Imaging-associated stress causes divergent phase transitions of RNA-binding proteins in the Caenorhabditis elegans germ line

Abstract

One emerging paradigm of cellular organization of RNA and RNA-binding proteins is the formation of membraneless organelles. Examples of membraneless organelles include several types of ribonucleoprotein granules that form via phase separation. A variety of intracellular pH changes and posttranslational modifications, as well as extracellular stresses, can stimulate the condensation of proteins into granules. For example, the assembly of stress granules induced by oxidative stress, osmotic stress, and heat stress has been well characterized in a variety of somatic cell types. In the germ line, similar stress-induced condensation of proteins occurs; however, less is known about the role of phase separation during gamete production. Researchers who study phase transitions often make use of fluorescent reporters to study the dynamics of RNA-binding proteins during live cell imaging. In this report, we demonstrate that common conditions of live-imaging Caenorhabditis elegans can cause an inadvertent stress and trigger phase transitions of RNA-binding proteins. We show that this imaging-associated stress stimulates decondensation of multiple germ granule proteins and condensation of several P-body proteins. Proteins within larger ribonucleoprotein granules in meiotically arrested oocytes do not appear to be as sensitive to the stress as proteins in diakinesis oocytes of young hermaphrodites, with the exception of the germ granule protein PGL-1. Our results have important methodological implications for all researchers using live-cell imaging techniques. The data also suggest that the RNA-binding proteins within large ribonucleoprotein granules of arrested oocytes may have distinct phases, which we characterize in our companion article.

Keywords: RNA-binding protein; RNP granule; condensation; live imaging; oogenesis; phase transition; stress.

Fig. 1.

Imaging conditions can inadvertently induce stress. a) Subcellular localization of PGL-1::GFP in the germ line: early during imaging process (within 25 min) or after extended time imaging (60 min); and after short (15 min) or extended (60 min) exposure to 6.25 mM levamisole prior to mounting worms on a slide. Proximal oocytes are oriented on the bottom left in all images. Scale bar is 10 µm. b) Subcellular localization of DAF-16::GFP in young hermaphrodites: imaging was done early (within 25 min) or late (after 60–90 min) of slides being prepared and after short (15 min) or extended (60 min) exposure to 6.25 mM levamisole. c) Graphs show the percentage of DAF-16::GFP worms with cytoplasmic distribution, intermediate distribution, or nuclear localization. Statistical significance was determined using the Fisher’s exact test. **** indicates P < 0.0001 and *** indicates P < 0.001. n = 15 (early), 16 (late), 28 (short), and 19 (extended).

Fig. 2.

Imaging-associated stress causes distinct phase transitions among RNA-binding proteins. a) Micrographs of GFP-tagged P granule protein reporter strains, PGL-1::GFP and GLH-1::GFP. Top row: distribution of GFP in germ line early during imaging; bottom row: distribution of GFP in germ line after extended imaging (late) with strongest phenotypes shown. Asterisk marks the most proximal oocyte in each germ line. Smaller distal pachytene nuclei are visible at top of each image. Scale bar is 10 µm. b) Graphs showing the number of GFP granules in a single Z-slice of proximal oocytes (see Methods). Statistical significance was determined using the Mann–Whitney test. **** indicates P < 0.0001; * indicates P < 0.05. n = 16–26. c) Micrographs of GFP-tagged RNA-binding protein reporter strains (MEX-3, CGH-1, and CAR-1) in most proximal oocytes. Top row: distribution of GFP early during imaging; Bottom row: distribution of GFP after extended imaging (late) with strongest phenotypes shown. Arrows indicate ectopic granules; arrowheads indicate cortical enrichment. Scale bar is 10 µm. d) Graphs showing the number of GFP granules in a single Z-slice of proximal oocytes. Statistical significance was determined using the Mann–Whitney test. **** indicates P < 0.0001, *** indicates P < 0.001, and ** indicates P < 0.01. n = 17–20. Error bars indicate mean ± SEM.

Fig. 3.

Imaging-associated stress preferentially affects PGL-1 within large RNP granules of meiotically arrested oocytes. a) Micrographs of GFP-tagged RNA-binding proteins (MEX-3, CGH-1, PGL-1, MEG-3) in a fog-2 background. Top row: distribution of GFP in arrested oocytes early during imaging. Bottom row: distribution of GFP in arrested oocytes after extended imaging (late). Asterisk marks the most proximal oocyte in each germ line. Scale bar is 10 µm. b) Graphs showing either the number of GFP granules or the integrated density of GFP in granules in a single Z-slice of proximal oocytes (see Methods). Statistical significance was determined using the Mann–Whitney test. ** indicates P < 0.01 and ns indicates not significant. n = 18–33. Error bars indicate mean ± SEM.

Fig. 4.

Extended imaging does not appear to induce oxidative or ER stress. a) Micrographs of gcs-1::GFP, hsp-16.2::GFP, and hsp-4::GFP young adult worms either early during imaging or after extended imaging. The positive control for each was heat stress (see Methods). b) Graphs show the percentage of worms with low, moderate, or high levels of expression of GFP. No significant differences were detected between early and late-imaged worms in any of the reporter strains. n = 27–54.

Fig. 5.

Imaging conditions can quickly trigger stress and modulate phase transitions. a–d) Time course of the effects of imaging-associated stress on DAF-16 nuclear translocation, and condensation of PGL-1 and CAR-1. For each reporter, micrographs of GFP-tagged strains are shown on the left with proximal oocytes oriented to the left, and graphs show the percentage of worms with nuclear translocation (DAF-16) or phase transitions (PGL-1, CAR-1). Whole hermaphrodites are shown in (a); diakinesis oocytes in young hermaphrodites are shown in (b) and (d); and arrested oocytes are in (b). All scale bars are 10 µm. See Methods and text for the description of thresholds used to categorize phase transitions. Statistical significance was determined using the Kruskal–Wallis test. **** indicates P < 0.0001, *** indicates P < 0.001, ** indicates P < 0.01, * indicates P < 0.05, and ns indicates not significant. n = 9–15.