Candida albicans is one of the most important fungal pathogens of humans. It causes systemic candidiasis, which is life-threatening for patients with deficient immunity^[1]. Its pathogenicity is tightly correlated with its virulence and host-pathogen interactions. To study the molecular mechanisms of its pathogenicity, it is very important to construct single or multiple genes deletion mutants for exploring their functional roles.

In recent years, many molecular tools for gene deletion and marker recycling were reported for Candida albicans^[2-4]. URA3 blaster method was the first and important one for gene deletion in Candida albicans^[5]. This method has been applied and adapted in many published papers and genetic toolkits for Candida albicans^[6-9]. However, the URA3 marker was reported to have a negative effect on its hyphal development and virulence^[10-11]. Its flanking repeats of hisG may also affect the expression of URA3 and subsequent multiple gene disruptions^[12-13]. In addition, 5-FOA which was used for URA3 excision may induce chromosome alterations in Candida albicans^[14], complicating the phenotypes and genotypes of mutant strains. To avoid the use of URA3, new reference strains (SN serial strains) and heterologous auxotrophic markers (i.e. Candida maltosa LEU2, Candida dubliniensis HIS1 and Candida dubliniensis ARG4) were applied and preferred in this field^[15].

Marker recycling is important for multiple gene deletions. In C. albicans, FLP-FRT system was adapted to excise URA3 or SAT1 selection marker, which was called URA flipper or SAT1 flipper ^[16-17]. The Cre-loxP system was also applied with traditional selection markers and an inducible promoter MET3p for marker loop-out, called Cre-loxP flipper and Clox flipper^{[6, 18]}. The famous genetic manipulation tool, CRISPR/Cas9 system has also been developed for C. albicans^[19]. Based on CRISPR, an efficient gene deletion and marker recycling system was constructed^[20-21].

In this study, we reported a fast, convenient and cost-effective strategy for gene deletion and marker recycling in C. albicans. This system uses loxP-flanked markers with short homology regions to delete genes. Cre was controlled by the Tet-on promoter which could be effectively induced to express by doxycycline (Dox, a derivative of tetracycline). The Tet-on promoter is widely used in mammalian cells and was also introduced into the Candida field for tetracycline-inducible gene expression^[22]. After induction, all selection markers will be excised and be ready for next genetic manipulation.

1 Materials and Methods 1.1 Strains and growth conditions

Candida albicans strains used in this study are listed in Table 1. SN152 was kindly provided by Suzanne M. Noble^[15]. The other strains were constructed following the illustration of Figure 3. Simply, the loxP-CmLEU2-loxP cassette was amplified from pCPC48 with primers which were designed following the introduction of Figure 1-C. This cassette was introduced into SN152 to delete one copy of the target gene. loxP-CdHIS1-loxP was amplified from pCPC49 with the same primers and was introduced into the heterozygote to delete the second copy of the target gene, generating the homozygous null mutant. Then HACreH cassette was amplified from pCPC51 and was introduced into the mutant to perform marker excision by Dox induction. After excision, only one loxP site was left at the original locus for each copy.

All Candida strains were grown in YPD (1% yeast extract, 2% peptone, 2% glucose) or SC (synthetic complete, 0.67% yeast nitrogen base with ammonium sulfate, 2% glucose, appropriate amino acid mix) medium. Doxycycline hydrochloride was bought from Sangon Biotech (China) and was added at a final concentration of 50 μg/mL. The Escherichia coli strain DH5α was used for molecular cloning and was selected in LB medium supplemented with 100 μg/mL of ampicillin or 50 μg/mL of kanamycin.

1.2 Plasmid construction and PCR conditions

pSN40, pSN52 and pSN69, which contains CmLEU2, CdHIS1 and CdARG4 respectively, were kindly provided by Suzanne M. Noble^[15]. CmLEU2, CdHIS1 and CdARG4 were amplified from pSN40/52/69 using loxP-F and loxP-R primers (Table 2, loxP sites was shown in bold and italic). The three loxP-flanked markers were then introduced into the backbone fragment amplified from pUC19 (TransGen, Beijing) using primers pUC18+762s and pUC18+2629a, via Exo Ⅲ-mediated ligation-independent cloning (LIC)^[23], generating pCPC48, pCPC49, and pCPC50 respectively.

The rTetR (reverse tetracycline repressor) element was chemically synthesized by QinlanBio, China. It was codon-optimized using OPTIMIZER^[24] according to the codon usage table of C. albicans (http://www.kazusa.or.jp/codon/), encoding the same protein of Tet-on 3G of pCMV-Tet3G Vector of ClonTech, TaKaRa, Japan. The rTetR element was amplified by primers TetR-F and TetR-R. The GAL4AD element was amplified by primers GAL4AD-F and cartTA-R. These two elements formed a fusion protein called cartTA (Candida albicans reverse tetracycline-controlled trans-activator). The tetracycline responsive element P_tet was amplified using primers Ptet-F and OP4-R. The P_tet element and GAL4AD were all amplified from pCaUme6-3^[25]. The caCre gene was amplified from pCAD^[6] using primers Cre-F and Cre-R. The ADH1 promoter was amplified using primers ADH1-735s and ADH1p-R. The ACT1 terminator was amplified using primers Tact1+1765s and Tact1+2070a. The HIS1 marker was amplified using primers HIS1-446s and HIS1+1149a. The ARG4 marker was amplified using primers ARG4-331s and ARG4+1635a. The ADH1 promoter, the ACT1 terminator, HIS1 and ARG4 were all amplified from genomic DNA of the wildtype Candida albicans strain SC5314. Finally, the ADH1 promoter, cartTA, the ACT1 terminator, ARG4, P_tet, and CaCre were sequentially assembled with the backbone amplified from pSN52 using primers CdHIS1-F and CdHIS1-R via LIC, generating pCPC51. Similarly, the ADH1 promoter, cartTA, the ACT1 terminator, HIS1, P_tet and CaCre were sequentially assembled with the backbone amplified from pSN69 using primers CdARG4-Fand CdARG4-R via LIC, generating pCPC52.

All PCR amplification used the KOD-plus enzyme (Toyobo, Japan) following the manufacturer's instruction. The integrity of all DNA cassettes and correct cloning of all constructs were verified by sequencing (MajorBio, Shanghai).

The complete reference sequences of these plasmids are available with GenBank Accession numbers: pCPC48 (MG874799), pCPC49 (MG874800), pCPC50 (MG874801), pCPC51 (MG874802), pCPC52 (MG874803).

1.3 Primers

Primers used in this study were all synthesized by Sangon, Biotech (China) with HAP purification and are listed in Table 2. loxP sites were shown in italic and bold.

1.4 C. albicans transformation and mutant validation

C. albicans are directly transformed with PCR products containing loxP-marker-loxP with 78 bp flanking homology region from target genes without purification. The transformation was performed by following an improved protocol^[26]. All mutants were validated via diagnostic PCR using gene-specific primers and universal primers VP9 & VP10 for CmLEU2, VP11 & VP12 for CdHIS1, VP13 & VP14 for CdARG4 (Table 2).

2 Results 2.1 Components and strategy for marker cassettes amplification

Our goal was to create gene deletion cassettes with short homology regions fast and cost-effectively. Therefore, we chose heterologous auxotrophic markers which could reduce the mistargeting effect of homologous markers^[15]. Besides, marker recycling was also required. We chose the Cre-loxP system for its high efficiency^[6].

As shown in Figure 1-A, CmLEU2, CdHIS1 and CdARG4 were amplified from pSN40, pSN52 and pSN69 respectively with primers containing 34 nt loxP sites and were then inserted into the same backbone of pUC19 via Exo Ⅲ-mediated cloning (LIC), generating ready-to-use loxP-marker-loxP cassettes. This DNA assembly method is convenient and efficient because only 15 bp homology is necessary for sequential multiple fragment assembling. The backbone of pUC19 here contained only the replication origin and ampicillin resistance gene. Since the three markers were introduced into the same backbone, universal primers could be used for marker cassette amplification.

These cassettes could be amplified using F1 and R1 primers in the Round 1 PCR. Because F1/R1 primer sets comprised a common binding region (20 nt, Figure 1-C) and target gene-specific region (39 nt), the PCR #1 product was flanked by 39 bp homology regions. F2/R2 primers sets comprised a 20 nt region homologous to F1/R1 and additional 39 nt homology regions. PCR #1 product could be used as templates for PCR #2 directly and generated final marker cassettes flanked by 78 bp homology regions (Figure 1-B). This successive PCR strategy avoids the need for long primer (> 60 nt) synthesis and generates enough PCR products for transformation directly.

For marker integration verification, we predesigned universal verification primers, i.e. VP9 & 10 were used for CmLEU2 integration verification, VP11 & 12 for CdHIS1 and VP13 & 14 for CdARG4 (Figure 1-A). Diagnostic PCR could be easily performed using gene-specific primers and these validation primers.

2.2 Components and concepts for marker recycling

To excise markers in the cells efficiently, Cre must be controlled by an inducible promoter. MET3p promoter was used in previous reports^[6]. However, the special medium was necessary to induce or inhibit the expression of Cre. To simplify this system, we chose tet-on promoter to control Cre expression since it could control gene expression efficiently and tightly in C. albicans^[22].

The Tet-on 3G promoter system is recommended by the Clontech company since it has lower background activity and higher sensitivity. To adopt this, we performed codon optimization of the core element of Tet-on 3G promoter, i.e. rTetR (reverse tetracycline repressor), according to the codon usage database (http://www.kazusa.or.jp/codon/), obtaining the carTetR element. Then we fused it to the codon-optimized GAL4AD element amplified from pCaUme6-3^[25], generating the fusion gene cartTA (Candida albicans reverse tet transactivator). The constitutive promoter, ADH1p, was used to control the expression of this fusion gene. The ACT1 terminator was placed downstream of the fusion gene to stop translation efficiently. Classical selection marker, ARG4, was placed upstream of the tetracycline-responsive promoter P_tet, which was also amplified from pCaUme6-3. The codon-optimized recombinase caCre was amplified from pCAD^[6] and placed downstream of the P_tet promoter. Then these fragments were introduced into the coding region of pSN52 via one-step reaction of LIC, generating pCPC51 (Figure 2-A). Similarly, HIS1 was used to replace ARG4 and these fragments were introduced into the coding region of pSN69 via LIC, generating pCPC52 (Figure 2-A).

Since the ADH1 promoter is constitutive in C. albicans, the fusion protein cartTA is expressed constitutively. In the absence of Dox, the fusion protein could not bind to the P_tet region, resulting in the silence of the recombinase Cre. In the presence of Dox, the fusion protein could interact with Dox and then binds to the P_tet region steadily, sustaining a very high expression of the recombinase Cre. Once the recombinase Cre was efficiently expressed, a recombination event can occur between the loxP sites, leaving one copy of loxP site.

For convenience, the core cassette of pCPC51, i.e. 5′ CdHIS1-P_ADH1-cartTA-T_ACT1-ARG4-P_tet- caCre-3′ CdHIS1, was named HACreH. Likewise, the core cassette of pCPC52, 5′ CdARG4-P_ADH1-cartTA- T_ACT1-HIS1-P_tet-caCre-3′ CdARG4, was named AHCreA. Predesigned primer sets HACreH-F/R (for pCPC51) or AHCreA-F/R (for pCPC52) could amplify the HACreH or AHCreA cassettes efficiently and the PCR products are ready-to-use for marker recycling.

When markers are required to be excised in a gene deletion mutant (e.g. CmLEU2 and CdHIS1 mutant), HACreH cassette was introduced into mutant cells via transformation and CdHIS1 would be then disrupted. Subsequently, transformants were treated with Dox to induce Cre expression. Finally, Cre could catalyze site-specific recombination between loxP sites, resulting in marker excision and only one loxP site left in the target gene locus (Figure 2-B).

2.3 A detailed protocol for this approach

Here we describe a detailed protocol for target gene deletion and marker excisions (Figure 3). This protocol was used many times in practice and generated sufficient positive transformants each time. The first step, to delete a gene, loxP-CmLEU2-loxP, and loxP-CdHIS1-loxP cassettes with 78 bp homology regions are amplified via two rounds of PCR. PCR products are transformed into competent cells directly via two rounds of transformation to generate a gene deletion mutant with the genotype of LEU2+, HIS1+, which could grow on SC medium without leucine and histidine. The second step, to excise all markers, PCR amplified HACreH cassette is transformed into this mutant to generate a mutant with the genotype of LEU2+, ARG4+, his1-, which could grow on SC medium without leucine and arginine. Transformants are selected on SC plates without arginine. Pick out 10 transformants randomly, pool them in 1 mL YPD containing 50 μg/mL Dox and incubate at 30 ℃ for at least 8 hours with gentle shaking. Then streak the mixture on 1–2 YPD plates and incubate at 30 ℃ for single colonies. Pick out 20–40 single colonies randomly and replicate them on YPD and SC-Leu/His/Arg plates. The clones will only grow on YPD plates are positive ones. According to our practice, 60%–90% single colonies are positive clones.

2.4 Gene deletions using the Tet-on promoter controlled Cre/loxP system

To validate the functionality of the Tet-on promoter controlled Cre/loxP system, we used this system to delete several genes that play roles in the hyphal development of C. albicans in SN152 strain^[15]. SET3 and SIF2 are two genes that coding components of histone deacetylase complex Set3C. Null mutants of these two genes were reported to have a phenotype of increased hyphal growth. By using CmLEU2-loxP and CdHIS1-loxP cassettes, we deleted SET3 gene in C. albicans (i.e. CPS52) and performed marker excision, generating CPS56. SIF2 gene was deleted by markers without loxP sites (i.e. CPS18). Then we checked the phenotypes of the two null mutants. set3 and sif2 mutants (CPS56 and CPS18) grew wrinkled colonies on YPD plates at 37 ℃ (Figure 4-A). In liquid YPD, these two mutants grew true hyphae at 37 ℃ while wild-type cells grew pseudohyphae (Figure 4-B). Under YPD, 25 ℃ condition, these two null mutants showed no differences between them and wildtype cells (Figure 4-C). This phenotype was insistent with previous reports^[27].

In addition, we also constructed another two null mutants and performed marker excision, i.e. flo8 (CPS146 and CPS147) and gcn5 (CPS50 and CPS55). Flo8 was essential for hyphal development and Gcn5 played important roles in hyphal elongation^[28-29]. Under hyphal induction condition (YPD+10% serum, 37 ℃), flo8 mutant failed to initiate hyphal growth and gcn5 mutant failed to form long hypha (Figure 4-D). These results showed that target genes could be deleted efficiently by using these loxP-flanked marker cassettes.

3 Discussion

C. albicans is an important fungal pathogen of humans and it is very important to study the molecular mechanisms of its pathogenicity. Identifying and characterizing relevant gene functions will improve our understanding about this pathogen. So multiple markers and marker recycling tools are required to construct multiple genes deletion mutants.

In recent twenty years, many genetic tools are developed for C. albicans. Many of them were based on URA3 blaster method. As the URA3 marker and 5-FOA were found to have disadvantages, URA3-independent strains and markers were preferable. For marker recycling, FLP/FRT system, Cre/loxP system and CRISPR/Cas9 system were all adapted in C. albicans.

In this study, we have built a system using the Tet-on promoter controlled Cre/loxP system to delete genes and excise markers. Three heterologous markers were flanked with loxP sites and 78 bp gene specific homology region could be generated after two rounds of conventional PCR (i.e. successive PCR). Heterologous auxotrophic markers, instead of conventional homologous markers, is used in this system to increase the gene targeting efficiency and to avoid the impact of URA3 on hyphal development and virulence of C. albicans.

There are several benefits of this system in comparison to previously established ones. Firstly, Tet-on promoter, instead of MET3p, is used to control the expression of Cre, which means there is no need to change special medium for induction. Secondly, considering the price differences between common primers (< 60 nt) and long primers (> 60 nt), this method is significant cost-effective. In addition, this method needs no further purification steps of PCR products, which could be used in transformation directly. This successive PCR strategy is more reliable and robust than fusion PCR and yields enough positive transformants because 78 bp homology regions are sufficient for homologous genomic integration in C. albicans.

For marker recycling, we constructed two P_Tet-on-Cre cassettes targeting CdHIS1 or CdARG4. To induce Cre expression, we only need adding Dox to the common growth medium. There is no need for primer design and special medium preparation. Besides, low-grade doxycycline which is very cheap is sufficient to generate enough positive clones. The efficiency of marker recycling would increase if change YPD to SC-Arg or SC-His medium for Dox induction since untransformed cells will not grow in the latter medium.

Using this system, we have constructed many gene deletion mutants efficiently. In summary, this system described in this study provides a simple, fast, convenient and cost-effective way to construct gene deletion mutant and excise markers in C. albicans.

Acknowledgments:

We thank Suzanne M. Noble for providing SN serial strains and pSN40/52/69. Also special thanks to Prof. Alistair J.P. Brown for providing pCAD.