Zones of cellular damage around pulsed-laser wounds

Abstract

After a tissue is wounded, cells surrounding the wound adopt distinct wound-healing behaviors to repair the tissue. Considerable effort has been spent on understanding the signaling pathways that regulate immune and tissue-resident cells as they respond to wounds, but these signals must ultimately originate from the physical damage inflicted by the wound. Tissue wounds comprise several types of cellular damage, and recent work indicates that different types of cellular damage initiate different types of signaling. Hence to understand wound signaling, it is important to identify and localize the types of wound-induced cellular damage. Laser ablation is widely used by researchers to create reproducible, aseptic wounds in a tissue that can be live-imaged. Because laser wounding involves a combination of photochemical, photothermal and photomechanical mechanisms, each with distinct spatial dependencies, cells around a pulsed-laser wound will experience a gradient of damage. Here we exploit this gradient to create a map of wound-induced cellular damage. Using genetically-encoded fluorescent proteins, we monitor damaged cellular and sub-cellular components of epithelial cells in living Drosophila pupae in the seconds to minutes following wounding. We hypothesized that the regions of damage would be predictably arrayed around wounds of varying sizes, and subsequent analysis found that all damage radii are linearly related over a 3-fold range of wound size. Thus, around laser wounds, the distinct regions of damage can be estimated after measuring any one. This report identifies several different types of cellular damage within a wounded epithelial tissue in a living animal. By quantitatively mapping the size and placement of these different types of damage, we set the foundation for tracing wound-induced signaling back to the damage that initiates it.

Fig 1. mCherry.NLS reports nuclear envelope integrity.

(A) An intact Drosophila pupa (left) and one prepared for imaging with the pupal case partially removed (right), exposing the pupal notum. (B) Interphase nuclei in unwounded tissue are labeled by mCherry.NLS driven by the Gal4/UAS system. During mitosis, the fluorescence signal evident at 0 min disperses when the nuclear envelope breaks down, seen within circles in panels at 15 min and 45 min, before reappearing as two distinct/punctate nuclei at 100 min. (C) A schematic showing the diffusion of mCherry.NLS out of the nucleus during mitosis and its concentration in nuclei again after the nuclear envelope reforms in the daughter cells. Thus, the localization of mCherry.NLS reports nuclear membrane integrity. Scale bar = 25 μm.

Fig 2. Several zones of damage are evident around pulsed-laser wounds.

(A) The region of laser-induced rupture is observed as the area of disrupted mCherry.NLS within the first two frames after wounding (here 2 seconds), annotated with an orange circle in A’. (B) The region of delayed cell lysis is observed as the complete loss of mCherry.NLS at 90 seconds, (red circle in B’) and the region of nuclear membrane damage is observed as a diffuse, non-nuclear mCherry.NLS signal (light blue circle in B’), maintained within the cells by the plasma membrane at 90 seconds after wounding. (C) The region of plasma membrane damage is observed as an increase in cytoplasmic calcium levels immediately after wounding (green circle in C’, as reported by GCaMP6m, a fluorescent calcium indicator, in the first frame after wounding (here 0 seconds). In this region, the wound creates microtears in the plasma membrane, allowing immediate influx of extracellular calcium. All images are of the same wound; the frame is shifted in B,B’. (D) Actual measurements of radii in four different samples demonstrates a consistent relationship between these regions; as any one of these regions increases in radius, the others likewise increase in radius. Panels E-J quantify this trend with 95% prediction bands. n = 17 pupae. Scale bar = 50 μm.

Fig 3. Histones reveal that within the region of nuclear membrane damage, some nuclei have intact chromatin whereas nuclei closer to the center have disrupted chromatin.

(A) His2Av-GFP, which labels histones in chromatin, reveals a damage region with misshapen chromatin and decreased fluorescence, indicated with a purple circle in A’. (B) mCherry.NLS in the same wound reveals the region of nuclear membrane damage, (blue circle in B’) and the region of delayed cell lysis (red circle in B’). (C) The overlay of His2Av-GFP with mCherry.NLS shows that the region of chromatin disruption is much smaller than the region of nuclear membrane damage, but similar in size to the region of delayed cell lysis. (D-E) The relationships between the region of chromatin disruption and the regions of nuclear membrane damage (D) and of delayed cell lysis (E) are linear. The 95% confidence interval is indicated. n = 17 pupae. Scale bar = 50 μm.

Fig 4. Cell borders are disrupted following wounding.

(A) Genetically encoded Ecadherin-GFP marks cell borders and reveals damage around a pulsed-laser wound. The region of immediate Ecadherin-GFP loss is indicated with a gold circle in A’. (B,C) The region of immediate Ecadherin loss is distinct from and within the regions of delayed cell lysis (red circle) and nuclear membrane damage (blue circle). (D) The relationship of the radius of delayed cell lysis to the radius of immediate Ecadherin loss is linear. (E) The relationship of the radius of nuclear membrane damage to the radius of immediate Ecadherin-loss is linear. 95% prediction interval is indicated in D,E. n = 32 pupae. Scale bar = 50 μm.

Fig 5. The first expansion of calcium extends about 20 μm from the region of plasma membrane damage, independent of wound size.

(A) The influx of calcium immediately after wounding identifies cells with plasma membrane damage. This region is observed by an immediate increase in GCaMP fluorescence, indicated by a green circle in A’. (B) The calcium radius increases over the next ~15 seconds [15], and the maximum first expansion region is indicated with a dark blue circle in panel B’. The relationship between the plasma membrane radius and the first expansion radius is linear. Because the slope is ~1, the y-intercept of 19 indicates that the 1^st expansion radius is expected to be ~20 μm larger (~2 to 4 cell diameters) than the radius of plasma membrane damage, regardless of wound size or laser energy. n = 92 pupae. Scale bar = 50 μm.

Fig 6. Each zone of damage can be estimated given one of them.

(A) Schematic of how to calculate each zone of damage value from a given one, with each of the eleven linear equations displayed below corresponding to the arrows above. Arrows go from independent variable (x) to dependent variable (y), corresponding to Eqs 1–11 derived in Figs 2–5. The complete set of 42 equations necessary to relate each of the 7 zones of damage were derived according to the Materials and Methods and are provided in S2 Dataset. (B) Three hypothetical wounds are indicated, each with a different initial laser-induced rupture radius (10, 20, or 30 μm). The equations in S2 Dataset were used to derive the radii of the six other zones of damage from the given laser-induced rupture radius. The 95% confidence intervals are displayed for all values. With a radius of 10 μm, the first four regions overlap (laser-induced rupture, immediate Ecadherin loss, chromatin disruption, delayed cell lysis). Because the average cell diameter in this tissue is ~7 μm, the overlap suggests that these four regions may be within one cell diameter. As the wound gets larger, the regions become more distinguishable.