Spatial and temporal control of expression with light-gated LOV-LexA

Abstract

The ability to drive expression of exogenous genes in different tissues and cell types, under the control of specific enhancers, has been crucial for discovery in biology. While many enhancers drive expression broadly, several genetic tools were developed to obtain access to isolated cell types. Studies of spatially organized neuropiles in the central nervous system of fruit flies have raised the need for a system that targets subsets of cells within a single neuronal type, a feat currently dependent on stochastic flip-out methods. To access the same cells within a given expression pattern consistently across fruit flies, we developed the light-gated expression system LOV-LexA. We combined the bacterial LexA transcription factor with the plant-derived light, oxygen, or voltage photosensitive domain and a fluorescent protein. Exposure to blue light uncages a nuclear localizing signal in the C-terminal of the light, oxygen, or voltage domain and leads to the translocation of LOV-LexA to the nucleus, with the subsequent initiation of transcription. LOV-LexA enables spatial and temporal control of expression of transgenes under LexAop sequences in larval fat body and pupal and adult neurons with blue light. The LOV-LexA tool is ready to use with GAL4 and Split-GAL4 drivers in its current form and constitutes another layer of intersectional genetics that provides light-controlled genetic access to specific cells across flies.

Keywords: binary expression system; light-gated expression; photosensitive.

Fig. 1.

Testing components for a light-gated expression system based on eLOV. a) The NLS-like sequence in LexA is mutagenized in mLexA. b) S2R+ cell line was used to test whether LexA-transactivator:tdTomato and mLexA-transactivator:tdTomato drive transcription of a LexAop-reporter, LexAop-myr:GFP, using the MET-GAL4 driver. c) The ratio of LexAop reporter myr:GFP expression in relation to expression of the different LexA- and mLexA-transactivator chimeras determined by tdTomato signal. Cotransfection of UAS-mCherry and UAS-CD8:GFP was used as an approximate measure of coexpression (first bar in the boxplot), whereas cotransfection of UAS-mCherry with the LexAop reporter LexAop-myr:GFP established the baseline (second bar in the boxplot). The constructs for GAD were LexA:GAD-tdTomato and mLexA:GAD-tdTomato, for p65 were LexA:p65-tdTomato and mLexA:p65-tdTomato, and for VP16 were LexA:VP16-tdTomato and mLexA:VP16-tdTomato. Mutagenizing NLS-like sequence reduces the transcriptional activity of mLexA-transactivator chimeras compared to LexA-transactivator chimeras. At least 200 cells with medium levels of expression of mCherry or tdTomato, from at least 2 transfections of S2R+ cells are represented for each condition. d) S2R+ cells were transfected with a reporter, LexAop-myr:GFP, together with the driver MET-GAL4 and the test construct, to examine light-gated transcription for test constructs. e) Expression of LexAop reporter myr:GFP in relation to expression of mLexA-transactivator chimeras combined with eLOV. Placement of eLOV-nls N-terminal followed by the fluorescent protein tdTomato and the mLexA-transactivator chimera yielded the best signal for cells exposed to pulses of blue light, while maintaining a low reporter signal in cells kept in the dark. As in (b), cotransfection of UAS-mCherry with UAS-CD8:GFP (second bar in plot) or LexAop-myr:GFP (first bar in plot) served as a positive and negative control, respectively. The test constructs were eLOV-nls-tdTomato-mLexA:GAD (GAD), eLOV-nls-tdTomato-mLexA:p65 (p65), and eLOV-nls-tdTomato-mLexA:VP16 (VP16). At least 200 cells with medium levels of expression of mCherry or tdTomato, from 2 to 5 transfections of S2R+ cells are represented for each condition. c, e) *** represents P-values <0.001; n.s. represents P-values >0.05, obtained with Student’s t-test. f) Representative examples of S2R+ cells expressing test constructs indicated above the images under the control of MET-GAL4. The eLOV-nls-tdTomato-mLexA:GAD forms clusters in the cytoplasm, whereas both eLOV-nls-tdTomato-mLexA:p65 and eLOV-nls-tdTomato-mLexA:VP16 are evenly distributed in the cytoplasm, and sometimes nucleoplasm, like mCherry. g) Schematic representation showing how eLOV-nls-tdTomato-mLexA:p65 (or LOV-LexA) works.

Fig. 2.

LOV-LexA is gated by light in vivo. a) Drosophila larvae expressing LOV-LexA in the fat body were exposed to blue light and examined for expression of the LexAop reporter, as well as the construct selected for LOV-LexA. b) Schematics showing the timeline of the experiment, light regime. Second or young third instar larvae were selected from vials kept at 18°C and transferred to 15% sucrose solution. The first light pulse was delivered immediately after this transfer, with a blue LED on an inverted microscope (Supplementary Table 3). Larvae were placed in the dark at 25°C between light pulses (see text for details), until dissection. Fat bodies expressing LOV-LexA with Cg-GAL4 for second to third instar larvae kept in the dark (c–e) or exposed to 3 30-s pulses of blue LED light at 1 Hz (f–h). Exposure to blue light appears to alter LOV-LexA cellular distribution (d, g) and leads to the expression of LexAop-CsChrimson:Venus in fat body cells as detected with anti-GFP antibody (e, h). i) Ratio of fluorescence, measured as pixel intensity in confocal-acquired images, of anti-GFP signal/anti-RFP signal for stained fat bodies from larvae with the genotype w, LexAop-CsChrimson:Venus; Cg-GAL4/+; UAS-LOV-LexA/+ that were kept in the dark (N = 7, representative example in (c)–(e), exposed to 6 light pulses and dissected after 7 h (N = 10), or 11 h (N = 11), or exposed to 3 light pulses and dissected 11 h later (N = 16, representative example in (f)–(h). Varying number of light pulses and the incubation period at 25°C before dissection led us to conclude that LOV-LexA gates expression with light in fat body and that LOV-LexA light-gated expression is highest with 3 light pulses and an 11-h incubation period at 25°C. *** represents P-values <0.001, n.s. represents P-values >0.05, 2-tailed Mann–Whitney tests. Exposure to blue light leads to an increase in the amount of Venus relative to LOV-LexA levels. j) Schematics showing the timeline of the experiment, light exposure, and functional imaging. k) Drosophila pupae expressing LOV-LexA and CD8:GFP in oenocytes were mounted on double-side sticky tape, and an opening in the pupal case that exposes oenocytes was created. Pupae expressed LOV-LexA and CD8:GFP in oenocytes with the following genotype: w; 109(2)-GAL4, UAS-CD8:GFP/+; UAS-LOV-LexA/+. l) Representative images of pupal oenocytes showing LOV-LexA before (before) and immediately following exposure to blue light (after), 60 min after exposure to blue light (recovery), and 120 min after light exposure (final scan). The final scan included the green channel to capture CD8:GFP, coexpressed with LOV-LexA, and used to delineate the cell body. The light used to capture GFP is blue and elicited another translocation of LOV-LexA to the nucleus, thereby demonstrating that the oenocytes were healthy after imaging. m) Mean nuclear tdTomato fluorescence over time, imaged live every 5 min. Shades represent standard error of the mean (SEM). LOV-LexA translocates to the nucleus upon exposure to blue light within minutes in oenocytes and slowly leaks out of the nucleus after exposure to blue light.

Fig. 3.

LOV-LexA gates expression with light in neurons. a, b) Schematic representation outlining the experiment. Pupae reared at 18°C aged 2–3 days APF were removed from vials, mounted on double side sticky tape on a cover slip and kept in the dark (a), or pasted onto a slide and exposed to blue light (b). Mounted pupae kept in the dark or exposed to light were shifted to 25°C until dissection. c) Schematic representation showing pupae lined on double side sticky tape for light delivery. Adult brains showing expression of LOV-LexA (red in d and f) and LexAop-CsChrimson:Venus (Venus in d and f and dedicated image in e and g), as detected by native fluorescence of tdTomato and Venus, from pupae kept in the dark (d, e) or exposed to light (f, g) at 3–4 days APF, as shown in (b). h) Ratio of Venus signal intensity over DAPI signal intensity for fru+ neuronal cell bodies located in the anterior brain. Pupae exposed to pulses of blue light (N=12) express the LexAop reporter Venus at higher levels compared to pupae kept in the dark (N = 13), demonstrating that exposure to light leads to higher LOV-LexA transcriptional activity. Adult brains showing expression of LOV-LexA (red in i and k) and LexAop-CsChrimson:Venus (Venus, green in i and k and dedicated image in j and l), as detected by the native fluorescence of tdTomato and Venus, from w, LexAop-CsChrimson:Venus;+/LC10a-SS1.AD; UAS-LOV-LexA/LC10a-SS1.DBD pupae kept in the dark (i, j) or exposed to light at 3–4 days APF (k, l), as shown in (b). m) Ratio of Venus signal intensity over DAPI signal intensity for LC10a neuronal cell bodies. Pupae exposed to pulses of blue light (N = 12) express the LexAop reporter Venus at higher levels compared to pupae kept in the dark (N = 6). n) Ratio of Venus over LOV-LexA native fluorescence from adult brains w, LexAop-CsChrimson:Venus;+; UAS-LOV-LexA/fru-GAL4 exposed to 0, 2, 4, or 8 pulses of blue light as 2–3 days APF pupae (N = 3, 4, 11, and 5, respectively). *** represents P-values <0.001, * represents P-values <0.05, n.s. represents P-values >0.05, 2-tailed Mann–Whitney tests.

Fig. 4.

LOV-LexA enables spatial and temporal control of transgene expression with light. a) Schematic representation outlining the experiment (top) and schematic representation showing pupae lined up on a slide, with exposed heads for live imaging and blue light delivery (bottom), shown in (c) and (d). Live image of 4 day APF pupal head after removal of the pupal case, with expression of Venus in fru+ neurons (w, LexAop-CsChrimson:Venus;+; UAS-LOV-LexA/fru-GAL4) before delivery of blue light (b) and 12 h after delivery of blue light (c). d) Change in the ratio of native Venus signal over LOV-LexA tdTomato native signal, before and after light delivery (N = 5). e) Timeline of the experiment. f) Schematic representation showing preparation to deliver spatially restricted light to immobilized adult flies, glued with low temperature melting wax to an opaque coverslip, with the head placed under a hole with a diameter between 300 and 400 µm. g-i, k-m, o-q) Representative images of adult brains expressing LOV-LexA in several LC neurons and spatially restricted LexAop-CsChrimson:Venus, after exposure to spatially restricted blue light to target visual projection neurons unilaterally and quantification. g–j) LC10-group neurons LC10s-SS2 (w, LexAop-CsChrimson:Venus; +/LC10s-SS2.AD; UAS-LOV-LexA/LC10s-SS2.DBD) with N = 8. k–n) LC10a neurons LC10a-SS1 (w, LexAop-CsChrimson:Venus; +/LC10a-SS1.AD; UAS-LOV-LexA/LC10a-SS1.DBD) with N = 11. o–r) LC6-OL77B neurons (w, LexAop-CsChrimson:Venus; +/OL77B.AD; UAS-LOV-LexA/OL77B.DBD) with N = 8. s–v) LC9-SS2651 neurons (w, LexAop-CsChrimson:Venus; +/SS2651.AD; UAS-LOV-LexA/SS2651.DBD) with N = 7. j, n, r, v) Ratio of native Venus over native LOV-LexA (tdTomato) signals between the side of the head that was illuminated compared to the side that was kept in the dark, with each plot corresponding to the genotypes shown in the same row. * represents P-values <0.05, ** represents P-values <0.01, Wilcoxon test.