Blog

13th April 2023

Salmonids, a group of fish including salmon, trout*, char, and their relatives such as whitefish and huchen, include many iconic and hugely ecologically and economically important species. Furthermore, the robust and cosmopolitan rainbow trout is a workhorse, the white laboratory rat of the fish physiology world (whilst the smaller and more genetically tractable zebrafish may be argued to be the C57BL/6 lab mouse). Rainbow trout have been studied for decades across the world for cardiovascular studies in particularly, including many investigations on how adrenaline regulates fish heart function.

Salmonid genetics are complicated because the lineage underwent a whole genome duplication approximately 100 million years ago, after their divergence from their closest relatives (other fishes such as pike). This means that the ancestor of salmonids had, more or less, double the number of genes compared to most other bony fish. Some of the duplicate genes with redundant functions have been since lost, but many remain and could have assumed new roles (or subdivided labour between themselves).

To better understand the foundations of adrenergic regulation in salmonids, I undertook a study (Joyce, 2023) to find all of the ß-adrenergic receptor genes (adrb genes) in genomes of several salmonid species, namely: rainbow trout (Oncorhynchus mykiss), Atlantic salmon (Salmo salar), brown trout (Salmo trutta), lake trout (Salvelinus namaycush), lake whitefish (Coregonus clupeaformis), and huchen (Hucho hucho), and then compared them to many other bony fish, including their cousin the Northern pike (Esox lucius). The genomes of these species were scoured using a variety of bioinformatic searches in major genomic databases (NCBI GenBank and Ensembl). 

Together these searches revealed each species of salmonid had seven adrb genes, more than any other bony fish studied, which typically had four or five adrb genes each (pike, for instance, had five adrb genes). Most strikingly, however, salmonids appear to be the only fish (or any known jawed vertebrate) that lack the adrb1 gene, which encodes the ß1-adrenergic receptor. In mammals the ß1-adrenergic receptor is synonymous with the stimulation of cardiac function and accelerated heart rate during stress or exercise. Previous work in salmonids had already suggested that the heart instead primarily relies on other ß-adrenergic receptors (namely ß2-adrenergic receptors), which accords perfectly with these latest findings. The problem arises, however, when this finding has been translated to other bony fish, which have also been widely presumed to mainly express ß2-adrenergic receptors in the heart. Instead, using transcriptomic databases, I showed many bony fish species (including pike as well as other familiar species such as cod) predominantly express adrb1 in the heart, much like mammals and other vertebrates. As such, the most consequential result is that, due to the unusual lack of adrb1, the wealth of studies on adrenergic regulation in salmonids such as rainbow trout should be generalised to other species carefully. Instead, it would be worthwhile to pay greater attention to other fish species. It will also be particularly interesting to try to now work out if there are any particular limitations on salmonids owing to the loss of adrb1.

Joyce, W. (2023). Evolutionary loss of the β1-adrenergic receptor in salmonids. General and Comparative Endocrinology, 338, 114279. https://doi.org/10.1016/j.ygcen.2023.114279

*the common names ‘salmon’ and ‘trout’ are not evolutionarily informative (i.e. monophyletic); a brown trout is more closely related to an Atlantic salmon (both of the genus Salmo) than it is to a rainbow trout, whilst rainbow trout are in turn closely related to Pacific salmon species including sockeye and coho salmon (all belonging to the genus Oncorhynchus). 

2nd January 2023

The contraction of striated muscle – whether it be cardiac muscle responsible for the beating heart or skeletal muscle that moves the body – underlies some of the most fundamental physiological functions in vertebrates. Striated muscle contraction is triggered by the binding of calcium ions to the troponin complex, which thereafter permits actin and myosin sliding filaments to generate force. Troponin is thus at the interface between rising calcium concentration and the molecular motors of the heart, where it is able to influence the speed and force of muscle contraction and relaxation. 

Troponin is made up of three components: troponin C, troponin T and troponin I. Troponin C binds to calcium, troponin T is associated with tropomyosin that gates the access of myosin to actin, and troponin I exerts an inhibitory effect whilst influencing protein interactions within the complex. In mammals such as humans, where the vast majority of research tends to focus, there are a number of different genes of troponin C (two copies), T (three copies) and I (three copies) that are expressed in different types of muscle fibre. For instances, troponin I 1 is expressed in slow twitch skeletal muscle, troponin I 2 is expressed in fast twitch skeletal muscle, and troponin I 3 is expressed in the heart. Each of the different proteins has specific properties that optimize calcium binding in their given muscle type. Troponin I 3, expressed in the heart, is characterised by a putatively unique ‘N-terminal extension’ peptide that makes it sensitive to protein kinase A phosphorylation, which partially controls the increased rate of cardiac relaxation when adrenaline rises (during the ‘fight-or-flight’ response).  

The evolution of troponin, especially the troponin I subunit with its N-terminal extension, has received a great deal of attention over the last 25 years (for example: Hastings, 1997; Rasmussen et al., 2022; Shaffer and Gillis, 2010). Bony fish, such as trout, lack the N-terminal extension in the troponin I expressed in their heart, leading researchers to previously conclude the molecular structure possibly evolved in the ancestors of lobe-finned fishes, and could have been essential for the first tetrapod vertebrates to transition to land. However, all of these studies critically lacked the inclusion of even more distantly related vertebrate groups, for instance sharks and lampreys, which are essential to understand the early stages of troponin evolution. 

In our latest study published in Genome Biology and Evolution, we sought to gain a clearer understanding of troponin evolution by studying expression and comparative genomics in a comprehensive set of vertebrates, including the previously neglected groups such as sharks (Joyce et al., 2023). 

Our first discovery was that there are five, not three, distinct families of troponin I genes, but two of these have been lost in the lineage leading to mammals, so were largely restricted to fishes and had previously been missed. The five families of troponin I genes originated during whole-genome duplications in the earliest vertebrates, which generated multiple copies of many types of genes.

One of the new gene families, which we named troponin I 5, was highly expressed in shark hearts. The biggest surprise was that it bears an N-terminal extension with a striking resemblance to troponin I 3, particularly that found in amphibians hearts. Yet troponin I 5 was clearly phylogenetically distantly related to troponin I 3. Either the N-terminal extension evolved twice in the two proteins independently, or else it was present in the ancestral troponin I protein but was lost in the other lineages. We strongly favour the latter view, because it is difficult to envisage the such a similar (at the molecular level) signature evolving twice and, moreover, because previous researchers had already found structurally similar N-terminal extensions in distantly related invertebrate species. 

Another important finding was that the hearts of bony fish (like trout) express a troponin I in the heart that lacks an N-terminal extension not because it had not yet evolved, but because they lost the troponin I 3 gene, and instead express the troponin I 1 protein (normally expressed in slow skeletal muscle). Interestingly this form of the protein is also found in the hearts of foetal mammals. Troponin I 1 was additionally expressed in variable proportions in the hearts of sharks, alongside troponin I 5. We predict the relative expression of the two forms in shark hearts could be changed to adapt to different environmental conditions, which we endeavour to explore in future studies.

Our study exemplifies the power of studying diverse species, such as sharks, to shed light on the evolutionary history of physiologically important molecules. The results re-shape our view on the origin of the N-terminal extension, and also lay the foundations for future studies to characterise the properties and expression patterns of the newly discovered troponin I proteins.

References:

Hastings, K. E. M. (1997). Molecular Evolution of the Vertebrate Troponin I Gene Family. Cell Struct. Funct. 22, 205–211.

Joyce, W., Ripley, D. M., Gillis, T., Black, A. C., Shiels, H. A. and Hoffmann, F. G. (2023). A revised perspective on the evolution of troponin I and troponin T gene families in vertebrates. Genome Biol. Evol. evac173.

Rasmussen, M., Feng, H.-Z. and Jin, J.-P. (2022). Evolution of the N-Terminal Regulation of Cardiac Troponin I for Heart Function of Tetrapods: Lungfish Presents an Example of the Emergence of Novel Submolecular Structure to Lead the Capacity of Adaptation. J. Mol. Evol. 90, 30–43.

Shaffer, J. F. and Gillis, T. E. (2010). Evolution of the regulatory control of vertebrate striated muscle: the roles of troponin I and myosin binding protein-C. Physiol. Genomics 42, 406–419.

17th October 2022

In the late 1800s, an Italian scientist, Giulio Fano, reported peculiar ‘tonus wave’ contractions in isolated atrium tissue dissected from the hearts of European pond turtles (Fano, 1887). The slow changes in tension were clearly distinct from the faster contractions produced by cardiac muscle in a normal heart beat. This was regarded as a significant discovery at the time, published as part of a ‘Festschrift’ dedicated to the famous German physiologist Carl Ludwig, who Fano had earlier trained with.

In the ensuing decades, numerous independent researchers established that the slow tonus waves were generated by dense smooth muscle in the atrial walls of pond turtles. Smooth muscle is normally distributed in various organs across the body (such as blood vessels, digestive organs, bladder and the uterus) but in most animals it is not typically found in such abundance in the walls of heart muscle. Indeed, this property appeared to be specific to aquatic turtles; Walter Gaskell remarked that he had not seen tonus waves in land tortoises (Gaskell, 1900) during his pioneering work on autonomic nervous control (Gaskell, 1883).

Research on the atrial smooth muscle waned for most of the 20th Century but was eventually revived for new audiences by Gannon et al. (2003) and Galli et al. (2006) who re-reported tonus waves in hearts from pond sliders, relatively close relatives to the European pond turtles that were the subjects of Fano’s work.

The phylogenetic distribution of atrial smooth muscle has recently been clarified by anatomical studies of the hearts from diverse turtle species (Joyce et al., 2020). It is sparse, but detectable is many lineages (such as snapping turtles and side-necked freshwater turtles) and, as pre-empted by Gaskell, effectively absent in land tortoises. Atrial smooth muscle is most prominent in emydid pond turtles, conveniently the focus of the historical studies.

Modern studies have also shown that, in pond sliders, contraction of atrial smooth muscle appears to regulate cardiac filling; smooth muscle contraction shrinks the dimensions of the atria with the power to impede venous return (Joyce et al., 2019). Given that sliders are capable of large changes in cardiac output during everyday diving behaviour, when heart rate characteristically plummets during submersion (dive bradycardia), we predicted the smooth muscle could kick into action under water, negating large increases in stroke volume when cardiac filling time is prolonged. This proposition was strengthened by the conspicuous absence of smooth muscle in land tortoise atria.

Until now, one prominent group of aquatic turtles has evaded study on atrial smooth muscle almost entirely. Sea turtles show similar cardiovascular control to pond turtles, including dive bradycardia for potentially even more extended periods. If atrial smooth muscle is indispensable during diving, surely the sea turtles should possess it also.

In our new study reported in the Journal of Experimental Biology, Costello and colleagues used histology and immunohistochemistry to search for smooth muscle in the hearts of loggerhead sea turtles. The fixed hearts were provided from colleagues at the Oceanogràfic (València, Spain) from animals that were unable to be rehabilitated after fatal injuries resulting from fishing accidents.

Much to our surprise, the sea turtle hearts essentially lacked atrial smooth muscle, unlike some of their freshwater counterparts. Instead, however, as reported in other turtle lineages, we saw abundant smooth muscle in the sinus venosus, the chamber that receives blood from major veins before it enters the right atrium. At the ‘gateway to the heart’, the smooth muscle could also regulate cardiac filling, including during diving.

This study not only adds to our understanding of sea turtle cardiac function, it forces us to reassess the role of atrial smooth muscle in pond turtles. It now appears less likely as an essential adaptation to diving, although it could still support smooth muscle in the sinus venosus to regulate cardiac filling. We do acknowledge, however, that many questions remain open, and will require further anatomical and physiological studies in diverse turtle species to help answer.

These experiments were led by Leah Costello, a PhD student at the University of Manchester (UK). The project relied on collaborations with Jonathan Codd and Holly Shiels (also UoM) as well as Daniel García-Párraga and Jose Luis Crespo-Picazo (Oceanogràfic, València, Spain).

References:

Costello, L. M., García-Párraga, D., Crespo-Picazo, J. L., Codd, J. R., Shiels, H. A. and Joyce, W. (2022). Absence of atrial smooth muscle in the heart of the loggerhead sea turtle (Caretta caretta): a re-evaluation of its role in diving physiology. J. Exp. Biol. jeb.244864.

Fano, G. (1887). Ueber die Tonusschwankungen der Atrien des Herzens von Emys europaea. J. B. Hirschfeld.

Galli, G. L. J., Gesser, H., Taylor, E. W., Shiels, H. A. and Wang, T. (2006). The role of the sarcoplasmic reticulum in the generation of high heart rates and blood pressures in reptiles. J. Exp. Biol. 209, 1956–1963.

Gannon, B. J., Campbell, G. D., Thomas, A. C. and Snyder, G. K. (2003). Endocardial smooth muscle in an Australian and two North American tortoises: cardiac tonus waves revisited 75 years on. Comp Biochem Physiol 134A, S112.

Gaskell, W. H. (1883). On the Innervation of the Heart, with especial reference to the Heart of the Tortoise. J. Physiol. 4, 43-230.14.

Gaskell, W. H. (1900). The Contraction of Cardiac Muscle. In Textbook of Physiology (ed. Schæfer, E. A.), pp. 169–227. Edinburgh and London: Young. J. Pentland.

Joyce, W., Axelsson, M. and Wang, T. (2019). Contraction of atrial smooth muscle reduces cardiac output in perfused turtle hearts. J. Exp. Biol. 222, jeb199828.

Joyce, W., Crossley, D. A., Wang, T. and Jensen, B. (2020). Smooth Muscle in Cardiac Chambers is Common in Turtles and Extensive in the Emydid Turtle, Trachemys scripta. Anat. Rec. 303, 1327–1336.

15th June 2022

ß-adrenergic receptors, which decorate the surface of heart cells as well as other tissues of the body, are responsible for translating increased extracellular levels of adrenaline (and related signalling molecules) into changes in physiological function. When adrenaline—circulating in the bloodstream or released from neurons innervating the heart—binds to cardiac ß-adrenergic receptors, it sparks an intracellular signalling cascade that culminates in the activation of protein kinases that modulate the activity of ion channels, pumps, and the molecular motors of the heart, ultimately increasing heart rate and contractility. The crucial actions of adrenaline on the heart are broadly similar between different species of vertebrates, from fish to frogs and humans, but we must also be aware of the potential underlying differences.

In recent years, the use of fish, especially the fecund and easily housed zebrafish, in research has proliferated. In the United Kingdom, for instance, fish represent the second most widely used group of research animals, only behind mice and ahead of the idiomatic lab rat or guinea pig. Zebrafish are particularly convenient for cardiovascular research because, especially during early development, the heart is readily visible with a simple microscope, so heart rate and other variables can easily be measured. The zebrafish genome is also well sequenced and annotated, so this species is particularly amenable to genetic manipulation, which provides an essential tool to understand molecular pathways. A problem arises, however, when researchers make a jump from mammals to zebrafish with an assumption that their hearts work in exactly the same way.

The ß-adrenergic receptors come in three general ‘flavours’, known as ß1, ß2 and ß3 sub-types. The categories were defined decades ago based on different sensitivity to specific drugs, but all are involved in responses to adrenergic stimulation. All three types can be expressed in human heart cells, but at least in the healthy state, the ß1 receptor is the most abundant. Early pharmacological work in fish, such as trout and carp, however, indicated ß2 receptors, with different affinities for some drugs, may take centre stage. Some modern cardiovascular studies in zebrafish have nevertheless employed ß1 adrenergic specific drugs, with mixed effects, and it has remained unclear which sub-types are most important in the heart in this burgeoning model species.

In our recent study, published in Acta Physiologica, we aimed to establish the importance of the ß1 adrenergic receptor in the control of heart rate in zebrafish by deleting the gene, using CRISPR-Cas9, and studying the ensuing effects.

Our standout results were as follows:

· Zebrafish larvae lacking the ß1 adrenergic receptor (‘knockouts’) exhibited lower heart rates than wild-type (control) larvae under routine conditions.

· Surprisingly, however, the knockouts still responded to treatment with adrenaline with a similar increase in heart rate to wild-types, indicating they were still capable of mounting a similar response to adrenergic stimulation.

· ß2 and ß3 genes were upregulated in the knockout line, suggesting that other ß-adrenergic receptors may be able to account for the responses to adrenaline in zebrafish.

· ß1 knockout larvae exhibited similar rates of oxygen consumption to wild-types and grew to be fertile adults, firmly establishing that the ß1-receptor is not strictly essential for survival in zebrafish, at least when they are raised in cushy lab conditions.

These results lead us to urge more cautious interpretation of the results of studies in zebrafish in their extrapolation to mammals. Of course, this does not negate the utility of zebrafish in cardiovascular research. Rather the opposite, it is a reminder that nature often showcases diverse solutions to common demands, and that by studying a range of species we can gain an appreciation of how different pathways may be effectively recruited in different circumstances.

This study was borne out of collaboration with Yihang (Kevin) Pan, Kayla Garvey, Vishal Saxena and Steve. F. Perry.

16th April 2022

When we exercise the heart has to deliver more blood to our oxygen-hungry muscles. The amount of blood that the heart pumps per minute, termed cardiac output, is the product of heart rate (number of beats per minute) and stroke volume (the volume of blood pumped per beat). Under most circumstances, heart rate increases to a greater degree than stroke volume, and the same is true for many animals during activity. 

Surprisingly, however, increasing heart rate alone does not increase cardiac output. Studies from humans to alligators have previously shown that increasing heart rate, even beyond the usual maximum bounds, is associated with a concomitant decrease in stroke volume that leaves cardiac output unchanged. Instead, we believe, cardiac output is ultimately determined by the regulation of blood vessels in the peripheral circulation, which determines the rate of venous return. If venous return is ultimately dictating cardiac output, why then, does heart rate even increase during activity? An increase in venous return could otherwise drive an increase in stroke volume, which would probably be more energy-efficient than increasing the pumping rate. 

In our recent study in snapping turtles, published in the American Journal of Physiology, we explored the hypothesis that something may be preventing the increase in stroke volume. We predicted that ‘something’ could be the pericardium, a tough sac that encloses and protects the heart, and is found across vertebrates, including in turtles and humans alike. Heart rate may have to increase during activity to keep stroke volume within the limits that can be accommodated within the pericardium. To investigate this, we devised a way to ‘prevent’ the normal increase in heart rate with a bradycardic drug, ivabradine, otherwise used to decrease heart rate in patients with heart failure. To see if the pericardium may be constraining the heart at artificially low heart rates, in half of the animals it was surgically opened (essentially a ‘pericardiectomy’, as it is known in medicine, which may be performed in patients with a pathologically fibrotic pericardium constricting the heart). Implanted blood flow probes allowed us to measure cardiac output during rest and exercise. 

The experiments revealed that, indeed, when heart rate was decreased with ivabradine, maximum cardiac output decreased, which may limit performance during exercise. To our surprise however, the pericardium was not responsible for the limited rise in cardiac output; turtles with the open pericardium performed just like those with a normal intact pericardium. We wonder if other anatomical constraints within the turtle shell prevent stroke volume from rising further. The story may also not stop here because there is already more compelling evidence for pericardial constraint in exercising mammals than reptiles, so it remains distinctly possible that the pericardium does impose a need to regulate heart rate during exercise in some species, potentially including humans, but this will require future work to validate.

This work was conducted in collaboration with Dane Crossley and Brandt Smith, the two of whom spearheaded the experiments, and Tobias Wang.