Shuttle vectors for facile gap repair cloning and integration into a neutral locus in Candida albicans

Abstract

Candida albicans is the most prevalent fungal pathogen of humans. The current techniques used to construct C. albicans strains require integration of exogenous DNA at ectopic locations, which can exert position effects on gene expression that can confound the interpretation of data from critical experiments such as virulence assays. We have identified a large intergenic region, NEUT5L, which facilitates the integration and expression of ectopic genes. To construct and integrate inserts into this novel locus, we re-engineered yeast/bacterial shuttle vectors by incorporating 550 bp of homology to NEUT5L. These vectors allow rapid, facile cloning through in vivo recombination (gap repair) in Saccharomyces cerevisiae and efficient integration of the construct into the NEUT5L locus. Other useful features of these vectors include a choice of three selectable markers (URA3, the recyclable URA3-dpl200 or NAT1), and rare restriction enzyme recognition sites for releasing the insert from the vector prior to transformation into C. albicans, thereby reducing the insert size and preventing integration of non-C. albicans DNA. Importantly, unlike the commonly used RPS1/RP10 locus, integration at NEUT5L has no negative effect on growth rates and allows native-locus expression levels, making it an ideal genomic locus for the integration of exogenous DNA in C. albicans.

Introduction

Candida albicans, a commensal of most individuals, is also one of the most prevalent fungal pathogens of humans. Most individuals are colonized with C. albicans by approximately 6 weeks of age. As a commensal, C. albicans inhabits a range of niches within the human body, including the skin, the oral-gastrointestinal tract and the uro-genital tracts. Normally, C. albicans causes non-lethal, superficial mucosal infections. However, conditions of immunosuppression, prolonged antibiotic treatment, or the use of invasive catheters, among other risk factors, can allow C. albicans to reach the bloodstream, leading to systemic infections that are associated with high mortality rates (values range from ~10–15 % to ~40 % mortality due to candidaemia in the USA) (Falagas et al., 2006; Gudlaugsson et al., 2003; Zaoutis et al., 2005). Thus, current antifungal therapies are not sufficiently effective and there is a great need to better understand the biology and pathogenicity of C. albicans.

C. albicans research routinely requires in vivo studies, and it is important to ensure that mutant strains are neither auxotrophic nor haploinsufficient for any genes required for virulence. For example, strains that express only an ectopic copy of URA3, a commonly used auxotrophic marker, can exhibit reduced virulence, a fact that has confounded the interpretation of experiments using this marker (Brand et al., 2004; Cheng et al., 2003; Lay et al., 1998; Staab & Sundstrom, 2003). Because C. albicans lacks a conventional meiotic sexual cycle and has a diploid genome, constructing strains with disrupted or replaced genes relies upon transformation with exogenous DNA. Furthermore, since neither an RNA interference (RNAi) system, nor stable extrachromosomal plasmids with useful markers are efficient for use in C. albicans (Huber & Rustchenko, 2001; Kurtz et al., 1987), genetic alterations must be done by integration of DNA into the fungal genome. Thus, major approaches to strain construction utilize DNA transformation by homologous recombination of PCR-amplified inserts or of plasmid DNA.

The number and variety of molecular tools available for genetic manipulation of C. albicans have increased greatly in the past decade with the development of auxotrophic and drug-resistance markers, epitope tags, fluorescent tags and constitutive or regulatable promoters (Backen et al., 2000; Barelle et al., 2004; Basso et al., 2010; Beckerman et al., 2001; Care et al., 1999; Doyle et al., 2006; Enloe et al., 2000; Finkel et al., 2011; Gerami-Nejad et al., 2001, 2009, 2004, 2012; Gola et al., 2003; Keppler-Ross et al., 2008; Lai et al., 2011; Lebel et al., 2006; Murad et al., 2000; Nobile & Mitchell, 2009; Oh et al., 2010; Park & Choi, 2002; Reijnst et al., 2011; Reuß et al., 2004; Sánchez-Martínez & Pérez-Martín, 2002; Shen et al., 2005; Srikantha et al., 1996; Staab et al., 2003; Stynen et al., 2010; Umeyama et al., 2002; Vieira et al., 2010; Wilson et al., 2000; Xu et al., 2011). Available vectors for the integration of genes into ectopic locations in the C. albicans genome target the HIS1 promoter (orf19.4026), the ENO1 promoter (orf19.935) or the widely used RPS1 ORF (orf19.3002, also referred to as RPS10 or RP10) (Care et al., 1999; Davis et al., 2002; Dennison et al., 2005; Murad et al., 2000; Spreghini et al., 2003; Staab et al., 2003; Vieira et al., 2010). However, there are several important limitations to these approaches. First, many of these vectors rely on markers useful only in auxotrophic strains (e.g. his1Δ/Δ or ura3Δ/Δ strains) and not in prototrophic clinical isolates. Second, gene expression at some genomic loci is subject to position effects that alter the level of expression relative to what is seen at the native locus (Pirrotta & Gross, 2005; Tartof, 1994). Third, insertion at RPS1 disrupts the ORF encoding the Rps1 ribosomal protein resulting in hemizygosity of RPS1, which has been reported to cause growth defects (Oh et al., 2010) (and see below). Thus, it would be ideal to have a ‘neutral’ locus with no coding or other known function and no effect on growth rates, for use as a conventional site for insertion of exogenous DNA in C. albicans.

Here, we describe the identification of a neutral locus, NEUT5L, which facilitates strain construction by allowing for high-frequency integration of exogenous DNA and efficient expression of the integrated constructs. To facilitate the cloning and integration at this locus, we constructed a series of integrative shuttle vectors that exploit homologous recombination in vivo (gap repair) in Saccharomyces cerevisiae and carry NAT1, the dominant drug resistance marker that can be used in clinical isolates, a URA3 counterselectable auxotrophic marker, or a recyclable URA3 auxotrophic marker (URA3-dpl200) (Wilson et al., 2000). Cloning into these vectors can also be achieved by classical restriction enzyme digestion and ligation. Furthermore, convenient restriction enzyme sites have been engineered such that most of the non-C. albicans vector sequence is excluded from the integrative fragment prior to transformation (this includes the Escherichia coli and S. cerevisiae selectable markers and the S. cerevisiae centromeric and autonomous replication sequences). Similar to shuttle vectors that target the RPS1 locus (Vieira et al., 2010), fragments cloned into these NEUT5L shuttle vectors can be PCR-amplified directly from S. cerevisiae DNA extracts, thereby obviating the amplification step in E. coli, which may be difficult for large plasmids and/or toxic inserts. Importantly, expression levels of genes inserted at NEUT5L are similar to levels of expression from the native locus, and do not appear to be subject to any position effects. Furthermore, no fitness cost is incurred when exogenous DNA is integrated at this locus. Therefore, we have identified NEUT5L, a neutral locus in the C. albicans genome, for the integration and expression of ectopic DNA, and we have generated shuttle vectors containing convenient selectable markers for facile and rapid cloning and integration into NEUT5L.

Methods

Strains, media and growth conditions.

C. albicans was routinely grown at 30 °C in rich YPAD medium (1 % yeast extract, 2 % peptone, 2 % glucose). Transformants of S. cerevisiae were selected in synthetic complete medium (SC) [0.17 % yeast nitrogen base without ammonium sulfate (Q-BioGene), 0.5 % ammonium sulfate, 2 % glucose, and supplemented with a dropout mix containing amino and nucleic acids except tryptophan or uracil, depending on the auxotrophic requirement for the selection (Adams & Stearns, 1997)]. All media were supplemented with uridine (80 µg ml⁻¹) except when used for selection of ura⁻ strains. For solid media, 2 % Bactoagar was used.

All C. albicans strains derive from strain SC5314 (YJB11562), and are listed in Table 1. All C. albicans transformations were performed following the LiAc protocol (Wilson et al., 1999). nat+ transformants were selected by plating the transformation mix on YPAD and replica plating the following day onto YPAD plates supplemented with 400 µg nourseothricin ml⁻¹ (Werner Bioagents).

Table 1. C. albicans and S. cerevisiae strains used in this study.

Strain (YJB)	Parent	Genotype	Reference
11562 (SC5314)		Wild-type (clinical isolate)	Kurtz et al. (1986)
3018 (CAI4)	SC5314	ura3 : : λimm434/ura3 : : λimm434	Fonzi & Irwin (1993)
1873 (CAF2-1)	SC5314	ura3 : : λimm434/ura3 : : λimm434	Fonzi & Irwin (1993)
12780	SC5314	RPS1/rps1 : : ENO1–GFP–NAT1	This study
11372	SC5314	ENO1/ENO1–GFP–NAT1	This study
12257	SC5314	NEUT5L/neut5l : : ENO1–GFP–NAT1	This study
12982 (YAG134)	CAI4	ura3 : : λimm434/ura3 : : λimm434 rbp1 : : MX3/RBP1	Cruz et al. (2001)
12543 (DAY1230)	CAI4	ura3 : : λimm434/ura3 : : λimm434 rbp1 : : MX3/rbp1 : : MX3	Zacchi et al. (2010)
12563	12543	ura3 : : λimm434/ura3 : : λimm434 rbp1 : : MX3/rbp1 : : MX3 NEUT5L/NEUT5L–NAT1–RBP1–NEUT5L	This study
12565	12543	ura3 : : λimm434/ura3 : : λimm434 rbp1 : : MX3/rbp1 : : MX3 NEUT5L/neut5l : : RBP1-NAT1	This study
12566	12543	ura3 : : λimm434/ura3 : : λimm434 rbp1 : : MX3/rbp1 : : MX3 NEUT5L/neut5l : : RBP1G62D-NAT1	This study
12984	CAI4	ura3 : : λimm434/ura3 : : λimm434 NEUT5L/NEUT5L–URA3–NEUT5L	This study
12985	CAI4	ura3 : : λimm434/ura3 : : λimm434 NEUT5L/neut5l : : URA3	This study
12046		MATα his3Δ200 trp1-901 leu2-3,-112 ade2 LYS2 : : (lexAop)4-HIS3 URA3 : : (lexAop)8-lacZ GAL4	Vojtek et al. (1993)

Strains YJB11372, YJB12780 and YJB12257 containing an ENO1–GFP–NAT1 cassette at ENO1 (orf19.395), at RPS1 (orf19.3002) or at NEUT5L, respectively, were constructed by transforming C. albicans strain SC5314 as follows. Strain YJB11372 was constructed by C-terminally tagging Eno1 with a GFP–NAT1 cassette PCR amplified from plasmid pMG2120 with primers 714 and 3185 (Tables 2 and 3), which contain 70 bp of homology to the ENO1 3′ region. The correct integration of the GFP–NAT1 cassette was verified by PCR of genomic DNA using the primers 716 and 658 (Table 2). Strains YJB12257 and YJB12780 were constructed by transforming the ENO1–GFP–NAT1 cassette, including 1000 bp of ENO1 promoter, PCR-amplified from strain YJB11372 using primers 5359 and 5360, which have 70 bp of homology to RPS1, or primers 4730 and 4731, which have 70 bp of homology to NEUT5L (Table 2). The correct integration of the ENO1–GFP–NAT1 cassette was verified by PCR of genomic DNA using the primer pairs 5361 and 968, and 2040 and 5401, for integration into RPS1 and 4577 and 968, and 2040 and 4580, for integration into NEUT5L (Table 2).

Table 2. Primers used in this study.

Lower case, primer tail homologous to pRS314 for in vivo recombination; underlined, SfiI–NgoMIV restriction sites; bold, point mutation (G62D). All primers were designed for this study.

Primer	Sequence 5′–3′
324	CTTCAACTGTAAAAGGGGCG
658	TTTGTACAATTCATCCATACCATG
714	GAGAATCGAAGAAGAATTAGGTTCTGAAGCTATCTACGCTGGTAAAGATTTCCAAAAGGCTTCTCAATTGGGTGGTGGTTCTAAAGGTGAAGAATT
716	GCTTTGTTGTTGAAGGTTAACC
986	AATAAAGGGAGTAAGGTGGT
2040	GCATCACCTGGAACAGAAGTTCTGTATC
3118	CCCAGATGCGAAGTTAAGTGCGCAG
3185	TTTAATTAGTTCATATATTCAAGATGTTCCTATAAAAGAAAAAAAAAGCACCAGCTTTTTTTTATTTAATCGTAAAACGACGGCCAGTGAATTC
3951	AGATCCAGATATTGAAGGTAAAAGGTA
3952	GAAAATCTCCCCCTTCACATT
4423	AATACCACGTTACCGTCATTGAT
4424	CCAGTGATCATATTCTTGATGAAATC
4574	cctaaagggagcccccgatttagagcttgacggggaaagcGGCCGGCTAGGCCTTAAACAAGTGGTATTCAAGCAC
4575	ctttcgctttcttcccttcctttctcgccacgttcgccggTAACCCACTGAATTCTACATCG
4576	accccaggctttacactttatgcttccggctcctatgttgGGCCGGCTAGGCCTGGAAGGACGATGAAGGAGAG
4577	atatcaagcttatcgataccgtcgacctcgagggggggccGTAATTGTAGTAAGAATGAC
4580	cctaaagggagcccccgatttagagcttgacggggaaagcGGCCGGCTAGGCCTTAACCCACTGAATTCTACATCG
4581	ctttcgctttcttcccttcctttctcgccacgttcgccggTAAACAAGTGGTATTCAAGCAC
4582	accccaggctttacactttatgcttccggctcctatgttgGGCCGGCTAGGCCTGTAATTGTAGTAAGAATGAC
4583	atatcaagcttatcgataccgtcgacctcgagggggggccGGAAGGACGATGAAGGAGAG
4658	Ttaagttgggtaacgccagggttttcccagtcacgacgtt
4659	ctgcagcccgggggatccactagttctagagcggccgccaTGTGGAATTGTGAGCGGATA
4730	GCTCGGAGGAGGCTCCCCAAAGGTTTTATCACCAATGGTGGTACCACTAACCCAAGAACAGAAAAAGCATTGTAGACATGATTGTGCAATACCGTG
4731	CAGAGTATGTGAAGCAATTGCAATTGCAATTATTAGAGATCCAGAAAACTGAATTGTGCTTGAATACCACGATATCAAGCTTGCCTCGTCCCCG
4926	AAAAGGCCTGATAAGGAGAGATCCATTAAGAGCA
4969	GGATTTAGTTCCATTATGG
5227	ggtcatagctgtttcctgtgtgaaattgttatccgctcacCATCAGTTAGTGTCATGGAG
5228	ggccccccctcgaggtcgacggtatcgataagcttgatatAGAAAATCCAACTCCGAGAG
5229	TGTTGGTCAAGTTATTAAAGATTGGGATATTTCTTTAACT
5230	AGTTAAAGAAATATCCCAATCTTTAATAACTTGACCAACA
5359	GGTAAAACTTTGATCAACAGATCTACCGGTTTAAAGAATGCCGCTGATGGCTTGAAAGGTAGAGTTTTCGGTAGACATGATTGTGCAATACCGTG
5360	ACCTTGGACTTCATCAACTCTCAATTTGATTTTTCTGTAAGAGTGGTCTTCGGAACCTTGTAAGTCGGCCGATATCAAGCTTGCCTCGTCCCCG
5361	GTGTGTGTGTTCCAAGTCCCAGCTCTCAC
5401	CTCAATTTAGAGGATTGAGCGTAAGTAG
5429	GTGCCGTAAAGCACTAAATCGG
5430	GTTACCTCACTCATTAGGCACCCC
5930	GTCTGAAGAACTTCCACAAATTG
5931	TGGTTACTGTATCACCTGGC

Table 3. Plasmids used in this study.

Plasmid	Alias	Characteristics	Reference
p2308	pRS314	Amp, CEN6/ARS4, TRP1	Sikorski & Hieter (1989)
p1374		pGEM-URA3	Wilson et al. (1999)
p1653	pDDB57	URA3-dpl200	Wilson et al. (2000)
p2105	pJK795	NAT1	Shen et al. (2005)
pMG2120	pMG2120	GFP-NAT1	Gerami-Nejad et al. (2012)
pDUP2	p2365	pRS314-NEUT5LDUP (for insertion/duplication)	This study
pDUP3	p2366	pRS314-NEUT5LDUP-NAT1	This study
pDUP4	p2367	pRS314-NEUT5LDUP-URA3	This study
pDUP5	p2368	pRS314-NEUT5LDUP-URA3dpl200	This study
pDUP3-RBP1	p2369	pRS314-NEUT5LDUP-NAT1-RBP1	This study
pDIS2	p2372	pRS314-NEUT5LDIS (for insertion/disruption)	This study
pDIS3	p2373	pRS314-NEUT5LDIS-NAT1	This study
pDIS4	p2374	pRS314-NEUT5LDIS-URA3	This study
pDIS5	p2375	pRS314-NEUT5LDIS-URA3dpl200	This study
pDIS3-RBP1	p2376	pRS314-NEUT5LDIS-NAT1-RBP1	This study
pDIS3-RBP1G62D	p2377	pRS314-NEUT5LDIS-NAT1-RBP1G62D	This study

Strain YJB12563 (Table 1) was constructed by transforming SfiI-digested pDUP3-RBP1 into the rbp1Δ/Δ strain YJB12543 (Zacchi et al., 2010). Similarly, strains YJB12565 and YJB12566 (Table 1) were constructed by transforming SfiI-digested pDIS3-RBP1 or pDIS3-RBP1^G62D, respectively, into the rbp1Δ/Δ strain YJB12543. The correct integration of the inserts was verified by performing PCR from genomic DNA. Because the insert from pDIS vectors integrates at NEUT5L in the opposite orientation to the insert from pDUP (Fig. 2c, d), the PCR analysis used NAT1 primer 3118 (Table 2) and two primers annealing outside NEUT5L: primer 4969 (Table 2) for strain YJB12563 (for pDUP insert, 1.1 kb PCR product), and primer 4926 (Table 2) for strains YJB12565-6 (for pDIS inserts, 1.6 kb PCR product) (Fig. 3a).

Fig. 2.

Maps of the S. cerevisiae–E. coli shuttle vectors for integration into the C. albicans NEUT5L locus. (a) pDUP shuttle vector; and (b) pDIS shuttle vector. Common features of the vectors include the Amp^R Ampicilin resistance gene for selection in E. coli, the TRP1 marker for selection on S. cerevisiae trp- strains, and a centromeric (CEN6) and autonomous replicating sequence (ARSH4) to ensure replicative stability in S. cerevisiae. The selectable markers for the resulting vectors are: pDUP3 and pDIS3, NAT1; pDUP4 and pDIS4, URA3; pDUP5 and pDIS5, URA3-dpI200. (c, d) Representation of SfiI/NgoMIV restriction enzyme digestion of inserts from pDUP and pDIS vectors and integration into NEUT5L. (e) Schematic representation of the primer design for gap repair. These primers work in both pDUP and pDIS vectors. Primers are ≥60 bp long, and consist of ≥20 bp with homology to the insert required for PCR amplification of the insert (grey line), followed by ≥40 bp with homology to pRS314 added as the primer tails (black line). The box contains the 40 bp tails from primers 5227 (top) and 5228 (bottom) that should be added to the primers to clone inserts into pDUP3-5/pDIS3-5 (Table 2). (f) Representation of the in vivo recombination between a SmaI-linearized pDIS vector and a PCR amplified insert containing 40 bp of homology to pDIS.

Fig. 3.

Inserts from pDUP/pDIS vectors are efficiently integrated and expressed from NEUT5L. (a) PCR from genomic DNA of NAT⁺ transformants showing proper integration into NEUT5L of the RBP1–NAT1 cassette from pDUP3/pDIS3-RBP1 and pDIS3-RBP1^G62D. (i) Primer set 3118+4969 was used to PCR amplify the genomic DNA of strain YJB12563 [rbp1Δ/Δ NEUT5L-RBP1 (pDUP)], while (ii) primer set 3118+4926 was used to PCR amplify the genomic DNA of strains YJB12565 [rbp1Δ/Δ NEUT5L-RBP1(pDIS)] and YJB12566 [rbp1Δ/Δ NEUT5L-RBP1^G62D (pDIS)]. (b) Phenotypic test showing that the RBP1 and RBP1^G62D constructs integrated via duplication (pDUP vector, column 3) or disruption (pDIS vectors, columns 4 and 5) of the NEUT5L region are expressed. Strains SC5314 (RBP1/RBP1), YJB12543 (rbp1Δ/Δ), and YJB12563-6 were plated onto YPAD medium and grown in the presence of discs embedded with 5 % DMSO (upper panels) or with 1 µg rapamycin ml⁻¹ in 5 % DMSO (bottom panels). Pictures showing the zone of growth inhibition were taken after 24 h of incubation at 30 °C. (c) and (d) Quantitative RT-PCR showing that RBP1 and RBP1^G62D (c) and URA3 (d) are expressed to native-locus levels when integrated at the NEUT5L region. RBP1 expression levels at its native locus in the wild-type SC5314 (RBP1/RBP1), the RBP1/rbp1Δ (YJB12982), and rbp1Δ/Δ (YJB12543) strains were compared with RBP1 expression at NEUT5L from strains YJB12563 [rbp1Δ/Δ NEUT5L-RBP1(pDUP)], YJB12565 [rbp1Δ/Δ NEUT5L-RBP1(pDIS)], and YJB12566 [rbp1Δ/Δ NEUT5L-RBP1^G62D (pDIS)]. URA3 expression levels at its native locus in the wild-type (CAF2-1) (URA3/URA3), the URA3/uraΔ (CAI4), and ura3Δ/Δ SC5314 strains were compared with URA3 expression at NEUT5L from strains YJB12984 [ura3Δ/Δ NEUT5L-URA3 (pDUP)] and YJB12985 [ura3Δ/Δ NEUT5L-URA3 (pDIS)]. qRT-PCR was performed on RNA samples obtained from the different strains grown to exponential phase in YPD at 30 °C. Shown is the mean±sd of the target gene (RBP1 or URA3)/reference gene (TEF1) ratio for each strain normalized to RBP1 or URA3 levels of expression of the wild-type strain of at least two independent experiments.

Strains YJB12984 and YJB12985 (Table 1) containing a URA3 cassette inserted at NEUT5L were constructed by transforming C. albicans strain CAI4 as follows. To construct strains YJB12984 and YJB12985, vectors pDUP4 and pDIS4 (Table 3) were digested with NaeI and transformed into CAI4 strain. The correct integration of the URA3 cassette was verified by PCR of genomic DNA using the primers 324 and 4969 (Table 2) for integration at NEUT5L using the pDUP vector, and 324 and 4926 (Table 2) for integration at NEUT5L using the pDIS vector.

Tests for rapamycin resistance were performed with single colonies picked from YPAD plates and resuspended in 300 µl sterile water, 200 µl of this dilution was spread on YPAD and plates were left to dry at room temperature for 1 h. Sterile Whatman paper discs spotted with 10 µl 5 % DMSO or 1 µg rapamycin (LC Laboratories) in 5 % DMSO were placed on the inoculated agar surface. Plates were incubated at 30 °C for 24 h.

Vector construction.

The shuttle vectors pDUP3–pDUP5 and pDIS3–pDIS5 for cloning through gap repair and integration into NEUT5L (Figs 1 and 2 and S1 and S2, available with the online version of this paper) were constructed as follows. The 550 bp region located in the middle of the intergenic region between orf19.1961 and orf19.1963 was chosen as the locus for ectopic DNA integration (Fig. 1a, white box). This 550 bp fragment (designated NEUT5L) was PCR amplified in two consecutive fragments of 300 bp (NEUT5L 3′) and 250 bp (NEUT5L 5′) each from C. albicans SC5314 DNA (strain YJB11562). Each PCR fragment was sequentially cloned into the S. cerevisiae shuttle vector pRS314 (GenBank: U03440.1) (Sikorski & Hieter, 1989) by in vivo recombination, such that the NEUT5L 5′ and 3′ fragments flanked the vector’s polylinker region (for more details on the in vivo recombination technique see the Supplementary Methods) (Figs 2 and S1). The PCR fragments were integrated into the vector in opposite orientations, to generate vectors that will integrate into the genome by duplicating NEUT5L (pDUP vectors) or by disrupting NEUT5L (pDIS vectors) (see Results and Discussion for more details). The 300 bp NEUT5L 3′ fragment was PCR amplified using the primers 4574 and 4575 (to construct pDUP1) or 4580 and 4581 (to construct pDIS1), and each fragment was in vivo recombined into NgoMIV-digested pRS314 (p2308, Table 3) in the trp⁻ S. cerevisiae strain L40 (YJB12046) (Fig. S1) (Sikorski & Hieter, 1989; Vojtek et al., 1993). The 250 bp NEUT5L 5′ fragment was PCR amplified using the primers 4576 and 4577 (for pDUP2) or 4582 and 4583 (for pDIS2), and each fragment was in vivo recombined into KpnI-digested pDUP1 or pDIS1, respectively (Fig. S1). The pDUP2/pDIS2 vectors containing the two fragments comprising the 550 bp NEUT5L region cloned in different orientations were verified by sequencing. To incorporate each of the auxotrophic markers URA3 and URA3dpl200 and the drug resistance marker NAT1 into pDUP2 and pDIS2, each marker was PCR amplified using primers 4658 and 4659 (Table 2) using as templates the plasmids p1374 (URA3), p1653/pDDB57 (URA3-dpl200) and p2105/pJK795 (NAT1) (Table 3) and was in vivo recombined into SacI/SacII digested pDUP2 and pDIS2 in S. cerevisiae strain L40 to generate pDUP3–pDUP5 and pDIS3–pDIS5, respectively (Table 3, Figs 2a, b, S2). All in vivo recombination ‘reactions’ were plated on SC-TRP medium, except for the construction of plasmids carrying the URA3-dpl200 cassettes (pDUP5 and pDIS5) which were plated on SC-URA to ensure the integrity of the recyclable cassette and to prevent the selection of plasmids carrying only the dpl200 portion of the marker. All PCRs were performed using high-fidelity polymerases (Pfx Platinum, Invitrogen; Pfu Turbo, Agilent). All vectors containing the selectable markers were sequenced using primers 5429 and 5430 (Table 2) that anneal to the pRS314 sequence flanking the NEUT5L fragments. The complete sequence and maps of pDUP3–pDUP5/pDIS3–pDIS5, as well as detailed protocols for transformation of S. cerevisiae and general procedures and important notes for performing a successful in vivo recombination are available in the Supplementary Material.

Fig. 1.

The intergenic region between orf19.1961 and orf19.1963 is a neutral locus for integration and expression of exogenous DNA. (a) Region of chromosome 5 showing CLN3/orf19.1960, orf19.1961, GDS1/orf19.1963, and the 550 bp NEUT5L region (white box, which corresponds to Chromosome 5L coordinates 232782–233331); (b) Western blot showing expression of Eno1–GFP integrated at ENO1 (strain YJB11372), NEUT5L region (YJB12257), or RPS1 (YJB12780) detected with anti-GFP antibody. The wild-type strain SC5314 was used as a negative control. Anti-H4 antibody was used as a loading control. Indicated below is the mean±sd of the GFP signal intensity normalized to the loading control and to the ENO1 native locus of two independent experiments. (c) Overnight YPAD cultures of strains SC5314 (WT, dashed and dotted black line), YJB11372 (ENO1 : : GFP, continuous grey line), YJB12257 (neut5l : : ENO1 : : GFP, dashed black line), and YJB12780 (rps1 : : ENO1 : : GFP, dotted black line) were diluted 1 : 1000 in YPAD, YPAD buffered at pH 3, 6 or 9, or YPAD containing 1 M NaCl or 4 mM H₂O₂, and incubated for 24 h 25 min at 30 °C with high-speed shaking. OD₅₉₅ was measured every 15 min using a Tecan Sunrise microplate. The growth curves show the combined results of two independent experiments in which each strain was tested in triplicate. (d) Doubling times calculated from the growth curves of two independent experiments in which each strain was tested in triplicate (shown is the mean±se, total n = 6). * P-value <0.0083.

Vectors pDUP3/pDIS3-RBP1 and pDIS3-RBP1^G62D were constructed as follows. The RBP1 (orf19.6452) ORF, including 914 bp of promoter and 463 bp of terminator region, was PCR amplified from C. albicans SC5314 DNA using primer pairs 5227 and 5228 (for RBP1), or 5227 and 5230, and 5228 and 5229 (for RBP1^G62D) (Fig. S4), and the PCR fragments were in vivo recombined into SmaI-digested and phosphatase-treated pDUP3 and pDIS3 in S. cerevisiae strain L40, to generate pDUP3/pDIS3-RBP1 and pDIS3-RBP1^G62D, respectively. Primers 5230 and 5229 carry the point mutation that leads to RBP1^G62D (bold type in Table 2).

Growth rate experiments.

Three independent colonies from each strain were inoculated into three culture tubes containing 2 ml YPAD and grown overnight at 30 °C. Fifty microlitres of each culture was washed with 1 ml water and resuspended in 1 ml water, and 3 µl were inoculated into 150 µl media using 96-well round-bottom sterile polystyrene plates (Corning). Growth conditions tested were: YPAD, YPAD buffered at pH 3 (100 mM HEPES), YPAD buffered at pH 6 (100 mM HEPES), YPAD buffered at pH 9 (100 mM HEPES), YPAD+1 M NaCl and YPAD+4 mM H₂O₂. The plate was incubated at 30 °C in a Tecan Sunrise microplate incubator/spectrophotometer with high-speed shaking. Readings of OD₅₉₅ were collected every 15 min (for at least 24 h). Data were processed with an in-house script written in Matlab software (Mathworks), in which the raw OD data were pre-processed by subtraction of a blank measurement and shortening of curves that end with long saturation periods; the time-dependent growth of the log-transformed OD was fitted to a piecewise linear function with three pieces: (1) constant OD over time, representing the lag phase; (2) a linear growth of the log-transformed OD; and (3) constant OD over time, representing saturation of growth. The free parameters of the function are: the starting OD during lag phase, the doubling time, the time at which half-maximum saturation is reached and the maximal OD at saturation. The fitting procedure minimized the sum of absolute differences between the experimental data and the estimated curve. The highest growth rate (lowest doubling time) of the culture corresponds to the slope of the inflection point of the resulting fitted sigmoid curve. Graphs were made using Microsoft Office Excel 2007. Statistical analyses were performed on two independent experiments with three replicates (total n = 6), using standard statistical operations manually implemented in Microsoft Office Excel. A t-test was used to determine if there were differences between experiments (no difference found, P-value >0.32). An ANOVA test was used to determine the likelihood that strains had identical doubling times (not identical, P-value <0.0061). A pair-wise t-test using Bonferroni correction was employed to identify differences between the doubling times of the strains within each experimental condition (six comparisons, significant P-value <0.05/6 = 0.0083).

Protein extraction and Western blot analysis.

Strains were grown in YPAD at 30 °C to exponential phase, and equal amounts of the cultures, as determined spectrophotometrically (OD₆₀₀), were resuspended in 400 µl Thorner Buffer (40 mM Tris pH 8, 5 % SDS, 8 M Urea, 100 µM EDTA) supplemented with 2 mM DTT. Three hundred microlitres of 500 µM acid-washed glass beads were added to the samples, and samples were heated at 95 °C for 8 min and vortexed for 30 min to lyse cells. 5× non-reducing sample buffer (0.5 M Tris pH 8, 50 % glycerol v/v, 0.0125 % Bromophenol blue, 5.4 % SDS) was added to the samples to a 1× final concentration with 5 % β-mercaptoethanol. Samples were heated at 95 °C for 8 min, centrifuged to pellet insoluble cell debris, and equal amounts of protein (as determined spectrophotometrically) were separated in a 12 % acrylamide gel and transferred to PVDF membranes for Western blotting. Mouse anti-GFP (Roche) and rabbit anti-S. cerevisiae histone H4 (which cross-reacts with C. albicans histone H4) (Glowczewski et al., 2004) were used to probe the membranes. Images of the membranes were obtained using a digital camera documentation system (Chemiimager 5500, ProteinSimple). GFP signal intensity was quantified using Image J 1.44d (Wayne Rasband, National Institutes For Health); final values displayed in Fig. 1(b) were normalized first to the loading control signal and later to the GFP signal intensity of Eno1-GFP integrated at ENO1 (value set to 1).

RNA extraction and real-time quantitative PCR (qPCR).

RNA was prepared from strains grown to exponential phase in YPAD at 30 °C by using the MasterPure RNA Purification kit (Epicentre). The samples were treated with DNase (Epicentre) to remove contaminating genomic DNA, and cDNA was prepared using the ProtoScript M-MuLV First Strand cDNA Synthesis kit (New England BioLabs), using oligo d(T) primers. cDNA was measured by qPCR using SYBR Green I (Roche Applied Sciences). Gene expression was calculated as the amount of RBP1 (primers 5930 and 5931, Table 2) or URA3 (primers 3951 and 3952, Table 2) cDNA relative to the amount of TEF1 (primers 4423 and 4424, Table 2) cDNA in the same sample using the LightCycler 480 software system (Roche Applied Science). Each RBP1 or URA3 experiment was performed at least twice. The data from each experiment were normalized to the RBP1 or URA3 expression of the wild-type strain, and the mean±sd of the normalized data of multiple experiments was plotted using Microsoft Excel.

Results and Discussion

Identification of a neutral locus for DNA integration into C. albicans genome

Integration of DNA into C. albicans at an exogenous locus has been classically done by targeting the HIS1 promoter (Davis et al., 2002), the ENO1 promoter (Staab et al., 2003) (less commonly) and the RPS1 ORF (Dennison et al., 2005; Murad et al., 2000; Vieira et al., 2010). However, integration at the HIS1 or ENO1 promoters using the available vectors requires specific auxotrophies (his- or ura-, respectively) (Davis et al., 2002; Staab et al., 2003). Integration at RPS1 disrupts one RPS1 allele, causing hemizygosity which is associated with a haploinsufficient growth defect (Oh et al., 2010). Thus, we sought an alternative locus in the C. albicans genome that permits the efficient integration and expression of exogenous DNA without a detectable effect of the integration on the strain fitness.

To this end, we identified a large region (approximately 13 kb) on Chromosome 5L between ORFs CLN3/orf19.1960 and GDS1/orf19.1963 that contains only one other ORF, the uncharacterized orf19.1961 (approximately 1.6 kb long) (Fig. 1a). The intergenic region between orf19.1961 and CLN3 is approximately 7.8 kb, and the intergenic region between orf19.1961 and GDS1 is approximately 3.8 kb, both much longer than the average size of C. albicans intergenic regions (865 bp) (A. Koren and J. Berman, unpublished results). The region between CLN3 and orf19.1961 appeared attractive as a locus for exogenous DNA integration due to its length and to the absence of other coding elements, but detection of spontaneous neocentromere formation at this location (L. S. Burrack and J. Berman, unpublished observation) suggested that the region may have uncharacterized features that could interfere with gene expression. Thus, we used a region located approximately midway between orf19.1961 and GDS1 (located 1.71 kb from the GDS1 start codon and 1.53 kb from the orf19.1961 stop codon), which has diverged extensively in Candida dubliniensis, suggesting that it is under minimal purifying selective pressure (unpublished data). We refer to this 550 bp long region as the NEUT5L region (Fig. 1a, white box), based on its position on Chr5L and its neutral character with respect to gene expression and strain fitness, as characterized below.

To determine the efficiency of expression of exogenous DNA integrated at NEUT5L, we compared the expression of an ENO1–GFP cassette integrated at three different loci: the ENO1 (native) locus (YJB11372), NEUT5L locus (YJB12257) and RPS1 locus (YJB12780). Integration of ENO1–GFP disrupted NEUT5L and RPS1 loci, but not ENO1, where GFP is integrated in-frame at the 3′ end of the ENO1 ORF (see Methods). Western blot analysis of Eno1–GFP expression revealed that normalized expression levels of Eno1–GFP at the ENO1 locus and at the NEUT5L locus were similar (signal ratio of 1 : 0.9±0.2 for ENO1 versus NEUT5L) (Fig. 1b). In contrast, expression levels of Eno1–GFP inserted at RPS1 were lower than at the ENO1 locus (signal ratio of 1 : 0.5±0.2 for ENO1 versus RPS1) (Fig. 1b). Similar results were obtained with an independently constructed set of strains with the same genotype (data not shown). This suggests that integration of ENO1–GFP at RPS1 had a position effect, reducing the level of ENO1–GFP expression relative to its native locus (Henikoff, 1992; Tartof, 1994), while integration at NEUT5L enabled expression levels similar to those observed at the native locus. In support of this conclusion, we observed that expression of RBP1 and of URA3 from the NEUT5L locus was also similar to native locus levels (Fig. 3c, d).

To determine if integration at NEUT5L caused any obvious growth defects in C. albicans, we compared growth rates of a wild-type strain (SC5314, no insertions) with the growth rates of the three strains described above expressing ENO1–GFP at the ENO1, NEUT5L, or RPS1 loci. Growth rates of the different strains were measured in rich medium (YPAD) and under stress conditions, including YPAD buffered at different pH (pH 3, pH 6 and pH 9), high salinity (1 M NaCl) and oxidative stress (4 mM H₂O₂) (Fig. 1c, d). These assays were performed at 30 °C. Strains with ENO1–GFP integrated at the ENO1 native locus (YJB11372) or at NEUT5L (YJB12257) exhibited similar growth curves (Fig. 1c) and similar doubling times (Fig. 1d) to the wild-type strain (YJB11562) under all the conditions tested. In contrast, the strain containing an ENO1–GFP cassette integrated at RPS1 (YJB12780) exhibited a more pronounced lag phase (average approximately 7 h, Fig. 1c) and a significantly longer doubling time (P-value <0.0083, Fig. 1d) than the other three strains under all the conditions tested (average approximately 5.0 h lag phase, Fig. 1c). Essentially similar results were obtained when using an independently constructed set of strains with the same genotype and when using a different cassette inserted at RPS1 (ENO1–GFP–2W1S–OVA_323–339–OVA_257–264–I-Eα_50–66–NAT1) [from pMG2271 (Igyártó et al., 2011); data not shown]. The reduced growth observed in the YJB12780 strain (RPS1/rps1 : : ENO1–GFP) is consistent with previous results showing that RPS1 is haploinsufficient (Oh et al., 2010). Therefore, the phenotypes of strains constructed by integration at the RPS1 locus should be compared with those of strains that carry control inserts in the RPS1 locus as well. Importantly, the ability of strains containing an integration of ENO1 : : GFP at NEUT5L to filament in solid or liquid Spider medium and in medium containing serum was similar to that of the wild-type strain (data not shown). Taken together, our results indicate that insertion at NEUT5L permits expression of inserts at near-native locus levels and that it incurs no fitness cost during growth in vitro.

Construction of shuttle vectors for gap-repair-mediated cloning and integration at the NEUT5L locus

To efficiently integrate exogenous DNA into the NEUT5L locus, we constructed two sets of integrative shuttle vectors: vectors pDUP (Fig. 2a) and pDIS (Fig. 2b), which only differ in how the construct integrates into NEUT5L [i.e. integration by duplication of NEUT5L (Fig. 2c) versus disruption of NEUT5L (Fig. 2d)]. Integration by duplication of the region targeted for recombination (Nunes et al., 1999) is a common approach used in diverse organisms, including the C. albicans vectors Cip10, Cip10-2μ, pHWP1-GFP3 and pDDB78 (Dennison et al., 2005; Spreghini et al., 2003; Staab et al., 2003; Vieira et al., 2010). Integration by disruption of the region targeted for recombination is also an approach commonly used in C. albicans (Sasse & Morschhäuser, 2012; Wilson et al., 1999). Here, vectors were designed so that either one of these approaches can be used to target integration into the NEUT5L locus.

The backbone of the pDUP and pDIS vectors (Fig. 2a, b) is the S. cerevisiae–E. coli shuttle vector pRS314 (Sikorski & Hieter, 1989) which has already been used for integration into the C. albicans genome (at the HIS1 locus) (Davis et al., 2002; Ganguly & Mitchell, 2012; Spreghini et al., 2003; Wolf & Davis, 2010; Zacchi et al., 2010). Important features of pRS314 include: an ampicillin resistance (AmpR) gene for selection in bacterial strains, the pBLUESCRIPT (Stratagene) polylinker region for convenient cloning and three elements from S. cerevisiae that permit replication, maintenance and selection in this budding yeast [an autonomous replicating sequence (ARSH4) centromeric DNA (CEN6) and an auxotrophic marker (TRP1) (Fig. 2)].

To construct the vectors, we utilized an in vivo recombination or gap repair approach (Muhlrad et al., 1992). Gap repair seals double-strand breaks in plasmid DNA by recombining DNA fragments that share homology at the junction regions. In S. cerevisiae, gap repair facilitates highly efficient cloning of DNA fragments into plasmid backbones, and has been utilized to generate vectors for use in diverse organisms, including C. albicans (Davis et al., 2002; Ma et al., 1987; Nunes et al., 1999; Spreghini et al., 2003; Vieira et al., 2010). There are two important requirements for cloning by in vivo recombination into any vector using S. cerevisiae. First, the DNA fragments to be recombined must share at least 40 bp of homology at the junctions. This region of homology can be added to the 5′ end of primers used to PCR amplify the insert of choice (Fig. 2e). For clarity, the sequences of all primers used for in vivo recombination in this work display the sequences with homology to pRS314 in lower-case type (Table 2). Second, the restriction enzyme site (RES) used to linearize the vector must be located between the regions homologous to the desired insertion site (e.g. within the DNA region flanked by the two 40 bp primer tails) (Figs 2f and S1). Using this methodology, six different vectors were constructed for integration into the NEUT5L locus in C. albicans, as described below (Figs 2 and S2).

To construct the pDUP/pDIS vectors, we sequentially integrated the NEUT5L region and one of three different selectable markers (NAT1, URA3 and URA3-dpl200) into pRS314 by in vivo recombination (Fig. S1). The 550 bp NEUT5L region was cloned in two fragments (NEUT5L 3′ and NEUT5L 5′) each flanking the pRS314 polylinker region (Figs 2 and S1). SfiI and NgoMIV RES (underlined in primers in Table 2) were incorporated during the PCR so that they flank the NEUT5L 5′ and NEUT5L 3′ fragments (Figs 2 and S1). Finally, three different markers were incorporated into the pDUP and pDIS backbones: NAT1 (pDUP3, pDIS3), URA3 (pDUP4, pDIS4) and the URA3-dpl200 recyclable marker (pDUP5, pDIS5) (Figs 2 and S2) (Shen et al., 2005; Wilson et al., 2000). These six shuttle vectors are designed for cloning any DNA of interest within the remaining RES of the polylinker located between NEUT5L 5′ and the selectable marker (Fig. 2). Cloning into these vectors can be achieved with both classical restriction enzyme digestion/ligation (at the NotI–ClaI RES) and by in vivo recombination (at the SmaI RES) (Fig. 2). To integrate a construct from pDUP/pDIS vectors into the C. albicans genome, the entire region to be integrated is isolated from the pDUP/pDIS backbone by digestion at the SfiI/NgoMIV RES that flank the two NEUT5L fragments (Fig. 2c, d). This last step has two advantages: first, it reduces the size of the integrating DNA fragment and, second, it reduces the likelihood of integration of non-C. albicans DNA (AmpR, CEN6, ARSH4 and TRP1) into the C. albicans genome (data not shown). Integration of inserts from pDUP and pDIS vectors not only results in the duplication or disruption of NEUT5L, respectively, but also leads to integration of the insert in opposite orientations (Fig. 3a). In sum, we have engineered six S. cerevisiae/E. coli shuttle vectors based on the pRS314 backbone that allow for rapid and facile cloning through in vivo recombination and target the insert for integration into the NEUT5L C. albicans locus.

Proof of principle: expression of RBP1 and RBP1^G62D from NEUT5L

To test the utility of the plasmids, we cloned and inserted into NEUT5L the gene RBP1 (orf19.6452), which encodes the rapamycin binding protein (known as FKBP12 in mammals and Fpr1 in S. cerevisiae). RBP1 confers a readily detectable phenotype to the transformed cells: RBP1⁺ cells are sensitive to rapamycin, while rbp⁻ cells (rbp1Δ/Δ) are resistant to rapamycin (Cruz et al., 2001). Rapamycin is a bacterial macrolide that, when in complex with Rbp1, blocks Tor kinase function, triggering a major cellular response through the inhibition of the TOR signalling pathway that leads to cell growth arrest (Cruz et al., 2001; Wullschleger et al., 2006). Mutations in RBP1, such as a spontaneous recessive point mutation in a highly conserved glycine residue (fpr1^G65D in S. cerevisiae) also confer rapamycin resistance (Cruz et al., 2001; Ferrara et al., 1992; Heitman et al., 1991). In C. albicans, the equivalent residue to S. cerevisiae FPR1-G65 is RBP1-G62 (Fig. S3, black box). Thus, if C. albicans Rbp1 has functions like S. cerevisiae Fpr1, reintroduction of RBP1, but not of RBP1^G62D, into rbp1Δ/Δ C. albicans cells is expected to restore the wild-type, rapamycin-sensitive phenotype.

The wild-type RBP1 gene, including the RBP1 promoter and terminator regions, was cloned into both pDUP3 and pDIS3 by in vivo recombination to produce pDUP3-RBP1 and pDIS3-RBP1. The NEUT5L 5′–RBP1–NAT1–NEUT5L 3′ insert was released from pDUP3-RBP1 and pDIS3-RBP1 by SfiI digestion and transformed into the C. albicans rbp1Δ/Δ strain (YJB12543), yielding approximately 50 nourseothricin-resistant colonies (NAT⁺) µg⁻¹ digested plasmid DNA. PCR screening of the transformants detected correct integration in 100 % (four out of four) of the independent transformants tested for each vector (Fig. 3a). Similar levels of efficiency of integration were obtained with other DNA inserts as well (data not shown), indicating that both vectors direct highly efficient integration at the NEUT5L locus.

Next, we tested the effect of integration at NEUT5L on RBP1 expression. Growth of the wild-type RBP1/RBP1 SC5314 strain is inhibited by rapamycin and not by solvent alone (DMSO) (Fig. 3b, first column, halo of growth inhibition), while rbp1Δ/Δ cells are rapamycin-resistant (Fig. 3b, second column; (Cruz et al., 2001)). If RBP1 is indeed expressed at the NEUT5L locus, the rbp1Δ/Δ strains transformed with pDUP3-RBP1 (Fig. 3b, third column) and pDIS3-RBP1 (Fig. 3b, fourth column) should become rapamycin-sensitive. As expected, these strains showed a halo of growth inhibition indicating that constructs from both pDUP and pDIS shuttle vectors are efficiently expressed from the NEUT5L locus. In order to directly verify that the level of RBP1 gene expression in these strains was similar when integrated at NEUT5L compared with its native locus, we extracted RNA from cells grown in YPD at 30 °C and performed quantitative RT-PCR to measure RBP1 mRNA levels (Fig. 3c). To aid in the comparison, we normalized all the data to the expression level of the RBP1/RBP1 wild-type strain. RBP1 expression level in the RBP1/rpb1Δ strain was approximately one half of the expression level in the wild-type strain (0.57±0.37) (Fig. 3c). When integrated at NEUT5L, RBP1 expression was similar to that of the RBP1/rpb1Δ strain (0.38±0.08 for the pDUP construct and 0.44±0.14, for the pDIS construct) (Fig. 3c). No RBP1 expression was detected in the rpb1Δ/Δ strain, as expected. These results indicate that RBP1 integrated at NEUT5L using the pDUP and pDIS vectors is indeed expressed at native-locus levels, supporting the conclusion that NEUT5L is a neutral locus for expression of exogenous DNA (see above, Fig. 1b). Furthermore, these results also suggest that gene expression at NEUT5L is similar independently of the direction of integration and of the duplication of the NEUT5L region caused by using either the pDUP or pDIS systems.

Because the auxotrophic marker URA3 is susceptible to position effects (Brand et al., 2004; Cheng et al., 2003; Lay et al., 1998; Staab & Sundstrom, 2003), we verified that integration at NEUT5L allowed for URA3 native-locus level of expression. To do this we integrated URA3 cassettes from pDUP4 and pDIS4 vectors (Fig. 2a, b) into NEUT5L and compared URA3 expression in these strains with the wild-type strain and an URA3/ura3Δ heterozygote strain grown in YPD at 30 °C (Fig. 3d). To aid in the comparison, we normalized all the data to the expression level of the URA3/URA3 wild-type strain. URA3 expression level in the URA3/ura3Δ strain was approximately one half of the expression level in the wild-type strain (0.64±0.26) (Fig. 3d). When integrated at NEUT5L, URA3 expression was similar to the expression levels of the URA3/ura3Δ heterozygote strain (0.44±0.14 for the pDUP4 construct and 0.35±0.17 for the pDIS4 construct) (Fig. 3d). Almost no detectable URA3 expression was observed in the ura3Δ/Δ strain, as expected. These results indicate that URA3 is efficiently expressed from NEUT5L, and that this auxotrophic marker can be safely used when integrating constructs at NEUT5L to perform virulence studies.

Cloning by in vivo recombination is particularly useful because desired mutations, including point mutations as well as fusion with epitope tags or fluorescent proteins, can be introduced into the insert in a single step, obviating the need to design specific RES and perform sequential cloning steps [e.g. (Wolf & Davis, 2010; Zacchi et al., 2010)]. For example, we utilized this approach to introduce the SfiI and NgoMIV RES bordering the NEUT5L fragments of the pDUP/pDIS vectors (Figs 2 and S1). To provide another example of the advantages of the gap repair technique, we constructed pDIS3-RBP1^G62D vectors in a single step. RBP1 was amplified as two fragments, using the same external primers employed to clone wild-type RBP1 into pDIS3 and new, internal, perfectly overlapping 40 bp primers carrying the GGT (glycine) to GAT (aspartic acid) mutation present in the RBP1^G62D allele (Fig. S4). The two PCR fragments were then in vivo recombined into SmaI-digested pDIS3 by co-transformation into S. cerevisiae. The NEUT5L-RBP1^G62D-NAT1 constructs were released from the pDIS vector by SfiI digestion and transformed into the rbp1Δ/Δ strain. PCR screening of transformants showed 100 % efficiency of integration at NEUT5L (four out of four nourseothricin-resistant strains contained NAT1 at the NEUT5L locus) (Fig. 3a). We had hypothesized that the RBP1^G62D mutation would confer rapamycin resistance, similar to the phenotype of the equivalent FPR1^G65D mutation in S. cerevisiae [(Heitman et al., 1991) and Fig. S3, black box]. Consistent with this, integration of RBP1^G62D into the C. albicans rbp1Δ/Δ strain at the NEUT5L locus did not confer a rapamycin-sensitivity phenotype (Fig. 3b, fifth column), while integration of wild-type RBP1 did (Fig. 3b, fourth column). Importantly, RBP1^G62D expression level at NEUT5L was similar to RBP1 native-locus levels (0.45±0.12 versus 0.57±0.37), indicating that the rapamycin resistance phenotype was caused by the G62D mutation and not by the lack of expression of the construct (Fig. 3c). Thus, in vivo recombination and integration into NEUT5L enables efficient hypothesis-testing through site-directed mutagenesis and cloning coupled in a single step.

In summary, the work presented here describes two new important features for efficient genetic manipulation of the C. albicans genome (Table 4). First, we identified a novel neutral region that permits the efficient integration of exogenous DNA and the expression of inserts to levels similar to their expression at their native locus. Importantly, no growth or filamentation defects have been detected following disruption of one copy of the NEUT5L locus under a range of growth conditions. Second, we constructed a series of six shuttle vectors for efficient one-step cloning and integration of constructs into C. albicans. The vectors are based on the pRS314 plasmid backbone, which is commonly used for expression and complementation studies in S. cerevisiae, and has been used for cloning and integration of constructs into the C. albicans HIS1 locus. These re-engineered vectors provide a number of advantages over other available plasmids (Table 4). First, the 550 bp region of homology to NEUT5L contained in the vectors yields highly efficient targeting to this specific locus. Second, the three markers in these vectors expand the range of selection schemes available by incorporating the counterselectable auxotrophic URA3 or the recyclable URA3-dpl200 markers, and the dominant NAT1 drug resistance marker that is ideal for use in the transformation of clinical isolates and other prototrophic strains. Third, the two alternative RES (SfiI and NgoMIV) bordering the insert release the construct from the vector prior to transformation, thereby preventing integration of non-C. albicans DNA into the C. albicans genome and reducing the size of the insert. Fourth, the vectors allow for integration of the insert such that the region of insertions is either duplicated or disrupted. Fifth, when using the pDUP vectors, duplication of the region of insertion can be used to drive the reversible integration of the construct. Sixth, inserts can be cloned into the vectors by in vivo recombination or by using classical RES digestion/ligation approaches. Seventh, fragments cloned into the pDUP/pDIS vectors can be PCR-amplified directly from S. cerevisiae DNA extracts, thereby skipping the requirement for subcloning in E. coli, which may be impractical for large plasmids and/or toxic inserts. Finally, the steps described here to generate the pDUP/pDIS vectors can be used to design new vectors that direct integration into any chosen region of the genome.

Table 4. Advantages and applications of pDUP/pDIS vectors.

Advantages of the pDUP/pDIS vectors
1) Intergenic region between orf19.1961 and orf19.1963: the NEUT5L locus
a. Efficient insertion and expression of exogenous DNA.
b. Insertion at this locus does not detectably affect growth or filamentation.
2) pDUP/pDIS vectors
a. Markers: counterselectable URA3; recyclable URA3-dpl200; dominant drug resistance marker NAT1.
b. Convenient RES for preventing introduction of foreign DNA into C. albicans.
c. High efficiency of integration into NEUT5L (550 bp total homology to integration locus).
d. Two modes of integration: duplication and disruption.
3) Gap repair in S. cerevisiae
a. Facilitates cloning of multiple overlapping fragments (Eckert-Boulet et al., 2012; Iizasa & Nagano, 2006; Marykwas & Passmore, 1995).
b. One-step cloning simultaneously incorporates desired mutations or epitope tags.
c. Does not require shared specific RES in the insert and vector.
d. Obviates E. coli cloning step thereby facilitating cloning of inserts that are toxic to bacteria.
e. Lower cost relative to site-directed mutagenesis.
Applications of the pDUP/pDIS vectors
1) Rapid and facile construction of mutant/chimeric inserts. Recombining inserts into pDUP/pDIS vectors while simultaneously incorporating:
a. Mutations: point mutations, insertions, deletions, etc.
b. Epitope and fluorescent tags.
c. Regulatable promoters, or other modifications.
2) Reversible integration.
a. Pop-in/pop-out inserts integrated at NEUT5L using pDUP4 vectors.
b. Test for lethality of a homozygous null deletion by reversible insertion of a third copy at NEUT5L prior to deleting both copies at the native locus.

Abbreviations:

RES

restriction enzyme site