Expanding the Caenorhabditis elegans auxin-inducible degron system toolkit with internal expression and degradation controls and improved modular constructs for CRISPR/Cas9-mediated genome editing

The auxin-inducible degron (AID) system has emerged as a powerful tool to conditionally deplete proteins in a range of organisms and cell-types. Here, we describe a toolkit to augment the use of the AID system in Caenorhabditis elegans. We have generated a set of single-copy, tissue-specific (germline, intestine, neuron, muscle, hypodermis, seam cell, anchor cell) and pan-somatic TIR1-expressing strains carrying an equimolar co-expressed blue fluorescent reporter to enable use of both red and green channels in experiments. We have also constructed a set of plasmids to generate fluorescent protein::AID fusions through CRISPR/Cas9-mediated genome editing. These templates can be produced through frequently used cloning systems (Gibson assembly or SapTrap) or through ribonucleoprotein complex-mediated insertion of PCR-derived, linear repair templates. We have generated a set of sgRNA plasmids carrying modifications shown to boost editing efficiency, targeting standardized transgene insertion sites on chromosomes I and II. Together these reagents should complement existing TIR1 strains and facilitate rapid and high-throughput fluorescent protein::AID* tagging of factors of interest. This battery of new TIR1-expressing strains and modular, efficient cloning vectors serves as a platform for facile assembly of CRISPR/Cas9 repair templates for conditional protein depletion.


Introduction 20
Conditional degrons have emerged as a powerful tool to rapidly destroy proteins of 21 interest in order to interrogate their function in cells and in multicellular animals (Natsume 22 allow up to five proteins to be produced from a single mRNA (Ahier and Jarriault 2014). 1 Transgenes were introduced in single copy through CRISPR/Cas9 editing and self-2 excising cassette (SEC) selection into neutral loci that support robust expression. We in crosses. The loxP-flanked SEC is then excised by heat shock, producing the final 10 wildtype-moving strain. As expected, this sun-1p construct drives nuclear-localized BFP 11 in the germline and embryos, confirming the expression of the transgene ( Figure 2B). We insertion lines. We have not pursued whether these issues are due to transgene toxicity, 20 silencing, or other potential issues. 21 22 Using our new TIR1 construct, we created a suite of strains with ubiquitous or tissue-1 specific TIR1 expression ( Figure 3A). We created chromosome I and II knock-ins 2 expressing TIR1 in the germline (mex-5p and sun-1p), hypodermis (dpy-7p and col-10p), 3 muscle (unc-54p), and intestine (ges-1p) ( Figure 3A, Figure S1). We also created 4 chromosome I knock-ins expressing TIR1 in neurons (rgef-1p), somatic cells (eft-3p), 5 body wall muscle (myo-3p), and excretory cell+hypodermis+gut (vha-8p) ( Figure 3A, 6 Figure S1). Our vha-8p strain also resulted in promoter expression in unidentified cells in 7 the head. We also generated a strain expressing TIR1 in the seam cells using a minimal 8 SCMp enhancer (gift from Prof. Allison Woollard) and a pes-10 minimal promoter ( Figure  9 3A). While we saw robust seam cell expression in this strain, we also detected 10 hypodermal expression (unpublished data). We are making this strain available to the 11 community, but encourage careful evaluation before interpretation. observed auxin-dependent depletion of NHR-25::GFP::AID*::3xFLAG in the AC, while no 23 depletion was observed in the adjacent VPCs ( Figure 4). Thus, even if the presence of 1 TIR1 is undetectable through BFP reporter expression, there may still be sufficient 2 amounts of TIR1 to deplete proteins of interest. We also made a strain designed to 3 express TIR1 in both the soma and germline (smu-1p), as the eft-3p driven TIR1 4 transgenes are typically silenced in the germline. We could not detect BFP expression in 5 this smu-1p strain, but have made it available for the community to test. The majority of 6 the TIR1 strains (17/19) Figure S1); the exceptions to this statement 9 are the previously discussed smu-1p and cdh-3p TIR1 strains. cassettes for the SapTrap system. We, therefore, created a set of vectors compatible with 17 the most common genome-editing pipelines we use in our lab. First, we took a set of 18 vectors that use Gibson assembly to generate SEC-selectable repair templates 19 developed by Dickinson et al. (2015) and introduced AID* sequences upstream of the 20 3xFLAG epitope. This set of vectors allows for tagging genes with GFP, YPET, mKate2, 21 and TagRFP-T along with AID*::3xFLAG epitopes ( Figure 5, Table S3). Methods in C. and we constructed vectors with GFP, TagRFP-T, and mKate2. 5 6 We also have shifted to frequently using Cas9 RNP-based editing with linear repair Modifying the large, SEC-based selection cassettes for Gibson assembly was technically 21 challenging as the size and repetitive ccdB sequences made these prone to 22 recombination. The modularity of the type II restriction enzyme-based SapTrap cloning 23 pipeline was appealing as a method to rapidly develop new repair templates for 1 CRISPR/Cas9-mediated genome editing ( Figure 6A). Recent modifications to the system 2 have made it compatible with SEC-based cloning (Dickinson et al. 2018), retaining an FP 3 cassette and adding an SEC selection cassette. We generated a series of constructs for 4 the SapTrap NT and CT slots, consisting of flexible linkers, various combinations of AID* 5 cassettes, and epitopes for protein purification or detection (3xMyc, 3xFLAG, BioTag 6 ( Figure 6B).
We first attempted to generate a 30 amino acid 7 linker::GFP^SEC^TEV::AID*::3xFLAG construct to tag lin-42 at the C-terminus. While we 8 were able to produce the construct and obtain a knock-in strain (unpublished), our 9 efficiency was very low, and we were unable to get correct assemblies by simply selecting 10 colonies and sequencing. We, therefore, turned to colony PCR screening. Our standard 11 practice of screening one assembly junction produced false positives, where one 12 homology arm was correctly connected to the backbone and desired SapTrap cassettes, 13 but the other arm was missing cassettes. We therefore screened both assembly junctions 14 with the vector by colony PCR to identify correct clones, finding one correct assembly out 15 of 48 colonies. As this efficiency was much lower than reported in the original SapTrap indicating that our efficiency issues were not due to our SapTrap reagents. In examining 20 our colony PCR data for the lin-42 construct, we occasionally noted the presence of bands 21 smaller than the expected product. Sequencing the plasmid from these strains revealed 22 partial assemblies of 2-3 blocks. We then PCR-amplified these partial assemblies to 23 restore the terminal SapI sites and connectors. Using these partial assembly blocks 1 dramatically improved efficiency. To facilitate an SEC-based SapTrap assembly pipeline, 2 we generated a series of "multi-cassettes," where we combined fragments that we 3 frequently use ( Figure 6C, Table S3). We re-created the 30 amino acid 4 linker::GFP^SEC^TEV::AID*::3xFLAG lin-42 targeting construct using a multi-cassette, 5 and our colony PCR hit rate jumped from 1/48 (2.1% (70.8%)) to 17/24. For our most 6 commonly used vectors, we have generated constructs containing full assemblies of the 7 knock-in epitope, lacking only the homology arms. PCR amplifying homology arms with 8 SapI sites and appropriate connectors allows high-efficiency generation of repair 9 templates. 10 11

Additional vectors to support genome editing and gene expression studies 12
We had previously shown that the "Flipped and extended (F+E)" sgRNA modifications  We frequently used a separate sgRNA vector to reduce the number of fragments in an 17 assembly and to allow us to increase the molar ratio of sgRNA vector to repair template. 18 We generated SapTrap sgRNA RNA expression vectors using both commonly used U6 19 promoters (pJW1838,pJW1839)), as it is currently unclear whether one promoter is 20 The AID system has allowed rapid, conditional, and tissue-specific depletion of tagged 19 proteins in a wide range of organisms and cell types. Since its introduction to C. elegans 20 (Zhang et al. 2015), it has been promptly adopted by the community. This system has 21 allowed for rapid depletion of proteins in tissues that are refractory to RNA interference to read out TIR1 activity opens the door to performing suppressor screens for phenotypes 23 of interest generated using the AID system. Mutations in the TIR1 transgene or auxin 1 transport factors could lead to unintended suppression of a mutant phenotype when 2 performing such a screen. Therefore, being able to monitor TIR1 activity provides a should simplify integration of new FPs and epitopes. As we typically perform edits in a 1 range of genetic backgrounds, we designed our constructs to use the SEC cassettes 2 designed for the SapTrap system (Dickinson et al. 2018). In our first tests requiring 3 assembly of nine pieces of DNA (two homology arms, sgRNA, four cassettes, two plasmid 4 backbone fragments), we had poor efficiency and required screening of two assembly 5 junctions by colony PCR to identify a single correct assembly out of 48 colonies. While 6 the SEC-based version of SapTrap was less efficient than Gibson cloning (~60% vs. 30%) 7 (Dickinson et al. 2018), it was still much higher than the efficiencies we initially observed. 8 In screening our reactions, we would occasionally identify partial assemblies (i.e. 5' 9 homology arm+CT cassette). By cloning out these partial assemblies to reduce the 10 number of fragments, we boosted our assembly efficiencies. We then shifted to create 11 multi-cassettes, which minimize the number of ligations required to successfully assemble 12 the final vector. Additionally, by expressing our sgRNA on a separate plasmid, only four 13 pieces of DNA need to be assembled in all, boosting efficiency (backbone, 5' homology 14 arm, 3' homology arm, multi-cassette). We also provide methods for the community to 15 build new multi-cassettes for highly used FP-epitope combinations. Finally, we generated For many applications, the AID system offers a powerful method to conditionally degrade 2 proteins in specific tissues and at specific points in development. However, as the system 3 has gained popularity, particular challenges have emerged. While they do not dampen 4 our enthusiasm for the AID system, it is important to be aware of them. Here, we also 5 discuss potential solutions to these issues. It has become clear that in certain cases there 6 can be auxin-independent, TIR1-dependent degradation of AID-tagged proteins. This in the presence of auxin. ARF19 is an AID interaction partner and is thought to shield the 22 tagged protein from TIR1 in the absence of auxin. It may be useful to test whether ARF19 23 improves performance of the AID system in C. elegans. One important caveat is that the 1 authors used a full-length AID tag. The miniAID and AID* tags frequently used in C. 2 elegans lack domains III and IV of the protein which are thought to be important for the 3 ARF19 interaction. Full-length AID is 229 amino acids, a substantially larger tag that 4 would necessitate further study to ensure it did not interfere with fusion proteins. Another 5 approach could be a recently described AID system comprised of Arabidopsis thaliana 6 AFB2 and a minimal degron from IAA7, which was reported to minimize basal degradation 7 (Li et al. 2019). Our set of vectors will allow modular assembly of any new AID system 8 component and facile integration of any new reagents. We note that engineering an 9 improved TIR1 that did not promote auxin-independent degradation of miniAID-tagged 10 proteins would be most desirable. A strong candidate is a recently described TIR1 (F79A) 11 mutation and modified auxin that had 1000-fold stronger binding, reducing the amount of 12 auxin required for target knockdown (Nishimura et al. 2020). This reagent would be 13 compatible with and improve the performance of the collection of miniAID-and AID*-14 tagged strains which the C. elegans community has already generated. 15 16 We previously used the AID system to deplete the nuclear hormone receptor NHR-23 17 and reported a larval arrest phenotype similar to a previously-described null allele, and germline (manuscript in preparation). When we use a germline-specific TIR1 to deplete 7 NHR-23, by microscopy we observed no detectable expression following auxin exposure 8 (manuscript in preparation). However, ~30% of protein remains as detected by western 9 blot (manuscript in preparation). A similar inability to obtain null phenotypes using the AID inhibition could test whether an endogenous ubiquitin ligase could interact with the AID 23 tag in the absence of TIR1. Additionally, more information is required to determine rules 1 for optimal AID tag placement in both structured and unstructured domains of proteins. 2 As a precaution, we tend to use long 10-30 amino acid flexible linker sequences to space 3 the AID* tag away from the protein of interest.  However, we also note that effective depletion is possible even when TIR1 levels are low 21 enough where the BFP reporter is undetectable (Figure 4). This result highlights the 22 importance of functionally testing new TIR1 transgenes with FP::AID*-tagged alleles of 1

interest. 2 3
The ability to rapidly deplete proteins with temporal and cellular resolution allows precise 4 dissection of the roles of gene products in developmental processes of interest. With the 5 ever-increasing efficiency of genome editing and continued refinement of the AID system, 6 one can envision creating libraries of FP::AID*-tagged genes covering the genome and a 7 bank of TIR1 strains to allow depletion in virtually all cell types.   systems. In the presence of the plant hormone auxin, TIR1 recognizes and binds the AID 7 sequence, leading to ubiquitination and subsequent degradation of the AID-tagged 8 protein.
In C. elegans, the system is frequently used with single-copy TIR1 transgenes 9 inserted into neutral loci, and AID* knock-ins into genes of interest, though 10 extrachromosomal arrays can also be used.  condition; **** indicates P < 0.0001 by a two-tailed unpaired Student's t-test. P < 0.05 was 10 considered statistically significant). Scale bars represent 5 µm. allowing for the generation of programable 3bp sticky ends. B) SapTrap cloning facilitates 12 single reaction cloning of multiple fragments, in the correct order, into a single repair 13 template plasmid. Specific sticky ends are used for specific cassettes. C) Table of new  14 vectors generated for the SapTrap CT and NT slots. Our initial assembly efficiencies were 15 sub-optimal, and we found that reducing the number of fragments assembled improved 16 our efficiencies. We have generated a set of multi-cassettes where partial assemblies 17