In Vivo Delivery and Activation of Masked Fluorogenic Hydrolase Substrates by Endogenous Hydrolases in C. elegans

Abstract

Protein expression and localization are often studied in vivo by tagging molecules with green fluorescent protein (GFP), yet subtle changes in protein levels are not easily detected. To develop a sensitive in vivo method to amplify fluorescence signals and allow cell-specific quantification of protein abundance changes, we sought to apply an enzyme-activated cellular fluorescence system in vivo by delivering ester-masked fluorophores to Caenorhabditis elegans neurons expressing porcine liver esterase (PLE). To aid uptake into sensory neuron membranes, we synthesized two novel fluorogenic hydrolase substrates with long hydrocarbon tails. Recombinant PLE activated these fluorophores in vitro. In vivo activation occurred in sensory neurons, along with potent activation in intestinal lysosomes quantifiable by imaging and microplate and partially attributable to gut esterase 1 (GES-1) activity. These data demonstrate the promise of biorthogonal hydrolases and their fluorogenic substrates as in vivo neuronal imaging tools and for characterizing endogenous C. elegans hydrolase substrate specificities.

Keywords: C. elegans; bioorthogonal; fluorescent probes; hydrolases; imaging agents.

Figure 1. Activation of lipophilic ester-masked fluorogenic substrates occurs in vivo in C. elegans

(A) Mechanism of hydrolase-mediated activation of fluorogenic substrates. The fluorescence of fluorescein can be masked by the addition of ester moieties (R groups), which allow them to serve as hydrolase substrates. See Figure S1 for previously published R groups used for ester masking of fluorescein in this study. Two new fluorescein derivatives esterified with the long hydrocarbon R groups, 1 (C₁₂) and 2 (C₁₈), were synthesized and characterized here. (B) PLE is expressed in worms carrying the PLE transgene under a panneuronal promoter. (Left panel) qRT-PCR of PLE mRNA levels in Psnb-1::ple expressing worms. Total RNA was isolated from mixed stage PLE transgenic (kjrEx19) worms. qRT-PCR was run on these samples and samples isolated from wild type worms (see text) using primers targeting 133 bp segments of ple cDNA and emb-30 cDNA as an internal control. (Right panel) Western blot of PLE protein levels in WT and Psnb-1::PLE expressing worms. Western blots were run on 100 worm lysates of WT or PLE trangenic (kjrEx19) worms or on 2 µg purified PLE. The PLE:actin ratio is the ratio of the integrated pixel intensities of PLE and actin bands. (C) Substrates 1 and 2 are activated in vitro by PLE. Substrates 1 and 2 (5 µM) and 0.1 mg/ml purified PLE were incubated in a 96 well plate and fluorescence intensity (485nm excitation, 528nm emission) was measured every minute for two hours on a microplate reader. All groups were run in triplicate; average intensities ± s.e.m. at each timepoint are shown (open circles, LAD+PLE; filled squares, SAD+PLE; open squares, No PLE). (D) Substrate 1 labels head neurons in both wild type and PLE-expressing animals. Images of the heads of L4 wild type (WT, left panels) or Psnb-1::PLE expressing transgenic C. elegans (right panels) following six hour exposure to 50 µM masked fluorogenic substrates (63X). Psnb-1::PLE worms also express a red fluorophore, mCherry, under the Psnb-1 promoter (red label, neurons) allowing identification of PLE-expressing animals. Yellow arrows mark the cell bodies of neurons labeled with either DiO or substrate 1. White asterisks (*) mark autofluorescent gut granules in several images. (E) Substrate 1 activation occurs in DiI-labeled sensory neurons. Images of the head (top row) and tail (bottom row) of L4 wild type C. elegans following six-hour exposure to 50 µM masked fluorogenic substrate 1 (63X) and DiI for two hours. Yellow arrowheads mark substrate 1, DiI, and co-labeled cells. For D and E, input levels of all equivalent images from a single experiment were processed identically in Adobe Photoshop for image presentation; DiO and DiI images were exceptions to this, as levels of these images were set at lower input levels than were those of the other fluorophores due to the brightness of the DiO and DiI labeling. Scale bar is 20 µm.

Figure 2. Substrate specificity of endogenous intestinal C. elegans hydrolases against fluorogenic substrates

(A) Images of the intestine of L3-L4 stage wild type (WT) C. elegans following six hour exposure to 50 µM masked fluorogenic substrates (100X). Scale bar is 10 µm. (B) Quantification of the average intestinal fluorescence intensity for each treatment shown in (A) ± std. dev. ** p ≤ 0.01, *** p ≤ 0.001, two-tailed t-tests or Wilcoxon non-parametric test (substrate 9) compared to DMSO (n = 8–15 images per treatment). (C) Microplate quantification of the average rate of substrate activation for each fluorophore by endogenous esterases. Sample assay shown with n = 3 wells per treatment; error bars show the standard deviation of each group.

Figure 3. Activated hydrolase substrates are localized in lysosomes in the C. elegans gut

Images of wild type L3-L4 C. elegans following 48-hour treatment with 2 µM Lysotracker dye (red) and six-hour exposure to 50 µM masked fluorogenic substrates 1 or 9 or DMSO (green) (100X). All lysotracker labeling overlapped with fluorophore labeling; however, a few green fluorophore puncta were not labeled with red Lysotracker dye (white arrows). Scale bar is 10 µm.

Figure 4. Activation of fluorogenic hydrolase substrates in C. elegans is due to endogenous serine hydrolase activity, including GES-1

(A-C) Results of sample microplate experiments in which 4 worms/well of synchronized populations of wild type (WT) animals at the L4 stage were incubated with the indicated fluorogenic substrate (substrates 1, 2, or 9) at a final concentration of 5 µM in the presence of heat-killed HB101 bacteria (hk+worms) or heat killed HB101 in the presence of 50 µM of the lipophilic esterase inhibitor, THL (hk+worms+thl), as compared to heat-killed bacteria and S-complete buffer alone (hk+buffer). (D) Results of sample microplate experiment testing the hydrolase activity of worm supernatants. WT worms were synchronized and allowed to grow in liquid culture with heat-killed bacteria to the L4 stage. The supernatant from growth solutions (~30 worm equivalent) was incubated with 5 µM fluorogenic substrates in a microplate to assess the presence of secreted esterase activity. Adjusted supernatant activation per 4-worm equivalent is shown to match worm numbers used in panels A-C. All experiments (A-D) were performed in triplicate at least twice. (E) (Upper panel) Images of L4 stage wild type and ges-1 (ca6ca7) loss of function C. elegans mutants following six-hour exposure to 50 µM masked fluorogenic substrates (100X). Scale bar is 10 µm. (Lower panel left) Quantification of the average intestinal fluorescence intensity fold activation over DMSO in imaging experiments with substrates 1 and 9 is shown above ± std. dev. ** p ≤ 0.01, Student’s t test (n = 10 images per treatment). (Lower panel right) Microplate quantification of the average rate of substrate activation ± std. dev. for substrates 1 and 9 by wild type and ges-1 worms (30 worms/well) exposed to 5 µM fluorogenic substrates. * p ≤ 0.05, Student’s t test (n = 4 wells per treatment).