Establishment of tetracycline-regulated bimolecular fluorescence complementation assay to detect protein-protein interactions in Candida albicans

To visualize protein-protein interactions in Candida albicans with the bimolecular fluorescence complementation (BiFC) approach, we created a Tet-on system with the plasmids pWTN1 and pWTN2. Both plasmids bear a hygromycin B-resistant marker (CaHygB) that is compatible with the original Tet-on plasmid pNIM1, which carries a nourseothricin-resistant marker (CaSAT1). By using GFPmut2 and mCherry as reporters, we found that the two complementary Tet-on plasmids act synergistically in C. albicans with doxycycline in a dose-dependent manner and that expression of the fusion proteins, CaCdc11-GFPmut2 and mCherry-CaCdc10, derived from this system, is septum targeted. Furthermore, to allow detection of protein-protein interactions with the reassembly of a split fluorescent protein, we incorporated mCherry into our system. We generated pWTN1-RN and pNIM1-RC, which express the N-terminus (amino acids 1–159) and C-terminus (amino acids 160–237) of mCherry, respectively. To verify BiFC with mCherry, we created the pWTN1-CDC42-RN (or pWTN1-RN-CDC42) and pNIM1-RC-RDI1 plasmids. C. albicans cells containing these plasmids treated with doxycycline co-expressed the N- and C-terminal fragments of mCherry either N-terminally or C-terminally fused with CaCdc42 and CaRdi1, respectively, and the CaCdc42-CaRdi1 interaction reconstituted a functional form of mCherry. The establishment of this Tet-on-based BiFC system in C. albicans should facilitate the exploration of protein-protein interactions under a variety of conditions.


Construction of Tet-on Plasmids
We aimed to set up a Tet-on-based BiFC assay; thus, we developed an additional Tet-on system with hygromycin B resistance in C. albicans following the construction of pNIM1 1 . To efficiently utilize the pNIM1 vector described in our previous study 2 , a pair of primers, CaGFP-N-MCS and CaGFP-C-MCS, was designed to incorporate multiple cloning sites to the 5' and 3' ends of the GFPmut2-coding sequence by PCR, followed by digesting with SalI/BclI and cloning back into SalI/BglII-linearized pNIM1 to obtain pNIM1-NC ( Figure S1). To construct the additional pWTN2 plasmid for red fluorescent validation, the assembly was divided into three steps. (1) Construction of pBSII-TDH3p-rtTA-ACT1t. The pWTN2 plasmid was designed for gene targeting at the CaTDH3 locus due to its promoter, which is capable of constitutively driving reverse tetracycline-controlled transactivator (rtTA) expression. This promoter was amplified with the TDH3 SacII F and TDH3 XbaI R primers from genomic wild-type DNA and subsequently cloned into pBluescriptII SK(+) (pBSII) digested with SacII/XbaI to create pBSII-CaTDH3p. The rtTA coding sequence was coupled to the CaACT1 transcription terminator released from pNIM1, subcloned into pBSII-TDH3p and then digested with XbaI/PstI to produce pBSII-TDH3p-rtTA-ACT1t. (2) Construction of pBSII-SFS-HygB.
To create a backbone vector with hygromycin B resistance for the assembly of the components in the tetracycline-regulated system, the pSFS2A plasmid 3 was digested with SalI to release a part of the DNA fragment containing the SpeI site within the CaSAT1 gene and ligated into itself to obtain pSFS2A-SpeI with one remaining SpeI site. To replace the backbone of pBSII, the DNA cassette of pSFS2A-SpeI between two FRT sequences was digested with KpnI/SacI and subcloned into pBSII to generate pBSII-SFS2A-SpeI. To incorporate more suitable restriction enzyme sites flanking the FRT sequences, two DNA duplexes formed by two pairs of primers, MCS3 F/MCS3 R and MCS4 F/MCS4 R, were sequentially cloned into pBSII-SFS2A-SpeI via SpeI/SacI and then via KpnI/XhoI digestion to become pBSII-SFS-MCS4. The synthetic hygromycin B gene (CaHygB) was amplified with the HygB BglII F and HygB NsiI R primers using pYM70 4 as a DNA template, cut out with BglII/NsiI, and cloned into pBSII-SFS-MCS4 digested with BamHI/PstI to release a part of the SAT1-flipper to generate pBSII-SFS-HygB. (3) Construction of pCR2-TDH3t-TetO7-miniOP4-GFPt. To construct seven copies of the tetracycline operator (tetO), a unit of the tetO adaptor formed by annealing with a pair of primers, TetO F and TetO R, was cloned into pBSII-HA-2xStrepII (our unpublished data) digested with PmeI/SacI to produce pBSII-TetO1. Then, another adaptor was digested with PmeI/SalI and cloned into pBSII-TetO1 digested with PmeI/XhoI to generate pBSII-TetO2. This step was repeated five times to construct pBSII-TetO7. We used the multiple cloning site (MCS) of pCR2.0 for plasmid building by digesting TetO7 with KpnI/SacI and subcloning into pCR2.0 to generate pCR2-TetO7. To fuse the minimal CaOP4 promoter (miniOP4) to TetO7, the promoter was PCR-amplified with the miniOP4 SpeI F and miniOP4 XhoI R primers using pNIM1 as a DNA template; the construct was then digested with XhoI/SpeI and cloned into pCR2-TetO7 digested with SalI/SpeI to become pCR2-TetO7-miniOP4. A DNA fragment homologous to CaTDH3 for gene targeting was amplified with the TDH3t PmeI F2 and TDH3t KpnI R2 primers, digested with PmeI/KpnI, and cloned into pCR2-TetO7-miniOP4 to generate pCR2-TDH3t-TetO7-miniOP4. Similar to pNIM1, a GFPmut2-coding sequence linked to the CaACT1 transcriptional terminator was amplified with the GFPt XbaI F and GFPt SpeI R primers from the pNIM1-NC template, digested with XbaI/SpeI, and cloned into pCR2-TDH3t-TetO7-miniOP4 to produce pCR2-TDH3t-TetO7-miniOP4-GFPt. To assemble the pWTN1 plasmid, a prototype of pWTN2, from these three constructs, the DNA fragment comprising TDH3p-rtTA-ACT1t was released from the pBSII-TDH3p-rtTA-ACT1t plasmid via PstI/SacII digestion and was cloned into pBSII-SFS-HygB linearized with PstI/SacII to create pBSII-TDH3p-rtTA-HygB; then, another DNA fragment containing TDH3t-TetO7-miniOP4-GFPt was released from pCR2-TDH3t-TetO7-miniOP4-GFPt via KpnI/AgeI digestion and cloned into pBSII-TDH3p-rtTA-HygB to generate pWTN1 ( Figure   S1). To replace the GFPmut2-coding sequence of pWTN1with yEmRFP (yeast enhanced red fluorescent protein, as well as mCherry) 5 , which is a codon-optimized RFP for C. albicans, a pair of primers, mRFP NotI F and mRFP AatII R and pFA-yEmRFP-CmLEU2-2µ (our unpublished data), was used as a DNA template to amplify a coding sequence of yEmRFP; the construct was digested with NotI/AatII and cloned into pWTN1 to generate pWTN2 ( Figure   S1). For fluorescent labelling of septin, the coding sequences of CaCdc10 and CaCdc11 were digested with NheI/XhoI and AatII/BglII from pTET25M-CDC11-GFP and pTET25M-RFP-CDC10 (our unpublished data) and cloned into pNIM1-NC and pWTN2, respectively, to generate pNIM1-CDC11-GFPmut2 and pWTN2-mCherry-CDC10 ( Figure   S2).

Design of the BiFC assay
To fit the Tet-on systems for the BiFC assay, mCherry was used as a reporter in this study.
Split-mCherry, which is capable of reassembling, was divided into an N-terminal fragment containing amino acids 1-159 and a C-terminal fragment containing amino acids 160-237 6 .
The triple repeats of the GGGGS amino acid sequence were used to increase structural flexibility between a target gene and fragments of the reporter. To create DNA segments encoding the GGGGS amino acid sequence, we designed a pair of primers, GS F and GS R, to form a single unit of GS adaptor by annealing to each other and subsequently ligating into pBSII digested with EcoRV/BamHI to become pBSII-GS1. Then, two additional GS adaptors were sequentially cloned into pBSII-GS1 digested with EcoRV/BglII to obtain pBSII-GS3, which was used as an intermediate vector for fusion with the targets and split fragments in the Tet-on system. To construct pWTN1-6H-MCS-RN, which is capable of expressing the C-terminal end of the target protein with a polyhistidine tag (6H) fused to the N-terminal fragment of mCherry (RN), both the DNA sequences encoding 6H and RN were incorporated into pWTN1. The DNA adaptor encoding six histidine residues made by a pair of primers, WTN1-His F and WTN1-His R, was cloned into pWTN1 digested with SpeI/EcoRV to obtain pWTN1-6H. Subsequently, pWTN1-6H was digested with KpnI/BglII and ligated into KpnI/BglII-digested pBSII-GS3 to become pBSII-WTN1-6H-GS3. The DNA fragment encoding the N-terminal fragment of mCherry was PCR-amplified with primers, mRFP N BamHI F and mRFP N NotI R, and with pWTN2 as the DNA template and HiFi DNA polymerase (Kapa Biosystems), the fragment was then cloned into BamHI/NotI-digested pBSII to obtain pBSII-RN. Next, the WTN1-6H-GS3 DNA fragment was released from pBSII-WTN1-GS3 by digesting with KpnI/BamHI and was cloned into pBSII-RN to create pBSII-WTN1-6H-GS3-RN. Finally, the WTN1-6H-GS3-RN DNA fragment was released from this plasmid digested with KpnI/StuI and was cloned into pWTN1 to generate pWTN1-6H-MCS-RN. To create a vector capable of expressing the N-terminal fragment of mCherry to use as a control, this encoding fragment was amplified with a pair of primers, RedN KpnI F and mRFP N NotI R, following SpeI/StuI digestion and cloned into pWTN1 to generate pWTN1-RN. To construct another vector, pNIM1-RC-MCS, which is capable of producing the N-terminal end of the target protein fused to the C-terminal fragment of mCherry (RC), the pNIM1-NC was used and modified by a similar strategy described above.
The DNA fragment released from pNIM1 by KpnI/SalI digestion was cloned into KpnI/SalI-digested pBSII-GS3 to obtain pBSII-NIM1-GS3. The DNA encoding the C-terminal fragment of mCherry was PCR-amplified with the primers using pWTN2 as a DNA template; after XhoI/BamHI digestion, the fragment was cloned into SalI/BglII-digested pBSII-NIM1-GS3 to produce pBSII-NIM1-RC-GS3. The NIM1-RC-GS3 DNA fragment was To generate a vector capable of expressing the C-terminal fragment of mCherry as a control, this encoding fragment was amplified and produced with a pair of primers, mRFP C XhoI F and mRFP C BamHI R, and with pWTN2 as a template; after XhoI/BamHI digestion, the fragment was cloned into SalI/BglII-digested pNIM1 to become pNIM1-RC. Then, the coding sequences of CaCDC42 and CaRDI1 were amplified with the CDC42 EcoRV F and CDC42 BglII R, and RDI1 SpeI F and RDI1 StuI R primers, respectively, and cloned into pWTN1-6H-MCS-RN and pNIM1-RC-MCS to produce pWTN1-6H-CDC42-RN and pNIM1-RC-RDI1 ( Figure S3). To construct a vector, pWTN1-RN-CDC42 ( Figure S3), capable of expressing the N-terminal mCherry fusion to CDC42, a PCR-amplified DNA fragment encoding amino acids 1-159 of the mCherry protein by a pair of primers, RedN KpnI F and RedN EcoRV R, was cloned into the pBSII-GS3 plasmid linearized with KpnI and EcoRV to obtain pBSII-RN-GS3. Next, a PCR-amplified coding fragment of CDC42 by a pair of primers, CaCDC42 BamHI F and CaCDC42 NotI R, was cloned into pBSII-RN-GS3 linearized with BamHI and NotI to obtain pBSII-RN-GS3-CDC42. Finally, the DNA fragment containing the RN-GS3-CDC42 region was released by SpeI/StuI and cloned into pWTN2 digested with SpeI/StuI to obtain pWTN1-RN-CDC42 ( Figure S3). Figure S1. The maps of pNIM1-NC, pWTN1, and pWTN2 used in this study.     Table S1. Statistical analyses proceeded with one-way ANOVA with Bonferroni's comparison test listed in Table S2. Figure S5. The maps of pNIM1-CDC11-GFP and pWTN2-mCherry-CDC10.

Supplementary figure legends
The coding sequence of CaCDC11 was cloned into pNIM1-NC with XhoI and NheI, and the coding sequence of CaCDC10 was cloned into pWTN2 with AatII and BglII.    Table S3.