutely to reorganize SBN-wide functional connectivity, which has been associated with differences in behavioral state (Goodson and Kabelik 2009; Johnson and Young 2017). This, alongside changes to anatomical connectivity necessary for behavioral changes (Inoue et al. 2019), suggest a capacity for molecular correlates of internal state to modulate pheromone-mediated behavior by changing how network nodes interact with one another. The interactions between these brain regions and others, their sensitivity to an animal’s physiological state (in the form of hormones), and their sensitivity to a nearby other’s physiological state (in the form of pheromones) ultimately allow an organism to integrate this information at the level of the CNS and motivate an optimal behavioral response.

Conclusions and future directions

The past several decades have seen enormous progress in characterizing pheromonal ligands, their corresponding receptors, and neural circuits that process olfactory information. However, many pheromone receptors and their ligands, particularly among fish, remain to be identified. Deorphanization of olfactory receptors is challenging with in vivo methods, and likewise it is difficult to identify pheromonal components from complex odor mixtures. High-throughput deorphanization of mammalian olfactory receptors can be achieved by using receptor transport proteins (RTPs) to drive olfactory receptor expression in heterologous cells (Zhuang and Matsunami 2008). Similar methods were recently applied in zebrafish to identify Ora5 and Ora6 as the receptors for lithocholic acid (LCA) (Cong et al. 2019), and in lamprey to identify Or320a and Or320b as the receptors for the male sex pheromone 3-keto petromyzonol sulfate (3kPZS), and Taar348 as the receptor for the male lamprey polyamine pheromone spermine (Scott et al. 2019; Zhang et al. 2020) (Table 1). Recent advances in transcriptomics with cellular resolution can characterize olfactory circuit activation and organization by mapping neural activity proxies, olfactory receptor expression, and downstream targets of OSNs (Zhu et al. 2022). Techniques such as LC–MS hold promise for rapid identification and quantification of fish pheromone ligands from odor mixtures (Behrens et al. 2014; Chen et al. 2022). Application of these methods in fish species will allow for targeted analyses of how pheromones and their receptors function in behavior.

Beyond expanding known pheromonal ligand–receptor pairs, there is a lack of understanding of the evolution of ligand–receptor pairs over time, and how they contribute to species-specific behaviors. Though cyprinids, including zebrafish, express Or114 and demonstrate preference for PGF_2α, most fish species lack this receptor and the few that have been tested do not sense PGF_2α (Stacey 2014; Li et al. 2023). This suggests that the loss of that receptor may cause the loss of preference for PGF_2α. In the case of cichlids, a metabolite of PGF_2α has likely emerged as an alternate signal of reproductive status (Li et al. 2023). Changes in pheromone processing systems could also drive the emergence of reproductive barriers and even new species. A pair of closely related corn borer moth species that can interbreed respond to subtly different sex pheromones, and this difference in responsiveness is due to a single amino acid replacement in the pheromone receptor (Leary et al. 2012). Likewise, closely related Drosophila species exhibit dramatically different behavioral responses when exposed to the same female sex pheromone (D. melanogaster males approach, while D. simulans males avoid). This difference in behavioral response is due to differential wiring of central olfactory processing circuits and is associated with a specific region of the X chromosome, although specific gene(s) involved have yet to be identified (Seeholzer et al. 2018; Shahandeh et al. 2020). Research in fish, taking advantage of the ease of genetic manipulation and species diversity, has great potential to reveal evolutionary mechanisms in a vertebrate system.

Importantly, these neural circuits that mediate diverse reproductive behaviors are sensitive to internal status, particularly hormonal state. How do hormones modulate olfactory neural circuits? Numerous studies demonstrate hormonal modulation of pheromone-guided behaviors in fish, but more work is needed to identify the underlying mechanisms and neural substrates for hormonal modulation. For many fish species, we know little regarding the suite of hormone receptors expressed in the olfactory system, and even less about the cell types that express them. High-throughput sequencing technologies are powerful tools for characterizing hormone receptor expression in the olfactory system, as single cell RNA-sequencing can be used to characterize molecular heterogeneity in peripheral and central olfactory neural circuits (Saraiva et al. 2015; Tepe et al. 2018; Chen et al. 2021). Further, expression profiling and chromatin profiling experiments can identify the genomic targets of hormone receptors in neural circuits (Gegenhuber et al. 2022). Finally, genome editing techniques such as CRISPR/Cas9 can be leveraged to understand the functional role of specific hormone and pheromone receptors in a variety of behaviors (Juntti et al. 2016; Alward et al. 2020). Together, these techniques provide exciting avenues for future study of hormonal modulation of pheromonal communication in fish.

Notes

From the symposium “Neuroethology in the age of gene editing: New tools and novel insights into the molecular and neural basis of behavior’’ presented at the annual meeting of the Society for Integrative and Comparative Biology, January 3–7, 2023.

Funding

This work was supported by grants to S.A.J. from the National Institutes of Health (R35GM142872); the National Science Foundation (IOS-1825723); the Human Frontiers in Science Program (RGY0079-2018); and a fellowship to C.G.P. from the National Science Foundation (DBI-2209257).

Conflicts of interest

The authors have no conflicts of interest to report.