Supplementary Text and Figures

Supplementary Materials and Methods

Mammalian cell culture.

HEK293T cells (ATCC # CRL-3216) were cultured in high-glucose DMEM supplemented with 10% FBS, 1 mM glutamine and 100 µg/mL penicillin/streptomycin (Gibco). A split-wrmScarlet_1-10 cDNA codon-optimized for mammalian expression was fused to the C-terminus of eGFP and cloned into a pCDH lentiviral expression vector (SFFV GFP-split-wrmScarlet_1-10). Lentivirus was prepared using standard protocols [Kamiyama 2016] and used to infect HEK293T cells. A polyclonal population of GFP-mScarlet_1-10 positive cells was isolated by FACS (using GFP fluorescence) and served as parental cell line for further experiments. For CLTA-N CRISPR engineering, S. pyogenes Cas9/sgRNA ribonucleoprotein complexes were prepared as in [Leonetti 2016] , mixed with HDR donor templates and electroporated into of GFP-mScarlet_1-10 cells by nucleofection.

CLTA-N split-wrmScarlet₁₁ donor library.

A cDNA pool of degenerate split-wrmScarlet₁₁ sequences was generated by oligonucleotide synthesis (GeneScript) and homology arms for HDR-mediated insertion at CLTA N-terminus were appended by PCR (Supplementary Material – Table S7 for sequences). Library diversity was verified by Illumina MiSeq deep-sequencing.

Supplementary Results

Split mScarlet screening in mammalian cells

We tested the applicability of the split-wrmScarlet_1-10 system for mammalian cell engineering but were surprisingly unsuccessful at detecting fluorescence. We designed a human codon-optimized split-wrmScarlet_1-10 cDNA and expressed it as a C-terminal GFP fusion in HEK293T cells by lentiviral transduction. Expression of GFP verified the successful expression of the fusion protein (Figure S10A). However, subsequent expression of split-wrmScarlet₁₁ fragments did not give rise to detectable red fluorescence despite numerous attempts. We reasoned that the split-wrmScarlet₁₁ amino-acid sequence might be sub-optimal for complementation in human cells and synthesized a library of degenerate split-wrmScarlet₁₁ sequences covering any possible single and double amino-acid mutants. Using an established assay for CRISPR-based knock-in of sequences at the CLTA N-terminus (a highly expressed gene in HEK293T cells [Leonetti 2016] , neither our original split-wrmScarlet₁₁ sequence nor its mutant library enabled detectable complementation (Figure S10B, left panels). By contrast, a control experiment using the GFP_1-10/GFP₁₁ system showed a high level of knock-in and complementation in HEK293T (Figure S10B, right panels). It is possible that split-wrmScarlet_1-10 is expressed in a non-functional form in human cells, or that its binding to split-wrmScarlet₁₁ is occluded by competing interactions (with cellular chaperones, for example). In addition, we did not attempt complementation on primary non-transformed cell lines, like WI-38 cells, whose different proteostasis network and chaperones could aid split mScarlet folding. At this point, more experiments will be required to fully test the portability of split-wrmScarlet to mammalian systems.

Experiments to investigate whether split-wrmScarlet₁₁ functions as a degron in C. elegans

After finalizing our experiments, a paper that shows that C-terminal gly-gly sequences might function as degrons in mammalian cells was brought to our attention [Koren 2018]. Since we were unable to obtain non-sterile positive clones of TOMM-20::split-wrmScarlet₁₁ that did not have a mutation on the last glycine, and we were also unable to obtain EAT-6 homozygotes, we were concerned that our split-wrmScarlet₁₁ might be recognized as a degron. To investigate this, we first labeled HIS-3, EAT-6, and TOMM-20 with the 24 a.a. split-wrmScarlet_11(MDELYK), which adds the sequence MDELYK to the C-terminus of split-wrmScarlet₁₁. These worms were fertile, and at least as bright as those labeled with split-wrmScarlet₁₁ (Figure S6). However, increased fluorescence could be due to increased molecular brightness rather than increased abundance. To address this, we compared the abundance of nuclear HIS-3, HIS-3::split-wrmScarlet₁₁, and HIS-3::split-wrmScarlet_11(MDELYK) by western blot, and were unable to detect a significant change in abundance (Figure S11). We also could not detect differences in abundance in S. cerevisiae, using a p416-GPD plasmid expressing a mTagBFP-mScarlet fusion or the same fusion truncated so that it ends with gly-gly (Figure S12). However, because HIS-3 is a nuclear protein, and expression in yeast was done from an overexpressing plasmid, we cannot exclude that a protein ending with two glycines might be recognized as a degron in other cellular compartments, or at different expression levels, nor that there is no DesCEND degron pathway in yeast and worms. For these reasons, we recommend using the 24 a.a. split-wrmScarlet_11(MDELYK) when labeling proteins at their C-termini.

Figure S1. Split-wrmScarlet sequence comparison to mScarlet.

Figure S2. Split-wrmScarlet brightness in S. cerevisiae.

Figure S3. Developmental toxicity in worms expressing split-sfCherry3 in somatic nuclei.

Figure S4. Split-wrmScarlet and split-sfCherry3 comparison.

Figure S5. Brood size and lifespan of split-wrmScarlet_1-10 and sfGFP_1-10 lines.

Figure S6. Proteins tagged at their C-terminus with the 24 a.a. split-wrmScarlet_11(MDELYK).

Figure S7. Tissue-specific split-wrmScarlet fluorescence in the germline is undetectable when split-wrmScarlet_1-10 is integrated using a single-copy transgene via MosSCI.

Figure S8. Generation and validation of germline-specific split-wrmScarlet_1-10 and sGFP2_1-10 strains.

Figure S9. Somatic sfGFP₁₁ compared to full-length eGFP at the endogenous vha-13 locus.

Figure S10. Screen for split-wrmScarlet fluorescence in mammalian cells.

Figure S11. Split-wrmScarlet₁₁ C-terminal amino acids did not affect H2A abundance.

Figure S12. mScarlet ending with GG or MDELYK yields similar protein abundance in yeast.

Supplementary Tables

Table S1. DNA Sequences of split-wrmScarlet_1-10, split-wrmScarlet₁₁, sfGFP_1-10, sfGFP₁₁, sGFP2_1-10 and sGFP2₁₁

Table S2. C. elegans lines expressing single-copy of split-wrmScarlet_1-10 and/or sfGFP_1-10

Table S3. C. elegans strains, genotypes and sources

Table S4. crRNAs, HDR templates and oligonucleotide sequences

S4A. Sequences of crRNA and HDR template used for split-wrmScarlet₁₁ and sfGFP₁₁ knock-in experiments

S4B DNA template for split-wrmScarlet tandems HDR donor sequence - plasmids

S4C. Primers used to PCR split-wrmScarlet tandems HDR donor sequence from plasmid

S4D. Sequences of crRNA and HDR template used to generate split-wrmScarlet_1-10 and sfGFP_1-10 strains

S4E. Primers long and short

Table S5. Plasmid sequences

Table S6. Adult lifespans of strains in this study

Table S7. Mammalian cell screen oligo pool sequences