Significance and Additional Interpretations

Though the story is not yet complete, Chapter 3.2 represents the most comprehensive study, to date, of the role of Neuregulin1 in any model organism. Phenotypic analysis of Nrg1 mutants lacking Nrg1-I/II, Nrg1-I/II/II or Nrg1-III isoforms indicates that Nrg1 is dispensable for cardiovascular

development and function under homeostatic conditions in zebrafish. However, loss of Nrg1-III causes malformations in the ventricular nerve plexus, which have later cardiovascular consequences, ultimately leading to mortality. In addition to these findings, the novel mutant lines generated in this study are a significant addition to the zebrafish community. Combined with existing lines targeting nrg1-III (nrg1z26) , erbb2 (erbb2st61), erbb3 (erbb3bst14), and erbb4 (erbb4bsa21550 ), these will be important

for deciphering the precise role of nrg1 gene products during development (Busch-Nentwich, 2013; Lyons et al., 2005; Perlin et al., 2011).

To our knowledge, though the physiological role of cardiac innervation in regulating heart rate and cardiac output has been well-studied historically, and the overall structure of the adult cardiac plexus has been defined, the potential role of non-cell autonomous role of cardiac nerves in regulating heart

development has been largely unexplored (Nilsson, 2011; Stoyek et al., 2015). Given that nrg1nc26

mutants show signs of ventricular plexus malformations as early as SL5-6 (15 dpf) when nrg1WT ventricles are just beginning to show signs of intrinsic innervation, and we observed structural changes in the hearts of nrg1nc26 mutants, there may be previously unappreciated interplay between cardiac-associated nerves and myocardial maturation. Recent work has demonstrated that nerves are involved in regulating cardiomyocyte proliferation and regeneration in mice and zebrafish, but the homeostatic role of interplay between these cells types is unknown (Mahmoud et al., 2015). Our nrg1 mutants could be a valuable tool for exploring this relationship.

Additional interpretations

As discussed in Chapter 3.2, further studies are necessary to understanding the phenotypes of nrg1 mutant fish. However, additional interpretations of nrg1 mutant phenotypes and future directions warrant discussion.

Mechanism of cardiac innervation

How nrg1-III regulates formation of the cardiac nerve plexus at the cellular level remains an area of active research. Pigmentation and jaw phenotypes observed in nrg1-III-deficient fish are consistent with a defect in neural crest (NC) cell migration, suggesting that a neural crest derived cell is the primary source of nrg1 mutant phenotypes. During juvenile metamorphosis, neural crest (NC)-derived

melanophores migrate to the site of stripe formation where they differentiate and produce pigment (Budi et al., 2008; Parichy and Spiewak, 2015; Parichy et al., 2003). This migration requires ErbB signaling as well as thyroid hormone (Budi et al., 2008; McMenamin et al., 2014; Parichy et al., 2003). Interestingly, zebrafish lacking functional erbb3b do not appropriately form trunk dorsal root ganglia (DRG) or sympathetic neurons, both of which originate in the NC (Honjo et al., 2008). Canonically, while these trunk DRG are important for spinal (sympathetic) innervation of the heart, the vagus nerve provides cranial (parasympathetic) innervation, particularly at the pacemaker (Nilsson, 2011). Thus, we suggest NC-derived DRG cells are likely the cell of origin for the innervation phenotypes in nrg1-III-deficient fish. However many questions remain as to the cell autonomous and non-autonomous roles for the nrg1/erbb in establishing cardiac innervation, particularly in the ventricle.

Previous studies demonstrate that Nrg1-III, which is detected by ErbB3/ErbB2 receptors on Schwann cells, stimulates co-migration of Schwann cells with neurons and myelination of the long axons by Schwan cells (Lyons et al., 2005; Perlin et al., 2011). Reduced Schwann cell migration in nrg1z26 or erbb3bst14 mutant fish leads to an excess of neuromasts at 5 dpf (Lyons et al., 2005; Perlin et al., 2011). Interestingly, ectopic expression of Nrg1-III in neurons can partially rescue Schwann cell migratory defects (Perlin et al., 2011). If outgrowth of DRG axons to form the cardiac plexus employs a comparable molecular mechanism, then we expect that ectopic expression may similarly rescue ventricular plexus formation. This could be tested by crossing the nrg1nc26 allele onto the Tg(UAS:hNrg1typeIII);

Tg(S1101:Gal4) transgenic background (Perlin et al., 2011), to produce fish where nrg1-III is the only nrg1 isoform expressed and expression is restricted to neurons. We expect that ventricular coverage of the nerve plexus at SL6-SL10 will be rescued by this genotype. Furthermore, additional studies are required to test the spatiotemporal requirements for nrg1 in establishing and maintaining the ventricle nerve plexus. Such studies would necessitate development of genetic tools for conditional, tissue-specific

deletion of nrg1. By conditionally deleting nrg1 in neurons, cardiomyocytes, or endothelial cells of adult fish at different zebrafish life stages, the developmental and cell autonomous roles for nrg1 could be explored.

In support of an essential role of Nrg1-III in neurons, we have produced two additional nrg1 mutant lines, nrg1nc30 and nrg1nc31,which code for truncation of all isoforms of nrg1 after the

transmembrane domain, have supernumerary neuromasts in the lateral line at 5 dpf (data not shown). The ventricular innervation status of this line is currently under investigation. If membrane targeting is intact, then this mutant produces a partially functional Nrg1 which is predicted to have intact forward signaling but defective reverse signaling. Reverse Nrg1 signaling regulates gene expression in cortical neurons, and could play a similar, cell autonomous role in the neuronal projections which innervate the heart (Pedrique and Fazzari, 2010).

Functional performance

Although cardiac malformations are associated with nrg1-III-deficiency in zebrafish, the observed emaciation and impaired innervation phenotypes suggest that metabolic distress could be a factor contributing to decline and mortality. Proper innervation of the myenteric plexus is important for

propagating contractile waves to promote bulk transit through the gastrointestinal system. If innervation of this plexus is reduced in a manner similar to the cardiac plexus, these fish may have impaired nutrient absorption. Under standard laboratory rearing conditions, nutrient intake is a major limiting factor for zebrafish growth rates after initial larval stages. When wild type and nrg1-III-deficient siblings are reared separately, growth delays between genotypes are not readily observed prior to 6 wpf (~SL15)

(unpublished observation). However, when reared together, differential growth is readily observed at 6 wpf (unpublished observation). Combined, this suggests that even if bulk transit is defective in these fish, sufficient nutrients are absorbed for growth during this time range and that nrg1 mutants are at a

competitive disadvantage to nrg1WT siblings.

Total metabolic and cardiovascular performance are tested in zebrafish using swim performance assays by measuring peak speed, oxygen consumption, and time to fatigue (Palstra et al., 2010; Pelster et al., 2003; Plaut and Gordon, 1994). Such studies are in progress to compare total cardiovascular efficiency in all nrg1 lines, and it will be interesting to identify the size range at which nrg1-III-deficient fish

begin to demonstrate performance defects. This could lead to insights as to the functional consequences of nrg1-III-deficiency.

Potential heart failure model

In mammals, chronic heart failure often features derangements in the neurohormonal system including norepinephrine and adrenergic signaling (Mann and Bristow, 2005; Reed et al., 2014). We suggest that nrg1nc26 mutants may have utility as a scalable model of heart failure where lack of a ventricular nerve plexus could model deranged neurohormonal inputs. Since most mammalian models of heart failure require labor-intensive surgical or pharmacological treatments, as genetic, aquatic model, nrg1nc26 zebrafish could be advantageous to rapidly evaluate efficacy of novel therapeutics (Brown et al., 2016). Some necessary early steps in establishing this model would be to explore the manifestation of known signs and symptoms of heart failure in nrg1nc26 mutants including elevated heart rate, reduced ejection fraction, elevated expression of nppa and nppb, and elongation of myocardial cells (Mann and Bristow, 2005; Patten and Hall-Porter, 2009). Ultimately, if gold standard therapeutics (such AR1 agonists) are effective at reducing some or all of these indicators, nrg1nc26 mutant fish could be transitioned into a scalable pre-clinical model. An additional early step to validating this model is to determine whether cardiac arrhythmia accounts for mortality independent of heart failure. As mutant fish age, they show signs of cardiovascular distress, but also demonstrate an increased sensitivity to

anesthesia suggestive of a propensity to cardiac arrhythmia. Thus, it is unclear whether mutants die ultimately die from heart failure or from acquisition of fatal arrhythmias. This question could be addressed via thorough characterization of the electrocardiographic profile of mutants.