Bioelectric-calcineurin signaling module regulates allometric growth and size of the zebrafish fin

The establishment of relative size of organs and structures is paramount for attaining final form and function of an organism. Importantly, variation in the proportions of structures frequently underlies adaptive change in morphology in evolution and maybe a common mechanism underlying selection. However, the mechanism by which growth is integrated within tissues during development to achieve proper proportionality is poorly understood. We have shown that signaling by potassium channels mediates coordinated size regulation in zebrafish fins. Recently, calcineurin inhibitors were shown to elicit changes in zebrafish fin allometry as well. Here, we identify the potassium channel kcnk5b as a key player in integrating calcineurin’s growth effects, in part through regulation of the cytoplasmic C-terminus of the channel. We propose that the interaction between Kcnk5b and calcineurin acts as a signaling node to regulate allometric growth. Importantly, we find that this regulation is epistatic to inherent mechanisms instructing overall size as inhibition of calcineurin is able to bypass genetic instruction of size as seen in sof and wild-type fins, however, it is not sufficient to re-specify positional memory of size of the fin. These findings integrate classic signaling mediators such as calcineurin with ion channel function in the regulation of size and proportion during growth.


Introduction 45
The establishment of relative proportion of structures and organs is essential for the normal 46 physiology and function of an organism. The study of differential growth of structures in 47 development and in the evolution of form has a rich history (Gayon, 2000). A key advance in 48 our understanding differential growth stems from the work of D'Arcy Thompson and his efforts 49 to define the underlying rules of coordinated transformations in form (Thompson, 1917). This 50 foundational work was leveraged by Huxley and Teissier who formalized scaling relationships 51 between structures within organisms as a power law that details the relative proportion of 52 structures (Huxley and Teissier, 1936;Julian Huxley, 1932)). This law provides a scale-53 independent means of comparing growth in development and among species. 54 A common mechanism for the establishment of organ size and the relative proportions of 55 structures is through differential sensitivity of organs to a systemic growth signal such as insulin-56 like growth factors (IGF) or growth hormone (GH) (Bryant and Simpson, 1984;Conlon and 57 Raff, 1999). However, there is substantial evidence for an organ-intrinsic capacity to establish 58 6 Modification of kcnk5b function by calcineurin. Kcnk5b could mediate FK506 growth effect 151 indirectly or through modification of calcineurin regulation. Increasing the levels of wild-type 152 kcnk5b locally in the fin is sufficient to induce fin overgrowth (Perathoner et al., 2014). As large 153 changes in transcription are observed in regenerating fins treated with FK506 (Kujawski et al.,154 2014), we hypothesized that FK506 mediated overgrowth could be due to upregulation of kcnk5b 155 expression levels. However, we find that expression of kcnk5b is not significantly altered in 156 wild-type regenerating fins after FK506 treatment ( Figure 3A). To test the potential for direct 157 modulation of channel activity by FK506/calcineurin, we assessed the change in conductance of 158 It has been shown that the activity of another KCNK two-pore potassium channel family 164 member, Kcnk18/TRESK, is regulated by direct binding of calcineurin to the intracellular loop 165 of the channel. This binding affects downstream dephosphorylation events on conserved serine 166 residues on the C-terminus of the channel (Czirják and Enyedi, 2006;Czirják et al., 2004). 167 Intriguingly, we identified a putative calcineurin binding site in the cytoplasmic C-terminus of 168 zebrafish Kcnk5b that is similar to that identified in KCNK18/TRESK ( Figure 3B). Czirják et al 169 previously demonstrated that an isoleucine to alanine mutation in this domain in TRESK 170 attenuated the effect of calcineurin and its binding to the channel (Czirják and Enyedi, 2006). We 171 mutated the co-responding residue in the presumptive binding site in Kcnk5b (I290A; Figure  172 3B) and assessed the effect on conductance of the channel in oocytes. Oocytes expressing 173 Kcnk5b/I290A had lower conductance than wildtype, but were similar to wild-type channels 174 treated with FK506 ( Figure 3C-E). Importantly, the Kcnk5b/I290A expressing oocytes were 175 also non-responsive to FK506 treatment. Thus, the effect of the altering the molecular 176 characteristics of this site on the Kcnk5b C-terminal intracellular domain is comparable in effect 177 to chemical inhibition of calcineurin by FK506. This result suggests that, similar to KCNK18, 178 the activity of Kcnk5b is modulated by calcineurin through interactions with the cytoplasmic C-179 terminus.

7
The alf mutation affects the last transmembrane domain of the channel located just prior 181 to the predicted calcineurin binding site (Perathoner et al 2014, Figure S3). In an effort to 182 screen for further variants affecting growth within this region, we used CRISPR-targeted Cas9 183 genetic editing with guides targeted to the fifth exon of kcnk5b, which contains the 184 transmembrane and C-terminus, and screened adults for somatic clones sufficient to cause 185 overgrowth. We recovered fish having overgrowth caused by guides targeting the 186 transmembrane region of the channel. Analysis of Kcnk5b in these clones revealed formation of 187 frameshift mutations leading to early termination ( Figure S3). Thus, loss of the last 188 transmembrane domain and the cytoplasmic tail of Kcnk5b is sufficient to lead to increased 189 proportion. fins and suggest that calcineurin inhibition leads to a re-specification of the regenerating tissue to 195 a proximal identity. However, these data are also consistent with the hypothesis that the 196 treatment leads to a sustained increase in the rate of growth that then is associated with broader 197 domains of regional markers in the enlarged regenerated tissue. To address the effect of We first approached this question by addressing how fins respond when FK506 is 203 removed. Pectoral fins regenerating under the influence of FK506 exhibited increased growth 204 (Figure 2, S1). However, after removal of the drug, either prior to achieving their original size, 205 or in fins that had surpassed their original size, regenerative growth quickly stopped regardless of 206 the extent of the fin regenerate ( Figure 4A). Thus, FK506 induces an acceleration of growth that 207 requires continual presence of the drug for sustained overgrowth. We extended these studies to 208 address if the identity of the regenerated tissue was altered by treatment with FK506. As 209 amputated fins will regenerate back to a size similar to that originally specified during 210 development, the fin tissue remaining after amputation retains information concerning size and 211 proportion that is then integrated in the regenerate. However, if overgrown fins caused by 212 FK506 inhibition of calcineurin are resected, either into the original fin tissue or within a 213 regenerated portion of the fin, these resulting regenerates grew only to the original size of the fin 214 prior to treatment with FK506 ( Figure 4B,C). Similarly, FK506-treated overgrown fins cut 215 beyond the original wild-type fin length failed to grow further upon amputation without the 216 presence of the drug ( Figure 4C). Similar effects were seen in when tested in both pectoral as 217 well as caudal fin regenerates ( Figure S2). Thus, Kcnk5b/calcineurin signaling does not lead to 218 lasting alterations in positional identity, but rather regulates rate of growth leading to enhanced 219 fin proportions. caused by calcineurin inhibition. Importantly, our work here demonstrates that 240 Kcnk5b/calcineurin signaling is sufficient to increase growth regardless of innate size and 241 genotype and thus override, but not re-specify, positional identity of the fin. 242 We demonstrate that Knck5b, like Kcnk18, can signal through calcineurin via a region in 243 its C-terminus. Altering the presumptive binding site on the C-terminus of Kcnk5b (I290A) is 244 sufficient to mirror the effect of calcineurin inhibition and to remove sensitivity to FK506 245 treatment for conductance. FK506 treatment of Kcnk5b in oocytes leads to decreased current. through interactions with its C-terminus, may act to adjust and/or respond to changes in voltage 255 to regulate growth rate. Interestingly, we have found that premature truncation mutations 256 leading to channels that lack the last transmembrane domain and the C-terminus of Kcnk5b 257 result in fin overgrowth comparable to that seen in the alf mutant, which affects the identical 258 region (F241Y, Figure S3). As null alleles of kcnk5b do not cause an overgrowth phenotype, the 259 truncation mutation appears to cause an increase in function of the channel. Alteration of the 260 pore structure or conformation/presence of the C-terminus may cause dysregulated channel 261 activity leading to overgrowth. The regulation of calcineurin may reflect downstream response to 262 changes in conductance or conformation and thus, link calcium signaling to changes in pore 263 conductance and/or conformation. 264

Coordination of growth and proportion 265
Regulation of overgrowth by Kcnk5b/calcineurin signaling is coordinated among diverse tissues 266 of the developing and regenerating fin to establish a larger, functional structure. This 267 coordination may be attained through a re-interpretation of positional information within the fin 268 such that regenerating tissues replace the missing components with tissue in register with the 269 distal edge of the remaining fin stump. Positional information within the fin may be set up and 270 driven by differential gene expression patterns in development (Rabinowitz et al., 2017;Tornini 271 and Poss, 2014). Rabinowitz et al. (2017) detail expression profiles of wild-type fins that 272 support dynamic regulation of gene expression across the proximal-distal aspect of the fin. 273 Interestingly, a significant class of genes differentially isolated in these studies was ion channels, 274 suggesting that bioelectric signaling may be a key factor of this asymmetry and establishment of 275 positional cues. The work on calcineurin by Kujawski et al. (2014) also would point to 276 calcineurin signaling as mediating position and being able to re-specify identity when altered. 277 This mechanism was founded on extended gene expression domains in the proximal component 278 of the fin and delayed distal bifurcation of the rays after calcineurin inhibition. However, an 279 alternate explanation for the observed changes in pattern is that the expanded proximal molecular 280 and anatomical identities observed in FK506-treated fins are a consequence of increased growth 281 rate, such that larger regional domains, interpreted as position, are generated rather than specific 282 changes in identity per se. This hypothesis would explain observed overgrowth as well as 283 extended branching of the fin, but would be independent of changes in positional memory. 284 Alternatively, positional information is independent of size instructive signals suggesting that 285 modification of identity markers stemming from Kcnk5b/calcineurin signaling is coincident, but 286 independent from, mechanisms of size determination and relative proportion. 287 Our data link the growth regulation of Kcnk5b and calcineurin and demonstrate that their 288 function is not sufficient to re-specify identity, rather to override it. Establishing, or re-289 establishing, identity in the fin may require signals present only during development and may not 290 be malleable during regeneration. Long-term treatment of fish during development with FK506 291 leads to global alterations in growth of the body that make comparison of proportion difficult 292 ( Figure S4). Thus, identification of these developmental cues to regulate proportion and size will 293 require dissociation by other means. Our findings suggest that the regulation of growth by 294 Kcnk5b/calcineurin fulfills the general requirements of a scaling property previously defined by 295 relative growth comparisons and represented by the rate constant of Huxley and Teissier (1939). 296 The encoding and specification of absolute size, however remains undefined.  The zebrafish AB strains and Tübingen (Tü) strains carrying albino mutation were used as 524 background for all experiments. Fish were bred and maintained as previously described 525 (Nüsslein-Volhard and Dahm, 2002). All experimental procedures involving fish conform to 526 AAALAC standards and were approved by institutional IACUC committees. A complete 527 description of the husbandry and environmental conditions in housing for the fish used in these 528 experiments is available as a collection in protocols.io dx.doi.org/10.17504/protocols.io.mrjc54n 529 For all experiments, adult stages defined by reproductively mature fish >3 months old were used 530 for analysis. Mutant alleles used in this work are alf dt30mh , sof dj7e2 , and kcnk5b j131x8 . 531

Fin Regeneration and Measurements 532
Fish were anesthetized by treatment with tricaine for measurements and pectoral fin amputation. 533 Unless otherwise indicated, pectoral fins of wildtype and mutants were resected to roughly 50% 534 their pre-cut length using standard surgical scissors. Fin length was measured before, during and 535 after regeneration using handheld calipers and measurements were normalized to the fish 536 standard length, determined as the length from the tip of the snout to the posterior end of the 537 caudal peduncle. 538 539 540

FK506 Treatment 541
Stock FK506 (Sigma) was dissolved in dimethyl sulfoxide (DMSO; Sigma). Fish water was 542 treated with 100nM FK506 unless otherwise indicated. Up to five individual zebrafish were 543 housed in 1L of FK506-or DMSO-treated fish water and were fed daily with live artemia. 544 Water and drug were refreshed every other day throughout the duration of the experiment. 545

3'. 558
Electrophysiology 559 The cDNA of Kcnk5b was subcloned into pSGEM for cRNA expression. The plasmid was 560 linearized using NheI and cRNA was in vitro synthesized using the T7 mMessage mMachine kit 561 (Ambion). Xenopus laevis oocytes were provided by Ecocyte Bioscience. Oocyte handling, 562 injection and electrophysiological recordings were as previously described (Seebohm et al., 563 2005). Briefly, stage V oocytes were injected with 4 ng of cRNA encoding wildtype or mutant 564