- Title
-
Imaging cytoplasmic lipid droplets in vivo with fluorescent perilipin 2 and perilipin 3 knock-in zebrafish
- Authors
- Wilson, M.H., Ekker, S.C., Farber, S.A.
- Source
- Full text @ Elife
(A,B) Representative images of whole-mount in situ hybridization (ISH) with probes against zebrafish plin2 (ENSDARG00000042332) and zgc: 77,486 (plin3/4/5) (ENSDARG00000013711) at 6 days post fertilization (dpf) either unfed or following feeding with a high-fat meal for 90 min. ISH was performed three times for each gene with n = 10 larvae per probe per experiment; scale = 500 µm (A), scale = 100 µm (B). Plin2 is expressed in the intestine only following a high-fat meal whereas plin3 is expressed in the intestine in unfed fish and has stronger expression following a high-fat meal. (C) Overview of the location and strategy used for TALEN-mediated genome editing. EGFP was fused in-frame at the N-terminus of plin2. TALEN targets in plin2 are located in exon 1 of the plin2-203 ENSDART00000175378.2 transcript and flank a FokI restriction site, loss of which was used to confirm cutting activity. A donor plasmid with the coding sequence for EGFP and plin2 homology arms was co-injected with TALEN mRNA into one-cell stage embryos to be used as a template for homology directed repair. mTag-RFP-t was fused in-frame at the C-terminus of plin3. TALEN targets were located in exon 8 of the plin3 ENSDART00000100473.5 transcript and flank the termination codon and an HphI restriction site, loss of which was used to confirm cutting activity. A donor plasmid with the coding sequence for mTagRFP-t and plin3 homology arms was co-injected with TALEN mRNA into one-cell stage embryos to be used as a template for homology directed repair. (D) Following identification of fluorescent embryos in the F1 generation, RT-PCR and sequencing of genomic DNA using the primers noted on the knock-in loci depicted in (C) were used to confirm successful in-frame integration of the fluorescent tags. The size of the amplicons expected for correct integration were as follows: F1–R1 1033 bp, F2–R2 1340 bp, F3–R3 440 bp for wild-type (WT) and 1224 bp for Fus(EGFP-plin2) fusion, F4–R4 1218 bp, F5–R5 1274 bp, F6–R6 401 bp for WT and 1187 for Fus(plin3-RFP). Arrows indicate the larger amplicon in heterozygous fish carrying the fusion alleles. (E) Imaging in live larvae (6 dpf) reveals expression of EGFP-Plin2 only in the intestine of larvae fed a high-fat meal (7 hr post-start of 2 hr meal) and Plin3-RFP is expressed in the intestine of both unfed and fed larvae (4.5 hr post-start of 2 hr meal, larvae are heterozygous for the fusion proteins; the lumen of the intestine has strong autofluorescence in WT and transgenic fish; see Figure 1—figure supplement 2 for images of whole fish). Scale = 500 µm. (F) Examples of larvae expressing both EGFP-Plin2 and Plin3-RFP (7 hr post start of meal). Scale = 500 µm. For (E and F), images are representative of at least 15 fish from three independent clutches. (G) EGFP-Plin2 (green) and Plin3-RFP (magenta) label the lipid droplet surface in the intestine of 6 dpf larvae fed with a high-fat meal containing either BODIPY 558/568-C12 (magenta) or BODIPY FL-C12 (green) to label the stored lipids. Note the 558/568-C12 is not fully incorporated into stored lipid and is also found diffuse in the cytoplasm. Scale = 10 µm. (H) EGFP-Plin2 and Plin3-RFP can decorate the same lipid droplets in the intestine. Arrows denote examples of dual-labeled droplets, scale = 10 µm.
|
(A) Representative images of Plin3-RFP and EGFP-Plin2 fluorescence in the intestine of 6 days post fertilization (dpf) larvae at noted time-points following the onset of a 90 min high-fat meal. Scale = 500 µm. (B) Fluorescence intensity was quantitated every hour for the first 8 hr following onset of the meal and subsequently every 2 hr for a total of 30 hr. Mean fluorescence intensity of wild-type siblings at each time-point was used to correct for gut autofluorescence. Black circles and errors bars indicate mean ± SD, individual data points are shown in gray, n = 14–23 total larvae per time-point from three independent clutches. Significant changes in fluorescence between unfed (0 hr) and subsequent time-points were calculated with Kruskal-Wallis with Dunn’s multiple comparisons tests. For Plin3-RFP, fluorescence is significantly different from 0 hr at 6 and 8–28 hr, p < 0.05. For EGFP-Plin2, fluorescence is significantly different from 0 hr at 5–18, 22, and 24 hr, p < 0.05. Refer to Figure 2—source data 1 file for p-values at each time-point. (C) Example brightfield images and corresponding EGFP-Plin2 fluorescence images from larvae imaged 10 hr after the onset of the meal. The degree of intestinal opacity reflects the amount of lipid consumed. Scale = 500 µm. (D) EGFP-Plin2 fluorescence intensity of individual larvae 10 hr after the onset of the meal was plotted as a function of the mean gray value of the intestine in the corresponding brightfield image. The amount of lipid consumed (intestine opacity) predicts much of the EGFP-Plin2 fluorescence (simple linear regression, y = –96314x + 130885876, R2 = 0.6828, p < 0.0001).
|
(A) Lateral views of the anterior intestine in unfed larvae and in larvae at different time-points following the start of feeding with a high-fat meal for 1 hr. Fish were heterozygous for both Fus(plin3-RFP) and Fus(EGFP-plin2). Images are representative of three independent experiments (15–25 fish per experiment); data presented are from one experiment. Scale = 50 µm. (B) Higher magnification micrographs of lipid droplets highlight the transition from Plin3-RFP to EGFP-Plin2 on the surface of lipid droplets over time after a high-fat meal. Scale = 10 µm. (C) Manders’ colocalization coefficients for a subset of images was quantified following Costes method for automatic thresholding. Mean ± SD, n = 4 fish per time-point.
|
(A) Representative confocal micrographs of Fus(plin3-RFP)/+, Fus(EGFP-plin2)/+, or Tg(fabp2:EGFP-PLIN2)/+ fish fed a high-fat meal containing either green or red BODIPY C12 fatty acid analog. Fish were fed for a total of 90 min and chased at room temperature for 30 hr after the onset of the meal. Scale = 20 µm. See also Figure 4—video 1, Figure 4—video 2, Figure 4—video 3. (B) Perilipin fluorescence associated with the lipid droplets and in the cytoplasm was assessed by first segmenting the lipid droplets based on the BODIPY C12 signal, the segmented regions were expanded by three pixels in all dimensions (360 nm) to create a mask which was then applied to corresponding Plin fluorescent image. Mean fluorescence intensity was assessed both in the masked regions (LDs) and outside (cytoplasm). (C–E) Mean fluorescence intensity of Plin3-RFP (C), EGFP-Plin2 (D), and over-expressed human EGFP-PLIN2 (E) in the cytoplasm and associated with lipid droplets over time following a high-fat meal (mean ± SD). Data represent two experiments per genotype, each experiment contained larvae from two clutches (n = 9–12 (C), 4–18 (D), and 3–15 (E) fish per time-point). Due to the complexity and length of the time-course, the 0–8.5 and 16–30 hr time-points represent data from different clutches. The linear correlation coefficients (R) for the 0.75–8.5 hr time-points are –0.3266 (y = –15.20x + 364), 0.7732 (y = 48.91x + (–34.40)), and 0.7277 (y = 52.33x + 36.64) for Plin3, Plin2, and PLIN2, respectively. For the 16–30 hr time-points, the R coefficients are 0.0349 (y = 0.3730x + 180.6), –0.6321 (y = –31.84x + 1113), and –0.2865 (y = –13.33x + 693.8) for Plin3, Plin2, and PLIN2, respectively. (F) Data from C, D, and E were normalized based on the overall mean fluorescence from all individual data points for each genotype and plots were overlaid to better show the relationship between genotypes. A LOWESS local regression line was applied to the normalized data-sets for visualization purposes only.
|
(A) Lateral view of the liver in a 7 days post fertilization (dpf) zebrafish larvae heterozygous for both Fus(plin3-RFP) and Fus(EGFP-plin2). Scale = 50 µm. (B, C) Brightfield and fluorescence liver micrographs from 15 dpf larval zebrafish fed a diet of Gemma +4% cholesterol for 10 days. Lipid droplets in hepatocytes can be labeled with EGFP-Plin2 (B) and with Plin3-RFP (C). Scale = 20 µm. (D) Liver micrographs from 6, 9, 15 (Fus(EGFP-plin2)) or 18 dpf (Fus(plin3-RFP)) zebrafish. LipidTox Green or Red (LTOX) labels the hepatic ducts and the lipid droplets, which are also visible in the brightfield image (BF). Where noted, fish were fed a Gemma +4% cholesterol diet starting at 5 dpf. Standard length of the imaged fish is noted on the upper left corner of each set of image, see source data for additional standard length data. Scale = 20 µm. Images are representative of at least 15 fish from three or more clutches.
|
(A) Cartoon of 15 days of post fertilization (dpf) larval zebrafish showing the general location (green box) of images in panels B–D. (B, C) Examples of Plin-positive adipocyte lipid droplets. Fish heterozygous for the noted transgene were fed Gemma for 10 days and then stained with either LipidTOX Red (B) or Green (C) dyes for a minimum of 2 hr prior to imaging. Scale = 10 µm; standard length of all fish imaged was 5.14 ± 0.26 mm (mean ± SD, n = 47 fish). (D) EGFP-Plin2 and Plin3-RFP co-label the surface of adipocyte lipid droplets. Fish are heterozygous for each transgene and were fed Gemma + 4% cholesterol for 10 days prior to imaging. Scale = 10 µm; standard length of similarly fed double-heterozygous fish at 15 dpf, 5.19 ± 0.42 mm (mean ± SD, n = 14 fish). For B–D, images are representative of at least 10 fish per genotype from three independent clutches. (E) Darkfield whole-mount image of a 21 dpf zebrafish, yellow box indicates region of images in panels F–H; scale = 1 mm. (F) Images of pancreatic visceral adipose tissue/abdominal visceral adipose tissue (PVAT/AVAT) adipose depots in wild type and Fus(EGFP-plin2)/+; Fus(plin3-RFP)/+ fish at 21 dpf. Adipocyte lipid droplets are co-labeled with Plin2 and Plin3. Note the substantial autofluorescence in the EGFP channel in wild-type fish. (G, H) Examples of PVAT/AVAT in either Fus(plin3-RFP)/+ fish stained with LipidTOX Green (G) or Fus(EGFP-plin2)/+ stained with LipidTOX Red (H) to more clearly show the relationship of the Plin fluorescence relative to the lipid content of the droplets. For (F–H), fish were fed standard Gemma diet for 15 days and fasted for 24 hr prior to imaging to decrease the fluorescence in the intestine resulting from both EGFP-Plin2 expression and/or LipidTOX-labeled lipid droplets in enterocytes. Scale = 500 µm; standard length of the fish shown is noted in upper left corner of the darkfield images. Images are representative of 18–27 fish from two independent clutches.
|
(A) Cartoon of 15 days of post fertilization (dpf) larval zebrafish showing the general location of lateral line neuromast images shown in panel B. The standard length of larvae imaged at this stage was 5.14 ± 0.26 mm (mean ± SD, n = 47 fish, see Figure 6—source data 1). (B) Examples of lipid droplets around neuromasts in fish heterozygous for Fus(EGFP-plin2) at 15 dpf. Fish were fed Gemma for 10 days and then stained with LipidTOX Red for a minimum of 2 hr prior to imaging. Insets show enlarged images of the lipid droplets in the boxed regions. Scale = 10 µm, standard length of the fish shown is noted in the brightfield image panel. Images are representative of at least 10 fish from three independent clutches.
|
All fish are heterozygous for the noted transgene. Heat shock transgenic lines were incubated at 37°C for 45 min prior to feeding. For whole-mount images, larvae were fed for 2 hr with a high-fat meal and imaged 3–4.5 hr (PLIN3 lines) or 5–8 hr (PLIN2 lines) following the start of the feed. Where appropriate, images of whole fish are included as insets. Scale = 500 µm for main images and insets. In the right column, confocal micrographs are included to show the fluorescent perilipin proteins labeling BODIPY-C12-positive lipid droplets in the various transgenic lines following a high-fat meal. Unless noted, images are from the intestine. Scale = 10 µm for each image. Images are representative of at least 10 fish from three independent clutches. |