Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Sep 1.
Published in final edited form as: Dev Dyn. 2017 Jul 12;246(9):691–699. doi: 10.1002/dvdy.24531

Cx43 suppresses evx1 expression to regulate joint initiation in the regenerating fin

Gabrielle Dardis a, Robert Tryon b, Quynh Ton a, Stephen L Johnson b, M Kathryn Iovine a,c
PMCID: PMC5548607  NIHMSID: NIHMS882182  PMID: 28577298

Abstract

Background

How joints are correctly positioned in the vertebrate skeleton remains poorly understood. From our studies on the regenerating fin, we have evidence that the gap junction protein Cx43 suppresses joint formation by suppressing the expression of the evx1 transcription factor. Joint morphogenesis proceeds through at least two discrete stages. First, cells that will produce the joint condense in a single row on the bone matrix (“initiation”). Second, these cells separate coincident with articulation of the bone matrix. We propose that Cx43 activity is transiently reduced prior to joint initiation.

Results

We first define the timing of joint initiation with respect to regeneration. We next correlate reduced cx43 expression and increased evx1 expression with initiation. Through manipulation of cx43 expression we demonstrate that Cx43 negatively influences evx1 expression and joint formation. We further demonstrate that Cx43 activity in the dermal fibroblasts is required to rescue joint formation in the cx43 mutant, short fin b123.

Conclusions

We conclude that Cx43 activity in the dermal fibroblasts influences the expression of evx1, and therefore the differentiation of the precursor cells that give rise to the joint-forming osteoblasts.

Keywords: skeletal regeneration, zebrafish, gap junction, interzone

Introduction

Proper skeletal development and joint placement are necessary for normal vertebrate structure and movement, yet the molecular mechanisms underlying these events are poorly understood. We utilize the zebrafish regenerating fin model to provide insights into mechanisms regulating bone growth and joint formation. The fin is comprised of multiple bony fin rays, each comprised of multiple bony segments separated by joints. Following amputation, the fin rapidly restores its size and pattern during regenerative growth, allowing for extended opportunities to study segment length and joint formation. New growth occurs at the distal end, therefore younger tissue is always located distal to older tissue (Brown et al., 2009).

Each fin ray is comprised of two hemirays of bone matrix that surround a central mesenchyme. During regeneration, the rapidly dividing cells of the blastema are located medially (and distally), while the differentiating skeletal precursor cells of the osteoblast lineage (i.e. precursors of osteoblasts and joint-forming cells) are located laterally and essentially surround the dividing cells. Clonal analysis of fin growth and regeneration reveals nine lineage classes that contribute to the adult fin (Knopf et al., 2011; Tu and Johnson, 2011). Two of these lineages are responsible for the skeletal elements of the fin. These are the osteoblast lineage including both bone-forming cells and joint-forming cells, and the dermal fibroblast lineage, which includes the loose mesenchymal cells. The first overt sign that a joint will be produced occurs when joint-forming osteoblasts condense as a single row on the surface of the bone matrix (“initiation”). These cells subsequently mature into two rows of joint-forming cells that surround the newly articulated joint (Sims et al., 2009). Earlier events of joint morphogenesis are poorly understood, including how joint-forming cells are instructed to differentiate and to give rise to a joint in the appropriate location.

Our research suggests that the gap junction protein Cx43 influences how these decisions are made (Sims et al., 2009; Ton and Iovine, 2013b). Hypomorphic mutations in cx43 are responsible for the short fin (sof b123) mutant phenotypes, which includes short fins, short bony fin ray segments, and reduced cell proliferation (Iovine et al., 2005). Both cx43 mRNA and Cx43 protein are expressed throughout the medial fin ray mesenchyme (including in dividing cells) and in the adjacent joint-forming cells (Hoptak-Solga et al., 2008; Sims et al., 2009). Our findings reveal that Cx43 function (in one or both of these cellular compartments) is responsible for coordinating skeletal growth with skeletal patterning. An additional mutant, another long fin (alf dty86, with a mutation in kcnk5b, Perathoner et al., 2014), exhibits increased cx43 mRNA levels and overlong segments and fins (Sims et al., 2009; van Eeden et al., 1996). However, because segment length is rescued in alf dty86 regenerating fins following cx43 knockdown, we have suggested that Cx43 suppresses joint formation. This has lead to models whereby the long segment phenotype of alf dty86 is due to stochastic joint failure, and that the short segment phenotype of sof b123 is due to premature joint formation.

The earliest known gene required for joint formation is the even-skipped transcription factor 1 (evx1) (Schulte et al., 2011). The evx1 gene is expressed in a single band of cells at the distal end of fin rays as they produce joints (Borday et al., 2001). Regenerating fins in alf dty86 mutants show stochastic evx1 expression (i.e. consistent with stochastic joint failure), which remarkably, is rescued by cx43-knockdown (Ton and Iovine, 2013b). Thus, we suggest that Cx43 suppresses joint formation in alf dty86 by suppressing evx1 expression. In contrast, sof b123 mutants exhibit evx1 expression in regenerating fin rays more distally than in wild-type (Ton and Iovine, 2013b), consistent with the hypothesis that joint formation is premature. Together, these findings suggest a model where the differentiation of joint-forming cells is suppressed by Cx43 function in a way that permits joint-forming cells to differentiate at the appropriate time and place. This model predicts that Cx43 function must be reduced for joint initiation to occur. Here, we provide support for this model by showing that cx43 expression is negatively correlated with joint initiation, and we demonstrate that manipulation of Cx43 activity influences both the expression of evx1 and joint formation. Moreover, we demonstrate that Cx43 function in the dermal fibroblast lineage, but not in joint-forming osteoblasts, is required to rescue both regenerate length and segment length in sof b123 fins. This study provides evidence that Cx43 function suppresses the joint-forming cell fate in a non-autonomous manner.

Results

Cx43 function is reduced when joint formation is initiated

In order to monitor Cx43 activity during joint formation, we first identified a timeline for joint initiation during regeneration. Our prior studies on joint morphogenesis identified ZNS5-positive cell condensations, similar to the interzone cells of synovial joints, that predict the location of the future joint (Sims et al., 2009). Distally, ZNS5-positive joint cells exhibited an elongated morphology in a single row of cells during joint “initiation”. More proximally, joint-forming cells appeared to separate away from each other and, finally, joint cells were located in two rows of cuboidal cells on either side of the articulation in the bone matrix. These prior studies were completed on a population of fin rays at 5 days post amputation (5 dpa). To establish a timeline for the initiation phase of joints, we monitored distal ZNS5-positive condensation of elongated cells from a single fin ray (i.e. the third fin ray from either the dorsal-most or ventral-most sides of the fin) over time. Since the rate of regeneration is more rapid following proximal amputations (Lee et al., 2005), we amputated fins slightly more distally to increase the amount of time between transitions in joint morphology (i.e. amputations were performed at the 30% level vs. the 50 % level). During this timecourse, ZNS5-positive cells were in initiation at about 87 hours post amputation (hpa) (Figure 1).

Figure 1.

Figure 1

Open in a new tab

Establishment of the joint initiation timeline. Fins were amputated at the 30% level and permitted to regenerate for 2–5 days before processing for ZNS5 immunofluorescence. Selected timepoints are shown. Fins at 87 hpa frequently exhibited “initiating” joints in the third fin ray from the dorsal or ventral side. The ratios in each image indicate the frequency of initiating joints at each time point. Arrowheads indicate the organizing joint-forming cells and the * in 96 hpa identify the separating rows of joint-forming cells. The scale bar is 20 μm.

Next, we wished to correlate Cx43 activity during the timeline. We followed cx43 mRNA levels pre-initiation (72 hpa), at initiation (87 hpa), and post-initiation (96 hpa) using both whole mount in situ hybridization and cryosection in situ hybridization (ISH) to visualize the qualitative changes in expression levels of cx43 mRNA. Importantly, we observed a modest but detectable reduction in cx43 mRNA levels at 87 hpa (Figure 2). Because evx1 is required for joint formation (Schulte et al., 2011), we correlated its expression during the joint formation timeline. In contrast to cx43, we observed an increase in evx1 mRNA levels at 87 hpa by both whole mount in situ hybridization and by in situ hybridization on cyrosections (Figure 2). Therefore, we find a negative correlation between cx43 and joint initiation and a positive correlation between evx1 and joint formation.

Figure 2.

Figure 2

Open in a new tab

Expression of cx43 and evx1 during the joint initiation timeline. Qualitative changes in cx43 and evx1 gene expression are shown. Whole mount in situ hybridization for cx43 is shown at 72 hpf (A), 87 hpf (B), 96 hpf (C). Cryo-in situ hybridization for cx43 is shown at 72 hpa (A′), 87 hpf (B′), 96 hpf (C′). Whole mount in situ hybridization for evx1 is shown at 72 hpf (D), 87 hpf (E), 96 hpf (F). Cryo-in situ hybridization for evx1 is shown at 72 hpa (D′), 87 hpf (E′), 96 hpf (F′). Asterisks identify regions of gene expression. Note that cx43 mRNA is detected medially in the dermal fibroblasts and laterally in a subset of cells in the osteoblast lineage, while evx1 mRNA is restricted to a subset of cells in the osteoblast lineage.

We additionally utilized qRT-PCR to further demonstrate the differences in cx43 and evx1 mRNA levels in a more quantitative manner. We compared ΔCT values between samples of different time points, where ΔCT represents the normalized cycle when the threshold for detection is reached. Therefore, higher ΔCT values indicate lower levels of template mRNA while lower ΔCT values indicate higher levels of template mRNA. Indeed, we found that ΔCT values were increased for cx43 at 87 hpa and decreased for evx1 at 87 hpa in three independent biological replicates (Table 1), consistent with our in situ hybridization results. We note that while the three replicates consistently indicate relatively higher evx1 expression at 87 hpa, the results are somewhat variable. We attribute this finding to the fact that relatively few cells express evx1 specific to the most distal joint within the regenerating tissue. Overall, the qRT-PCR results confirm our findings by in situ hybridization and provide evidence that cx43 mRNA levels are reduced during joint initiation while evx1 mRNA levels are increased.

Table 1.

aΔCT values for cx43 and evx1 during the joint initiation timecourse.

Biological replicates cx43 72 hpa cx43 87 hpa cx43 96 hpa evx1 72 hpa evx1 87 hpa evx1 96 hpa
 1 3.27 5.41 4.23 9.84 9.28 9.68
 2 2.88 4.48 3.39 8.11 3.31 8.19
 3 3.16 4.24 3.98 11.37 10.35 12.76
Open in a new tab
a

ΔCT values are determined by subtracting the average actin CT value from the average CT value for either cx43 or evx1.

Increased Cx43 reduces the frequency of evx1-positive rays and influences segment length

Based on the above correlative studies, we predict that the manipulation of Cx43 activity will influence both evx1 expression and joint initiation. For example, increasing Cx43 function prior to joint initiation should inhibit evx1 expression and delay joint initiation, leading to joint failure or overlong segments. To overexpress Cx43, we took advantage of a transgenic line created to inhibit miR-133a, which targets cx43 mRNA for degradation (Yin et al., 2012). Inhibition of miR-133 is accomplished by inserting several miR-133 binding sequences upstream of egfp and downstream of an inducible heat shock promoter hsp70. This line, referred to as the “sponge” line (Tg(hsp70:miR-133sp)pd48), sequesters endogenous miR133a and therefore increases cx43 expression (note that evx1 does not have the predicted miR-133a binding site, Yin et al., 2012). Indeed, we find that cx43 levels are increased in Tg(hsp70:miR-133sp)pd48 at 24 hours following a one-hour heat pulse (Figure 3). Moreover, in a prior study we showed that Cx43 protein levels are increased by 50 % in Tg(hsp70:miR-133sp)pd48 treated with a heat pulse compared to similarly treated non-transgenic siblings (Banerji et al., 2016).

Figure 3.

Figure 3

Open in a new tab

Increased cx43 at the time of joint formation reduces the frequency of evx1-positive rays. (A) Whole mount ISH shows increased cx43 and decreased evx1 expression at the time of joint initiation (87 hpa) in the Tg(hsp70:miR-133sp)pd48 line following a heat pulse at 72 hpa. Transgene-positive fins, sp(+) (n = 9 fins), were compared to non-transgenic sp(-) siblings (n = 7 fins). (B) The sp(+) transgenic fish showed reduced frequency of evx1 expression compared to the sp(-) non-transgenic sibling. Both sp(+) and sp(-) fish were similarly treated for heat shock. Statistically significant differences were determined by the student’s t-test where p < 0.01 (**). Error bars represent the standard deviation.

We tested if increased cx43 influences evx1 expression at the time of joint initiation. Fins were amputated at the 30 % level, followed by a one-hour heat pulse at 72 hpa (pre-initiation) in Tg(hsp70:miR-133sp)pd48. We harvested fins at 87 hpa (initiation) and evaluated evx1 expression compared to non-transgenic siblings that received the same treatment. Since elevated cx43 in alf dty86 fins causes stochastic evx1 expression (Ton and Iovine, 2013b), we calculated the percentage of rays expressing evx1 (evx1-positive) rather than evaluate expression qualitatively. We compared the percent evx1-positive fin rays across the entire fin between Tg(hsp70:miR-133sp)pd48 and non-transgenic siblings. As expected, we found a significant decrease in the percentage of evx1-positive fin rays in Tg(hsp70:miR-133sp)pd48 fins, from 87.9% (109/126 fin rays) in the non-transgenic sibling to 57.2% (94/179 fin rays) in the sponge line (Figure 3).

Following the same timing of amputation and heat treatment we next measured regenerate length and segment length (i.e. as a proxy for joint formation), by harvesting at 96 hours post heat shock-induction. We found a significant increase in both regenerate length and in segment length in the Tg(hsp70:miR-133sp)pd48 line compared to the non-transgenic siblings (Figure 4).

Figure 4.

Figure 4

Open in a new tab

Increased cx43 in the Tg(hsp70:miR-133sp)pd48 line leads to an increase in regenerate length (A) and segment length (B) in sp(+) (transgene-positive, n = 12) compared to similarly treated sp(-) siblings (n = 10). Both sp(+) and sp(-) fish were similarly treated for heat shock. Statistically significant differences were determined by the student’s t-test where p < 0.05 (*) or p <0.001 (***). Error bars represent the standard deviation.

To supplement our transgenic manipulation of Cx43 expression levels, we additionally treated wild type zebrafish with FK506 (tacrolimus), a calcineurin inhibitor that results in increased fin length and increased cx43 expression (Kujawski et al., 2014). FK506-treated populations were compared to control populations treated with DMSO. To evaluate changes in cx43 and evx1 expression, fish were treated at 72 hpa with FK506 for 15 hours and harvested 87 hpa (initiation). Whole mount in situ hybridization was conducted to visualize expression levels of cx43 and evx1 mRNA. We observed an increase in cx43 expression in regenerating fins from FK506-treated fish, and a reduction in the percentage of evx1-positive fin rays from 81.3% (48/59 fin rays) in DMSO-treated to 4.1% (2/49 fin rays) in FK506 treated (Figure 5). To evaluate regenerate length and segment length, fish were treated with FK506 for three consecutive days (i.e. starting at 48 hpa and harvesting at 120 hpa). Calcein staining was used to detect bone matrix and joints between segments. We found a significant increase in regenerate length in fish treated with FK506 (Figure 6). Additionally, while control DMSO-treated fish developed 1–2 joints per fin ray, FK506-treated fish failed to develop any joints during the observed time period (i.e. and therefore segment length could not be measured, Figure 6). These results indicate that increased cx43 levels decreases the number of fin rays expressing evx1, decreases the number of fin rays that produce joints, and causes joint delay/overlong segments. These findings support our model that overexpression of cx43 causes the failure of both evx1 expression and of joint initiation.

Figure 5.

Figure 5

Open in a new tab

Inhibition of calcineurin (FK506-treatment) increases cx43 and reduces the frequency of evx1-positive rays. (A) Whole Mount ISH shows increased cx43 and decreased evx1 expression at the time of joint initiation (87 hpa) in FK506-treated fish versus DMSO (negative control). (B) FK506-treated fish (n= 3 fins) showed reduced frequency of evx1 positive fin rays compared to DMSO-treated fish (n = 4 fins). Statistically significant differences were determined by the student’s t-test where p < 0.01 (**). Error bars represent the standard deviation.

Figure 6.

Figure 6

Open in a new tab

Inhibition of calcineurin (FK506-treatment) influences joint formation and regenerate length. (A, B) Calcein staining detects bone matrix and reveals joint failure and increased regenerate length in FK506-treated wild-type fish compared to control DMSO-treated wild-type fish. Panels on the left show fluorescence, while panels on the right show fluorescence plus bright field to better illustrate the end of the regenerating fin (the dotted line indicates the amputation plane). Arrowheads point to joints. Double-headed arrows identify regenerate length. (C) Regenerate length is significantly increased FK506-treated (n = 18) versus DMSO-treated (n = 18, negative control) fish. Statistically significant differences were determined by the student’s t-test where p < 0.001 (***). Error bars represent the standard deviation.

Decreased Cx43 activity leads to premature evx1 expression and reduced segment length

Decreasing cx43 function is predicted to cause premature evx1 expression and premature joint initiation, leading to short segments. Indeed, in a prior study we found that sof b123 mutants exhibit these phenotypes (Ton and Iovine, 2013b). In order to study the effects of cx43 depletion independently, a transgenic line that overexpresses miR-133 was utilized (i.e. Tg(hsp70:miR-133a1)pd47, Yin et al., 2012). This line reduces cx43 expression. Indeed, following a one-hour, 37°C heat shock, we find that cx43 levels are reduced in the Tg(hsp70:miR-133a1)pd47 (Figure 7).

Figure 7.

Figure 7

Open in a new tab

Reduced cx43 prior to joint formation increases the frequency of evx1-positive rays. (A) Whole mount ISH shows decreased cx43 expression and increased evx1 expression in the Tg(hsp70:miR-133a1)pd47 line at the time of joint initiation (87 hpa) in a1(+) (transgene-positive) versus a1(-) (non-transgenic). Bottom panels in A show evx1-positive fin rays across entire fins. (B) A higher percentage of fin rays in a1(+) transgenic fins showed expression of evx1 (n= 9 fins) compared to the a1(-) non-transgenic sibling (n = 9 fins). Both sp(+) and sp(-) fish were similarly treated for heat shock. Statistically significant differences were determined by the student’s t-test p < 0.01 (**). Error bars represent the standard deviation.

We tested if decreased Cx43 influences evx1 expression at the time of joint initiation. Fins were amputated at the 30 % level, followed by a one-hour heat pulse at 72 hpa (pre-initiation) in the Tg(hsp70:miR-133a1)pd47. To confirm activation after heat shock, as there is no egfp encoded in the miR-133a transgene, miR-133sp fish with the EGFP indicator were amputated along the same timeline and added to heat shock tanks. GFP was observed in the miR-133sp indicator fish post heat shock, indicating that the heat shock was effective. To determine whether decreased cx43 expression influences evx1 expression, fins were harvested at 87 hpa and processed for evx1 in situ hybridization. In contrast to the miR-133a sponge line, the percentage of evx1-positive fin rays increased in Tg(hsp70:miR-133a1)pd47 regenerating fins, from 51 % (87/170 fin rays) in the non-transgenic siblings to 79.5 % (132/166 fin rays) in the miR-133 overexpression line (Figure 7).

We next measured regenerate length and segment length using calcein staining at 96 hours post heat shock induction. While regenerate length appears shorter in the Tg(hsp70:miR-133a1)pd47 line compared to non-transgenic siblings, this difference is not significant. Therefore, transiently reducing cx43 expression at 72 hours is not sufficient to impact the rate of regeneration over several days. In contrast, we found a significant decrease in segment length in Tg(hsp70:miR-133a1)pd47 compared to non-transgenic siblings (Figure 8). These results indicate that decreased cx43 levels in wild type leads to premature expression of evx1, which in turn leads to premature joint initiation and therefore shortened segments. These findings support our model that reduced expression of cx43 permits premature evx1 expression and joint initiation.

Figure 8.

Figure 8

Open in a new tab

Decreased cx43 in the Tg(hsp70:miR-133a1)pd47 line influences segment length. Regenerate length (A) is not significantly decreased while segment length (B) is significantly decreased in a1(+) (transgene-positive, n = 16) compared with non-transgenic a1(-) siblings (n = 12). Both a1(+) and a1(-) fish were similarly treated for heat shock. Statistically significant differences were determined by the student’s t-test p < 0.001 (***). Error bars represent the standard deviation.

Cx43 function in dermal fibroblasts is sufficient for Cx43-dependent skeletal growth and patterning

Because Cx43 is expressed in both the medial dermal fibroblasts and in the lateral joint-forming osteoblasts, one possibility is that Cx43 functions autonomously in each population to coordinate skeletal growth (cell proliferation) and patterning (joint formation). Alternatively, Cx43 function in a single cell population could be responsible for both Cx43-dependent phenotypes. We chose to distinguish these possibilities by performing clonal analyses in sof b123 regenerating fins (i.e. as we have done before to demonstrate that expression of kitlga in dermal fibroblasts rescues the kit ligand a mutant phenotype, Tryon and Johnson, 2014). The cx43 coding sequence was inserted into the hsp-PT2 transposon plasmid behind the hsp70l promoter and together with the ef1a-egfp cassette. The latter element drives ubiquitous expression of GFP and therefore permits detection of clones in the absence of heat shock. This construct, plus transposase, was injected into one-cell sof b123 embryos. All GFP-positive fish were raised and screened for caudal fin clones in adults. GFP-positive fin clones were assigned to one of the lineage classes based on morphology (as described in Tu and Johnson, 2011). Clones were amputated in the middle of the clone and regeneration proceeded with a daily heat pulse to induce transient over-expression of the cx43 transgene. Remarkably, while GFP-positive cells were identified in multiple cell lineages, only GFP-positive clones in the dermal fibroblasts reliably rescued the sof b123 growth phenotypes (Table 2). The occasional rescue of the sof b123 phenotypes by epidermal clones is ascribed to these clones masking underlying dermal fibroblast clones. The level of rescue of regenerate length and segment length phenotypes were measured by comparing the ratios of fin ray length or segment length from fin rays within the GFP-positive clone to the contralateral GFP-negative fin rays at the same position (Figure 9). The rescue of segment length by Cx43 requires that both evx1 expression and joint formation are delayed in the context of sof b123 regenerating fins, although only joint formation was evaluated. Although we propose that Cx43 suppresses evx1/joint formation, the transient nature of Cx43 overexpression prevents complete inhibition of joint formation, but rather delays joint formation so that segment length is increased. These data demonstrate that Cx43 activity in dermal fibroblasts both promotes fin ray length (i.e. increased ray length) and suppresses joint formation (i.e. increased segment length), and further suggests that medially located Cx43-positive cells communicate with cells in both the medial and lateral mesenchyme to influence cell proliferation and cell differentiation.

Table 2.

GFP clone types and rescue of sof b123 fin phenotypes.

Clone type Rescue Fail
osteoblast 1 20
epidermis 3 16
fibroblast 17 9
lateral line 0 9
vasculature 0 5
pigment 0 1
intraray glia 0 1
blood 0 2
Open in a new tab

Figure 9.

Figure 9

Open in a new tab

Expression of Cx43 in dermal fibroblasts rescues both regenerate length and segment length. (A) Regenerate length is significantly increased in fins where Cx43 is induced in dermal fibroblasts but not in fins where Cx43 is induced in other lineages. (B) Segment length is significantly increased in fins where Cx43 is induced in dermal fibroblasts but not in fins where Cx43 is induced in other lineages. Both regenerate and segment length measurements were taken as the ratio of the GFP-positive clone length over the analogous fin ray on the opposite side of the fin. Statistically significant differences were determined by the student’s t-test where p < 0.001 (***). Error bars represent the standard deviation. (C) Representative image of a fin containing a GFP-positive clone in the dermal fibroblasts. The amputation plane is indicated by a dotted line. Scale bar is 150 μm. Insets show higher magnification views of a region of both the GFP-positive fin ray and in the contralateral GFP-negative fin ray (bright field is shown to better visualize the joints). Joints are indicated as arrowheads in both the GFP-positive fin ray and in the GFP-negative fin ray. (D) Higher magnification view of the GFP-positive clone to show GFP expression in dermal fibroblasts. Dermal fibroblast clones are notable for the ability to see the unlabeled artery running down the middle of ray (arrow pointing to middle of ray) (i.e. whereas osteoblast clones obscure the artery). Dermal fibroblast clones also show distinct GFP-positive fibroblast cells that are adjacent to, but not contained within the ray itself (arrows pointing to GFP-positive cells flanking the ray) (i.e. whereas osteoblast clones remain strictly associated with the hemirays).

Discussion

How the precise location of joints in the skeleton is determined is poorly understood. In prior studies, we showed that joint formation in fin ray joints “initiates” with the condensation of joint-forming cells (i.e., analogous to the interzone in synovial joints, Pacifici et al., 2006), and that reducing Cx43 levels in alf dty86 mutants rescues both segment length and evx1 expression (Sims et al., 2009; Ton and Iovine, 2013b). Here, we connect joint initiation with cx43 and evx1 expression by revealing a time line for joint initiation, by negatively correlating changes in cx43 with joint initiation, by positively correlating changes in evx1 expression with joint initiation, and by showing that manipulation of Cx43 influences both evx1 expression and joint initiation. These studies, together with our earlier studies, suggest that a threshold of Cx43 activity determines whether or not evx1 will be expressed and joint initiation will occur (Figure 10). At Cx43 levels above the threshold, evx1 is not sufficient for joint initiation and a joint is not formed. At Cx43 levels at/below the threshold evx1 expression is less suppressed by Cx43 so evx1 expression increases and can initiate joint formation. We suggest that in sof b123 mutants (which exhibit reduced cx43 mRNA, Iovine et al., 2005), or when Cx43 levels are otherwise reduced (Figure 7), fin rays achieve the threshold of Cx43 activity sooner and trigger premature joint initiation. In contrast, when Cx43 levels are too high, joint initiation fails completely. If the level of Cx43 activity is only modestly higher than wild-type (i.e. as in alf dty86 mutants), then cx43 levels are less likely to be reduced to the required threshold, causing evx1 expression and joint formation to become stochastic (Sims et al., 2009). Note that because miR-133a has multiple targets (Yin et al., 2008), we cannot rule out the possibility that another targeted gene mediates the observed effects on evx1 expression and joint formation. However, our prior findings on the role of Cx43 in suppressing joint formation (Hoptak-Solga et al., 2008; Sims et al., 2009) are consistent with the proposed model.

Figure 10.

Figure 10

Open in a new tab

Cx43 threshold model for joint initiation. Cx43 activity levels decrease to a threshold level, allowing for evx1 expression and, therefore, joint initiation. In the sof b123 mutant, where cx43 mRNA levels are reduced with respect to WT, the threshold is reached sooner, inducing premature joint initiation and resulting in short segments.

Cx43 is expressed in cells of two distinct lineages in the regenerating fin: in the dermal fibroblasts and in the joint-forming osteoblasts (Sims et al., 2009). Our findings reveal that Cx43 activity in the dermal fibroblasts is responsible for rescuing both regenerate length and segment length in sof b123 fins. Regenerate length is influenced by the number of proliferating cells, which Cx43 may influence autonomously (Hoptak-Solga et al., 2008). Segment length is regulated by the timing of joint formation, which we show here is determined by a transient reduction in Cx43 that permits evx1 expression, and therefore differentiation of joint-forming osteoblasts from among the lateral skeletal precursor cells. Interestingly, joint-forming cells and bone-forming cells are both derived from a common precursor population (Tu and Johnson, 2011). Therefore, we suggest that the level of Cx43 activity in the dermal fibroblasts is responsible for influencing joint-forming cell differentiation. This mechanism ensures that joint-forming cells initiate a joint at the appropriate time and place, and in a manner that is coordinated with regenerate length.

Several questions remain. For example, by what mechanism(s) is Cx43 activity reduced? Our findings suggest that Cx43 activity cycles in concert with joint formation, but the identity of the factor(s) controlling these oscillations is unknown. To what extent must Cx43 activity be reduced to achieve the proposed threshold for evx1 expression? Here, we find a modest reduction in cx43 mRNA correlates with joint initiation. However, this does not rule out the possibility that Cx43 activity is influenced through additional mechanisms, such as an increase in turnover at the plasma membrane or inactivation of gap junction channels. Finally, by what mechanism(s) do the dermal fibroblasts communicate with the precursors of the osteoblast lineages? In a prior study we showed that Cx43 activity is required for the expression of the growth factor, sema3d, in cells of the lateral osteoblast lineage (i.e. which also appears to act upstream of joint formation) (Ton and Iovine, 2012). We have proposed that these cells may communicate directly, via heterotypic gap junction channels. Alternatively, communication may be indirect, through the secretion of a growth factor by the dermal fibroblasts that is recognized by a cell surface receptor in the precursors of the osteoblast lineages (Ton and Iovine, 2013a). Studies to distinguish these possibilities are ongoing.

Future studies are aimed at revealing the mechanisms for regulating Cx43 activity during skeletal regeneration, and at revealing how changes in Cx43 activity influence changes in gene expression. Given that so little is understood regarding the appropriate positioning of joints, we anticipate that our findings will continue to provide important insights into this question.

Experimental procedures

Fish maintenance

Zebrafish were raised at constant temperature of 28°C in a 14 light: 10 dark photoperiod. C32, sof b123, alf dty86 Tg(hsp70:miR-133sp)pd48 and Tg(hsp70:miR-133a1)pd47 were used. The transgenic lines were generously provided by V.P. Yin. Research was performed as described to the IACUC for Lehigh University (protocol #187, 3/14/2016 approval). For induction of transgene expression, fish were placed in a tank of fish water heated to 37°C for one hour, followed by a gradual cooling back to 28°C.

In situ hybridization

Antisense digoxigenin-labeled probes were generated as described (Iovine et al., 2005 for cx43 and Ton and Iovine, 2013b for evx1). Whole mount in situ hybridization was completed as described (Sims et al., 2009). To evaluate the relative level of gene expression, whole mount in situ hybridization was completed on four fins in each of three independent experiments (12 fins per group). For in situ hybridization on cryosections fins were fixed following the same fixation conditions, methanol dehydration, and gradual rehydration as in whole mount in situ hybridization, harvested fins were then embedded in 1.5% agarose/5% sucrose and equilibrated overnight in 30% sucrose. Those blocks were mounted in OCT and cryosectioned using the Reichert-Jung 2800 Frigocut cryostat. 15 μM sections were mounted on Superfrost Plus slides (Fisher) and air-dried overnight at room temperature. In situ hybridization on cryosections was completed as described (Ton and Iovine, 2013b). To evaluate the relative level of gene expression, cryo-in situ hybridization was evaluated from 5 sections from each of 3 different fins.

Joint morphology time course

To establish the time course for joint initiation, fins were amputated at the 30% level (just proximal to the cleft) and allowed to regenerate for 2–5 days. Fins were harvested and processed for ZNS5 immunoflourescence (below) and examined using the 40x N.A. 1.4 PlanApo objective on a Zeiss LSM510META confocal microscope to identify joint cell morphology as described (Sims et al., 2009).

ZNS5 immunofluorescence

Fins were harvested and fixed overnight in 4% paraformaldehyde/PBS at 4°C then stored in methanol at −20°C. Fins were rehydrated in successive washes in decreasing methanol/PBS solutions. After block (2% BSA/PBS), fins were treated with 1:200 ZNS5 (ZIRC) in block at 4°C overnight. Following three washes in block, fins were treated with secondary antibody (Alexa 488 or Alexa546) at 1:200 overnight at 4°C, and washed in block again before mounting on slides. Fins were mounted on Superfrost slides in glycerol. Fins were examined on a Zeiss confocal microscopy to evaluate joint morphology and using a Nikon Eclipse E80 microscope with a Nikon digital camera to measure segment length.

qRT-PCR analysis

Fins were amputated at the 30% level and allowed to regenerate for 72, 87, or 96 hpa. Total RNA was isolated from 5–10 harvested fins using Trizol reagent (Gibco) and cDNA was synthesized using oligo-dT and reverse transcriptase. Diluted cDNA (1:100 for cx43 and 1:10 for evx1) was combined with Qiagen SybrGreen PCR master mix and with either cx43 or actin (control) primers. Samples were run using the RotorGene Real Time PCR system, and average cycle number (CT) was determined for each amplicon. ΔCT represents the normalized gene level with respect to the actin control. Three biological replicates of each sample were prepared. Within each biological replicate, two technical replicates for cx43 and actin reactions were completed. The average values for the technical replicates are shown for each biological replicate (Table 1).

FK506 treatment

For evaluation of segment length and regenerate length, fins were amputated at the 50% level (these measurements are taken at 5 dpa and do not rely on the timing of the formation of the first joint). Fish were treated for 3 consecutive days starting 48 hpa with 0.1 μg/mL FK506 (fresh each day). At 5 dpa fish were treated for either ZNS5 staining (fixed tissue) or for calcein staining (live tissue, see below). The third fin rays of both the dorsal and ventral sides (V+3, D+3) were examined for regenerate and segment length measurements, as established (Iovine and Johnson, 2000). Student’s t-tests were performed to determine if data sets were statistically different (p<0.05).

For evaluation of gene expression at the time of joint initiation, fins were amputated at the 30% level, treated with 0.1 μg/mL FK506 (n = 5) or DMSO (n = 5) at 72 hpa for 15 hours, and harvested at 87 hpa. These fins were processed for in situ hybridization to detect evx1 expression.

Calcein staining

Calcein staining was performed to detect calcified bone matrix (Du et al., 2001). A 0.2% calcein solution was prepared by dissolving 2 g calcein powder (Sigma) in 1 L deionized water, and 0.5M NaOH was added to bring the pH to neutral. Zebrafish were immersed in the solution for 15 minutes, followed by several rinses with clean water and 10 minutes in fresh water to allow unbound calcein to diffuse out of the tissue. Zebrafish were then anesthetized in tricaine, viewed at 4X using a Nikon Eclipse 80i microscope with green fluorescence filters, and imaged using a Nikon digital camera.

Chimeric analysis

The cx43 open reading frame was amplified from a 96 hpf cDNA library using primers cx43-F 5′ GGCGGCGATCGCTTAATTAAATGGGTGACTGGAGTGC and cx43-R 5′ TCGACCTGCAGGTTAATTAAGACGTCCAGGTCATCAGG. The PCR product was cloned via Gibson Assembly into the hsp-PT2 transposon plasmid, downstream of the hsp70l promoter using the Pac I site. The resulting molecule contains two functional cassettes: (1) a ubiquitously expressed GFP driven by the Xenopus EF1a promoter for identifying cells that have integrated the transposon and (2) the hsp70l promoter driving cx43 for overexpressing the wild type protein when fish are exposed to heat shock.

sof b123 embryos were injected at the 1 cell stage with a mixture of hsp:cx43-pT2 plasmid DNA at 50ng/uL and transposase mRNA at 15 ng/uL (as described in Tryon and Johnson, 2012). All GFP-positive fish were grown to adulthood (5–6 months) and screened for caudal fin clones. GFP-positive clones were assigned to the appropriate lineage class based on the cellular and clone morphologies as described previously (Tu and Johnson, 2011). Caudal fin clones were subsequently amputated across the GFP positive clone, such that the labeled cells would contribute to the blastema and ultimately the regenerated fin tissue. Fish were grown for 4 weeks while being exposed to a daily heat pulse of 38° C for 30 minutes, and were imaged weekly to identify atypical growth during regeneration. The Nikon SMZ 1500 microscope with 1.6x WD 24 Nikon HR Plan Apo objective was used to visualize clones, and images were collected using a Jenoptik L.O.S. GmbH camera and ProgResC14 software.

Adobe Illustrator was used to measure the relative lengths of segments and rays in fins. A line, or series of lines in the case of a curved regenerated ray, were superimposed over the photographed image for both the GFP+ labeled ray and its contralateral counterpart. The measure tool in Adobe Illustrator was subsequently used to determine the length of each feature in points (pts).

Bullet points.

  • A time line for the initiation of joint formation during fin regeneration was established.

  • Manipulation of Cx43 levels is negatively correlated with evx1 expression and with joint formation, consistent with our model that Cx43 must be abrogated for joint formation to occur.

  • Cx43 function in dermal fibroblasts is responsible for both Cx43-dependent regenerate length and Cx43-dependent suppression of joint formation.

Acknowledgments

The authors wish to thank Rebecca Bowman for care of the animal facilities and Voot Yin for the miR133a transgenic lines. The authors also thank members of the Iovine and Johnson labs for thoughtful comments on the manuscript. We thank the Zebrafish International Research Center for providing the monoclonal ZNS5 antibody for reasonable cost. This work was supported in part through funding to MKI (R15HD080507) and SLJ (R01GM056988).

References

  1. Banerji R, Eble DM, Iovine MK, Skibbens RV. Esco2 regulates cx43 expression during skeletal regeneration in the zebrafish fin. Developmental dynamics : an official publication of the American Association of Anatomists. 2016;245:7–21. doi: 10.1002/dvdy.24354. [DOI] [PubMed] [Google Scholar]
  2. Borday V, Thaeron C, Avaron F, Brulfert A, Casane D, Laurenti P, Geraudie J. evx1 transcription in bony fin rays segment boundaries leads to a reiterated pattern during zebrafish fin development and regeneration. Developmental dynamics : an official publication of the American Association of Anatomists. 2001;220:91–98. doi: 10.1002/1097-0177(2000)9999:9999<::AID-DVDY1091>3.0.CO;2-J. [DOI] [PubMed] [Google Scholar]
  3. Brown AM, Fisher S, Iovine MK. Osteoblast maturation occurs in overlapping proximal-distal compartments during fin regeneration in zebrafish. Developmental dynamics : an official publication of the American Association of Anatomists. 2009;238:2922–2928. doi: 10.1002/dvdy.22114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Du SJ, Frenkel V, Kindschi G, Zohar Y. Visualizing normal and defective bone development in zebrafish embryos using the fluorescent chromophore calcein. Developmental biology. 2001;238:239–246. doi: 10.1006/dbio.2001.0390. [DOI] [PubMed] [Google Scholar]
  5. Hoptak-Solga AD, Nielsen S, Jain I, Thummel R, Hyde DR, Iovine MK. Connexin43 (GJA1) is required in the population of dividing cells during fin regeneration. Developmental biology. 2008;317:541–548. doi: 10.1016/j.ydbio.2008.02.051. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Iovine MK, Higgins EP, Hindes A, Coblitz B, Johnson SL. Mutations in connexin43 (GJA1) perturb bone growth in zebrafish fins. Developmental biology. 2005;278:208–219. doi: 10.1016/j.ydbio.2004.11.005. [DOI] [PubMed] [Google Scholar]
  7. Iovine MK, Johnson SL. Genetic analysis of isometric growth control mechanisms in the zebrafish caudal Fin. Genetics. 2000;155:1321–1329. doi: 10.1093/genetics/155.3.1321. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Knopf F, Hammond C, Chekuru A, Kurth T, Hans S, Weber CW, Mahatma G, Fisher S, Brand M, Schulte-Merker S, Weidinger G. Bone regenerates via dedifferentiation of osteoblasts in the zebrafish fin. Developmental cell. 2011;20:713–724. doi: 10.1016/j.devcel.2011.04.014. [DOI] [PubMed] [Google Scholar]
  9. Kujawski S, Lin W, Kitte F, Bormel M, Fuchs S, Arulmozhivarman G, Vogt S, Theil D, Zhang Y, Antos CL. Calcineurin regulates coordinated outgrowth of zebrafish regenerating fins. Developmental cell. 2014;28:573–587. doi: 10.1016/j.devcel.2014.01.019. [DOI] [PubMed] [Google Scholar]
  10. Lee Y, Grill S, Sanchez A, Murphy-Ryan M, Poss KD. Fgf signaling instructs position-dependent growth rate during zebrafish fin regeneration. Development. 2005;132:5173–5183. doi: 10.1242/dev.02101. [DOI] [PubMed] [Google Scholar]
  11. Pacifici M, Koyama E, Shibukawa Y, Wu C, Tamamura Y, Enomoto-Iwamoto M, Iwamoto M. Cellular and molecular mechanisms of synovial joint and articular cartilage formation. Annals of the New York Academy of Sciences. 2006;1068:74–86. doi: 10.1196/annals.1346.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Perathoner S, Daane JM, Henrion U, Seebohm G, Higdon CW, Johnson SL, Nusslein-Volhard C, Harris MP. Bioelectric signaling regulates size in zebrafish fins. PLoS genetics. 2014;10:e1004080. doi: 10.1371/journal.pgen.1004080. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Schulte CJ, Allen C, England SJ, Juarez-Morales JL, Lewis KE. Evx1 is required for joint formation in zebrafish fin dermoskeleton. Developmental dynamics : an official publication of the American Association of Anatomists. 2011;240:1240–1248. doi: 10.1002/dvdy.22534. [DOI] [PubMed] [Google Scholar]
  14. Sims K, Jr, Eble DM, Iovine MK. Connexin43 regulates joint location in zebrafish fins. Developmental biology. 2009;327:410–418. doi: 10.1016/j.ydbio.2008.12.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Ton QV, Iovine MK. Semaphorin3d mediates Cx43-dependent phenotypes during fin regeneration. Developmental biology. 2012;366:195–203. doi: 10.1016/j.ydbio.2012.03.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Ton QV, Iovine MK. Determining how defects in connexin43 cause skeletal disease. Genesis. 2013a;51:75–82. doi: 10.1002/dvg.22349. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Ton QV, Iovine MK. Identification of an evx1-dependent joint-formation pathway during FIN regeneration. PloS one. 2013b;8:e81240. doi: 10.1371/journal.pone.0081240. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Tryon RC, Johnson SL. Clonal and lineage analysis of melanocyte stem cells and their progeny in the zebrafish. Methods in molecular biology. 2012;916:181–195. doi: 10.1007/978-1-61779-980-8_14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Tryon RC, Johnson SL. Clonal analysis of kit ligand a functional expression reveals lineage-specific competence to promote melanocyte rescue in the mutant regenerating caudal fin. PloS one. 2014;9:e102317. doi: 10.1371/journal.pone.0102317. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Tu S, Johnson SL. Fate restriction in the growing and regenerating zebrafish fin. Developmental cell. 2011;20:725–732. doi: 10.1016/j.devcel.2011.04.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. van Eeden FJ, Granato M, Schach U, Brand M, Furutani-Seiki M, Haffter P, Hammerschmidt M, Heisenberg CP, Jiang YJ, Kane DA, Kelsh RN, Mullins MC, Odenthal J, Warga RM, Nusslein-Volhard C. Genetic analysis of fin formation in the zebrafish, Danio rerio. Development. 1996;123:255–262. doi: 10.1242/dev.123.1.255. [DOI] [PubMed] [Google Scholar]
  22. Yin VP, Lepilina A, Smith A, Poss KD. Regulation of zebrafish heart regeneration by miR-133. Developmental biology. 2012;365:319–327. doi: 10.1016/j.ydbio.2012.02.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Yin VP, Thomson JM, Thummel R, Hyde DR, Hammond SM, Poss KD. Fgf-dependent depletion of microRNA-133 promotes appendage regeneration in zebrafish. Genes & development. 2008;22:728–733. doi: 10.1101/gad.1641808. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES