1. Developmental Biology
  2. Stem Cells and Regenerative Medicine

Insulin-producing β-cells regenerate ectopically from a mesodermal origin under the perturbation of hemato-endothelial specification

  1. Ka-Cheuk Liu
  2. Alethia Villasenor
  3. Maria Bertuzzi
  4. Nicole Schmitner
  5. Niki Radros
  6. Linn Rautio
  7. Kenny Mattonet
  8. Ryota L Matsuoka
  9. Sven Reischauer
  10. Didier YR Stainier
  11. Olov Andersson  Is a corresponding author
  1. Department of Cell and Molecular Biology, Karolinska Institutet, Sweden
  2. Department of Developmental Genetics, Max Planck Institute for Heart and Lung Research, Germany
  3. Department of Neuroscience, Karolinska Institutet, Sweden
  4. Dermatology and Venereology Division, Department of Medicine (Solna), Karolinska Institutet, Sweden
  5. Department of Cardiovascular and Metabolic Sciences, Lerner Research Institute, Cleveland Clinic, United States
  6. Cardio-Pulmonary Institute, Frankfurt, Germany; Medical Clinic I, (Cardiology/Angiology) and Campus Kerckhoff, Justus-Liebig-University Giessen, Germany
https://doi.org/10.7554/eLife.65758
Share this article
Cite this article
  1. Ka-Cheuk Liu
  2. Alethia Villasenor
  3. Maria Bertuzzi
  4. Nicole Schmitner
  5. Niki Radros
  6. Linn Rautio
  7. Kenny Mattonet
  8. Ryota L Matsuoka
  9. Sven Reischauer
  10. Didier YR Stainier
  11. Olov Andersson
(2021)
Insulin-producing β-cells regenerate ectopically from a mesodermal origin under the perturbation of hemato-endothelial specification
eLife 10:e65758.
https://doi.org/10.7554/eLife.65758
  2. Download BibTeX
  3. Download .RIS

Abstract

To investigate the role of the vasculature in pancreatic β-cell regeneration, we crossed a zebrafish β-cell ablation model into the avascular npas4l mutant (i.e. cloche). Surprisingly, β-cell regeneration increased markedly in npas4l mutants owing to the ectopic differentiation of β-cells in the mesenchyme, a phenotype not previously reported in any models. The ectopic β-cells expressed endocrine markers of pancreatic β-cells, and also responded to glucose with increased calcium influx. Through lineage tracing, we determined that the vast majority of these ectopic β-cells has a mesodermal origin. Notably, ectopic β-cells were found in npas4l mutants as well as following knockdown of the endothelial/myeloid determinant Etsrp. Together, these data indicate that under the perturbation of endothelial/myeloid specification, mesodermal cells possess a remarkable plasticity enabling them to form β-cells, which are normally endodermal in origin. Understanding the restriction of this differentiation plasticity will help exploit an alternative source for β-cell regeneration.

Introduction

The concept of embryonic development and cell fate determination was illustrated by the famous Waddington landscape model decades ago (Waddington, 1957). Waddington’s model not only shows the importance of spatiotemporal precision in cell differentiation but also metaphorises cell fate determination as a sequential and irreversible event. In this hierarchical model, the endoderm follows the lineage paths downwards and progressively differentiates into multiple endodermal cell types, including pancreatic β-cells. Likewise, the mesoderm stays in the mesodermal lineage paths and differentiates into endothelial cells and other mesodermal cell types. However, in recent decades, multiple studies have suggested that committed cells are capable of differentiating across the germ layer boundaries by converting embryonic and/or adult mesodermal fibroblasts into ectodermal neuronal cells (Vierbuchen et al., 2010), multipotent neural stem cells (Ring et al., 2012), endodermal hepatocyte-like cells (Huang et al., 2011; Sekiya and Suzuki, 2011), or pancreatic β-like cells (Zhu et al., 2016) in vitro. These studies highlight the feasibility of converting mesodermal cells into ectodermal or endodermal cells in vitro.

Despite the extensive studies on cell fate conversion across germ layers in vitro, the number of in vivo studies is limited. Ectopic expression of Xsox17β in Xenopus embryos relocated cells normally fated to become ectoderm to appear in the gut epithelium, suggesting a possible change in cell fate in vivo (Clements and Woodland, 2000). Similarly, ectopic expression of sox32/casanova in presumptive endothelial cells switched their cell fate to endoderm in zebrafish embryos (Kikuchi et al., 2001). Furthermore, aggregated morulae and chimeric mouse embryos of β-catenin mutants provided evidence of precardiac mesoderm formation in endodermal tissues in vivo (Lickert et al., 2002). These studies suggest that the classical in vivo germ layer boundaries may not be as clear-cut as previously thought.

In this study, we aimed to elucidate the importance of the vasculature in pancreatic β-cell regeneration, which plays a crucial role in potential therapeutic strategies against diabetes. We employed cloche zebrafish mutants as an avascular model. The mutation of npas4l, a master regulator of endothelial and hematopoietic cell fates, is responsible for the severe loss of most endothelial and blood cells in cloche mutants (Parker and Stainier, 1999; Reischauer et al., 2016; Stainier et al., 1995). Unexpectedly, the npas4l mutation induced ectopic β-cell formation in the mesenchymal region outside of the pancreas after β-cell ablation. Lineage-tracing mesodermal cells expressing draculin (drl), npas4l, and etsrp (previously named etv2) validated the mesodermal lineage of the ectopic β-cells, which are normally endodermal in origin. These findings offer novel insights into cell fate determination and an alternative source of β-cells.

Results

Ectopic β-cell formation in npas4l mutants

To determine the importance of vasculogenesis and vascularisation for β-cell regeneration, we examined β-cell formation in zebrafish carrying the cloche mutation (npas4ls5 mutation) after β-cell ablation, i.e., in the Tg(ins:Flag-NTR);Tg(ins:H2BGFP;ins:DsRed) model. Nitroreductase (NTR), expressed under the insulin promoter, converts the prodrug metronidazole (MTZ) to a cytotoxin to specifically ablate insulin-producing β-cells (Curado et al., 2007). The homozygous mutation of npas4l significantly increased the number of ins:H2BGFP-positive cells during the β-cell regeneration period (Figure 1A–C). In addition, we observed a distinctive ectopic β-cell population in the mesenchymal region outside of the pancreas in the npas4l−/− group, an ectopic location that was very rarely observed in the sibling controls (including both wild-type and heterozygous siblings). This ectopic population of β-cells contributed to the major increase in the number of ins:H2BGFP-positive cells during β-cell regeneration (Figure 1C). Moreover, the comparable and sparse numbers of ins:DsRed-positive cells in the controls and mutants indicate that the npas4l mutation did not enhance the survival of β-cells during the ablation (Figure 1A,B) because the extended maturation time of DsRed (Baird et al., 2000) restricted the detection of DsRed to any surviving β-cells.

Figure 1 with 1 supplement see all
Download asset Open asset
Ectopic β-cell formation in npas4l mutants.

(A, B) Representative confocal projections of the pancreas and neighbouring tissues of control siblings and npas4l−/− Tg(ins:Flag-NTR);Tg(ins:H2BGFP;ins:DsRed) zebrafish larvae at 3 dpf after β-cell ablation by MTZ at 1–2 dpf, displaying regenerated β-cells in green and older β-cells that likely survived the ablation in yellow overlap as the DsRed fluorescence driven by the insulin promoter remained in these cells, at the same time as DsRed had not had enough time to mature in the regenerated β-cells (arrowheads). The ectopic β-cells are indicated by white dashed rectangle. Pancreata are outlined by solid white lines. (C) Quantification of the pancreatic or ectopic β-cells per larva at 3 dpf. ****p<0.0001 (Šidák’s multiple comparisons test); n = 24 (control) and 19 (npas4l−/−). Quantification data are represented as the mean ± SEM. (D, E) Representative image projections of the pancreas and neighbouring tissues in control siblings and npas4l-/- Tg(ins:Flag-NTR);Tg(ptf1a:EGFP) larvae at 3 dpf after β-cell ablation by MTZ at 1–2 dpf, displaying β-cells in red with immunostaining for insulin and ptf1a:EGFP+ exocrine pancreas in green. Pancreata are outlined by solid white lines. Dashed line outlines ectopic β-cells in the mesenchyme (E). (F–K) Representative images and projections of the pancreas and neighbouring mesenchyme of control siblings and npas4l-/- Tg(ins:Flag-NTR);Tg(hand2:EGFP) zebrafish larvae at 3 dpf after β-cell ablation by MTZ at 1–2 dpf, displaying β-cells in red with immunostaining for insulin and hand2:EGFP+ mesenchyme in green. White arrowheads point to β-cells in the pancreas (F, I). Dashed rectangles indicate the ectopic β-cells intermingling with the mesenchyme between the pronephros and the pancreas, without co-expressing insulin and hand2:EGFP (J, K). Selected area in dashed ovals shows other ectopic β-cells intermingling with the mesenchyme ventral to the pancreas (I, K). Scale bars = 20 μm. Anatomical axes: D (dorsal), V (ventral), A (anterior), and P (posterior).

To visualise the location of the ectopic β-cells better, we labelled the pancreas with ptf1a:EGFP expression and observed not only a drastic reduction in the pancreas size (Figure 1D,E, Figure 1—figure supplement 1) but also the regeneration of β-cells clearly outside of the ptf1a-expressing exocrine pancreas in npas4l mutants (Figure 1E). By labelling the mesenchyme with hand2:EGFP expression (Figure 1F–K), we further revealed that the majority of the ectopic β-cells formed in npas4l mutants intermingled with hand2:EGFP-positive mesenchymal cells between the pronephros and the pancreas (Figure 1J,K). In addition, we occasionally observed ectopic β-cells intermingled with hand2:EGFP-positive mesenchymal cells ventral to the pancreas (Figure 1I,K). Although the ectopic β-cells were located among the mesenchymal cells, they did not express hand2:EGFP.

The ectopic β-cells co-express insulin and endocrine markers in npas4l mutants

Next, we examined multiple pancreatic endocrine and β-cell markers, including Isl1, neurod1, pdx1, mnx1, pcsk1, and ascl1b (the functional homologue to Neurog3 in mammals), to validate the β-cell identity of the ectopic insulin-producing cells. The majority of the ectopic β-cells co-expressed insulin as well as these markers during β-cell regeneration (Figure 2). The high co-expression of pcsk1 (Figure 2R,S, Table 1), which encodes an enzyme necessary for insulin biosynthesis, indicates that most of the β-cells in the ectopic population are likely functional. Consistent with previous findings in pancreatic β-cells, not all ectopic β-cells expressed ascl1b:EGFP (Figure 2V,W, Table 1), which suggests that ascl1b works as a transient endocrine cell fate regulator (Flasse et al., 2013). In contrast with Isl1, mnx1, pcsk1, and ascl1b, we observed lower co-expression levels of neurod1 and pdx1 in ectopic β-cells compared with the pancreatic population in npas4l mutants (Table 1). In addition to the reduction in pancreas size (Figure 1—figure supplement 1), the pdx1-expressing pancreatic duct was also reduced in the npas4l mutant (Figure 2—figure supplement 1), indicating that the pancreas and its duct did not expand to form the ectopic β-cells. These observations together suggest that the pancreatic and ectopic β-cells are similar, yet they are two distinct β-cell populations.

Figure 2 with 1 supplement see all
Download asset Open asset
The ectopic β-cells co-express insulin and endocrine markers in npas4l mutants.

Representative confocal images of the tissues adjacent to the pancreas of control siblings and npas4l−/− Tg(ins:Flag-NTR) zebrafish larvae at 3 dpf after β-cell ablation by MTZ at 1–2 dpf, displaying cells expressing pancreatic endocrine cell markers Isl1 (A–B’), neurod1 (E–F’), pdx1 (I–J’), mnx1 (M–N’), pcsk1 (Q–R’), and ascl1b (U–V’) in green, and ectopic β-cells in red with immunostaining for insulin (except Q–R’ with the mCherry fluorescence driven by the insulin promoter). Arrowheads point to ectopic β-cells that express corresponding markers. Arrows point to β-cells that do not express ascl1b (V and V’). B’, F’, J’, N’, R’, and V’ are magnified from the areas indicated by the white dashed square in B, F, J, N, R, and V, respectively. Quantification of β-cells with or without corresponding marker expression in the ectopic location (C, G, K, O, S, and W) or in the pancreas (D, H, L, P, T, and X) per larva at 3 dpf. *p=0.0310, **p=0.0039, and ****p<0.0001 (Šidák’s multiple comparisons test); (C, D) n = 30 (control) and 31 (npas4l−/−); (G, H) n = 8 (control) and 8 (npas4l−/−); (K, L) n = 25 (control) and 24 (npas4l−/−); (O, P) n = 21 (control) and 23 (npas4l−/−); (S, T) n = 40 (control) and 32 (npas4l−/−); (W and X) n = 13 (control) and 13 (npas4l−/−). Data are represented as the mean ± SEM. Scale bars = 20 μm except B’, F’, J’, N’, R’, and V’ (10 μm). Anatomical axes: D (dorsal), V (ventral), A (anterior), and P (posterior).

Table 1
Percentages of pancreatic or ectopic cells co-expressing insulin and corresponding marker gene or protein in npas4l mutants.
Marker gene/proteinPancreatic co-expression (%)Ectopic co-expression (%)
Isl197.996.8
neurod197.517.5
pdx172.132.8
mnx175.370.2
pcsk185.174.1
ascl1b21.217.8

Ectopic β-cells can respond to glucose by displaying calcium oscillations

We further assessed the functionality and maturity of the ectopic β-cells by examining the calcium ion oscillation in vivo to determine their ability in responding to glucose to secrete insulin in the Tg(ins:GCaMP6s);Tg(ins:mCherry);Tg(ins:Flag-NTR) npas4l zebrafish mutants. After β-cell ablation by MTZ and one day of β-cell regeneration, we found comparable proportions of pancreatic (66%) and ectopic (67%) β-cells displaying mild and basal levels of calcium activity at the baseline (Figure 3 and Video 1). Upon glucose treatment, 50% of pancreatic β-cells and 33% of ectopic β-cells elicited calcium oscillations, suggesting that a portion of the pancreatic and ectopic β-cells have the ability to respond to glucose to secrete insulin in the npas4l mutants during β-cell regeneration.

Figure 3
Download asset Open asset
Ectopic β-cells can respond to glucose by displaying calcium oscillations.

Representative images captured from a live calcium imaging video of npas4l−/−Tg(ins:GCaMP6s);Tg(ins:mCherry);Tg(ins:Flag-NTR) zebrafish larvae at 3 dpf after β-cell ablation by MTZ from 1 to 2 dpf, showing β-cells expressing ins:mCherry in red and displaying calcium activity in green in the pancreatic (A, B) or ectopic (C, D) domain at the baseline (A, C) or after the addition of 200 mM of glucose to the E3 medium (B, D). (E) Sequential images captured from a live calcium imaging video to show the oscillation of calcium signalling at the baseline and after the glucose treatment, displaying the GCaMP6s signal in green. (F) Quantification of pancreatic or ectopic β-cells displaying calcium activity at the baseline or after glucose treatment in 11 (pancreatic) or 17 (ectopic) npas4l zebrafish mutants. Scale bars = 10 μm.

Video 1
Download asset
Live calcium imaging of β-cells in a npas4l mutant.

An example video of live calcium imaging of both pancreatic and ectopic β-cells in a Tg(ins:GCaMP6s);Tg(ins:mCherry);Tg(ins:Flag-NTR) npas4l mutant at 3 dpf after β-cell ablation from 1 to 2 dpf, displaying GCaMP6s signal in green. Glucose (200 mM) was added to the E3 media at around 9 min in the video.

The ectopic β-cells are of mesodermal origin in npas4l mutants and etsrp morphants

We have previously shown that npas4l expression is first initiated in the lateral plate mesoderm at the tailbud stage by in situ hybridisation (Reischauer et al., 2016). In this study, we examined npas4l expression at 20 hr postfertilisation (hpf) and found that npas4l was severely reduced in the lateral plate mesoderm in the npas4l mutants (Figure 4—figure supplement 1A,B), whereas normal expression levels were observed in the tailbud and brain. The cells with reduced npas4l expression were still present in the lateral plate mesoderm as demonstrated by the embryos incubated overnight to further develop the npas4l expression signal (Figure 4—figure supplement 1B’). On the contrary, there were no significant differences in the expression of the examined early mesodermal or endodermal transcription factors foxa2, gsc, and mixl1, or the pancreatic endocrine progenitor marker ascl1b (Figure 4—figure supplement 1C–J). Because the ectopic β-cells induced by the npas4l mutation also resided in the mesenchymal region, and npas4l can act cell-autonomously to affect the hematopoietic and endothelial lineages (Parker and Stainier, 1999), we hypothesised that the ectopic β-cells originated from a mesodermal lineage.

To determine whether the mesoderm was the origin of the ectopic β-cells, we genetically traced the mesodermal cells using drl:CreERT2, a tamoxifen-inducible Cre transgene driven by a drl promoter (Mosimann et al., 2015). The spatial expression pattern of drl in the npas4l mutants resembled that in the sibling controls (Figure 4—figure supplement 2), suggesting that npas4l mutation did not induce any ectopic expression of drl to disrupt the lineage-tracing approach. In combination with a −3.5ubb:LOXP-EGFP-LOXP-mCherry (ubi:Switch) reporter (Mosimann et al., 2011), the drl-expressing mesodermal cells were labelled in red in Tg(drl:CreERT2);Tg(ubi:Switch);Tg(ins:Flag-NTR) (drl-tracing) zebrafish larvae after 4-hydroxytamoxifen (4-OHT) induction (Figure 4A, Figure 4—figure supplement 3). We treated the transgenic embryos with 4-OHT at 10–12 hpf. We chose to label the mesodermal cells during this period as neither endothelial/hematopoietic cells nor β-cells have developed at that stage, i.e., to exclude confounding effects of endothelial/hematopoietic cells or possible ectopic expression of the lineage tracer in the β-cells of the npas4l mutant. To ablate the β-cells, we incubated the 4-OHT-treated transgenic embryos in MTZ at 1–2 days postfertilisation (dpf). We allowed the β-cells to regenerate for 30 hr before we fixed the larvae at 3 dpf for immunostaining (Figure 4B).

Figure 4 with 4 supplements see all
Download asset Open asset
The ectopic β-cells are of mesodermal origin in npas4l mutants and etsrp morphants.

(A) Constructs of −6.35drl:Cre-ERT2 (drl:CreERT2) and −3.5ubb:LOXP-EGFP-LOXP-mCherry (ubi:Switch). Upon 4-OHT induction between 10 and 12 hpf, Cre recombinase expressed by the drl promoter cleave the loxP sites to allow ubb:mCherry expression in the cells that once expressed drl. (B) Scheme for tracing the mesodermal lineage of ectopic β-cells in control siblings and npas4l−/− Tg(ins:Flag-NTR);Tg(ubi:Switch);Tg(drl:CreERT2) zebrafish larvae. (C, D, I, and J) Quantification of the pancreatic or ectopic β-cells with or without mesodermal lineage in npas4l mutants (C, D) or etsrp morpholino (MO)-injected larvae (I, J) at 3 dpf. *p=0.0227 and ****p<0.0001 (Šidák’s multiple comparisons test); (C, D) n = 21 (control) and 14 (npas4l−/−) or (I, J) n = 21 (control) and 23 (etsrp MO). Data are represented as the mean ± SEM. Standard control MO (4 ng) and etsrp MO2 (4 ng) were injected to the control group and etsrp MO group respectively at the one-cell stage (I, J, K–L’’’). (E–H’’’ and K–L’’’) Representative confocal images of pancreatic (E, F) or ectopic β-cells (G, H) of control siblings and npas4l−/−, or ectopic β-cells in control or etsrp MO-injected (K, L) Tg(ins:Flag-NTR);Tg(ubi:Switch);Tg(drl:CreERT2) zebrafish larvae at 3 dpf after β-cell ablation by MTZ at 1–2 dpf, displaying β-cells in cyan with immunostaining for insulin and lineage-traced cells derived from drl-expressing mesodermal cells in red from the Cre-recombined ubi:Switch. Pancreata are outlined by solid white lines (E, F). Arrowheads point to ectopic β-cells derived from the mesoderm (H–H’’’ and L–L’’’). Selected areas in dashed squares in (H) and (L) are magnified in split (H’, H’’, L’ and L’’) and merged (H’’’ and L’’’) channels, respectively. Scale bars = 20 μm (E–H, K, and L) or 10 μm (H’–H’’’ and L’–L’’). Anatomical axes: D (dorsal), V (ventral), A (anterior), and P (posterior).

Immunostaining against insulin displayed a normal set of β-cells in the pancreas of the drl-tracing larvae with or without npas4l mutation after 30 hr of regeneration (Figure 4C,E, and F). In line with the findings shown in Figure 1, the npas4l mutation induced the formation of ectopic β-cells in the mesenchymal region (Figure 4D,G, and H). Furthermore, 98.9% of the ectopic β-cells in the mesenchymal region were mCherry-positive (Figure 4H’–H’’’), indicating that they derived from the drl-expressing mesodermal cells.

With a similar setting, we injected the drl-tracing embryos (without any npas4l mutation) with control or etsrp morpholino at the one-cell stage. Npas4l is essential for the expression of etsrp, which is a key regulator of endothelial/myeloid cell specification and vasculogenesis (Reischauer et al., 2016; Sumanas and Lin, 2006). Similar to the npas4l mutation, knocking down etsrp led to the formation of ectopic β-cells (Figure 4I–L’’’). Co-injecting lower doses of two different etsrp morpholinos (MO1 and MO2) exerted a synergistic effect on ectopic β-cell emergence (Figure 4—figure supplement 4), suggesting that the phenotype was not due to an off-target effect. The majority of the ectopic β-cells (94.3%) in etsrp morphants was also lineage-traced back to the drl-expressing mesodermal cells, suggesting that the ectopic β-cell formation was also of mesodermal origin following etsrp knockdown.

The mesodermal cells lose npas4l expression after differentiating into ectopic β-cells

We then further examined the autonomous role of the progenitors of endothelial and hematopoietic cells in ectopic β-cell emergence in npas4l mutants by tracing the npas4l cell lineage with a zebrafish knock-in line, npas4lPt(+36-npas4l-p2a-Gal4-VP16)bns423. This knock-in line not only carries a +36 bp mutation in exon 2 of npas4l (npas4lbns423 mutation) but also expresses the transcription activator Gal4 under control of the endogenous npas4l promoter (npas4l:Gal4), allowing us to trace the npas4l lineage after crossing it into Tg(UAS:Cre);Tg(ubi:Switch) (Figure 5A).

Figure 5
Download asset Open asset
The mesodermal cells lose npas4l expression after differentiating into ectopic β-cells.

(A) Schematics of npas4lPt(+36-npas4l-p2a-Gal4-VP16)bns423 (npas4l:Gal4), UAS:Cre and −3.5ubb:LOXP-EGFP-LOXP-mCherry (ubi:Switch). (B, C) Quantification of the pancreatic or ectopic β-cells with or without npas4l-positive mesodermal origin in control or transheterozygous npas4ls5/bns423 mutants at 3 dpf. ***p=0.0001 and ****P<0.0001 (Šidák’s multiple comparisons test); n = 20 (control) and 21 (npas4ls5/bns423). Data are represented as the mean ± SEM. (D–E’’’) Representative confocal images of ectopic β-cells and npas4l-positive lineage-traced cells in control siblings and npas4ls5/bns423 Tg(ins:Flag-NTR);Tg(ubi:Switch);Tg(npas4l:Gal4);Tg(UAS:Cre) zebrafish larvae at 3 dpf after β-cell ablation by MTZ from 1 to 2 dpf, displaying β-cells in cyan with immunostaining for insulin and lineage-traced cells derived from npas4l-expressing mesodermal cells in red from the Cre-recombined ubi:Switch. The selected area in a dashed square in (E) is magnified in split (E’ and E’’) and merged (E’’’) channels, respectively. Arrowheads point to ectopic β-cells derived from the mesoderm (E–E’’’). (F) Schematics of npas4lPt(+36-npas4l-p2a-Gal4-VP16)bns423 (npas4l:Gal4) and UAS:EGFP. (G–G’’’) Representative confocal images of ectopic β-cells losing npas4l expression in npas4ls5/bns423 Tg(ins:Flag-NTR); Tg(npas4l:Gal4);Tg(UAS:EGFP) zebrafish larvae at 3 dpf after β-cell ablation by MTZ from 1 to 2 dpf, displaying β-cells in magenta with immunostaining for insulin and cells expressing npas4l in green. The selected area in a dashed square in (G) is magnified in split (G’ and G’’) and merged (G’’’) channels, respectively. Arrowheads point to ectopic β-cells without npas4l expression (G–G’’’). Scale bars = 20 μm (D, E, and G) or 10 μm (E’–E’’’ and G’–G’’). Anatomical axes: D (dorsal), V (ventral), A (anterior), and P (posterior).

We obtained a transheterozygous npas4l mutant (npas4ls5/bns423) by crossing this npas4l tracer carrying the npas4lbns423 mutation into the npas4ls5 mutant used in the studies above, and observed similar ectopic β-cell formation (Figure 5B–E’’’) as found in the homozygous npas4ls5 mutant (Figure 1). Some of these ectopic β-cells (32.6%) could be traced back to the npas4l lineage as they were labelled by ubb:mCherry after the Cre-recombination driven by npas4l:Gal4 and UAS:Cre (Figure 5C and E–E’’’), which further supports a mesodermal origin of the ectopic β-cells.

By crossing Tg(npas4l:Gal4) into Tg(UAS:EGFP), the endogenous promoter activity of npas4l could be revealed by EGFP (Figure 5F). The ectopic β-cells emerged in close proximity to mesodermal cells with active npas4l expression in the transheterozygous npas4ls5/bns423 mutants (Figure 5G–G’’’). However, we did not observe any ectopic β-cells with EGFP expression, suggesting that these mesodermal cells have lost the identity of the lateral plate mesoderm with npas4l expression and started to express insulin as in the endodermal pancreatic β-cells instead.

The ectopic β-cells derive from the etsrp-expressing mesodermal lineage in etsrp morphants

To confirm the origin of the ectopic β-cell using a different lineage-tracing approach, we generated Tg(etsrp:Cre) zebrafish, based on the promoter of −2.3estrp:GFP that has been demonstrated to closely recapitulate the endogenous etsrp expression in the lateral plate mesoderm and vasculature (Veldman and Lin, 2012). We then crossed Tg(etsrp:Cre) into Tg(ubi:Switch);Tg(ins:Flag-NTR), and thereby labelling descendants of the etsrp lineage in red (Figure 6A). We validated the efficiency of the etsrp-lineage tracer and revealed that the majority of kdrl-expressing endothelial cells were being traced in the intersegmental vessels (86.6%) and other vasculature (Figure 6—figure supplement 1). At the one-cell stage, we injected the etsrp-tracing embryos with control or etsrp morpholinos. After β-cell ablation by MTZ treatment at 1–2 dpf and β-cell regeneration for 30 hr ectopic β-cells formed in the etsrp morphants, and 73.9% of the ectopic β-cells were labelled in red (Figure 6B–E’’’), illustrating that the etsrp-expressing lineage gave rise to a significant portion of the ectopic β-cells.

Figure 6 with 1 supplement see all
Download asset Open asset
The ectopic β-cells derive from the etsrp-expressing mesodermal lineage in etsrp morphants.

(A) Constructs of etsrp:Cre and −3.5ubb:LOXP-EGFP-LOXP-mCherry (ubi:Switch). (B, C) Quantification of the pancreatic or ectopic β-cells with or without etsrp-positive mesodermal origin in control or etsrp morpholino (MO)-injected larvae at 3 dpf. *p=0.0169, ***p=0.0002, and ****p<0.0001 (Šidák’s multiple comparisons test); n=33 (control) and 30 (etsrp MO). Data are represented as the mean ± SEM. (D–E’’’) Representative confocal images of ectopic β-cells and etsrp-positive lineage-traced cells in control or etsrp MO-injected Tg(ins:Flag-NTR);Tg(ubi:Switch);Tg(etsrp:Cre) zebrafish larvae at 3 dpf after β-cell ablation by MTZ at 1–2 dpf, displaying β-cells in cyan with immunostaining for insulin and lineage-traced cells derived from etsrp-expressing mesodermal cells in red from the Cre-recombined ubi:Switch. The selected area in a dashed square in (E) is magnified in split (E’, E’’) and merged (E’’’) channels, respectively. Arrowheads point to ectopic β-cells derived from the mesoderm (E–E’’’). (F) Constructs of etsrp:Cre and ins:LOXP-mCherry-LOXP-Hsa.HIST1H2BJ-GFP (ins:CSH). (G–G’’’) Representative confocal images of ectopic β-cells derived from the etsrp-expressing lineage in etsrp MO-injected Tg(ins:Flag-NTR);Tg(ins:CSH);Tg(etsrp:Cre) zebrafish larvae at 3 dpf after β-cell ablation by MTZ at 1–2 dpf, displaying β-cells in magenta with immunostaining for insulin and lineage-traced cells derived from etsrp-expressing mesodermal cells in nuclear green from the Cre-recombined ins:CSH. The selected area in a dashed square in (G) is magnified in split (G’ and G’’) and merged (G’’’) channels, respectively. Scale bars = 20 μm (D, E, and G) or 10 μm (E’–E’’’ and G’–G’’’). Anatomical axes: D (dorsal), V (ventral), A (anterior), and P (posterior).

Moreover, we replaced ubi:Switch with ins:LOXP-mCherry-LOXP-Hsa.HIST1H2BJ-GFP (ins:CSH) in the etsrp-tracing zebrafish larvae to directly trace insulin-expressing cells originating from the etsrp-expressing mesodermal lineage (Figure 6F). The co-localisation of insulin staining and the nuclear green tracer further confirms the mesodermal lineage of the ectopic β-cells (Figure 6G–G’’’).

Together, we used three different lineage-tracing models as well as three different loss of function models, that is using the promoter of drl, npas4l, or etsrp to drive Cre in npas4ls5/s5 mutants, npas4ls5/bns423 mutants, or etsrp morphants. This suggests that the ectopic β-cell formation is not restricted to the loss of a specific gene, but rather due to the perturbation of endothelial/myeloid specification.

Discussion

In this study, we first examined the role of blood vessels in β-cell regeneration in the cloche zebrafish mutant, which carries a homozygous npas4l mutation (Reischauer et al., 2016). We then unexpectedly revealed β-cells regenerating ectopically in the mesenchymal area. The ectopic β-cells were likely functional because they expressed several endocrine and β-cell markers, including Isl1, mnx1, and pcsk1, and were capable of responding to glucose to induce calcium oscillations during β-cell regeneration, although we do not know if they possess all the features of bona fide β-cells. By combining in situ hybridisation, lineage tracing, and confocal microscopy, we successfully traced the origin of the ectopic β-cells to the mesodermal lineage. A recent study has reported the conversion of Etsrp-deficient vascular progenitors into skeletal muscle cells and highlighted the plasticity of mesodermal cell fate determination within the same germ layer (Chestnut et al., 2020). Our data demonstrated the plasticity of β-cell differentiation across the committed germ layers in vivo, i.e., switching from a mesodermal to an endodermal fate in a regenerative setting, while gastrulation and cell fate commitment in the germ layers are considered to be irreversible in development. Ectopic pancreata have been observed before, e.g., in Hes1 mutant mice (Fukuda, 2006; Sumazaki et al., 2004), although that has been shown to be through an expansion of the pancreas rather than through changes in cell fate determination across organs or germ layers (Jørgensen et al., 2018). Our discovery is, to our knowledge, the first demonstration of ectopic β-cells with a mesodermal origin in vivo.

The recent genome-wide study with zebrafish embryos has confirmed the crucial role of npas4l in the early specification of endothelium and blood as the expressions of some endothelial and hematopoietic genes like tal1, etsrp, lmo2, gfi1aa, and gata1a were downregulated in homozygous npas4l mutants (Marass et al., 2019), suggesting that some populations of mesodermal cells may not have acquired their designated cell fates and remained in a more plastic state. In line with the important role of etsrp in the endothelial cell fate determination, Chestnut et al., 2020 have reported a significant reduction in the cell clusters of endothelial cells and endothelial progenitor cells in homozygous etsrp mutants in a single-cell RNA-seq analysis. Interestingly, they have also shown a remarkable increase in the cluster of lateral plate mesoderm, which points to the possibility that a larger pool of more versatile mesodermal cells remained in the homozygous etsrp mutants. However, it is difficult to deduce why those mesodermal cells with perturbed cell fates would be competent to differentiate into ectopic insulin-expressing β-cells from these transcriptomic data; because the population of these competent mesodermal cells might be very small according to the number of ectopic β-cells that we observed and their single-cell RNA-seq was not performed after β-cell ablation.

Since we could trace the origin of ectopic β-cells back to drl, npas4l, and etsrp lineages, we believe that the versatile lateral plate mesoderm close to the pancreas could contribute to the β-cell regeneration in npas4l mutants or etsrp morphants. Tracing the lineages with early mesodermal, endodermal, and endocrine markers, and sorting these lineage-traced cells could be necessary for a single-cell transcriptomic study to fully understand the cellular status of this small population of versatile cells and further elucidate the underlying molecular mechanisms. Before deciphering these mechanisms, we cannot formally rule out the possibility that the mechanisms of ectopic β-cell formation in npas4l mutants and etsrp morphants could be different.

The mutated gene in the cloche mutant was named npas4l because its encoded protein shares homology with human NPAS4 (Reischauer et al., 2016). Although injecting human NPAS4 mRNA or zebrafish npas4l mRNA into zebrafish cloche mutant embryos at the one-cell stage could rescue the mutants, Npas4 knockout mice are unlikely to share the same severe vascular and hematopoietic defects as zebrafish npas4l mutants because Npas4 knockout mice survive to adulthood (Lin et al., 2008). This discrepancy suggests that other members of the mammalian NPAS protein family or other proteins may be functionally redundant with NPAS4 in vascular and hematopoietic development. Mammalian NPAS4 has been shown to have important cell-autonomous functions in β-cells (Sabatini et al., 2018; Speckmann et al., 2016). In zebrafish, npas4a is the main npas4 paralogue expressed in β-cells (Tarifeño-Saldivia et al., 2017), meaning that it is unlikely the phenotype we identified early in development in npas4l mutants is related to the functions of Npas4 in β-cells. Further studies on NPAS4, related bHLH transcription factors, and ETV2 in mammals should elucidate whether inactivating such factors promotes β-cell formation with or without significantly perturbing the development of blood cells and vessels.

In summary, we have shown that the npas4l mutation and etsrp knockdown each induces ectopic regeneration of β-cells from the mesoderm. Our findings suggest a plasticity potential of mesodermal cells to differentiate into endodermal cells including β-cells (Figure 7). Future studies on the restriction of this plasticity would not only increase our understanding of the gating role of Npas4l and Etsrp in cell fate determination but also help to exploit an alternative source for β-cell regeneration.

Figure 7
Download asset Open asset
Npas4l/Etsrp restricts the plasticity of the mesoderm.

Mesoderm and endoderm normally follow Waddington’s landscape model to further differentiate into cells with mesodermal fates and endodermal fates, respectively, during development. However, mutating npas4l or knocking down etsrp not only abolishes the endothelial/myeloid specification but also induces plasticity of mesodermal cells to enable their differentiation into β-cells across the germ layer boundary.

Materials and methods

Key resources table
Reagent type
(species) or
resource		DesignationSource or
reference		IdentifiersAdditional
information
Genetic reagent (Danio rerio)npas4ls5PMID:7588049ZFIN: ZDB-ALT-010426–6
Genetic reagent (Danio rerio)npas4lPt(+36-npas4l-p2a-Gal4-VP16)bns423OtherWill be described in detail elsewhere (K.M. and D.Y.R.S., Manuscript in preparation)
Genetic reagent (Danio rerio)Tg(ins:Hsa.HIST1H2BJ-GFP;ins:DsRed)s960PMID:25117518ZFIN: ZDB-ALT-131001–6
Genetic reagent (Danio rerio)Tg(ins:FLAG-NTR,cryaa:mCherry)s950PMID:22608007ZFIN: ZDB-ALT-130930–5
Genetic reagent (Danio rerio)Tg(ptf1a:EGFP)jh1PMID:16258076ZFIN: ZDB-ALT-070531–2
Genetic reagent (Danio rerio)TgBAC(hand2:EGFP)pd24PMID:21397850ZFIN: ZDB-ALT-110128–40
Genetic reagent (Danio rerio)TgBAC(neurod1:EGFP)nl1PMID:18305245ZFIN: ZDB-ALT-080701–1
Genetic reagent (Danio rerio)TgBAC(pdx1:EGFP)bns13PMID:31142539ZFIN: ZDB-ALT-191212–6
Genetic reagent (Danio rerio)Tg(mnx1:GFP)ml2PMID:16162647ZFIN: ZDB-ALT-051025–4
Genetic reagent (Danio rerio)Tg(pcsk1:EGFP)KI106PMID:27516442ZFIN: ZDB-ALT-161115–10
Genetic reagent (Danio rerio)TgBAC(ascl1b:EGFP-2A-Cre-ERT2)ulg006PMID:26329351ZFIN: ZDB-ALT-160205–2
Genetic reagent (Danio rerio)Tg(ins:GCaMP6s,cryaa:RFP)PMID:28939870ZFIN: ZDB-ALT-181221–2
Genetic reagent (Danio rerio)Tg(−3.5ubb:LOXP-EGFP-LOXP-mCherry)cz1701PMID:21138979ZFIN: ZDB-ALT-110124–1
Genetic reagent (Danio rerio)Tg(−6.35drl:Cre-ERT2,cryaa:Venus)cz3333PMID:26306682ZFIN: ZDB-ALT-160129–4
Genetic reagent (Danio rerio)Tg(5xUAS:EGFP)nkuasgfp1aPMID:18202183ZFIN: ZDB-ALT-080528–1
Genetic reagent (Danio rerio)Tg(kdrl:EGFP)s843PMID:16251212ZFIN: ZDB-ALT-050916–14
Genetic reagent (Danio rerio)Tg(ubb:LOXP-CFP-LOXP-zgc:114046-mCherry)jh63PMID:25773748ZFIN: ZDB-ALT-151007–31
Genetic reagent (Danio rerio)Tg(ins:LOXP-mCherry-LOXP-Hsa.HIST1H2BJ-GFP,cryaa:Cerulean)s934PMID:21497092ZFIN: ZDB-ALT-111031–2
Genetic reagent (Danio rerio)Tg(etsrp:iCre;cryaa:Venus)KI114This paperSee Materials and methods, section Zebrafish
Genetic reagent (Danio rerio)Tg(UAS:Cre, cryaa:Cerulean)bns382This paperSee Materials and methods, section Zebrafish
AntibodyAnti-GFP (chicken polyclonal)Aves LabsGFP-1020(1:500)
AntibodyAnti-RFP (rabbit polyclonal)Abcamab62341(1:500)
AntibodyAnti-tdTomato (goat polyclonal)MyBioSourceMBS448092(1:500)
AntibodyAnti-insulin (rabbit polyclonal)Cambridge Research BiochemicalsCustomised(1:100)
AntibodyAnti-pan-cadherin (rabbit polyclonal)Sigma-AldrichC3678(1:5000)
AntibodyAnti-islet-1-homeobox (mouse monoclonal)DSHB40.3A4 supernatant(1:10)
Recombinant DNA reagentp5E-MCS (plasmid)Tol2kit228
Recombinant DNA reagentp5E-etsrp (plasmid)This paper−2.3etsrp promoter inserted into p5E-MCS
Recombinant DNA reagentpME-iCre (plasmid)OtherFrom Kristen M. Kwan’s lab
Recombinant DNA reagentp3E-polyA (plasmid)Tol2kit302
Recombinant DNA reagentpDestTol2gY (plasmid)OtherFrom Naoki Tsuji
Recombinant DNA reagent−2.3etsrp:iCre,cryaa:Venus (plasmid)This paperSee Materials and methods, section Zebrafish
Sequence-based reagent−2.3etsrp_FWDThis paperPCR primerTATAGGGCGAATTGggtaccTTCAGTAAGCAGACTCCTTCAATCA
Sequence-based reagent−2.3etsrp_REVThis paperPCR primerAGCTGGAGCTCCAccgcggTTCGGCATACTGCTGTTGGAC
Sequence-based reagentStandard control morpholinoGene ToolsMorpholinoCCTCTTACCTCAGTTACAATTTATA
Sequence-based reagentRandom control morpholinoGene ToolsMorpholinoMixture of many oligo sequences
Sequence-based reagentetsrp morpholino MO1Gene ToolsMorpholino
ZFIN: ZDB-MRPHLNO-060407–2		TTGGTACATTTCCATATCTTAAAGT
Sequence-based reagentetsrp morpholino MO2Gene ToolsMorpholino
ZFIN: ZDB-MRPHLNO-060407–3		CACTGAGTCCTTATTTCACTATATC
Sequence-based reagentnpas4l_ISH_FWDThis paperPCR primerACTCGGGCATCAGGAGGATC
Sequence-based reagentnpas4l_ISH_REVThis paperPCR primer(CCTAATACGACTCACTATAGGG)GACACCAGCATACGACACACAAC
Sequence-based reagentdrl_ISH_FWDThis paperPCR primerATGAAGAATACAACAAAACCC
Sequence-based reagentdrl_ISH_REVThis paperPCR primer(CCTAATACGACTCACTATAGGG)TGAGAAGCTCTGGCCGC
Sequence-based reagentnpas4ls5_FWDPMID:27411634PCR primerTTCCATCTTCTGAATCCTCCA
Sequence-based reagentnpas4ls5_REVPMID:27411634PCR primerGGACAGACCCAGATACTCGT
Sequence-based reagentnpas4ls5_SEQThis paperSequencing primerTTTCTGCCGTGAATGGATGTG
Commercial assay or kitGateway LR Clonase II Enzyme MixInvitrogen11791–020
Commercial assay or kitPhusion High-Fidelity DNA PolymeraseThermo ScientificF-530L
Commercial assay or kitIn-Fusion HD Cloning KitTakara Bio639648
Commercial assay or kitMAXIscript T7/T3 Transcription KitInvitrogenAM1324
Chemical compound, drugMetronidazoleSigma-AldrichM3761
Chemical compound, drug4-HydroxytamoxifenSigma-AldrichH7904
Software, algorithmImageJPMID:22930834
Software, algorithmFijiPMID:22743772
Software, algorithmLAS X Version 3.5.5.19976Leica
Software, algorithmZEN 3.1Zeiss
Software, algorithmGraphPad Prism 9GraphPad Software
OtherDAPIThermoFisher ScientificD1306(1 µg/mL)

Zebrafish

The following previously published mutant and transgenic zebrafish (Danio rerio) lines were used: clocheS5 (Field et al., 2003) as the npas4ls5 mutant, Tg(ins:Hsa.HIST1H2BJ-GFP;ins:DsRed)s960 (Tsuji et al., 2014) abbreviated as Tg(ins:H2BGFP;ins:DsRed), Tg(ins:FLAG-NTR,cryaa:mCherry)s950 (Andersson et al., 2012) abbreviated as Tg(ins:Flag-NTR), Tg(ptf1a:EGFP)jh1 (Godinho et al., 2005), TgBAC(hand2:EGFP)pd24 (Kikuchi et al., 2011), TgBAC(neurod1:EGFP)nl1 (Obholzer et al., 2008), TgBAC(pdx1:EGFP)bns13 (Helker et al., 2019), Tg(mnx1:GFP)ml2 (Flanagan-Steet et al., 2005), Tg(pcsk1:EGFP)KI106 (Lu et al., 2016), TgBAC(ascl1b:EGFP-2A-Cre-ERT2)ulg006 (Ghaye et al., 2015) abbreviated as Tg(ascl1b:EGFP), Tg(ins:GCaMP6s,cryaa:RFP) (Singh et al., 2017) generated by reinjecting the plasmid and abbreviated as Tg(ins:GCaMP6s), Tg(−3.5ubb:LOXP-EGFP-LOXP-mCherry)cz1701 (Mosimann et al., 2011) abbreviated as Tg(ubi:Switch), Tg(−6.35drl:Cre-ERT2,cryaa:Venus)cz3333 (Mosimann et al., 2015) abbreviated as Tg(drl:CreERT2), Tg(5xUAS:EGFP)nkuasgfp1a (Asakawa et al., 2008) abbreviated as Tg(UAS:EGFP), Tg(kdrl:EGFP)s843 (Jin et al., 2005), Tg(ubb:LOXP-CFP-LOXP-zgc:114046-mCherry)jh63 (Wang et al., 2015), and Tg(ins:LOXP-mCherry-LOXP-Hsa.HIST1H2BJ-GFP,cryaa:Cerulean)s934 (Hesselson et al., 2011), which is referred to Tg(ins:mCherry) in Figures 2 and 3 and Video 1 and Tg(ins:CSH) in Figure 6.

The npas4lPt(+36-npas4l-p2a-Gal4-VP16)bns423 line (abbreviated as npas4l:Gal4 or npas4lbns423) was generated using the GeneWeld system (Wierson et al., 2019) and will be described in detail elsewhere (K.M. and D.Y.R.S., Manuscript in preparation).

The Tg(etsrp:iCre;cryaa:Venus)KI114 line, abbreviated as Tg(etsrp:Cre), was generated by the Tol2 transposon system, and the construct was made by MultiSite Gateway Cloning (Invitrogen). The amplicon of the −2.3etsrp promoter was synthesised from zebrafish genomic DNA with a forward primer 5’-TATAGGGCGAATTGggtaccTTCAGTAAGCAGACTCCTTCAATCA-3’ and a reverse primer 5’-AGCTGGAGCTCCAccgcggTTCGGCATACTGCTGTTGGAC-3’ by Phusion High-Fidelity DNA Polymerase (Thermo Scientific) as an insert for In-Fusion Cloning (Takara Bio) with p5E-MCS using restriction sites KpnI and SacII to yield p5E-etsrp. Subsequently, p5E-etsrp, pME-iCre, p3E-polyA, and pDestTol2gY were used for the LR reaction to generate the construct −2.3etsrp:iCre,cryaa:Venus. Expressing iCre (improved Cre recombinase) by an etsrp promoter would cleave the loxP sites of reporters, e.g. ubi:Switch, which allows us to trace the descendants differentiated from an etsrp lineage.

The Tg(UAS:Cre, cryaa:Cerulean)bns382 line, abbreviated as Tg(UAS:Cre), was generated via Tol2-mediated transgenesis by injecting a construct that placed the Cre recombinase coding sequence downstream of five tandem copies of the upstream activation sequence (UAS). The eye-marker cryaa:Cerulean cassette (Hesselson et al., 2009) was inserted downstream of Cre in the reverse orientation using 0.56 kb of the cryaa promoter (Kurita et al., 2003).

Males and females ranging in age from 3 months to 2 years were used for breeding to obtain new offspring for experiments. Individuals were sorted into the control sibling group (npas4l+/+, npas4ls5/+, or npas4lbns423/+) and the npas4l mutant group (npas4ls5/s5 or npas4ls5/bns423) based on the characteristic pericardial oedema and blood-cell deficit. Zebrafish larvae were allocated into different experimental groups based on their phenotypes and genotypes in experiments involving cloche mutants. In morpholino knockdown experiments, zebrafish embryos were randomly assigned to each experimental condition for injection. Experimental procedures were performed on the zebrafish from 10 hpf to 3 dpf before the completion of sex determination and gonad differentiation. All zebrafish, except npas4l mutants and etsrp morphants, appeared healthy and survived to adulthood. The npas4l mutants exhibited pericardial oedema, bell-shaped hearts, and blood deficits, as previously reported (Stainier et al., 1995). The etsrp morphants had similar phenotypes. All studies involving zebrafish were performed in accordance with local guidelines and regulations and approved by regional authorities.

Chemical ablation of β-cells

Request a detailed protocol

As in our previous report (Schulz et al., 2016), we ablated β-cells by incubating the β-cell ablation model Tg(ins:Flag-NTR) zebrafish in E3 medium supplemented with 10 mM metronidazole (MTZ, Sigma-Aldrich), 1% DMSO (VWR), and 0.2 mM 1-phenyl-2-thiourea (Acros Organics) for 24 hr from 1 to 2 dpf.

Microinjection of morpholinos

Request a detailed protocol

Standard control morpholino (5’-CCTCTTACCTCAGTTACAATTTATA-3’), random control morpholino, etsrp morpholino MO1 (5’-TTGGTACATTTCCATATCTTAAAGT-3’), and MO2 (5’-CACTGAGTCCTTATTTCACTATATC-3’) (Sumanas and Lin, 2006) were purchased from Gene Tools, LLC. Morpholinos were injected into the one-cell-stage zebrafish embryos at the doses specified in the figure legends.

Lineage tracing by tamoxifen-inducible cre recombinase

Request a detailed protocol

To genetically trace the mesodermal lineage, we treated Tg(ins:Flag-NTR);Tg(ubi:Switch);Tg(drl:CreERT2) zebrafish embryos with 10 μM 4-OHT (Sigma-Aldrich) in E3 medium in 90 mm Petri dishes, with approximately 60 individuals per dish, from 10 to 12 hpf. Upon induction by 4-OHT, cytoplasmic CreERT2 would be translocated to the nucleus to excise the loxP-flanked EGFP to enable mCherry expression in drl-expressing cells and their descendants, indicating a mesodermal lineage.

Sample fixation for immunostaining

Request a detailed protocol

Before fixing the zebrafish larvae, we confirmed the presence of the transgenes by determining the corresponding fluorescent markers and subsequently examined them under a widefield fluorescence microscope LEICA M165 FC (Leica Microsystems). We then euthanised the zebrafish larvae with 250 mg/L tricaine (Sigma-Aldrich) in E3 medium followed by washing in distilled water three times. We fixed the samples in 4% formaldehyde (Sigma-Aldrich) in PBS (ThermoFisher Scientific) at 4°C overnight. After washing away the fixative with PBS three times, we removed the skin and crystallised yolk of the zebrafish larvae by forceps under the microscope to expose the pancreas and mesenchyme for immunostaining.

Immunostaining and confocal imaging

Request a detailed protocol

As in our previous report (Liu et al., 2018), we started immunostaining by incubating the zebrafish samples in blocking solution (0.3% Triton X-100, 4% BSA and 0.02% sodium azide from Sigma-Aldrich in PBS) at room temperature for 1 hr. We then incubated the samples in blocking solution with primary antibodies at 4°C overnight. After removing the primary antibodies, we washed the samples with washing buffer (0.3% Triton X-100 in PBS) eight times at room temperature for 2 hr. Afterwards, we incubated the samples in blocking solution with fluorescent dye-conjugated secondary antibodies and the nuclear counterstain DAPI (ThermoFisher Scientific) if applicable at 4°C overnight. Next, we removed the secondary antibodies and nuclear counterstain and washed the samples with washing buffer eight times at room temperature for 2 hr. The following primary antibodies were used: anti-GFP (1:500, Aves Labs, GFP-1020), anti-RFP (1:500, Abcam, ab62341), anti-tdTomato (1:500, MyBioSource, MBS448092), anti-insulin (1:100, Cambridge Research Biochemicals, customised), anti-pan-cadherin (1:5000, Sigma, C3678), and anti-islet-1-homeobox (1:10, DSHB, 40.3A4 supernatant).

Before confocal imaging, we mounted the stained samples in VECTASHIELD Antifade Mounting Medium (Vector Laboratories) on microscope slides with the pancreas facing the cover slips. We imaged the pancreas and neighbouring mesenchyme of every zebrafish sample that we had mounted with the confocal laser scanning microscopy platform Leica TCS SP8 and LAS X (Leica Microsystems). We analysed the images by Fiji (Schindelin et al., 2012) and classified a β-cell as pancreatic when it was located in the pancreas, as delineated by pan-cadherin, DAPI, or ptf1a:EGFP labelling. The insulin-positive cells outside of the pancreas were defined as ectopic β-cells.

Live calcium imaging of zebrafish β-cells

Request a detailed protocol

Imaging of pancreatic and ectopic β-cells was performed on 3 dpf Tg(ins:GCaMP6s);Tg(ins:mCherry);Tg(ins:Flag-NTR) npas4l mutants on a ZEISS LSM 980 confocal microscope equipped with a W Plan-Apochromat ×20/1 NA water correction lens and operated by ZEN. The GCaMP6s and mCherry signals from β-cells were simultaneously acquired using the 488 nm and 587 nm laser lines. The GCaMP6s signal was rendered in green, and the mCherry signal was rendered in red. Time series recordings were taken with an in-plane resolution of 1024 × 1024 pixels and a fully open pinhole to maximise light capture. The videos were recorded for 400 cycles, with approximately 2 s acquisition time per frame. Videos were analysed in ImageJ (Schneider et al., 2012).

Whole-mount in situ hybridisation

Request a detailed protocol

Zebrafish embryos at 10 and 20 dpf were fixed with 4% paraformaldehyde in PBS at 4°C overnight. Whole-mount in situ hybridisation was performed according to the method in a previous report (Thisse and Thisse, 2008). Probes against npas4l and drl were synthesised from transcription templates from PCR using bud-stage zebrafish cDNA, Phusion High-Fidelity DNA Polymerase (Thermo Scientific) and primer pairs 5’-ACTCGGGCATCAGGAGGATC-3’ plus 5’-(CCTAATACGACTCACTATAGGG) GACACCAGCATACGACACACAAC-3’ for npas4l, and 5’-ATGAAGAATACAACAAAACCC-3’ plus 5’-(CCTAATACGACTCACTATAGGG) TGAGAAGCTCTGGCCGC-3’ for drl, respectively. Plasmids carrying probe templates of foxa2, gsc, mixl1, and ascl1b were linearised by corresponding restriction enzymes. T7 (except foxa2 probe with T3) was employed for transcription, and digoxigenin (Roche) was used for labelling. To genotype the npas4ls5 mutants, PCR was performed using gDNA from the imaged samples and primers 5’-TTCCATCTTCTGAATCCTCCA-3’ plus 5’-GGACAGACCCAGATACTCGT-3’ at the conditions previously reported (Reischauer et al., 2016). The PCR products were then sent for sequencing with the following primer 5’-TTTCTGCCGTGAATGGATGTG-3’ (Eurofins Genomics).

Statistical analysis

Request a detailed protocol

Similar experiments were performed at least twice independently. The number of cells in the confocal microscopy images was all quantified manually with the aid of the Multipoint Tool from Fiji (Schindelin et al., 2012). Data were then analysed with Prism (GraphPad). Statistical analyses were carried out by two-tailed t-tests when two groups were analysed and by ANOVA when more than two groups were analysed. We have presented the results as the mean values ± SEM and considered p-values ≤ 0.05 to be statistically significant. The n number represents the number of zebrafish individuals in each group of each experiment.

Data availability

All data generated or analysed during this study are included in the manuscript and source data files.

References

  48. Book
    1. Waddington CH
    (1957)
    The Strategy of the Genes; a Discussion of Some Aspects of Theoretical Biology
    Allen & Unwin.

Decision letter

  1. Elke Ober
    Reviewing Editor; University of Copenhagen, Denmark
  2. Marianne E Bronner
    Senior Editor; California Institute of Technology, United States
  3. Elke Ober
    Reviewer; University of Copenhagen, Denmark
  4. Valerie Gouon-Evans
    Reviewer

Our editorial process produces two outputs: i) public reviews designed to be posted alongside the preprint for the benefit of readers; ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.

Acceptance summary:

This is an elegant study demonstrating the emergence of mesoderm-derived β-like cells following β-cell ablation in the context of compromised endothelial/myeloid lineage specification. These findings will be of interest to scientists in the areas of regeneration and reprogramming, as they reveal a previously unknown degree of germ layer plasticity in the embryo. In the long term the study has potential impact in the diabetes field, as it reveals a potentially novel path for redirecting somatic cells into insulin-producing cells in an in vivo context.

Decision letter after peer review:

Thank you for submitting your article "Regenerating insulin-producing β-cells ectopically from a mesodermal origin in the absence of endothelial specification" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, including Elke Ober as the Reviewing Editor and Reviewer #1, and the evaluation has been overseen by Marianne Bronner as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Valerie Gouon-Evans (Reviewer #3).

Essential Revisions:

1) Improve characterization of β-cell differentiation derived from mesoderm-endoderm conversion. The manuscript provides strong evidence that mesoderm-derived ectopic β-cells generated after β-cell ablation express insulin, nevertheless the characterization is too superficial to conclude there is a full conversion to endoderm. Indeed, the authors describe that in contrast to pancreatic β cells, most ectopic β-cells lack the pan-endocrine marker neurod1 and the key β cell gene pdx1 (Figure 2 and Figure 2-S1), suggesting these cells clearly differ from bona fide β-cells. In addition, analysis of endodermal or mesodermal markers is lacking, making it difficult to understand the exact nature of these ectopic cells. Therefore, a more comprehensive analysis of relevant key endodermal and mesodermal genes is required to better describe the β-cell state of these ectopic cells and determine the degree of germ layer conversion.

2) Address the competence of mesodermal cells converting into β-like cells in larvae with impaired npas4l and etv2 function. The manuscript does not sufficiently discuss the identity of the mesodermal cell population that in the absence of endothelial specification (npas4l-/- and etv2 MO) can adopt a β-cell like identity following β-cell ablation. Specifically, the following points should be discussed:

i) why are they competent to differentiate into insulin+ cells in response to β cell ablation? The authors could make use of the recent studies cited in the present paper and of the respective transcriptomic data (Marass, Development, 2019; Chestnut, Nat Commun, 2020).

ii) is there a temporal window of competence for extra β-cells to form in the mesenchyme? Is this restricted to continued development? Could this be tested in adults?

iii) is the competence to form extra β-cells restricted to the vicinity of the pancreas? If so, which signals control this, given that normally endothelial cells form along the anteroposterior axis, however do not adopt a β-cell like fate in npas4l mutants or etv2morphants?

iv) what is the possible mechanism of this specification?

3) The mesodermal lineage tracing experiments, which are a central part of the study, require additional support. Please define efficiency and specificity of the tracing approach by showing mCherry+ mesodermal cells co-stained with a mesodermal marker (e.g. drl, or alternative mesodermal gene expressed at the stages of ectopic β-cell analysis) in additional areas of the larvae. This should include control embryos for both mesodermal Cre-drivers (drl:CreERT2 and etv2:iCre) to show their exclusive mesoderm expression.

4) The etv2 knockdown requires further validation. Elongated cells resembling endothelial cells marked with the etv2 lineage mapping present in etv2 morphants suggests that endothelial cells still form. Related to this, using only a single etv2 and single standard control morpholino is surprising given the known problems with this approach. Please validate the use of the etv2 morpholino knockdowns according to Stainier et al., 2017. Moreover, endothelial cell identity of above cells should be tested.

In case these results are confirmed, it would suggest that the mechanism of ectopic β-cells emergence in etv2 morphants and npas4l mutants does not depend on the absence of endothelial cells per se. It may also suggest that the mechanism of ectopic β cells formation in etv2 morphants vs npas4l mutant is different. This should be discussed as the role of vasculogenesis and vascularization for β cell regeneration was the first motivation for this study, as introduced in the result section.

5) Clarification of assessment criteria for different insulin+ cell populations and their localization to the principle islet or ectopic tissues.

The regenerated β cells seem to be broadly distributed and not always seem to form a distinct islet, but are nevertheless assigned to the pancreatic domain (see Figure 1E). To clarify how the pancreatic and the ectopic domains have been discriminated in these experiments, please provide additional details how both populations have been identified.

Moreover, lineage labelled Ins+Cherry+ cells from the mesoderm are also found within the pancreas; for instance a substantial fraction of regenerated β-cells defined as pancreatic seems to derive from etv2+ mesodermal cells in etv2 morphants (Figure 4B). This raises the question of the identity and of these "pancreatic" cells and where they arise from, and requires additional characterization and explanation.

The β-like cells appear to come from the lateral plate mesoderm (supplement figure4), please specify at least once in the text for clarity of lineage relationship.

6) The interpretation of glucose measurements needs to be revisited, given the relevance of measuring free tissue glucose as proxy for the effect of insulin released by ectopic cells in npas4l mutants is questionable. npas4l mutants have no capillaries nor blood to vehicle the hormone to target tissues in mutants. It is likely that the glucose values obtained for the mutants are not the result of insulin action, and hence may not provide a reliable estimation of ectopic cell function. This would be more convincing, for example, if a transient increase is observed after ablation before normalization. If this is the case, the authors should experimentally test whether increasing glucose levels in non-ablated npas4l mutants would trigger a similar formation of ectopic β-cells.

Therefore, please provide further support or adjust and discuss the conclusions.

7) The significance of the conversion of other endocrine cell types, specifically somatostatin+ cells, from mesoderm in npas4l mutants needs to be clarified. The increase in the number of ectopic sst2 cells after δ-cell ablation presented in Figure 1 Supplement 2F is very mild and in addition ectopic sst2 cell seem to frequently form in controls outside the pancreas. Please clarify and discuss general implications for the competence of the mesenchyme for the appearance of endocrine cells outside the pancreas, as in its current form, this result is more confusing than supporting the other results.

Has the normal distribution of data points in Figure 1 supplement 2F been tested for control and npas4l datasets?

8) A number of clarifications and additions to the text and figures are required:

- The title requires attention, it seems grammatically incorrect, e.g. lacking a verb.

- Figure 2 indicates that the extra b-cell phenotype is not fully penetrant, please include this in the manuscript.

- Clarify and specify in all figure legends what the red staining represents, e.g. DsRed fluorescence driven by the insulin promoter or immunofluorescence staining for insulin?

- Figure 1, supplement 1: Please delineate the pancreas in panels C and D to reveal the ectopic sites of emerging β-like cells.

- p-Cadherin: please spell out PanCadherin in the figures to avoid confusion with a phosphorylated Cadherin form.

- Please include short explanation for iCre in the text or methods.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Insulin-producing β-cells regenerate ectopically from a mesodermal origin under the perturbation of hemato-endothelial specification" for further consideration by eLife. Your article has been reviewed by 2 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Marianne Bronner as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Valerie Gouon-Evans (Reviewer #3).

The manuscript has been improved but there are some remaining issues that need to be addressed, as outlined below.

The manuscript does not require any further experimental data, nevertheless, we ask you to address some important additions to the text, as suggested by the reviewers. Including the following changes will add essential information and will help to make your work more accessible to the reader:

– mention once in the manuscript the previous name of etsrp (etv2) for the sake of visibility and clarity.

– precise when 4OHT was applied in the control experiment shown in Figure 4 – Supplement figure 3 (drl:CRE-ERT in the heart).

– include the characterization of specificity and efficiency of the lineage tracing etsrp:Cre in the manuscript as described in the rebuttal letter (response to point 3).

– summarize lines 289-294 (Discussion) to simply suggest that the possibly versatile mutant mesodermal cells could be the origin of the ectopic β cells.

Reviewer #2 (Recommendations for the authors):

The revised version of this manuscript and the rebuttal letter carefully address all the concerns and the remarks raised by the first version:

The title and the general conclusions have been adequately adapted so that they more fairly correspond to the observed phenomenon of ectopic pancreatic β cell formation in the mesoderm. The manuscript now presents some interesting discussion about the plasticity of mesodermal cells.

Importantly, the authors have better characterized their lineage tracing tools. Also, they include a third tool, a GAL4-VP16 knockin in the npas4l locus to be used with UAS:CRE;ubb:switch. Together, these three lineage tracing systems convincingly support the conclusion that many ectopic pancreatic cells are from mesodermal origin. It is to note, however, that this third tool is only minimally described as it is said it will be presented in another paper submitted by colleagues.

The study now provides in vivo Calcium imaging of ectopic (compared to pancreatic) β cells, which is an accurate way to assess the capacity of these cells to respond to glucose. This also adds to the asked characterization of the ectopic β cells.

The use of a secondo morpholino against etv2/etsrp is now presented, as well as the synergy between both (with respect to the number of ectopic β cells that form during regeneration). This analysis complies with previously published guidelines for MO.

In addition, the figures and Methods provide clarification about the way cells have been assigned a pancreatic or ectopic localization.

Source data for the quantifications shown in each figure are provided.

Overall, the revised version is greatly improved.

Additional recommendations:

– mention once in the manuscript the previous name of etsrp (etv2) for the sake of visibility and clarity.

– precise when 4OHT was applied in the control experiment shown in Figure 4 – Supplement figure 3 (drl:CRE-ERT in the heart).

– summarize lines 289-294 (Discussion) to simply suggest that the possibly versatile mutant mesodermal cells could be the origin of the ectopic β cells.

Reviewer #3 (Recommendations for the authors):

Overall, the revisions are satisfactory and have greatly improved the manuscript.

Point #1: I agree that the new data using the npas4lPt(+36-npas4l-p2a-Gal4-VP16)bns423 fish crossed either with Tg(UAS:Cre);Tg(ubi:Switch) or Tg(UAS:EGFP) fish provide additional evidence for the mesodermal origin of the de novo β cells that are characterized with loss of the mesodermal marker expression npas4l.

Additional discussion addresses well point #2.

Point #3: Please include the characterization of specificity and efficiency of the lineage tracing etsrp:Cre in the manuscript as described in the rebuttal letter.

Point #4: MO2 data support well the previous MO1 data.

Point #5 is well addressed.

Point #6: Evaluation of calcium signaling in the ectopic β-cells by live imaging in vivo is useful to examine cell maturation.

Point #7: Removal of data related to somatostatin+ cells is appropriate.

Point #8: Well addressed.

https://doi.org/10.7554/eLife.65758.sa1

Author response

Essential Revisions:

1) Improve characterization of β-cell differentiation derived from mesoderm-endoderm conversion. The manuscript provides strong evidence that mesoderm-derived ectopic β-cells generated after β-cell ablation express insulin, nevertheless the characterization is too superficial to conclude there is a full conversion to endoderm. Indeed, the authors describe that in contrast to pancreatic β cells, most ectopic β-cells lack the pan-endocrine marker neurod1 and the key β cell gene pdx1 (Figure 2 and Figure 2-S1), suggesting these cells clearly differ from bona fide β-cells. In addition, analysis of endodermal or mesodermal markers is lacking, making it difficult to understand the exact nature of these ectopic cells. Therefore, a more comprehensive analysis of relevant key endodermal and mesodermal genes is required to better describe the β-cell state of these ectopic cells and determine the degree of germ layer conversion.

We agree with the reviewers that the ectopic β-cells are not identical to the bona fide β-cells and more characterization of the ectopic β-cells would be necessary to fully understand their cellular state. Therefore in the revision we have included a new zebrafish tool npas4lPt(+36-npas4l-p2a-Gal4-VP16)bns423, which not only allows us to trace the mesodermal npas4l-positive lineage of the endothelial and hematopoietic progenitors after crossing into Tg(UAS:Cre);Tg(ubi:Switch), but also reports the expression of this important regulator of the mesodermal cell fate after crossing into Tg(UAS:EGFP). The details and characterization of this new fish line are reported in another manuscript by K.M. and D.Y.R.S (submitted).

In the new Figure 5, we have shown that some of the ectopic β-cells could be traced back to the mesodermal npas4l lineage while none of the ectopic β-cells were still actively expressing npas4l as reported by EGFP at 3 dpf. This indicates that at least some of these ectopic β-cells from the npas4l lineage had lost this mesodermal marker and begun to express insulin and other endodermal endocrine markers (Figure 2).

In addition, we examined the expression patterns of some early mesodermal or endodermal markers (foxa2, gsc and mixl1) and we did not see any significant differences in the npas4l mutants (Figure 4—figure supplement 1), indicating that npas4l mutation did not significantly perturb the early formation of mesoderm and endoderm.

2) Address the competence of mesodermal cells converting into β-like cells in larvae with impaired npas4l and etv2 function. The manuscript does not sufficiently discuss the identity of the mesodermal cell population that in the absence of endothelial specification (npas4l-/- and etv2 MO) can adopt a β-cell like identity following β-cell ablation. Specifically, the following points should be discussed:

i) why are they competent to differentiate into insulin+ cells in response to β cell ablation? The authors could make use of the recent studies cited in the present paper and of the respective transcriptomic data (Marass, Development, 2019; Chestnut, Nat Commun, 2020).

We thank the reviewers’ suggestion of making use of these two studies to discuss the competence of the mesodermal cell population differentiating into ectopic β-cells. We have now substantially expanded this discussion point and referred to the suggested references (Lines 277-305).

ii) is there a temporal window of competence for extra β-cells to form in the mesenchyme? Is this restricted to continued development? Could this be tested in adults?

The formation of ectopic β-cells, unfortunately, cannot be tested in adult cloche mutants because their health conditions start to deteriorate from 3 days post-fertilisation (dpf) and they die at 4-5 dpf. Therefore, they can never reach the adult stage. We analysed the cloche mutants at 3 dpf in the current manuscript but we also tried to examine them at 4 dpf, but there were worsened health conditions of the cloche mutants in general that prohibited a meaningful analysis. In addition, the first insulin-expressing β-cell does not show up before 15 hr post-fertilisation (hpf). Thus, there is not much room for us to move the ablation procedure earlier than what we did in the current manuscript (with a start at ~24 hpf).

iii) is the competence to form extra β-cells restricted to the vicinity of the pancreas? If so, which signals control this, given that normally endothelial cells form along the anteroposterior axis, however do not adopt a β-cell like fate in npas4l mutants or etv2morphants?

We believe that the vicinity of the pancreas could be an important factor for the competence of some mesodermal cells to form ectopic β-cells. However, this is difficult to prove before we precisely know which population of mesodermal cells can differentiate across the germ layers to endodermal β-cells. If we had knowledge of this, we could have isolated those cells and transplanted them into different positions to study the importance of the vicinity of the pancreas.

In addition, our new lineage-tracing data has shown that some of the ectopic β-cells could be traced back to the npas4l lineage (Figure 5), suggesting that the npas4l-expressing lateral plate mesoderm not far from the pancreas (Figure 4—figure supplement 1) could be the source of the ectopic β-cells. We have added this point to the newly expanded discussion (Line 296).

iv) what is the possible mechanism of this specification?

As supported by the transcriptomic data (Marass, Development, 2019; Chestnut, Nat Commun, 2020), we propose that a small population of cells from the lateral plate mesoderm have lost their mesodermal cell fate in npas4l mutants or etsrp (etv2) morphants, which maintains the plasticity of these cells and allows them to differentiate into ectopic β-cells during β-cell regeneration. We have discussed this possible mechanism in the expanded discussion (Lines 277-305), although the detailed molecular mechanisms remain elusive.

3) The mesodermal lineage tracing experiments, which are a central part of the study, require additional support. Please define efficiency and specificity of the tracing approach by showing mCherry+ mesodermal cells co-stained with a mesodermal marker (e.g. drl, or alternative mesodermal gene expressed at the stages of ectopic β-cell analysis) in additional areas of the larvae. This should include control embryos for both mesodermal Cre-drivers (drl:CreERT2 and etv2:iCre) to show their exclusive mesoderm expression.

We thank the reviewers for pointing out the importance of characterizing the Cre-drivers for lineage-tracing, especially with regards to the etsrp:Cre (which we previously called etv2:iCre), which has been newly made for this manuscript. We understand the significance of having an efficient and specific lineage-tracing approach, while we also realize that no approach is absolutely perfect. Therefore, in this revised manuscript, we have included yet another lineage tracer npas4lPt(+36-npas4l-p2a-Gal4-VP16)bns423;UAS:Cre;ubi:Switch. Together with the other two mesodermal lineage tracers drl:CreERT2;ubi:Switch and etsrp:Cre;ubi:Switch, we hope these three lineage tracers would complement each other to provide convincing evidence to show that the ectopic β-cells came from an mesodermal origin.

The Cre-driver drl:CreERT2 has been extensively characterized by Mosimann and others [Nat Commun 6, 8146 (2015)] and it can genetically trace the lateral plate mesoderm descendants including cardiac mesoderm, pectoral find mesoderm, red blood cells, kidney, vasculature and trunk muscle cells. In the revised manuscript, we have added images to show that this tracer could respond to tamoxifen induction and genetically trace cardiac cells (Figure 4—figure supplement 3). Moreover, we have constructed etsrp:Cre based on the promoter of -2.3estrp:GFP, which has been demonstrated to closely recapitulate the endogenous etsrp expression in lateral plate mesoderm and vasculatures by Veldman and Lin [Circ Res 110, 220-9 (2012)]. We have shown that etsrp:Cre can trace over 80% of kdrl:GFP-expressing intersegmental vessels (Figure 6—figure supplement 1). Furthermore, the new fish model npas4lPt(+36-npas4l-p2a-Gal4-VP16)bns423 has been characterised in another manuscript by K.M. and D.Y.R.S (submitted).

4) The etv2 knockdown requires further validation. Elongated cells resembling endothelial cells marked with the etv2 lineage mapping present in etv2 morphants suggests that endothelial cells still form. Related to this, using only a single etv2 and single standard control morpholino is surprising given the known problems with this approach. Please validate the use of the etv2 morpholino knockdowns according to Stainier et al., 2017. Moreover, endothelial cell identity of above cells should be tested.

In case these results are confirmed, it would suggest that the mechanism of ectopic β-cells emergence in etv2 morphants and npas4l mutants does not depend on the absence of endothelial cells per se. It may also suggest that the mechanism of ectopic β cells formation in etv2 morphants vs npas4l mutant is different. This should be discussed as the role of vasculogenesis and vascularization for β cell regeneration was the first motivation for this study, as introduced in the result section.

We agree with the reviewers that the formation of ectopic β-cells does not require the complete absence of endothelial cells and it is common that the phenotypes in morphants does not have complete penetrance. Sumanas and Lin have also shown the dose-dependent and incomplete effects of etsrp (etv2) morpholinos [PLoS Biol 4(1):e10 (2006)]. Therefore, we have changed the title of the manuscript to “Insulin-producing β-cells regenerate ectopically from a mesodermal origin under the perturbation of hemato-endothelial specification”.

In the revision, we have added another widely used etsrp morpholino (MO1) in addition to the MO2 included in the original submission. Both MO1 and MO2 could induce the formation of ectopic β-cells independently compared to the random control morpholino (instead of the standard control). Importantly, co-injecting lower doses of MO1 and MO2 could exert a synergistic effect on ectopic β-cell formation (Figure 4—figure supplement 4), suggesting that both morpholinos are specifically targeting etsrp (in line with the suggestions in Stainier et al., 2017). Interestingly, similar to the higher effectiveness of MO2 on inducing the complete loss of circulation reported by Sumanas and Lin [PLoS Biol 4(1):e10 (2006)], MO2 also tended to be more potent than MO1 in inducing ectopic β-cell formation. We have also tried to block the ectopic β-cell formation by co-injecting etsrp mRNA with MO2. However, etsrp mRNA did not significantly stop the ectopic β-cell formation, which might be due to mRNA degradation.

It is true that we cannot absolutely rule out the possibility that the mechanisms of ectopic β-cell formation in npas4l mutants and etsrp morphants could be different. Therefore, we have added a statement in the discussion part (Lines 302-305). However, the lack of vasculatures is unlikely the major driving force for ectopic β-cell formation because we did not observe the same phenotype in zebrafish embryos treated with chemicals that inhibit angiogenesis.

5) Clarification of assessment criteria for different insulin+ cell populations and their localization to the principle islet or ectopic tissues.

The regenerated β cells seem to be broadly distributed and not always seem to form a distinct islet, but are nevertheless assigned to the pancreatic domain (see Figure 1E). To clarify how the pancreatic and the ectopic domains have been discriminated in these experiments, please provide additional details how both populations have been identified.

Moreover, lineage labelled Ins+Cherry+ cells from the mesoderm are also found within the pancreas; for instance a substantial fraction of regenerated β-cells defined as pancreatic seems to derive from etv2+ mesodermal cells in etv2 morphants (Figure 4B). This raises the question of the identity and of these "pancreatic" cells and where they arise from, and requires additional characterization and explanation.

The β-like cells appear to come from the lateral plate mesoderm (supplement figure4), please specify at least once in the text for clarity of lineage relationship.

We thank the reviewers for expressing concerns about the criteria for defining the pancreatic and ectopic β-cells. Now we have clarified in the Methods part how we classified these two domains during the various analyses (Lines 443-447). It was completely based on the location of the β-cells—i.e., whether they were in the pancreas or not, while the region of the pancreas was defined by the signals of ptf1a:EGFP, DAPI or pan-cadherin staining.

We agree that the ectopic β-cells were not very clearly illustrated in the previous Figure 1E when we tried to keep the dashed rectangle intact. In this revised version, we have modified the way of illustration and delineated the whole pancreatic domains to make things clearer (Figure 1D and E). Newly regenerated β-cells usually appear more dispersed than those under basal development without ablation and it often takes a few more days of recovery before they cluster together.

Without the inducible ERT2 temporal control, some low- or miss-expression of etsrp:Cre (etv2:iCre) in early progenitors from mesendoderm might be sufficient enough to trigger the Cre-recombination ubi:Switch and genetically traced a few descendants of mesendoderm, including the endodermal pancreatic islet. Therefore, in this revised manuscript, we have included three different lineage-tracing models to complement each other to minimise the influence of any confounding effect.

In addition, we have added the point of the lateral plate mesoderm being a potential origin of the ectopic β-cells in the expanded discussion (Lines 277-305).

6) The interpretation of glucose measurements needs to be revisited, given the relevance of measuring free tissue glucose as proxy for the effect of insulin released by ectopic cells in npas4l mutants is questionable. npas4l mutants have no capillaries nor blood to vehicle the hormone to target tissues in mutants. It is likely that the glucose values obtained for the mutants are not the result of insulin action, and hence may not provide a reliable estimation of ectopic cell function. This would be more convincing, for example, if a transient increase is observed after ablation before normalization. If this is the case, the authors should experimentally test whether increasing glucose levels in non-ablated npas4l mutants would trigger a similar formation of ectopic β-cells.

Therefore, please provide further support or adjust and discuss the conclusions.

We thank the reviewers for pointing out the potential problems of the glucose experiment. Hence, we tried to collect the zebrafish samples earlier at different time points to see if there was a transient increase in free glucose levels in the npas4l mutants after β-cell ablation. However, we only sometimes saw a mild but insignificant increase in the glucose levels, suggesting that there was likely no severely elevated glucose level for the ectopic β-cells to tackle with and increased glucose level was not a crucial factor for the formation of the ectopic β-cells.

Instead, during the revision of the manuscript, we have examined the calcium signalling in the ectopic β-cells by live imaging in vivo to see how mature they are as secretory cells. Similar to some of the pancreatic β-cells, a portion of ectopic β-cells also displayed certain basal calcium activity at the baseline and responded to glucose treatments with calcium oscillations in the npas4l mutants (Figure 3), suggesting that some of the ectopic β-cells are glucose-responsive secretory cells.

7) The significance of the conversion of other endocrine cell types, specifically somatostatin+ cells, from mesoderm in npas4l mutants needs to be clarified. The increase in the number of ectopic sst2 cells after δ-cell ablation presented in Figure 1 Supplement 2F is very mild and in addition ectopic sst2 cell seem to frequently form in controls outside the pancreas. Please clarify and discuss general implications for the competence of the mesenchyme for the appearance of endocrine cells outside the pancreas, as in its current form, this result is more confusing than supporting the other results.

Has the normal distribution of data points in Figure 1 supplement 2F been tested for control and npas4l datasets?

We agree with the reviewers that the data of somatostatin+ cells may not be very helpful. Therefore, we have removed this part of the data from the revised manuscript.

8) A number of clarifications and additions to the text and figures are required:

– The title requires attention, it seems grammatically incorrect, e.g. lacking a verb.

We have changed the title to “Insulin-producing β-cells regenerate ectopically from a mesodermal origin under the perturbation of hemato-endothelial specification”.

– Figure 2 indicates that the extra b-cell phenotype is not fully penetrant, please include this in the manuscript.

The quantifications of Figure 2 are not optimal for describing the penetrance of the phenotype because an individual with all the Ins+ cells carrying a marker (e.g. Ins+Isl1+ cells) would result in zero Ins+ only cells in the same individual but it does not mean the individual carries no ectopic Ins+ cells. Figure 1 is more suitable for showing the penetrance with the total ectopic β-cell number. We did not get 100% penetrance in every single experiment (although the penetrance of ectopic β-cells is very high), which could be dependent on different β-cell reporters/immunostaining, slight variations of developmental stages of individuals, or mild variations of the β-cell ablation.

– Clarify and specify in all figure legends what the red staining represents, e.g. DsRed fluorescence driven by the insulin promoter or immunofluorescence staining for insulin?

We thank the reviewers’ suggestion and we have clarified the red fluorescence in the figure legends in the revised manuscript.

– Figure 1, supplement 1: Please delineate the pancreas in panels C and D to reveal the ectopic sites of emerging β-like cells.

We have added white lines to delineate the pancreas in Figure 1—figure supplement 1.

– p-Cadherin: please spell out PanCadherin in the figures to avoid confusion with a phosphorylated Cadherin form.

We have changed the term p-Cadherin to pan-Cadherin.

– Please include short explanation for iCre in the text or methods.

The iCre used in the study is improved Cre-recombinase. We have abbreviated iCre as Cre to avoid confusion and added a short description of it in the methods (Lines 364-366).

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

The manuscript does not require any further experimental data, nevertheless, we ask you to address some important additions to the text, as suggested by the reviewers. Including the following changes will add essential information and will help to make your work more accessible to the reader:

– mention once in the manuscript the previous name of etsrp (etv2) for the sake of visibility and clarity.

Now we have mentioned the previous name of etsrp in the introduction (line 78 in the word document / line 79 after conversion to the pdf document).

– precise when 4OHT was applied in the control experiment shown in Figure 4 – Supplement figure 3 (drl:CRE-ERT in the heart).

The information has been added to the figure legend (line 946 in word/961 in pdf).

– include the characterization of specificity and efficiency of the lineage tracing etsrp:Cre in the manuscript as described in the rebuttal letter (response to point 3).

The description of the characterisation has been further expanded (line 229-231 in word/234-236 in pdf).

– summarize lines 289-294 (Discussion) to simply suggest that the possibly versatile mutant mesodermal cells could be the origin of the ectopic β cells.

Text changes have been made to make a simpler summary (line 291-293/298-300 in pdf).

https://doi.org/10.7554/eLife.65758.sa2

Article and author information

Author details

  1. Ka-Cheuk Liu

    Department of Cell and Molecular Biology, Karolinska Institutet, Stockholm, Sweden
    Contribution
    Data curation, Formal analysis, Investigation, Visualization, Methodology, Writing - original draft, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-6082-8801

  2. Alethia Villasenor

    Department of Developmental Genetics, Max Planck Institute for Heart and Lung Research, Bad Nauheim, Germany
    Contribution
    Formal analysis, Validation, Investigation, Visualization, Writing - review and editing
    Competing interests
    No competing interests declared

  3. Maria Bertuzzi

    Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden
    Contribution
    Formal analysis, Investigation, Methodology, Writing - review and editing
    Competing interests
    No competing interests declared

  4. Nicole Schmitner

    Department of Cell and Molecular Biology, Karolinska Institutet, Stockholm, Sweden
    Contribution
    Formal analysis, Investigation, Visualization, Methodology
    Competing interests
    No competing interests declared

  5. Niki Radros

    Dermatology and Venereology Division, Department of Medicine (Solna), Karolinska Institutet, Stockholm, Sweden
    Contribution
    Formal analysis, Investigation
    Competing interests
    No competing interests declared

  6. Linn Rautio

    Department of Cell and Molecular Biology, Karolinska Institutet, Stockholm, Sweden
    Contribution
    Formal analysis, Investigation
    Competing interests
    No competing interests declared

  7. Kenny Mattonet

    Department of Developmental Genetics, Max Planck Institute for Heart and Lung Research, Bad Nauheim, Germany
    Contribution
    Resources, Investigation, Methodology, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-9705-8086

  8. Ryota L Matsuoka

    1. Department of Developmental Genetics, Max Planck Institute for Heart and Lung Research, Bad Nauheim, Germany
    2. Department of Cardiovascular and Metabolic Sciences, Lerner Research Institute, Cleveland Clinic, Cleveland, United States
    Contribution
    Resources, Methodology, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-6214-2889

  9. Sven Reischauer

    1. Department of Developmental Genetics, Max Planck Institute for Heart and Lung Research, Bad Nauheim, Germany
    2. Cardio-Pulmonary Institute, Frankfurt, Germany; Medical Clinic I, (Cardiology/Angiology) and Campus Kerckhoff, Justus-Liebig-University Giessen, Giessen, Germany
    Contribution
    Investigation, Methodology, Writing - review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-6955-9481

  10. Didier YR Stainier

    Department of Developmental Genetics, Max Planck Institute for Heart and Lung Research, Bad Nauheim, Germany
    Contribution
    Supervision, Funding acquisition, Investigation, Methodology, Project administration, Writing - review and editing
    Competing interests
    Senior editor, eLife
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-0382-0026

  11. Olov Andersson

    Department of Cell and Molecular Biology, Karolinska Institutet, Stockholm, Sweden
    Contribution
    Conceptualization, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Writing - original draft, Project administration, Writing - review and editing
    For correspondence
    Olov.Andersson@ki.se
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-6715-781X

Funding

Ragnar Söderbergs stiftelse

  • Olov Andersson

Max Planck Society

  • Didier YR Stainier

Vetenskapsrådet

  • Olov Andersson

Novo Nordisk Fonden

  • Olov Andersson

Karolinska Institutetvia SRP Diabetes & StratRegen

  • Olov Andersson

H2020 European Research Council (772365)

  • Olov Andersson

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We are very thankful to Abdeljabbar El Manira for making equipment available for calcium imaging; Dirk Meyer for plasmids carrying probe templates of foxa2, gsc, mixl1, and ascl1b; Christian Mosimann for sharing the Tg(−6.35drl:Cre-ERT2,cryaa:Venus)cz3333 line and information; and Naoki Tsuji for the pDestTol2gY plasmid.

Ethics

Animal experimentation: This study was performed in accordance with the recommendations of Karolinska Institutet. All of the animals were handled according to approved institutional animal care and in accordance with rules in L150 from the Swedish Board of Agriculture. The protocol was approved by Stockholms Djurförsöksetiska nämnd (Permit Number: N145/15 and 6848-2020).

Senior Editor

  1. Marianne E Bronner, California Institute of Technology, United States

Reviewing Editor

  1. Elke Ober, University of Copenhagen, Denmark

Reviewers

  1. Elke Ober, University of Copenhagen, Denmark
  2. Valerie Gouon-Evans

Version history

  1. Received: December 15, 2020
  2. Preprint posted: February 2, 2021 (view preprint)
  3. Accepted: August 6, 2021
  4. Version of Record published: August 17, 2021 (version 1)

Copyright

© 2021, Liu et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,483
    Page views
  • 144
    Downloads
  • 3
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Ka-Cheuk Liu
  2. Alethia Villasenor
  3. Maria Bertuzzi
  4. Nicole Schmitner
  5. Niki Radros
  6. Linn Rautio
  7. Kenny Mattonet
  8. Ryota L Matsuoka
  9. Sven Reischauer
  10. Didier YR Stainier
  11. Olov Andersson
(2021)
Insulin-producing β-cells regenerate ectopically from a mesodermal origin under the perturbation of hemato-endothelial specification
eLife 10:e65758.
https://doi.org/10.7554/eLife.65758

Categories and tags

Research organism

Further reading

    1. Computational and Systems Biology
    2. Developmental Biology

    A comprehensive model of Drosophila epithelium reveals the role of embryo geometry and cell topology in mechanical responses

    Mohamad Ibrahim Cheikh, Joel Tchoufag ... Konstantin Doubrovinski
    Research Article

    In order to understand morphogenesis, it is necessary to know the material properties or forces shaping the living tissue. In spite of this need, very few in vivo measurements are currently available. Here, using the early Drosophila embryo as a model, we describe a novel cantilever-based technique which allows for the simultaneous quantification of applied force and tissue displacement in a living embryo. By analyzing data from a series of experiments in which embryonic epithelium is subjected to developmentally relevant perturbations, we conclude that the response to applied force is adiabatic and is dominated by elastic forces and geometric constraints, or system size effects. Crucially, computational modeling of the experimental data indicated that the apical surface of the epithelium must be softer than the basal surface, a result which we confirmed experimentally. Further, we used the combination of experimental data and comprehensive computational model to estimate the elastic modulus of the apical surface and set a lower bound on the elastic modulus of the basal surface. More generally, our investigations revealed important general features that we believe should be more widely addressed when quantitatively modeling tissue mechanics in any system. Specifically, different compartments of the same cell can have very different mechanical properties; when they do, they can contribute differently to different mechanical stimuli and cannot be merely averaged together. Additionally, tissue geometry can play a substantial role in mechanical response, and cannot be neglected.

    1. Developmental Biology
    2. Evolutionary Biology

    Hierarchical morphogenesis of swallowtail butterfly wing scale nanostructures

    Kwi Shan Seah, Vinodkumar Saranathan
    Research Article

    The study of color patterns in the animal integument is a fundamental question in biology, with many lepidopteran species being exemplary models in this endeavor due to their relative simplicity and elegance. While significant advances have been made in unraveling the cellular and molecular basis of lepidopteran pigmentary coloration, the morphogenesis of wing scale nanostructures involved in structural color production is not well understood. Contemporary research on this topic largely focuses on a few nymphalid model taxa (e.g., Bicyclus, Heliconius), despite an overwhelming diversity in the hierarchical nanostructural organization of lepidopteran wing scales. Here, we present a time-resolved, comparative developmental study of hierarchical scale nanostructures in Parides eurimedes and five other papilionid species. Our results uphold the putative conserved role of F-actin bundles in acting as spacers between developing ridges, as previously documented in several nymphalid species. Interestingly, while ridges are developing in P. eurimedes, plasma membrane manifests irregular mesh-like crossribs characteristic of Papilionidae, which delineate the accretion of cuticle into rows of planar disks in between ridges. Once the ridges have grown, disintegrating F-actin bundles appear to reorganize into a network that supports the invagination of plasma membrane underlying the disks, subsequently forming an extruded honeycomb lattice. Our results uncover a previously undocumented role for F-actin in the morphogenesis of complex wing scale nanostructures, likely specific to Papilionidae.

    1. Developmental Biology
    2. Neuroscience

    The differentiation and integration of the hippocampal dorsoventral axis are controlled by two nuclear receptor genes

    Xiong Yang, Rong Wan ... Ke Tang
    Research Article

    The hippocampus executes crucial functions from declarative memory to adaptive behaviors associated with cognition and emotion. However, the mechanisms of how morphogenesis and functions along the hippocampal dorsoventral axis are differentiated and integrated are still largely unclear. Here, we show that Nr2f1 and Nr2f2 genes are distinctively expressed in the dorsal and ventral hippocampus, respectively. The loss of Nr2f2 results in ectopic CA1/CA3 domains in the ventral hippocampus. The deficiency of Nr2f1 leads to the failed specification of dorsal CA1, among which there are place cells. The deletion of both Nr2f genes causes almost agenesis of the hippocampus with abnormalities of trisynaptic circuit and adult neurogenesis. Moreover, Nr2f1/2 may cooperate to guarantee appropriate morphogenesis and function of the hippocampus by regulating the Lhx5-Lhx2 axis. Our findings revealed a novel mechanism that Nr2f1 and Nr2f2 converge to govern the differentiation and integration of distinct characteristics of the hippocampus in mice.