KEM1/XRN1 and RAT1 are two known exoribonuclease genes in Saccharomyces cereivsiae and encode a cytoplasmic and nuclear exoribonuclease, respectively. CaKEM1/CaXRN1 and CaRAT1, the Candida albicans homologs of 5′→3′ exoribonuclease genes, were identified by protein sequence comparisons and by functional complementation of the S. cerevisiae kem1/xrn1 null mutation. The deduced amino acid sequences of CaKEM1 and CaRAT1 show 51% and 55% identities to those of the S. cerevisiae KEM1 and RAT1, respectively. The exonuclease motifs were found to be highly conserved in CaKem1p and CaRat1p. We disrupted two chromosomal copies of CaKEM1 in a diploid C. albicans strain and demonstrate that C. albicans kem1/kem1 mutants are defective in filamentous growth on filamentous-inducing media. These results imply that CaKEM1 is involved in filamentous growth of C. albicans.
Candida albicans is the major fungal pathogen in humans. Upon various environmental signals, it undergoes reversible morphogenetic transitions between budding yeast, pseudohyphal, and hyphal forms [1,2]. The yeast-to-hypha transition is considered the major cause of Candida pathogenicity. The hyphal form is more adherent to mammalian cells and seems better adapted for tissue penetration or macrophage cell lysis than the yeast form [3,4]. A network of multiple signaling pathways controlling the yeast-to-hypha switch has been identified in C. albicans. These include the MAP kinase pathway (Cst20, Hst7, and Cek1), the cAMP/PKA pathway (Bcy1, Tpk1, and Tpk2), and the pH-responsive pathway [1,5,6]. These multiple pathways in conjunction with the pathway-specific transcription factors regulate the expressions of hypha-specific genes, such as ECE1, HWP1, HYR1, ALS3, ALS8, RBT1, and RBT4 [6,7]. Many of these hypha-specific genes encode either cell wall or secreted proteins.
The elucidation of the regulatory elements and the transcription factors in the hyphal development of C. albicans has been mainly based on the strong molecular conservation between C. albicans and Saccharomyces cerevisiae. In S. cerevisiae, the filamentation regulatory pathways are better characterized [8,9]. Most of the C. albicans signal transduction genes have been isolated using genetic screens in S. cerevisiae or by searching for their homologous counterparts in the C. albicans genome sequence.
In S. cereivsiae, two 5′→3′ exoribonucleases have been identified. Kem1p/Xrn1p, a major cytoplasmic exoribonuclease, is not essential for cell growth and functions in mRNA turnover [10,11]. The nuclear exoribonuclease, Rat1p, is essential for cell viability and is involved in the turnover and the processing of nuclear RNA, and in the export of poly(A)+ RNA from the nucleus . It has been shown that Rat1p and Kem1p/Xrn1p are functionally interchangeable proteins. The dominant RAT1 alleles with mutations in the nuclear localization sequence (NLS) restore KEM1-like functions. Conversely, targeting of Kem1p to the nucleus by the addition of an NLS rescues the temperature-sensitivity of the rat1-1 mutant .
KEM1 was initially identified because of its functions in microtubule-mediated processes, including nuclear fusion during mating, chromosome transmission, and spindle pole body separation . The Kem1 protein was subsequently reported to be a microtubule-associated protein . Purified Kem1p promoted the in vitro assembly of tubulin into microtubules. The exoribonuclease motifs present in Kem1p are significant in the exoribonuclease activity and the benomyl phenotype of Kem1p . However, the mutant Kem1p, which lacks exoribonuclease activity, still showed microtubule-assembly activity as well as the meiosis-specific functions such as, nucleic acid binding and homologous pairing activities. Accordingly, these functions of Kem1p are considered separable from its exoribonuclease activities.
Recently, it was reported that KEM1 is involved in the filamentatous growth of Σ1278b strains . Both haploid invasive growth and diploid pseudohyphal growth were found to be greatly impaired in kem1 mutant strains. Transcripts of FLO11, a filamentation-related gene, are greatly diminished in the kem1 strain. However, the kem1 mutation does not disturb the general expression of filamentation-specific genes, because this effect of kem1 is not observed at the TEC1 or STE12 transcript level .
In the present communication, we asked whether the filamentation-associated functions of KEM1 are conserved in the pathogenic yeast, C. albicans. We have isolated the C. albicans KEM1 gene and characterized its function in the filamentous growth of this organism.
2Materials and methods
2.1Strains and growth conditions
The yeast strains used in this study are listed in Table 1. Standard methods of yeast transformation and genetic crosses were used for all strain constructions [17,18]. Escherichia coli and yeast media were prepared using established procedures. Solid media containing 5-fluroorotic acid (5-FOA) were prepared as described previously and used for the selection of C. albicans URA3 pop-outs . Assays for haploid invasive growth and diploid pseudohyphal growth in S. cerevisiae have been described previously [20,21]. For the invasive growth test, cells were patched on SC-Ura plates and grown for 3 days at 30 °C. Plates were photographed before and after washing the cells off the agar surface. Pseudohyphal growth testing was conducted by streaking cells on SLAD media for single colonies . These cells were incubated for 3–5 days at 30 °C, and photographed at 4× or 40× magnification under a differential interference contrast (DIC) microscope. The filamentation phenotype of C. albicans cells was tested with 10% fetal bovine serum containing media, Spider medium, and modified Lee's media, as described previously, with the addition of 2% agar for solid media . Overnight liquid cultures were diluted in water, plated for single colonies, and incubated for 2–7 days at 37 °C. To induce hyphal growth in liquid media, cells were pre-grown in YEPD media overnight at 30 °C, and then diluted (1 × 106 cells/ml) into the inducing medium indicated. Cells were grown at 37 °C.
The DNA sequence of CaKEM1 was obtained from the Stanford DNA sequencing and Technology Center web site. The oligonucleotides JM01 (5′-GGGAGCTC TTGATAGTCTTTGGAAGACG-3′) and JM02 (5′-GAATTGACTAGT AATTTTAG-3′) were designed, and used to amplify a 4.1-kb N-terminal fragment from C. albicans SC5314 genomic DNA. The oligonucleotides JM03 (5′-AAAATTACTAGT CAATTCAT-3′) and JM04 (5′-GGGAAGCTT AATACTGGGACAAGTTCCGG-3′) were designed and used to amplify a 1.7-kb C-terminal fragment. Each primer sequence contains a useful restriction site and these are underlined (SacI for JM01, SpeI for JM02 and JM03, and HindIII for JM04). The two resulting PCR fragments were cleaved with the restriction enzymes SacI–SpeI or SpeI–HindIII, and then cloned into the SacI and HindIII sites of S. cerevisiae plasmids pRS316 and pRS426, and T-vector by the three-fragment ligation method [22,23]. The resulting plasmids were named pRS316-CaKEM1, pRS426-CaKEM1, and T-CaKEM1, respectively. pRS426-CaKEM1-1 was constructed by inserting the NotI fragment of T-CaKEM1 into the NotI site of pRS426. Plasmid pRC18-CaKEM1 was constructed by inserting the CaKEM1-containing PvuII fragment of pRS316-CaKEM1 into the SmaI site of pRC18 .
The DNA fragment of CaRAT1 was isolated from a Candida genomic library constructed on plasmid pRS202 by using kem1 complementation test. The S. cerevisiae kem1 mutant strain (JK351) was transformed with the Candida library . Transformants were selected on SC-Ura plates and screened for colonies that restored the invasive growth phenotype.
2.3Disruption of the C. albicans KEM1 gene
To disrupt two chromosomal copies of CaKEM1 sequentially, we constructed two different deletion cassettes. A 3.4-kb BamHI fragment of hph-CaURA3-hph from plasmid pQF86  was ligated into the BglII–BamHI sites of p316-CaKEM1 to generate plasmid pkem1::hph. For the construction of pkem1::hisG, pRS426-CaKEM1del was generated first. The 0.7-kb internal fragment of CaKEM1 was PCR-amplified with the primers JM105 (5′-GGGAGATCT GATTTTGAATGTGA-3′) and JM102. The primer JM105 carried the BglII restriction site. The BglII–SpeI digests of this PCR product was inserted into the BglII–SpeI sites of pRS426-CaKEM1-1 to generate pRS426-CaKEM1del. A 4.0-kb BglII–BamHI fragment of hisG-CaURA3-hisG from plasmid pCUB-6  was ligated into the BglII site of p426CaKEM1del to generate pkem1::hisG.
The pkem1::hph DNA was digested with SfuI and HindIII, and transformed into C. albicans strain CAI4 to obtain a kem1 deletion (KEM1/kem1::hph-CaURA3-hph). Transformants were screened by PCR and by Southern blotting for the presence of a one-copy disruption construct at the KEM1 locus. The KEM1/kem1::hph-CaURA3-hph Ura3+ cells were grown on 5-FOA media to select for the loss of URA3. The second copy of KEM1 was disrupted by transforming KEM1/kem1::hph cells with pkem1::hisG DNA digested with SfuI and SpeI. The kem1::hph/kem1::hisG isolates were screened by PCR and Southern blotting.
3Results and discussion
3.1Candida albicans KEM1 complements the filamentation defect of the kem1 null mutation in S. cerevisiae
Screening of the C. albicans genomic library was carried out to identify the functional homologs that rescued the filamentation growth defect of the kem1 null mutation in S. cerevisiae. One of 12,000 transformants showed the haploid invasive growth even after plasmid DNA recovery and retransformation. DNA sequence analysis of the 4.0-kb insert revealed that the cloned DNA fragment encoded a 968 amino acid protein with 55% identity to the S. cerevisiae Rat1 protein, a nuclear exoribonuclease. To identify a KEM1 homolog in C. albicans, we searched the Stanford C. albicans databases for sequences similar to that of Kem1p. A 4527-bp open reading frame (ORF) was identified, whose amino acid sequences showed 51% identity with S. cerevisiae Kem1p. The exonuclease motifs were found to be highly conserved in CaKem1p and CaRat1p, just as in ScKem1p and ScRat1p.
Based on this sequence information, primers were designed and CaKEM1 was isolated from Candida genomic DNA by PCR amplification. To determine whether CaKEM1 is a functional homolog of the S. cerevisiae gene with respect to its filamentation phenotype, it was cloned into S. cerevisiae plasmids pRS316 and pRS426, and these constructs were transformed into a S. cerevisiae kem1 mutant strain with a Σ1278b background. As shown in Fig. 1(a), CaKEM1 complemented weakly the haploid invasive growth defect of the kem1 mutation. A high-copy or low-copy number of the CaKEM1 plasmid made no significant difference in these complementation phenotypes (data not shown). The complementation activity of CaKEM1 was found to be weaker than that of CaRAT1, which would explain the initial identification of CaRAT1 in our screening of the C. albicans genomic library.
The diploid kem1/kem1 strain of S. cerevisiae was transformed with CaKEM1 or CaRAT1 plasmids and grown on low-nitrogen medium. The kem1/kem1 mutant strain has a pseudohyphal growth defect, but CaKEM1 or CaRAT1 transformants partially restored pseudohyphal growth (Fig. 1(b)). These results demonstrate that CaKEM1 and CaRAT1 can replace the function of Kem1p in S. cerevisiae and promote pseudohyphal growth.
The kem1-complementing activity of CaRAT1 is interesting because S. cerevisiae Rat1p, a major nuclear exoribonuclease, is able to replace the cytoplasmic exoribonuclease Kem1p/Xrn1p when mislocalized to the cytoplasm by the mutations in its NLS . C. albicans RAT1 showed even stronger complementation activity than CaKEM1. These results imply that CaRat1p could somehow stay in the cytoplasm and functionally replace the Kem1p function in S. cerevisiae. The amino acid sequence alignment of CaRat1p and ScRat1p located the putative CaRat1 NLS at amino acids 465–488. The putative CaRat1 NLS (TKANLANSDAAAELKKLIDSKKTQ) differs from ScRat1 NLS (KKHRLEKDNEEEIAKDSKKVK) and matches only in part the consensus bipartite NLS (two basic residues, a 10-amino-acid spacer of any sequence, followed by a cluster of basic residues in which three of five are basic). Whether CaRat1p and CaKem1p are differentially localized, and thus are functionally separable in C. albicans, is not clear at this point, but their heterologous expressions in S. cerevisiae reveal that CaKEM1 and CaRAT1 have analogous kem1-complementing activity.
3.2CaKEM1 complements the slow growth defect and nuclear fusion defect of the kem1 null mutation
In addition to the defect in filamentation growth, the S. cerevisiae kem1 mutant exhibited a slow-growth phenotype on rich media and a nuclear fusion defect during mating . We conducted complementaion assays with the kem1 mutant strain transformed with CaKEM1 or CaRAT1 plasmids. As shown in Fig. 2, the growth curves of the CaKEM1 or CaRAT1 transformants were as similar as that of ScKEM1. Therefore, the significant growth defect of kem1 was fully suppressed by CaKEM1 and CaRAT1.
The nuclear fusion phenotypes of the kem1 strains were assayed using the plate mating method. The S. cerevisiae kem1 mutant strain showed the mating defect and the kem1 mutant strain containing ScKEM1 recovered this mating ability. Compared with the kem1 strain, carrying an empty vector, the kem1 strain containing CaKEM1 or CaRAT1 recovered the mating phenotype (Fig. 3). These results indicate that CaKEM1 and CaRAT1 can functionally replace the S. cerevisiae KEM1 gene.
3.3C. albicans kem1/kem1 mutants show defects in hyphal morphology
To determine the role of CaKEM1 in the cell growth and morphology of a diploid C. albicans strain, both copies of CaKEM1 were sequentially deleted by replacing their coding sequences with the hph-CaURA3-hph and hisG-CaURA3-hisG deletion cassettes. Growth rates of the wild type, heterozygous mutant (KEM1/kem1) and homozygous mutant (kem1/kem1) strains were similar at 30 °C in YEPD media (data not shown). During infection of the mammalian host, C. albicans undergoes morphogenetic switching from the yeast to the pseudohyphal and hyphal forms. This process of filamentation can also be induced in culture using various media such as serum-containing medium, Lee's medium, or Spider medium. KEM1/KEM1, KEM1/kem1, and kem1/kem1 strains were tested on solid filamentation-inducing media (Fig. 4(a)). In Spider and in Lee's media, both KEM1/kem1 and kem1/kem1 null mutants showed strong defects in hyphal growth. However, in 10% serum containing media, which is a strong inducing condition, these hyphal defects were not so evident as the effects on the colony morphology. The filamentaion phenotypes of these strains were also tested in liquid media at 37 °C (Fig. 4(b)). The germ-tube formation of the kem1/kem1 mutant was severely reduced in Lee's media and in Spider media. Moreover, the reintroduction of the wild-type KEM1 to the kem1/kem1 strain fully restored filamentation, indicating that the original filamentous growth defects are attributable to the loss of KEM1. These results show that KEM1 is important for filamentous growth under several specific environmental conditions.
The S. cerevisiae Kem1p is known to have diverse biochemical activities, including exoribonuclease, microtubule-binding, and DNA-strand exchange activities, and thus to be involved in a number of cellular processes [13,14]. The filamentous growth of the Σ1278b strains is one of these KEM1-mediated processes. Deletion of KEM1 in S. cerevisiae strains causes a slow-growth phenotype on rich media and significant defects in filamentous growth . Our current results show that the kem1/kem1 mutant strain of C. albicans has a wild-type growth rate on rich media but has a pronounced defect in filamentatous growth on various filamentation-inducing media. It appears that the filamentation-associated functions of KEM1 are well conserved in S. cerevisiae and C. albicans. The slow-growth phenotype of the S. cerevisiae kem1 mutation is not observed in the C. albicans kem1/kem1 mutant strain (data not shown). Considering the results that the CaKEM1 and CaRAT1 genes, when expressed in S. cerevisiae, fully complemented the growth defect of kem1 (Fig. 2), it is suggested that the growth-associated biochemical functions of ScKem1p are conserved in CaKem1p and CaRat1p.
We are grateful to G.R. Fink and H. Liu for kindly supplying plasmids, strains, and the C. albicans library. This work was supported by a grant from the Molecular Medicine Research Group Program of the Korean Ministry of Science and Technology (M10106000070–01A20001700). H.-S.A. and K.L. were supported by the BK21 program administered by the Ministry of Education, Republic of Korea, 1999.