Live imaging of cutaneous wound healing after rotary tool injury in zebrafish

Abstract

Cutaneous wounds are common afflictions that follow a stereotypical healing process involving hemostasis, inflammation, proliferation, and remodeling phases. In the elderly and those suffering from vascular or metabolic diseases, poor healing following cutaneous injuries can lead to open chronic wounds susceptible to infection. The discovery of new therapeutic strategies to improve this defective wound healing requires a better understanding of the cellular behaviors and molecular mechanisms that drive the different phases of wound healing and how these are altered with age or disease. The zebrafish provides an ideal model for visualization and experimental manipulation of the cellular and molecular events during wound healing in the context of an intact, living vertebrate. To facilitate studies of cutaneous wound healing in the zebrafish, we have developed an inexpensive, simple, and effective method for generating reproducible cutaneous injuries in adult zebrafish using a rotary tool. We demonstrate that our injury system can be used in combination with high-resolution live imaging to monitor skin re-epithelialization, immune cell recruitment and activation, and vessel regrowth in the same animal over time. This injury system provides a valuable experimental platform to study key cellular and molecular events during wound healing in vivo with unprecedented resolution.

INTRODUCTION

Cutaneous wound healing is a complex process requiring coordination of different cell types to ensure wound closure and tissue repair. It is characterized by four overlapping phases: hemostasis in which a blood clot is formed, inflammation where immune cells are recruited to the site of injury, proliferation including skin re-epithelialization, angiogenesis, and fibrogenesis, and remodeling, in which vessels are pruned and collagen remodeled, usually resulting in scar formation (Schrementi et al., 2015). Poor healing is common in the elderly and diabetics, leading to chronic wounds susceptible to infection (Guo and Dipietro, 2010, Schrementi et al., 2015). Understanding how healing is altered in aging or disease is critical to generate better therapeutics to ameliorate defective healing.

Mammalian models have yielded important insights into the cell types and signals driving healing. Although a few impactful studies in rodents have used high resolution live imaging of healing wounds (Kienle et al., 2021, Lammermann et al., 2013, Ng et al., 2011, Paterson and Lammermann, 2022), these experiments can be quite challenging in mammals. Zebrafish provide an attractive vertebrate model for studying tissue injury and repair in vivo, with their high regenerative capacity (Johnson and Weston, 1995, Poss et al., 2002), and conservation of major stages of cutaneous wound healing (Richardson et al., 2013). Unlike mammals, zebrafish have optically clear embryos and larvae, and pigment-free adult zebrafish lines are available (White et al., 2008), facilitating high-resolution optical imaging of cellular and even subcellular features in intact animals. In addition, we and others have developed methods for long-term optical imaging of intubated living adult fish (Castranova et al., 2022, Xu et al., 2014, Xu et al., 2015). These features are complemented by a plethora of transgenic and mutant fish lines available for direct visualization and functional manipulation of the different epithelial, immune, vascular, and other cell populations contributing to wound healing.

Previous attempts to study cutaneous injury in the adult zebrafish have used either a specialized, expensive (>$40,000) dermatology laser to create wounds (Richardson et al., 2016, Richardson et al., 2013), a biopsy punch, difficult to use for making controlled-depth wounds on thin-skinned, scale-covered zebrafish (Kim et al., 2023), descaling or exfoliation, which only create superficial or partial thickness wounds (Chen et al., 2016, Peterman et al., 2023, Rasmussen et al., 2018), or fine forceps, which make small and inconsistent superficial lacerations (Noishiki et al., 2019, Yuge et al., 2022). We have developed an inexpensive, simple method for generating reproducible cutaneous injuries in adult zebrafish using a rotary tool. We demonstrate the usefulness of this method for studying wound healing by generating cutaneous injuries in transgenic zebrafish and using high-resolution live imaging to monitor skin re-epithelialization, neutrophil recruitment, macrophage activation, and blood and lymphatic vessel regrowth. This method will facilitate imaging and experimental analysis of cutaneous wound healing in zebrafish,

RESULTS

We developed an easy to use, inexpensive and reproducible injury method for adult zebrafish using a rotary tool (Figure 1a, Supplementary Table S1). The rotary tool is mounted on a workstation comprised of a drill press, rotary tool holder, and mounting stand, allowing the rotary tool to be mounted at an angle and smoothly and incrementally advanced towards the sample. A square nose 1/16” or 1/8” end mill cutting bit (Figure 1b–c) is mounted on the end of the tool, and a custom 3D printed plastic cowling (Figure 1d, Supplementary File S1) is attached to the tip of the rotary tool (Figure 1e) around the bit, with just the tip of the cutting bit protruding from the cowling. This custom printed cowling helps limit and control the depth of the injury (see methods). Fish are mounted in a custom 3D printed plastic holding platform with a rotating plastic insert (Figure 1f, Supplementary File S2) that permits precise control of injury placement with the rotary tool positioned perpendicular to the flank regardless of fish size. The holding platform is placed on top of the dissecting microscope base adjacent to the rotary tool (Figure 1g).

Figure 1. Cutaneous wounding using a rotary tool.

(a) Image of cutaneous wounding setup. (b) 1/16^th and 1/8^th inch square end mill cutting bits used for generating cutaneous wounds. (c) Magnified side and front views of 1/16^th inch cutting bit tip. (d) 3D rendering of the plastic printed cowling. (e) Schematic diagram showing attachment of the cowling to the rotary tool. (f) 3D rendering of the plastic printed holding platform with rotating insert. (g) Higher magnification image of the rotary tool with cutting bit, cowling, and holding platform on a dissecting microscope. (h-j) Images of mounting fish onto the holding platform, showing a platform with wetted filter paper (h), a fish placed on the platform on top of the paper (i), and a mounted fish immobilized on the platform with anchoring rods (j).

Adult fish being subjected to injury are anesthetized in system water containing 0.5X Tricaine for 4 minutes, then transferred to system water containing 1X Tricaine for up to a minute longer, until they stop moving. Using a plastic spoon, the fish is placed on a strip of filter paper soaked in 1X Tricaine system water on the holding platform and immobilized using anchoring rods that fit securely between pegs in the holding platform (Figure 1h–j, Movie 1; rods are also custom 3D printed, Supplementary File S3–4). The holding platform is then placed on the stereoscope and positioned to ensure proper placement of the injury site on the left flank of the fish between the pelvic and anal fins to avoid any major organs. The rotary tool is turned on at speed 3 (8500 RPM), the lowest speed that produces reproducible injury without causing excess damage, and pressed gently against the side of the fish for ten seconds, creating either 1/16” or 1/8” diameter (2.4 × 10⁶ μm² or 7.7 × 10⁶ μm² area) injuries depending on the bit used (Figure 2a–c, Movie 1). After injury, fish are moved from the holding platform into fresh system water, moving water over the fish until it awakens. 16/16 fish given 1/16” wounds recovered quickly and were swimming normally shortly after revival without obvious signs of distress (see Movie 1). 13/16 fish were still alive and appeared healthy 2 weeks after injury, while 3/16 fish died a week or more after injury from repeated prolonged Tricaine exposure during subsequent imaging. 8/8 fish given 1/8” wounds awoke after injury but two died by the next day and a third showed buoyancy issues 2 weeks after injury, suggesting this size wound may be too severe and thus only wounds made with the 1/16” drill bit were used for further cutaneous wound healing analysis. To determine the best depth of injury, different cowlings were used to create shallow or deep wounds averaging 197μm and 661μm in depth respectively (Figure 2d–f). Since shallow wounds only caused slight abrasions, deeper wounds were generated to ensure full thickness injuries for all remaining experiments.

Figure 2. Reproducible cutaneous wounds can be made with a rotary tool.

(a-b) Images of a male and female fish before and after cutaneous wounding with a 1/16^th inch drill bit (small wound) (a) or 1/8^th inch drill bit (large wound) (b), showing consistent wound formation using a rotary tool. (c) Quantification of wound area for small (n=13) and large wounds (n=8), with each dot representing a single fish. (d-e) Confocal images of fixed and DAPI stained shallow (d) and deep (e) cutaneous wounds showing XY, XZ, and YZ planes. (f) Quantification of wound depth for shallow (n=4) and deep wounds (n=6), with each dot representing a single fish. Error bars in C and F indicate mean +/− SD. Images are oriented anterior left and dorsal up. Scale bars = 5000μm (a-b), and 1000μm (d-e).

To analyze the extent of tissue damage we collected both top view stereoscope images and Masson Trichrome stained transverse section side views of the wound before and up to 20 days post injury (dpi) (Figure 3a–d). In uninjured tissue, layers of epidermis, scale, fat, and muscle are easily identifiable (Figure 3b, uninjured). After injury, blood clotting and cessation of bleeding occurs rapidly, and all tissue layers up to the muscle are severely damaged (Figure 3a–b, after injury). By 2dpi, the blood clot has resolved and a new layer of thick epidermis forms over the wound site (Figure 3a–b, 2dpi). By 4dpi granulation tissue has formed with an increase in collagen deposition by 7dpi (Figure 3d, 4–7dpi). Scales have regrown and granulation tissue starts to resolve by 10dpi, with tissue layers reforming by 15dpi, and normal tissue thickness by 20dpi (Figure 3d, 10–20dpi). Pigment cells gradually reform stripes between 4–20dpi (Figure 3c), although this process seems to be slower in males than in females (Supplementary Figure S1a–c), suggesting sex-specific differences in pigment reformation which may be due to hormones since hormonal differences have been shown to affect pigment development in zebrafish and humans (Guillot et al., 2016, Natale et al., 2016). These data show that the rotary tool injury creates a full thickness cutaneous wound that is largely repaired by 20dpi.

Figure 3. Skin and tissue repair after cutaneous injury.

(a-b) Magnified stereoscope images of the flank of the same fish (a) or representative Masson Trichrome transverse cross sections (b) of fish epidermis, scales, fat and muscle in uninjured tissue, after injury tissue, or 2 days post injury (dpi) tissue. (c-d) Magnified stereoscope images of the flank of the same fish (c) or representative Masson Trichrome transverse cross sections (d) of cutaneous injury tissue in intervals from 4dpi to 20dpi. (e) Maximum projection confocal images of XY (top view) and XZ (side view) planes of adult Tg(krt4:lyneGFP)^sq18 transgenic fish with keratinocytes in green, showing skin re-epithelialization occurs within the first 24 hours after wounding. Images are oriented anterior left and dorsal up, except for Masson Trichrome images which are oriented dorsal left and lateral up. Scale bars = 2000μm (a,c), 100μm (b,d), 1000μm (e).

Our histology data and previous studies (Richardson et al., 2016, Richardson et al., 2013) suggest re-epithelialization of cutaneous wounds occurs within a few days. To examine this directly we imaged keratinocytes in Tg(krt4:lyneGFP)^sq18 transgenic fish after injury (Figure 3e). Keratinocytes initially form a thin layer of skin covering the scales (Figure 3e, before injury), which is lost after injury induction (Figure 3e, after injury). By 3 hours post injury (3hpi), keratinocytes cover a large portion of the wound site, with the wound mostly closed by 6hpi and fully sealed within one day (Figure 3e). Time-lapse confocal imaging of an injured Tg(actb2:GFP) transgenic adult fish from 0.5–17.5hpi confirms that new skin streams rapidly into the wound site from the surrounding area after injury (Movie 2). This agrees with previous reports showing that unlike in mammals, where re-epithelialization occurs later, zebrafish re-epithelialize wounds early in the repair process (Richardson et al 2013, Richardson et al 2016). Although skin re-epithelializes quickly, the underlying scales take about 10 days to span the injury site (Figure 3d, Supplementary Figure S1c), suggesting that skin remodeling occurs over a more extended time period. Our histology images show the thick skin layer over the wound site only returns to normal thickness by 20dpi (Figure 3a,c).

Since inflammation is important for wound healing, we examined immune cell recruitment to cutaneous injuries. We performed confocal time-lapse imaging of neutrophils in injured Tg(lyz:DsRed2)^nz50 transgenic, pigment-deficient casper mutant adult fish soaked in far-red fluorescent BODIPY to visualize all cell membranes (Lumaquin et al., 2021), observing rapid influx of neutrophils to the injury site (Movie 3). To quantify neutrophil behavior, we tracked individual transgene-positive neutrophils in high magnification time-lapse images (Movie 4–5, Figure 4). Like the neutrophil swarming observed after dermal injury in mammals (Lammermann et al., 2013, Ng et al., 2011) and tailfin transection in larval zebrafish (Isles et al., 2021), after cutaneous injury in adult zebrafish neutrophils showed (i) initial migration towards the wound site, (ii) amplified recruitment of neutrophils from distant locations, (iii) clustering, and an additional fourth step of (iv) wound infiltration where groups of neutrophils move deeper into the wound (Movie 4, Figure 4a–b). Neutrophil numbers increased in an exponential fashion within the wound site over 16hpi (Figure 4c), with neutrophils migrating towards wounds moving faster than those in uninjured control fish (0.104μm/s vs. 0.031μm/s, respectively) (Movie 4–5, Figure 4d). To quantify neutrophil clustering, we measured the average distance from each cell to its nine nearest neighbors (Figure 4e–f). In uninjured fish, neutrophils average 98μm from each other, but after injury neutrophils were initially 150μm apart migrating into the wound then began to cluster in the wound at a spacing of 57μm (Figure 4f). This suggests that neutrophils in zebrafish follow similar swarming patterns observed in mammals with small (area=1000 μm²) dermal wounds (Lammermann et al., 2013, Ng et al., 2011), but because we can track these behaviors over longer timescales (17 hours vs 1hour) we can show this also occurs in the larger cutaneous wounds we create (area = 2.4 × 10⁶ μm²). A confocal time series of Tg(lyz:DsRed2)^nz50 fish before and up to 15 days after wounding shows neutrophils accumulating at the wound periphery within a few hours with peak neutrophil accumulation in the wound occurring at 1dpi (Supplementary Figure S2). Wound neutrophil numbers rapidly diminish over the next several days, returning essentially to baseline by 7dpi (Supplementary Figure S2). These data show zebrafish have an acute neutrophil response to cutaneous wounds that dissipates quickly.

Figure 4. Neutrophils are rapidly recruited to cutaneous wounds.

(a-b) Confocal images of adult Tg(lyz:DsRed2)^nz50 transgenic casper fish labeling neutrophils (magenta) (a) or 3D rendering of neutrophil spots (magenta) with dragon tails (cyan) from last 15 minutes of movement (b) demonstrating stages of recruitment to the wound. (c) Quantification of neutrophil number in the wound over 17 hours. Black dotted line represents the exponential best fit. (d) Graph showing the mean track speed of neutrophils in uninjured and injured fish. (e) Confocal images and track heatmap of neutrophil clustering in uninjured and injured fish. Bluer tracks indicate closer cells. (f) Graph depicting distance of neutrophils to nine nearest neighbors over time in uninjured and injured fish. Error bars indicate mean +/− SD. Welch’s t-test, ****p<0.0001. Images are oriented anterior left and dorsal up. Scale bars = 400μm (a-b, e).

Macrophages also play a critical role in the wound healing process, initially driving inflammation and then providing anti-inflammatory signals promoting tissue repair (Guo and Dipietro, 2010). We imaged macrophages in Tg(mpeg1:eGFP)^gl22 transgenic fish before and up to 30dpi (Figure 5, Movie 6). In uninjured skin resident macrophages known as Langerhans cells maintain highly asymmetric morphology as they surveil their surroundings. Upon injury, macrophages appear to stream passively into the wound site along with regrowing epithelium, unlike active neutrophil migration (Movie 6). Macrophages continued to be recruited for over a week, peaking at 10dpi then gradually declining in number, still not reaching uninjured levels by 30dpi (Figure 5a,c). Since macrophages have been previously shown to change from more irregular to rounder shapes upon activation (McWhorter et al., 2013, Sipka et al., 2022), we measured macrophage ellipticity in 1dpi-30dpi healing wounds and compared to uninjured controls (Figure 5b–d). Ellipticities of 0–0.4 represent asymmetric inactive macrophages, 0.4–0.6 represent more symmetric macrophages, and 0.6–1 represent the roundest and most activated macrophages (Figure 5b). At 1dpi there was a higher proportion of activated macrophages in wound sites compared to uninjured controls, and this was maintained through 7dpi when the proportion of activated macrophages started to decrease, becoming similar to uninjured controls by 10dpi (Figure 5d). This suggests macrophage activation may be highest during the first week after injury. The pro-inflammatory cytokine Tumor necrosis factor alpha (TNF-α) is upregulated during the inflammation stage of cutaneous wound healing (Krizanova et al., 2022). We used hybridization chain reaction in situs of paraffin histology sections from healing wounds to examine TNF-α expression (Supplementary Figure S3). By 4dpi, TNF-α is more strongly expressed in both the skin and granulation tissue, declining by 10dpi. This timing correlates with macrophage activation, suggesting inflammation peaks around 4dpi.

Figure 5. Macrophage recruitment and activation in cutaneous wounds.

(a) Confocal images of adult Tg(mpeg:eGFP)^gl22 transgenic fish labeling macrophages (green) before and after cutaneous injury and at intervals thereafter up to 15dpi. (b) Representative images of macrophages with varying degrees of ellipticity with the higher number showing a rounder shape indicating a more activated phenotype. (c) Graph showing the total number of macrophages, and the number of macrophages with ellipticity values between 0–0.4, 0.4–0.6, and 0.6–1 within the wound site over 30dpi (n=6 fish). Dotted green line depicts average number of macrophages in uninjured controls. (d) Graph depicting the percent of macrophages with indicated ellipticity phenotype out of total number of macrophages at indicated timepoint from 1–30dpi including uninjured control (n=6 fish). Chi-square test, *p<0.05, **p<0.01. Images are oriented anterior left and dorsal up. Scale bars = 1000μm (a).

Revascularization is critical for wound healing. We examined vessel regrowth after injury in Tg(mrc1a:egfp)^y251, Tg(kdrl:mcherry)^y206 double-transgenic, casper mutant adult fish with fluorescently labeled lymphatic and blood vessels (Figure 6). Uninjured superficial trunk lymphatic vessels have a stereotypic pattern, with one lateral lymphatic vessel along the trunk midline and intersegmental lymphatic vessels branching dorsal and ventrally along myotomal boundaries (Figure 6a–b,e,k). Skin blood vessels have a more irregular pattern, although underlying muscle blood vessels are thinner and align longitudinally across myotomes (Figure 6a,c,e,q). Immediately after injury, wounds contain only severed fragments of blood and lymphatic vessels (Figure 6f,l,r) that disappear by 2dpi (Figure 6g,m,s). Sprouting angiogenesis initiates by 2dpi, continuing at a fast rate until 4dpi when greater than 75% of the wound area is revascularized (Figure 6d,g-h,m-n,s-t). Vessel growth continues at a slower pace for several more days, with most wounds approximately 90% covered by blood and lymphatic vessels by 10dpi (Figure. 6d,i-j,o-p,u-v), although remodeling of vessels continues for several more months (data not shown). These data show that cutaneous wounds are efficiently revascularized after rotary tool injury.

Figure 6. Revascularization of cutaneous wounds.

Confocal images of uninjured (a-c) and injured (e-v) Tg(mrc1a:eGFP)^y251, Tg(kdrl:mcherry)^y206 double transgenic, casper mutant adult fish with ossified structures blue (autofluorescence), lymphatic vessels green (mrc1a:eGFP), and blood vessels magenta (kdrl:mCherry). Whole-fish overview (a) and magnified flank images of lymphatic (b) or blood vessels (c). White dotted box in A depicts injury area for E-V. (d) Quantification of percent (%) vessel regrowth for lymphatic (green) and blood (magenta) vessels at time points following cutaneous injury (n=4 fish). Error bars indicate mean +/− SEM. (e-v) Images showing one fish before cutaneous injury (e,k,q), after injury (f,l,r), and at several time points thereafter through 10 dpi (g-v). Merged images (e-j) or lymphatic (k-p) and blood (q-v) vessel only are shown. Images are oriented anterior left and dorsal up. Scale bars = 5000μm (a), and 1000μm (b-c, e-v).

DISCUSSION

We have developed a method to generate full thickness cutaneous wounds in adult zebrafish. Using a rotary tool mounted on a drill press and custom 3D printed items including a cowling to control injury depth and an adjustable holding platform for the fish, we can easily and inexpensively create reproducible cutaneous injuries. We include 3D CAD files for the cowling, holding platform, and anchoring rods, allowing other labs to easily reproduce our methods (Supplementary Files S1–4). Our method produces consistent wounds of different widths and depths based on the bit size and cowling length utilized, however a 1/16” width bit is the smallest commercially available. In addition, it is important that during injury the bit speed is fast enough, and the fish pulled taut enough, so the drill bit can easily penetrate the scales and skin without creating damage to scales and tissue beyond the indicated injury site and forming an inconsistent wound. Injuries made following these guidelines are quite consistent, as shown in Fig. 2. Although injury can be limited to superficial layers without damaging underlying muscle, this method may not allow as precise control over wound depth as descaling, which produces very superficial minor injuries with limited reproducibility, or dermatology lasers, which are quite reproducible and tunable for different depths but are expensive (>$40,000) and use infrared light (2940 nm) to create heat, potentially triggering different healing responses than injuries produced using our method.

We can monitor healing via continuous imaging of intubated adult fish (Castranova et al., 2022, Xu et al., 2014, Xu et al., 2015) or serial imaging of the same fish. As also noted using other zebrafish injury methods (Richardson et al., 2016, Richardson et al., 2013), re-epithelialization occurs rapidly, within hours after injury, with full tissue repair by 20dpi, suggesting our method appropriately triggers wound healing responses. Neutrophils are recruited soon after injury, with swarming patterns similar to those previously described in mammalian dermal wounds (Lammermann et al., 2013, Ng et al., 2011) and resected larval zebrafish tailfins (Isles et al., 2021), with the added step of wound infiltration in which neutrophils move deeper into the wound peaking at 1–2dpi. Macrophage recruitment and activation occurs next, with peak macrophage activation between 1–4dpi and maximum number of macrophages at 10dpi, possibly due to a shift from pro-inflammatory to anti-inflammatory macrophages, as further supported by the decrease in the pro-inflammatory cytokine TNF-α after 4dpi. Wound revascularization occurs rapidly over the first four days after injury, with slower vessel growth thereafter transitioning into an extended phase of vascular remodeling.

Our cutaneous injury model and the live imaging capabilities of zebrafish can provide important new insights into the cell biology of wound healing. Thousands of zebrafish transgenic lines are available for an enormous range of cell types (Choe et al., 2021, Gore et al., 2011, Hall et al., 2007, Jung et al., 2017) and signaling pathways (Okuda et al., 2021, Schiavone et al., 2014) and these lines can all be combined for real-time direct in vivo observation, making the zebrafish a powerful model to study the coordination and dynamics of the cellular and molecular processes driving cutaneous wound healing.

MATERIALS AND METHODS

Wound healing apparatus

CAD designs for custom 3D-printed plastic holding platform, cowling, and anchoring rods are provided (Supplementary Files S1–4). The holding platform consists of a holder with a 3D printed rotating insert for easy, adjustable fish mounting and positioning for injury (Supplementary File S2). Silicone grease (SG-ONE 24708) is applied before inserting the rotating insert. Printed anchoring rods secure fish to the platform (Supplementary File S3–4). The cowling attaches to the end of rotary tool with only a small part of the bit tip exposed to limit injury depth (Supplementary File 1). All 3D-printed items were designed using SketchUp 3D modeling software and printed with polylactic acid (PLA) on an Original Prusa i3 MK3S+ 3D printer (Prusa Research). These CAD designs could also be printed by commercial suppliers such as Xometry (xometry.com). A Dremel 400 XPR rotary tool was attached to a Dremel drill press on a workstation stand (Dremel 220–01 workstation). The drill bit was a 1/16-inch or 1/8-inch micro grain carbide square end mill with 4 flutes from Speed Tiger Precision Technology (X002VBA0QV). A Leica M60 stereomicroscope was used for magnified visualization of wounding.

Cutaneous wound preparation

3 to 13 month old zebrafish were anesthetized in 84mg/l Tricaine (0.5X Tricaine) in system water (Tricaine S by Syndel ver. 121718, buffered to pH7 with 1M Tris-pH9), then moved to 168mg/l Tricaine (1X Tricaine) in system water for final brief anesthetization before injury, as described in the results. Gradual anesthetization using Tricaine is better for adult fish survival during live imaging. Fish were placed on the holding platform on a strip Whatman filter paper soaked in 1X Tricaine. 3D-printed anchoring rods were inserted into their holders to immobilize the fish, and mounted fish were moved to the stereomicroscope positioned next to the rotary tool. The tool bit was lined up below the midpoint of the fish between the pelvic and anal fin, the rotary tool was turned on at speed 3 (8500 RPM) and the bit gently pushed into the fish for 10 seconds to create a small cutaneous wound. After injury, fish were returned to system water and revived.

All experiments were performed in an AAALAC accredited facility under an active research project overseen and approved by the NICHD Animal Care & Use Committee (ACUC) within the National Institutes of Health (NIH) following the NIH Guide for the Care and Use of laboratory Animals, Animal Study Proposal 21–015 for zebrafish.

Abbreviations:

CAD

computer aided design

dpi

days post injury

DATA AVAILABILITY STATEMENT

All raw image files and program parameters can be provided upon request. No large datasets were generated or analyzed during the current study. CAD files of 3D printed items can be found as supplementary files.