A Novel Scale Biopsy Method to Collect Zebrafish Malignant and Non-Malignant Lymphocytes

Summary

A novel scale biopsy method allows the collection of acute lymphoblastic leukemia cells from living zebrafish. Results show abundant lymphocytes in wild-type (WT) fish scales, and dramatically increased lymphocytes in scales from fish with ALL (acute lymphoblastic leukemia). This technique allows minimally invasive study of epidermal normal and malignant lymphocytes.

Abstract

Zebrafish (Danio rerio) are a powerful model for studying lymphopoiesis and lymphoid cancers, since zebrafish and humans share similar adaptive immune systems, including B and T cells. Several D. rerio T-cell acute lymphoblastic leukemia (T-ALL) models are described, including fish that express lymphoblast-specific human MYC (hMYC). We reported that rag2:hMYC fish also develop B-cell acute lymphoblastic leukemia (B-ALL), making them useful to study both ALL types. A limitation of zebrafish ALL models is that no non-lethal methods exist for serial sampling of ALL. To address this, we pioneered a novel ‘scale biopsy’ method to collect ALL cells from live zebrafish. However, knowledge about which lymphocytes reside in normal zebrafish scales is scarce. Thus, to define normal epidermal lymphocyte identities, we performed scale biopsies on lymphocyte-labeled transgenic lines lacking rag2:hMYC. We used conventional and confocal microscopy to analyze scales, revealing abundant epidermal lymphocytes. We also analyzed scales from fish with T- and B-ALL, which demonstrated markedly increased lymphocytes in confluent sheets of ALL cells in many scales. We next used fluorescence-activated cell sorting (FACS) to purify non-malignant and ALL lymphocytes from scales for ex vivo studies. Flow cytometric and expression analyses of ALL cells from scales or other tissues of the same fish showed that scale ALL cell gene expression resembled that of ALL cells elsewhere. We also quantified the number of normal T, B, T-ALL, and B-ALL cells that could be purified per scale. Overall, scale biopsy provides a non-lethal, minimally invasive technique to study zebrafish epidermal lymphocytes, enabling ex vivo studies, including analyses of malignant samples.

Introduction

Mammalian skin’s barrier function and roles in immunity are established^1,2,3. Mammalian lymphocytes contribute to cutaneous immunity, with cytotoxic T cells residing in the epidermis, and B, T, and natural killer (NK) cells present in the dermis^1,3,4,5. Zebrafish have these lymphocytes, with surface immunoglobulins, T cell receptors, major histocompatibility complex (MHC) receptors, cytokines and their receptors, and other immune molecules, thus sharing most mammalian immunological features^6,7,8,9. Recent studies of zebrafish skin and scales demonstrated various cutaneous lymphocyte populations, including epithelial T and B cells^10,11, and a lymphoid network facilitating T cell trafficking and antigen surveillance¹². Similarities between lymphocytes of teleost fish and mammals make D. rerio a powerful in vivo model to study lymphopoiesis, lymphocyte function, and lymphoid diseases, including acute lymphoblastic leukemia (ALL)^{8,10,13,14,15,16,17,18,19,20}.

Our prior work showed that T- and B-lineage ALL (T-ALL, B-ALL) both arise in transgenic rag2:hMYC zebrafish, making this line useful to study both ALL types^13,15,21,22. A major limitation of D. rerio ALL models, particularly in longitudinal expression studies or drug-testing studies, is the lack of methods for collecting serial ALL samples from individual animals. Previous methods, such as retro-orbital blood collection, are technically challenging due to the small size of D. rerio, which makes serial blood sampling difficult²³. Thus, most investigators euthanize zebrafish to obtain ALL cells. This has been common practice in the laboratory and that of most in the field, but it precludes the ability to perform longitudinal experiments on live animals. To overcome this, we recently described epithelial T and B lymphocytes in zebrafish scales that were collected using a non-lethal scale biopsy strategy¹⁰. Another recent study demonstrated a T cell reservoir on/near the zebrafish surface, between adjacent scales¹². D. rerio harbors hundreds of scales, which regenerate rapidly just days after removal²⁴. Together, these findings and features indicate zebrafish scales are a source of lymphocytes that do not require euthanasia, enabling longitudinal studies in living fish.

Here, we describe a scale biopsy method for collecting zebrafish epidermal lymphocytes from fish with lymphocyte-specific fluorophore markers, with or without ALL. To study T-lineage cells, we used well-established lck:GFP fish and a similar lck:mCherry line; both label T and T-ALL cells^10,13,15,20. To study B-lineage cells, we used cd79a:GFP and cd79b:GFP transgenic fish with fluorescently labeled B and B-ALL cells^10,13,17. Each of these lines allows visualization by fluorescent microscopy and flow cytometric cell-sorting purification¹³. We examined scales post-biopsy, generating high-resolution images of lymphocytes in scales, identifying epidermal lymphocyte gene expression by qRT-PCR, comparing it with that of lymphocytes from other lymphoid tissues of the same fish, and quantifying lymphocytes per scale by flow cytometry/fluorescence-activated cell sorting. In summary, the protocol describes a simple and practical method for obtaining lymphocytes from live zebrafish, including sampling individual animals longitudinally across multiple time points in an experimental sequence. This method is intended to study zebrafish lymphocyte biology, leukemia progression, drug responses, or other longitudinal experimental designs that require repeated sampling of individual zebrafish. We developed the protocol using adult transgenic zebrafish with fluorescently labeled lymphocyte populations and validated it for downstream applications, including fluorescence microscopy, flow cytometry/fluorescence-activated cell sorting, RNA expression analyses, and other ex vivo studies requiring viable lymphocytes.

Protocol

All animal procedures were approved by the Institutional Animal Care and Use Committee of the University of Oklahoma Health Sciences Center (protocols 24-028-EAH and 25-068-EACHIR). This protocol presents an approach to isolate and analyze normal and/or malignant lymphocytes from the scales of live zebrafish, with details on the purpose, execution, and principles underlying these procedures.

CAUTION: MS-222 is a hazardous chemical anesthetic and must be handled with appropriate personal protective equipment (PPE), such as laboratory gloves, eye protection, and a laboratory coat. Dispose of MS-222 solutions in compliance with institutional chemical waste guidelines. Exercise caution when handling needles and forceps to prevent sharp injuries. Consider all biological samples potentially biohazardous and dispose of waste in accordance with institutional biosafety procedures.

1. Preparation of sorting media

1. Supplement 1× RPMI 1640 media solution with 1% fetal bovine serum (FBS) and 1% penicillin/streptomycin.
   NOTE: RPMI 1640 media maintains lymphocyte viability ex vivo. RPMI 1640 can be replaced with 0.9× PBS. Fetal bovine serum (FBS) supplies growth factors to promote cellular viability. Penicillin and streptomycin limit bacterial contamination.
2. Aliquot media into 1.5 mL microcentrifuge tubes in 500 µL volumes, with each tube intended for 10–20 scales. Maintain sorting media tubes on ice.

2. Zebrafish anesthesia, fluorescence microscopy screening, and imaging

1. Anesthetize D. rerio with buffered 0.02% tricaine methanesulfonate (MS-222) prepared according to institutional animal care guidelines using water from the fish system.
2. Once sedated, gently place the fish in a cell-culture dish with sufficient system water (plus 0.02% MS-222) to maintain moisture during subsequent handling.
3. Examine and image anesthetized fish using an epifluorescent microscope to detect fluorescence from the transgene(s). This step identifies fluorescent scales for biopsy.
   NOTE: Lines lacking fluorescent transgenes can be viewed via standard light microscopy.
4. Use fluorescence to identify T cells (lck:GFP or lck:mCherry fish) or T-ALL (rag2:hMYC + lck:GFP or lck:mCherry fish). T-lineage cells are brightly fluorescent in these lines; B-ALL cells are dimly fluorescent in the lck:GFP and lck:mCherry backgrounds^13,15,21.
5. Screen B-lineage–labeled cd79a:GFP or cd79b:GFP fish (with or without rag2:hMYC) for epidermal B cells and/or B-ALL, using GFP to select scales for biopsy.
6. Practice positioning zebrafish and adjusting microscope settings before biopsy, if imaging is desired
   NOTE: Imaging and biopsy require 5–7 min of sedation via 0.02% MS-222. Monitor opercular movement continuously. If opercular movement slows markedly or ceases, immediately transfer the fish to an anesthetic-free system water and abort the biopsy. Reapply ~50–100 µL of anesthetic-containing fish water to the skin, operculum, and gills every 1–2 min during imaging and biopsy to maintain surface moisture.

3. Scale collection and preparation

1. To position the fish for biopsy, place the sedated fish on its side in a cell-culture dish.
2. To prevent the fish from sliding during biopsy, use a damp sponge with a linear or V-shaped cutout and place it in the cell-culture dish to cradle the fish.
3. Alternatively, ensure the system water level barely covers the cell-culture dish bottom (enough to keep the surface moist) to allow surface tension to stabilize the fish.
4. Periodically apply ~50–100 µL of 0.02% MS-222 prepared in system water to the operculum and gills to maintain moisture and sustain anesthesia during imaging and biopsy.
5. Perform biopsy using epifluorescence microscopy to select highly fluorescent scales enriched with lymphocytes or ALL cells.
6. Collect 10–20 adjacent scales from a single highly fluorescent region, preferably within the first or second horizontal pigmented stripe.
   NOTE: One scale is generally sufficient for imaging. A collection of 10–20 scales typically yields >1,000 lymphocytes from WT fish and is recommended for downstream analyses such as flow cytometry and qRT-PCR.
7. Dominant hand (needle): Use a 31-gauge syringe needle, gently place the needle tip under the posterior edge of a scale, and flip it forward (i.e., cephalad). Use a different sterile needle for each fish. Clean forceps between fish; sterilize if required by institutional procedures.
   NOTE: Elevating the scale with a needle is preferable to using forceps alone, as it lifts the scale to minimize surface scraping. This ensures that the delicate epidermal layer beneath the scale remains adherent to the scale surface, which is critical for maximizing cell yield.
8. Non-dominant hand (forceps): Holding the scale in the flipped-forward position with the needle, grasp the rigid outer surface of the scale with forceps. In a smooth motion, pluck the scale from its pocket.
   NOTE: Minor bleeding may occur at sites of scale extraction, which is typically self-limited.
9. If excess bleeding occurs, apply gentle pressure with forceps and rinse briefly with 0.02% MS-222 in system water.
   NOTE: Hemostasis can be achieved within seconds before proceeding to the next scale.
10. To biopsy multiple scales, collect 10–20 scales per fish, removing adjacent scales from a contiguous region in one horizontal pigmented stripe, rather than multiple body regions or both sides of the animal.
11. Pluck individual scales sequentially and retain them on the forceps tips during collection with the medial scale surface facing upward.
12. To avoid interference of accumulated scales with subsequent scale removal, flip each scale forward first and then grasp it at its posterior edge. Transfer scales to a tube containing sorting media after every 5 scales are collected, or whenever additional scales cannot be securely retained on the forceps tips.
13. After scale collection, process samples as rapidly as possible. Immediately transfer scales from the forceps into 1.5 mL tubes containing 500 µL of cold sorting media on ice to preserve cell viability. Decontaminate forceps between fish using 70% ethanol before collecting subsequent samples. Process scales from each fish in separate tubes throughout collection, dissociation, and for downstream analyses.
14. After scale collection, immediately transfer the fish to a recovery tank with fresh system water without MS-222.
15. Monitor opercular movement and swimming behavior until normal respiration and posture are restored. Observe the biopsy site for persistent bleeding or signs of distress before returning the fish to the recirculating system.
16. Return the fish to its home tank after normal equilibrium and swimming resume.

4. Cell dissociation and filtering

1. Check scales at the bottom of the tube. Scales are easily visible to the naked eye before dissociation. Dissociate scales using a sterile plastic microtube pestle homogenizer by applying gentle, controlled manual pressure for 30–60 s.
2. Use slow rotational and up-and-down motion to release epidermal cells while avoiding excessive shearing of scales.
3. Continue dissociation until the sorting media becomes visibly turbid.
   NOTE: The success of the procedure is indicated by the media becoming cloudy, indicating a high concentration of liberated cells in suspension.
4. Pass homogenized samples through a 35 µm nylon mesh filter to remove debris and generate single-cell suspensions for flow cytometry/fluorescence-activated cell sorting.
5. If visible cell aggregates are present after the initial filtration, pass the sample through a fresh 35 µm mesh filter once more immediately before flow cytometric analysis. Repeat filtration no more than 2x total to minimize cell loss.

5. Flow cytometric analysis

1. Analyze samples by flow cytometry.
   NOTE: Flow cytometry can be performed before fluorescence-activated cell sorting purification or RNA extraction to assess sample composition.
2. Analyze single-cell suspensions using a flow cytometer equipped with a 488 nm laser for GFP detection and a 561 nm laser for mCherry detection.
3. Exclude debris using forward scatter (FSC) and side scatter (SSC) parameters. Identify lymphocytes using the previously described lymphocyte gate^10,13,21.
4. Exclude cell aggregates via singlet gating (i.e., SSC-area [SSC-A] versus SSC-height [SSC-H]) parameters.
5. Assess cell viability using propidium iodide (PI) or an equivalent live/dead dye. Exclude PI-positive cells from subsequent analyses and sorting procedures.
6. Establish fluorescence gates using non-transgenic zebrafish controls and single-color fluorescent transgenic controls. Define GFP-positive and mCherry-positive populations relative to negative controls.
7. Export flow cytometry standard (FCS) files and analyze data using flow cytometry analysis software.

6. Fluorescence-activated cell sorting

1. To sort cells, use a fluorescence-activated cell sorting equipped with 488 nm and/or 561 nm lasers for GFP and mCherry detection.
2. Sequentially gate lymphocytes (FSC/SSC), singlets (SSC-A versus SSC-H), and viable cells (PI-negative) prior to fluorescence-based sorting.
3. Define GFP^lo, GFP^hi, and mCherry-positive populations using fluorescence gates established from negative control samples.
4. Sort GFP^lo, GFP^hi, and mCherry-positive populations into collection tubes containing chilled sorting media.
   NOTE: In this work with the lck:GFP line, high GFP (GFP^hi) versus low GFP (GFP^lo) cells are distinct lymphocyte lineages, as previously described^10,13,15,21.
5. Maintain sorted samples on ice and proceed immediately to downstream applications, such as RNA isolation.

Results

Following the steps above, we anesthetized and screened adult zebrafish for fluorescent lymphocytes or ALL via microscopy. Wild-type (WT) lck:GFP fish exhibited fluorescence primarily in the thymic region, while cd79b:GFP fish displayed fluorescence in the head/gill region, consistent with prior observations¹⁰ (Figure 1A, left). In contrast, double-transgenic rag2:hMYC; lck:GFP/lck:mCherry fish with T-ALL or rag2:hMYC + cd79b:GFP fish with B-ALL showed bright, disseminated fluorescence throughout the body (Figure 1A, right), which affords many highly-fluorescent scales for biopsy.

After removal, scales remained intact, with epidermal tissue adherent to their surfaces. Conventional and confocal imaging of mounted scales revealed scattered GFP⁺ lymphocytes in the epidermis of WT lck:GFP and cd79b:GFP scales, while fish with leukemia frequently harbored dense clusters or confluent sheets of GFP⁺/mCherry⁺ cells on scale surfaces (Figure 1B). To validate lymphoid morphologies of the cells recovered from scales, Wright-Giemsa staining was performed on cells obtained from different anatomic sites of a WT lck:GFP fish. Scale-derived lymphocytes exhibited morphologies similar to lymphocytes from thymus, marrow, and blood, including high nuclear-to-cytoplasmic ratios and densely staining nuclei (Figure 1C, Top). Wright-Giemsa staining of a scale from a fish with T-ALL further demonstrated many lymphoblast-appearing cells near the scale’s surface (Figure 1C, bottom). These morphologic findings corroborate the fluorescence microscopy data and confirm that dense accumulations of fluorescent cells on T-ALL scales are composed of lymphoblasts. These microscopy findings provide a qualitative readout, with morphological confirmation that scales contain epidermal lymphocytes (or leukemic lymphoblasts) available for scale dissociation. WT double-transgenic lck:mCherry + cd79a:GFP fish allowed the simultaneous visualization of mCherry⁺ T-lineage and GFP⁺ B-lineage cells (Figure 1D, 1E).

For flow cytometry analyses and fluorescence-activated cell sorting, scales were dissociated and filtered to generate single-cell suspensions. The lymphoid gate was identified^10,25 and filtered for lymphocytes/blasts via forward scatter and side scatter. Propidium iodide (PI) staining demonstrated that the majority of cells within the lymphoid gate remained viable following scale dissociation and processing, with a mean viability of ~92% (n = 16 fish). Fluorescence gates were established using fluorophore-negative non-transgenic controls and GFP⁺ and mCherry⁺ transgenic lines (Figure 2A). In WT lck:GFP (i.e., fish lacking rag2:hMYC) scales, most fluorescent cells were GFP^lo (Figure 2B, upper panel), whereas scales of rag2:hMYC + lck:GFP fish with T-ALL contained prominent GFP^hi populations, consistent with malignant lymphoblasts (Figure 2B, middle panel). B-lineage labeled cd79b:GFP scales yielded both GFP^lo and GFP^hi populations (Figure 2B, lower panel), enabling isolation of distinct B-lineage subsets.

To validate lineage identities of fluorescent cells from scales, GFP^lo cells from WT lck:GFP scales, GFP^hi cells from rag2:hMYC + lck:GFP T-ALL scales, and GFP⁺ cells from cd79b:GFP scales were fluorescence-activated cell sorting purified, RNA extracted, and then analyzed by qRT-PCR. Cells from cd79b:GFP scales expressed B-lineage transcripts (pax5, cd79b, ighm, and ighz; Figure 2C, top panel), while lck:GFP scale cells expressed T-lineage genes (cd4, cd8, and lck; Figure 2C, middle panel), consistent with our prior profiles of zebrafish epidermal lymphocytes obtained by scale biopsy, where WT scale T cells expressed more cd4 than cd8¹⁰. T-ALL scale cells expressed T-lineage transcripts with higher rag2 expression than non-leukemic scale cells (Figure 2C, bottom panel), consistent with their lymphoblast identity. T-ALL cells also expressed near-equal levels of cd4 and cd8, consistent with their double-positive immature T-lymphoblast status.

To determine whether scale ALL cells are representative of cancer cells elsewhere in the animal, we compared flow cytometric cell-sorting-purified lymphocytes from scales or body tissue from the same fish, and analyzed them by qRT-PCR. GFP⁺ cells from scales or body tissue of rag2:hMYC + lck:GFP T-ALL fish showed near-identical gene expression profiles, with higher expression of T cell genes (cd4, lck), lymphoblast-enriched genes (rag1, rag2), and the hMYC oncogene relative to B cell genes (pax5, ighm, ighz) (Figure 2D). Overall, these data and our prior work¹⁰ support the use of scale biopsies to collect non-malignant and ALL cells for downstream study. To assess the suitability of scale-derived leukemic cells for downstream molecular analyses, RNA was isolated from matched body- and scale-derived T-ALL cells after flow cytometric cell-sorting purification. After sorting 50,000 GFP^hi cells, scale-derived T-ALL samples yielded an average RNA yield of ~82 ng, compared with ~106 ng from body-derived T-ALL cells (n = 4 fish). Despite the slightly lower yield from scale-derived samples, successful amplification of lineage- and leukemia-associated transcripts by qRT-PCR demonstrated that scale-derived RNA is suitable for downstream gene-expression analyses.

Biopsies from non-leukemic fish typically yielded 50–65 fluorescent lymphocytes per scale (Figure 3), allowing collection of >1,000 lymphocytes from 20 scales, which is sufficient for flow cytometric analysis and targeted expression assays. In fish with ALL, yields were dramatically higher, with T-ALL scales containing ~20,000 GFP^hi cells/scale (WT lck:GFP fish had ~65 cells/scale; Figure 3, top). B-ALL from rag2:hMYC + cd79b:GFP scales harbored ~4,000 GFP⁺ cells/scale (WT cd79b:GFP fish had ~50 GFP⁺ cells/scale; Figure 3, bottom). The large number of cells recoverable from fish with ALL provides a practical basis for longitudinally sampling ALL cells from individual fish across multiple time points. Supporting the feasibility of repeated sampling, all 24 zebrafish in a prior longitudinal study survived serial biopsies, which is uncommon, and the biopsies could be controlled and performed on the same side repeatedly over two weeks¹⁰. Furthermore, longitudinal analyses of eight fish showed recovery of scales, scale lymphocytes, and fluorescence at the biopsy site by 10 days post-sampling. These findings suggest that a 10-day interval is adequate for repeated sampling of the same biopsy site. Collectively, these findings support the regenerative capacity of biopsied scales and establish scale biopsy as a practical approach for longitudinal studies of live zebrafish.

Suboptimal outcomes are often characterized by low cell recovery and excess debris in biopsy samples. Low recovery can occur if scales are collected from low-fluorescence body areas or if the epidermis is inadvertently stripped during scale extraction. To avoid this, we flip scales forward (cephalad) prior to forceps removal. Removing scales from highly fluorescent contiguous regions also appears to improve yield. Regarding excess debris, aggressive homogenization can increase cell shearing, reducing the number of cells available for fluorescence-activated cell sorting or flow cytometry. Gently dissociating until no epidermis remains visibly attached to scale fragments, followed by filtration through a 35 µm mesh, generally yields a suitable single-cell suspension. Poor viability is most often associated with prolonged processing or excess mechanical dissociation; therefore, scales should be transferred immediately into ice-cold sorting media and processed promptly. Cell clumping can interfere with flow cytometric analysis and sorting, but can be minimized by repeated filtration through a 35 µm mesh immediately before analysis. Persistent bleeding at biopsy sites is uncommon and can be controlled by applying pressure with forceps and rinsing with an anesthetic-containing solution. Weak fluorescence signals may result from sampling scales outside regions enriched with fluorescent lymphocytes/ALL cells; thus, fluorescent imaging of the entire fish prior to biopsy is recommended to identify optimal sampling regions.

Figure 1: Fluorescence microscopy and confocal images of T-lineage (wild-type [WT] lck:GFP, T-cell acute lymphoblastic leukemia [T-ALL] lck:GFP + rag2:hMYC, and T-ALL lck:mCherry + rag2:hMYC), B-lineage (WT cd79b:GFP and B-cell acute lymphoblastic leukemia [B-ALL] cd79b:GFP + rag2:hMYC), and dual-lineage (WT lck:mCherry;cd79a:GFP) labeled transgenic fish and scales. (A) Brightfield (top row) and fluorescence (bottom row) images of adult zebrafish from WT T-lineage labeled (lck:GFP) and WT B-lineage labeled (cd79b:GFP) lines, and example T-ALL (lck:GFP or lck:mCherry + rag2:hMYC) and B-ALL (cd79b:GFP + rag2:hMYC) fish. Yellow circle denotes thymic region. Scale bars, 2 mm. (B) Representative scales for each genotype. Top row: brightfield images; Middle row: fluorescent microscopy images of GFP⁺/mCherry⁺ cells in scale epidermis. Bottom row: confocal images of boxed regions in the middle row showing individual lymphocytes in WT scales and clusters/sheets of cells in ALL scales. Scale bars, 200 µm (brightfield and low-magnification rows) and 20 µm (high-magnification row). (C) Wright-Giemsa-stained images of cells obtained from thymus, marrow, scale, and blood of a WT lck:GFP zebrafish and Wright-Giemsa-stained scale images obtained from a T-ALL (lck:GFP + rag2:hMYC) zebrafish. Top four panels under lck:GFP: high-magnification images of stained cells from lymphoid organs of the same zebrafish. Upper panel under T-ALL: low-magnification image of a stained scale showing adherent cells. Lower panel under T-ALL: higher-magnification image of the boxed region in the upper panel demonstrating lymphoblast-appearing cells with high nuclear-to-cytoplasmic ratios and densely-staining nuclei. Scale bars: 10 µm (Top, individual lymphocyte images), 200 µm (middle; low-magnification), and 20 µm (Bottom; high-magnification). (D) Brightfield (BF), GFP, mCherry, and merged images of adult WT double-transgenic lck:mCherry;cd79a:GFP zebrafish. (E) Brightfield, GFP, mCherry, and merged images of a scale from the same fish, simultaneously demonstrating T-lineage mCherry⁺ cells and B-lineage GFP⁺ cells. Scale bar, 2 mm in (D) and 200 µm in (E). Fluorescence images were acquired using identical exposure settings across all channels, except in panel E, where they were adjusted to facilitate simultaneous visualization of both GFP and mCherry signals. Please click here to view a larger version of this figure.

Figure 2: Flow cytometry and qRT-PCR of scale lymphocytes and comparison between body and scale ALL cells. (A) Example flow cytometry workflow showing lymphoid (forward-scatter, FSC; side-scatter, SSC) and GFP⁻, GFP^lo, and GFP^hi gates of cells from a lck:GFP scale. (B) Example fluorescent images (left) and GFP histograms of scale cells (right) from WT lck:GFP (top), T-ALL lck:GFP + rag2:hMYC fish (middle), and WT cd79b:GFP (bottom) fish. The yellow circle denotes the thymic region. Scale bars, 2 mm. (C) qRT-PCR of fluorescence-activated cell sorting purified GFP⁺ scale cells for lineage- and stage-specific B-lineage (pax5, cd79b, ighm, ighz; top), T-lineage (cd4, cd8, lck; middle), and lymphoblast-enriched (igic1s1, rag2; bottom) transcripts. (D) qRT-PCR of matched body- and scale-derived lymphocytes isolated from the same T-ALL lck:GFP + rag2:hMYC zebrafish (n = 2). Expression of B cell (pax5, ighm, ighz), T cell (cd4, lck), and lymphoblast-enriched (igic1s1, rag1, rag2) genes, plus hMYC, is shown for body and scale samples. The 2^-ΔCt method (ΔCt = Ct_experimental − Ct_housekeeping) was used to calculate expression relative to the two housekeeping genes: eef1a1l1 (ef1a) and rpl13a. Results shown as means ± SD Each group represents data from two biological replicates. Data in panels C and D are descriptive illustrations demonstrating lineage-specific gene expression patterns. Body-versus-scale comparisons were not statistically tested. Please click here to view a larger version of this figure.

Figure 3: Scales from D. rerio with ALL yield markedly higher lymphocytes. Top: Fluorescent cells per scale from wild-type (WT) lck:GFP (GFP^lo) and T-cell acute lymphoblastic leukemia (T-ALL) lck:GFP;rag2:hMYC and cd79b:GFP;rag2:hMYC(GFP^hi) fish. Bottom: Fluorescent cells per scale from WT cd79b:GFP and B-cell acute lymphoblastic leukemia (B-ALL) cd79b:GFP;rag2:hMYC fish. Each point represents one scale. Scales were collected from two fish for each group (T-lineage groups: n = 10 scales total, 5 scales per fish; B-lineage groups: n = 8 scales total, 4 scales per fish). Symbols denote scales collected from different fish. Bars reflect group means with SD(standard deviation). These data represent example yields of fluorescent lymphocytes recovered from scales of non-leukemic and leukemic fish. Many scales derive from the same fish and may be viewed as technical, rather than biological, replicates. No statistical comparisons were performed because we non-randomly selected the brightest ALL scales for biopsy. Please click here to view a larger version of this figure.

Discussion

This protocol describes a non-lethal scale-biopsy method for collecting zebrafish epidermal lymphocytes, including both non-malignant lymphocytes and acute lymphoblastic leukemia (ALL) cells. Zebrafish are well-established models for lymphopoiesis and lymphoid malignancies, but most approaches to obtain lymphocytes require euthanasia and tissue dissections, preventing longitudinal studies of single animals^8,10,13,15. Scales offer an experimentally accessible epithelial compartment with abundant lymphocytes, and can be sampled repeatedly since adult zebrafish have hundreds of scales that rapidly regenerate after removal^10,24. Thus, scale biopsies are a practical strategy for serially monitoring lymphocytes (such as leukemic burden) and generating single-cell suspensions for downstream testing without euthanizing experimental animals.

Several steps are crucial to achieve reproducibly high cell recoveries. First, adequate anesthesia and the maintenance of surface moisture during imaging and biopsy (0.02% MS-222 in system water) reduce stress and limit drying. Second, collecting 10–20 adjacent scales from a single pigmented stripe improves reproducibility and minimizes regional variability. Third, flipping scales forward (cephalad) before extraction and immediately transferring them into ice-cold sorting media preserves epidermal tissue and lymphocyte viability. Although scale lymphocytes exhibited high viability and were validated for endpoint assays such as microscopy, flow cytometry/fluorescence-activated cell sorting, and qRT-PCR, the maximum duration for which scales can remain on ice and their suitability for long-term ex vivo culture were not examined and warrant future study. Finally, dissociation must be sufficiently forceful to dislodge epidermal cells, yet not so aggressive as to pulverize the scale matrix, thereby increasing debris and compromising flow cytometric sorting. Despite these precautions, the procedure likely recovers only a subset of the lymphocytes present in each scale, as scale microscopy (Figure 1B) shows that each scale contains many more than the 50–65 fluorescent cells/scale we can detect by flow cytometry/fluorescence-activated cell sorting (Figure 3). We have not tested trypsinization or other digestion methods to liberate cells from scales, and such or other protocol modifications could conceivably increase cell yields even further. Even so, this method provides a straightforward technique for collecting >1,000 lymphocytes from WT fish and tens to hundreds of thousands of malignant cells from fish with disseminated ALL.

Our method can be readily adapted to other experimental plans and transgenic backgrounds. Fluorescently labeled lines enable both microscopy-based readouts and fluorescence-activated cell sorting and purification of distinct cell types, such as GFP^lo vs. GFP^hi subsets. If longitudinal sampling is needed, scales can be collected from alternate flanks across time points to allow scale regrowth and reduce inflammation, which we previously implemented to serially sample scales of fish treated with dexamethasone¹⁰. Although the present study focused on lymphocytes, the scale epithelium contains additional immune cell populations, including macrophages and neutrophils¹¹, that may also be amenable to investigation using this approach.

Biopsy limitations should be considered in experimental design. In WT fish, the number of lymphocytes per scale is modest (<100 cells per scale), which may be insufficient for some assays unless many scales and/or multiple fish are used¹⁰. Also, biopsy perturbs the epidermis and may cause inflammation or tissue remodeling; thus, serial sampling studies should incorporate mock-treated biopsy controls and, when feasible, spatial separation of sites. In fish with ALL, scales offer a convenient, enriched source of malignant cells. However, investigators should consider that disease burden can vary across body regions and over time. We recommend fluorescence imaging of the entire fish prior to biopsy to identify representative sampling regions and minimize potential sampling bias. A limitation of this study is that comparisons between scale- and body-derived normal and malignant B cells were not performed. This protocol demonstrates efficient recovery of B-lineage cells from scales, but a more comprehensive evaluation of B-lineage cell populations across multiple anatomic sites is warranted in a dedicated future study. The number of scales that can be collected is also likely to depend on animal size and developmental stage. We developed the present protocol using adult-sized fish aged 4–12 months, which have hundreds of scales and readily tolerate the collection of 10–20 scales from a single body region. When using smaller or juvenile fish, investigators may need to reduce the number of scales collected. In our experience, repeated scale biopsies are well tolerated¹⁰, but future studies should evaluate potential procedure-associated events more systematically. Relevant endpoints could include cortisol levels, recovery behavior (e.g., abnormal swimming or altered feeding), and local inflammation (e.g., infection) at biopsy sites.

Despite these considerations, the scale-biopsy method expands experimental options in zebrafish immunology and leukemia research. This approach allows minimally invasive collection of viable lymphocytes for cytometry, fluorescence-activated cell sorting, expression analyses, functional assays, and other applications. Biopsies can be integrated into drug response studies and other longitudinal study designs¹⁰. By permitting repeated sampling from the same animal, this method can reduce animal use, improve statistical power for within-subject comparisons, and facilitate time-resolved studies of both normal epidermal immunity and malignant lymphocyte biology.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Reagents & Consumables:
Tricaine methanesulfonate (MS-222)	Sigma-Aldrich	E10521	Anesthetic agent
RPMI 1640 medium	ThermoFisher	11875093	Media for lymphocyte viability
Fetal Bovine Serum	Sigma-Aldrich	F2442	For cellular viability
Penicillin-Streptomycin	ThermoFisher	15140122	Limit bacterial contamination
70% ethanol	ThermoScientific	T038181000	To sterilize foreps
CellPro 1x PBS	VWR	104027-280	Dilute to 0.9x with sterile water
Propidium Iodide	Invitrogen	P1304MMP	Live/dead cell discrimination for flow/FACS
Cut Goods Misc-Pn; Size: 12 x 12 IN. 35um filter paper Lab Pak 03-35/16	Sefar		For cell sorting Reference: Burroughs-Garcia, J., Hasan, A., Park, G., Borga, C., Frazer, J. K. Isolating malignant and non-malignant b cells from lck:Egfp zebrafish. J Vis Exp. 10.3791/59191 (144), (2019).
Altrnative for filter paper: EASYstrainer Cell Sieves, Greiner Bio-One	Greiner	89508-342	For cell sorting
Eppendorf® Flex-Tubes® Microtubes, 1.5 ml	Eppendorf®	20901-551	microcentrifuge tube for cell sorting
UltiCare VetRx U-100 Insulin Syringes 3/10cc 31G x 5/16" Half Unit Marking	UltiCare	9436	For scale biopsies
100 x 25 mm Deep Petri Dish	USA Scientific	8609-0625	For scale biopsies
Bel-Art® Disposable Pestles	Bel-Art®	BAF199230001	For scale dissociation
E-Z Pik 6-Piece Tweezer Set	Aven	18480EZ	For scale biopsies
Corning® microscope slides, frosted one side, one end	Sigma-Aldrich	CLS294875X25	For imaging
Cover glasses	Sigma-Aldrich	C8181	For imaging
SlowFade™ Glass Soft-set Antifade Mountant, with DAPI	ThermoFisher	S36920	For mounting scales on slide
RNeasy Mini Kit	Qiagen	74104	For RNA extraction
Equipment:
Nikon AZ100 microscope and DS-Qi1MC camera	Nikon		Fluorescence microscope & camera
CFX96 Touch-PCR System	Bio-Rad	3600037
CytoFLEX	Beckman-Coulter		Flow cytometry analyses
Leica SP8	Leica		Confocal microscope
Software:
Kaluza Analysis software version 2.1	Beckman-Coulter		Flow cytometry analyses Reference: https://www.beckman.com/flow-cytometry/software/kaluza/downloads
LAS-X software version 3.7.4.23463	Leica		Confocal microscope Reference: https://www.leica-microsystems.com/products/microscope-software/p/leica-las-x-ls/downloads/