Characterization of Agglutinin-like Sequence Genes From Non-albicans Candida and Phylogenetic Analysis of the ALS Family

Corresponding author: Lois L. Hoyer, 2522 VMBSB, 2001 S. Lincoln Ave., Urbana, IL 61802., lhoyer{at}uiuc.edu (E-mail)

Communicating editor: M. E. ZOLAN

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The ALS (agglutinin-like sequence) gene family of Candida albicans encodes cell-surface glycoproteins implicated in adhesion of the organism to host surfaces. Southern blot analysis with ALS-specific probes suggested the presence of ALS gene families in C. dubliniensis and C. tropicalis; three partial ALS genes were isolated from each organism. Northern blot analysis demonstrated that mechanisms governing expression of ALS genes in C. albicans and C. dubliniensis are different. Western blots with an anti-Als serum showed that cross-reactive proteins are linked by ß1,6-glucan in the cell wall of each non-albicans Candida, suggesting similar cell wall architecture and conserved processing of Als proteins in these organisms. Although an ALS family is present in each organism, phylogenetic analysis of the C. albicans, C. dubliniensis, and C. tropicalis ALS genes indicated that, within each species, sequence diversification is extensive and unique ALS sequences have arisen. Phylogenetic analysis of the ALS and SAP (secreted aspartyl proteinase) families show that the ALS family is younger than the SAP family. ALS genes in C. albicans, C. dubliniensis, and C. tropicalis tend to be located on chromosomes that also encode genes from the SAP family, yet the two families have unexpectedly different evolutionary histories. Homologous recombination between the tandem repeat sequences present in ALS genes could explain the different histories for co-localized genes in a predominantly clonal organism like C. albicans.

CANDIDA albicans is an opportunistic pathogenic fungus that causes mucocutaneous and disseminated forms of disease. Two well-characterized gene families of C. albicans are believed to produce proteins that function in pathogenesis. The first characterized family, the SAP family, encodes secreted aspartyl proteinases (HUBE et al. 1998 ). Disruption of SAP genes or inhibition of SAP gene products reduces pathogenicity of C. albicans, providing evidence for the role of aspartyl proteinases in the disease process (SANGLARD et al. 1997 ; HOEGL et al. 1998 ; BORG-VON ZEPELIN et al. 1999 ; CASSONE et al. 1999 ; GRUBER et al. 1999 ; SCHALLER et al. 1999 ). The second large gene family in C. albicans is called ALS (agglutinin-like sequence) due to the resemblance of domains of its encoded proteins to -agglutinin, a cell-surface adhesion glycoprotein in Saccharomyces cerevisiae (LIPKE et al. 1989 ; HOYER et al. 1995 ). Presently, eight genes in the ALS family have been reported in the literature, although a small number of additional genes are found in the C. albicans genome (HOYER et al. 1995 , HOYER et al. 1998A , HOYER et al. 1998B ; GAUR and KLOTZ 1997 ; HOYER and HECHT 2000 , HOYER and HECHT 2001 ). ALS genes conform to a basic three-domain structure that includes a relatively conserved 5' domain of 1299 to 1308 nucleotides (433 to 436 amino acids), a central domain of variable length consisting entirely of a tandemly repeated 108-bp motif, and a 3' domain of variable length and sequence that encodes a serine-threonine-rich protein (HOYER et al. 1998B ). Heterologous expression of ALS genes in S. cerevisiae confers an adherence phenotype on the organism, suggesting Als proteins function in adhesion to host surfaces, a property that is positively correlated with Candida pathogenesis (CALDERONE and BRAUN 1991 ; GAUR and KLOTZ 1997 ; FU et al. 1998 ). In addition to the potential for Als proteins to function in pathogenesis, ALS genes are differentially expressed under a variety of conditions that include morphological form, growth medium composition, growth phase, and strain of C. albicans, similar to the SAP family (HUBE et al. 1994 ; HOYER et al. 1998A , HOYER et al. 1998B ). Association of Als proteins with pathogenicity mechanisms and differential expression of ALS genes suggest that, similar to the SAPs, the ALS family is important in C. albicans pathogenesis.

If these gene families play an important role in C. albicans pathogenesis, it is possible that they also contribute to the pathogenicity of clinically relevant non-albicans Candida. Previous studies identified SAP genes in other Candida species including C. dubliniensis, C. tropicalis, C. parapsilosis, and C. guilliermondii (MONOD et al. 1994 ; GILFILLAN et al. 1998 ). These organisms are among non-albicans Candida species that are isolated with increasing frequency from clinical specimens (WINGARD et al. 1979 ; WINGARD 1995 ; FRIDKIN and JARVIS 1996 ; VAN'T WOUT 1996 ; HOPPE et al. 1997 ; KUNOVA et al. 1997 ; SULLIVAN and COLEMAN 1997 ; WEINBERGER et al. 1997 ; DARWAZAH et al. 1999 ; RANGEL-FRAUSTO et al. 1999 ). These observations led us to question whether an ALS gene family was also present in clinically important non-albicans Candida.

In this study, we present evidence on the DNA, RNA, and protein level that ALS genes exist as a family in C. dubliniensis and C. tropicalis. We isolate multiple ALS gene sequences from each organism using a PCR-based strategy and demonstrate that, although the basic structure of ALS genes is likely to be conserved in these organisms, there is little conservation of individual gene sequences across the different species. Using ALS and SAP gene probes, we also demonstrate that ALS and SAP genes are co-localized on the same chromosomes in each organism. Data from these studies demonstrate conservation of basic cell wall architecture between C. albicans and the non-albicans species and highlight significant differences in ALS gene expression patterns between the two most closely related organisms, C. albicans and C. dubliniensis. Finally, ALS gene sequence data from C. albicans, C. dubliniensis, and C. tropicalis are used to present a phylogenetic analysis of the ALS family. The data presented here indicate that the ALS family is younger than the SAP family. The presence of genes on the same chromosome with different evolutionary histories is expected under sexual recombination and provides indirect evidence that C. albicans has mated throughout its evolutionary past (BARTON and WILSON 1996 ; HULL et al. 2000 ; MAGEE and MAGEE 2000 ).

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Candida strains:
Multiple strains of each organism were used in initial Southern blotting studies to detect ALS genes. As studies progressed, two strains of each organism were chosen as representative of results and used in the figures in this article. Strains listed here include all those used in the study. C. albicans strain SC5314 was a gift from W. A. Fonzi; strain B311 was purchased from the American Type Culture Collection (ATCC; Manassas, VA), and strain 1177 was a gift from Stewart Scherer. C. tropicalis strains CAPG3 and T60700 were a gift from Patricia Kammeyer; strains 13803, 201380, and 201381 were purchased from ATCC. C. dubliniensis strains CD36 (type strain), CM1, and 16F were provided by David Coleman; strain LY261 was a gift from Richard Barton. C. parapsilosis strain SB was a gift from Carrie Frey; Patricia Kammeyer provided strains 44 and X36406. The identity of the C. tropicalis and C. parapsilosis strains was verified using either the API 20C AUX or API 32C system. Cellular morphology of each organism was examined following growth on corn meal-Tween agar plates (Remel, Lenexa, KS) and matched descriptions provided in standard sources (LARONE 1995 ; SULLIVAN et al. 1995 ). Strains were maintained as glycerol stocks at -80° and streaked on YPD agar plates as needed.

ALS gene probes:
All methods for making ALS gene-specific probes were published previously (HOYER et al. 1995 , HOYER et al. 1998A , HOYER et al. 1998B ; HOYER and HECHT 2000 , HOYER and HECHT 2001 ); Table 1 summarizes these probes. To date, eight ALS genes were reported in the literature (HOYER et al. 1995 , HOYER et al. 1998A , HOYER et al. 1998B ; GAUR and KLOTZ 1997 ; HOYER and HECHT 2000 , HOYER and HECHT 2001 ). The sequences of ALS3 and ALS8 are essentially identical and are detected by the same probe (HOYER and HECHT 2000 ). To avoid redundancy, ALS8 was omitted from certain figures in this article.

Nucleic acid gels and blotting:
Protocols for genomic DNA extraction, running contour-clamped homogeneous electrical field (CHEF) gels, and Southern blotting were described previously (HOYER et al. 1995 , HOYER et al. 1998A , HOYER et al. 1998B ). All Southern blots were performed with the digoxigenin nonradioactive nucleic acid labeling and detection system (Roche Molecular Biochemicals, Indianapolis). Separation of C. albicans total RNA on formaldehyde gels and subsequent Northern blotting were described; detection of specific messages utilized radiolabeled DNA fragments (HOYER et al. 1995 , HOYER et al. 1998A , HOYER et al. 1998B ). Hybridization conditions for individual blots are included in the figure legends. ALS cross-hybridizing fragments were detected in Southern blots of C. parapsilosis genomic DNA after 40° hybridization and washing at 50° in 0.5x SSC/0.1% SDS.

Growth of Candida for Northern blot analysis:
A single colony each of C. dubliniensis CD36 and C. tropicalis 13803 was inoculated into separate flasks of YPD (yeast extract, peptone, dextrose) medium and grown overnight (16 hr) at 30° and 200 rpm shaking. Cells from each culture were counted and inoculated into a variety of growth media at a density of 5 x 10⁶ cells/ml. Growth media included fresh YPD, RPMI 1640 (catalog no. 11875-085; Life Technologies, Rockville, MD), and Lee medium (LEE et al. 1975 ) adjusted to pH 4.5, 5.5, 6.5, or 7.5. Cultures were grown for various lengths of time ranging from 2 to 8 hr. Cells were harvested, washed in pyrocarbonic acid diethyl ester-treated sterile water, flash-frozen in an ethanol-dry ice bath, and stored at -80° until RNA was extracted.

Cell wall fractionation and analysis of cell wall proteins:
Cells for cell wall fractionation were grown using the same conditions as for Northern analyses (see above). Methods for cell wall fractionation and protein analysis were previously described (KAPTEYN et al. 2000 ). In brief, Als proteins were released by ß1,6-glucanase digestion of isolated, SDS-extracted cell walls, separated by electrophoresis, and electrophoretically transferred onto polyvinylidene difluoride membranes. Membranes were treated for 30 min with 50 mM periodic acid, 100 mM sodium acetate (pH 4.5) to abolish any cross-reactivity of the serum to N- and O-linked glycan. Als proteins were visualized by treating the membranes with a polyclonal anti-Als antiserum raised by immunization of a New Zealand White rabbit with the purified N-terminal domain of C. albicans Als5p (HOYER and HECHT 2001 ). A serum dilution of 1:5000 in phosphate-buffered saline (PBS), containing 5% (w/v) nonfat milk powder, was used. Binding of the anti-Als antiserum was assessed with goat anti-rabbit IgG peroxidase (Pierce Chemical Co.) at a dilution of 1:10,000 in PBS/5% (w/v) milk powder. The blots were developed using enhanced chemiluminescence (ECL) Western blotting detection reagents (Amersham Pharmacia Biotech). The anti-Als serum did not show any signal on a Western blot of S. cerevisiae ß1,6-glucanase-released cell wall proteins (J. C. KAPTEYN, unpublished data).

PCR amplification of ALS gene fragments from C. dubliniensis and C. tropicalis:
Nucleotide sequences from the 5' domain of ALS1 through ALS7 were aligned using the PILEUP program of the GCG sequence analysis package (DEVEREUX et al. 1984 ) and regions of conserved sequence were defined. These regions were used to design consensus oligonucleotide primers where degenerate bases were included in positions of ambiguity. The resulting primers were 5' GCH ART SCN GGD GAY ACA TTY AYR TT 3' (forward) and 5' GGM AYA TCA AYR AHA ASA GTW GCW GTK YCH CC 3' (reverse). PCR reactions including these primers used genomic DNA from C. dubliniensis strain CD36 and from C. tropicalis strain ATCC 13803, a 52° annealing temperature, and Pfu polymerase (Stratagene, La Jolla, CA). Each strain produced the predicted PCR product of 1 kb. The products were cloned into pCRBlunt (Invitrogen, Carlsbad, CA) and transformed into Escherichia coli TOP10 (Invitrogen). Plasmid DNA from the resulting clones was analyzed by DNA sequencing. Open reading frames that resembled ALS sequences were given the accession nos. AF201685 (ALSD1), AF202529 (ALSD2), AF202530 (ALSD3), and AF201686 (ALST1). A second forward PCR primer was designed to amplify additional ALS-like sequences from C. tropicalis genomic DNA. The primer (5' GCH GGT TAT CGW CCW TTT DTK GA 3') was paired with the reverse primer above; amplification, cloning, and DNA sequencing followed the previous methods. DNA sequences isolated using this procedure were assigned accession nos. AF211865 (ALST2) and AF211866 (ALST3). All C. dubliniensis and C. tropicalis DNA sequences were translated with the alternate yeast genetic code tables because, like C. albicans, these species decode CUG as serine instead of leucine (SUGITA and NAKASE 1999 ).

C. tropicalis SAP gene probe:
Four SAP gene sequences from C. tropicalis have been reported in the GenBank database (accession nos. X61438, AF115320, AF115321, and AF115322). Coding regions from these sequences were aligned using the PILEUP program of GCG and consensus regions were identified. Primers were made to these regions using degenerate bases in positions of ambiguity. The resulting primers were 5' GTT DTB RTW GAY ACY GGW TCH TCY GAT 3' (forward) and 3' CCD GTA TAY TTR GCA TKR TCA AYV CC 3' (reverse). A consensus SAP gene probe of 460 nucleotides was amplified from genomic DNA of strain ATCC 13803. This fragment was purified from an agarose gel and labeled by random priming using the Genius nonisotopic system (Roche Molecular Biochemicals). The resulting probe was hybridized to Southern blots at 65° and washed in 0.5x SSC/0.1% SDS at the same temperature.

Phylogeny analysis:
The predicted amino acid sequences for C. albicans, C. dubliniensis, and C. tropicalis Als proteins were aligned using the PILEUP program of GCG software (Wisconsin Package Version 10, Genetics Computer Group, Madison, WI). While the full sequence of each gene is not known, the missing sequences are repetitive regions and would be excluded from phylogenetic analyses were they known (ORTI et al. 1997 ). Three maximum parsimony trees were constructed using PAUP in GCG, with all characters assigned equal weights, branches added stepwise, and bootstrap values computed. Bootstrap values were assigned from heuristic searches for topologies found with exhaustive searches. Heuristic and exhaustive searches produced the same topology.

The first tree was found by an exhaustive search and included seven C. albicans sequences, three C. dubliniensis sequences, and ALST1 from C. tropicalis and spanned amino acids 22 through 357. In this first tree, ALST2 and ALST3 were omitted to take advantage of the longer sequence available for the remaining genes. The second tree was found by an exhaustive search of the seven C. albicans amino acid sequences. Only the N-terminal and C-terminal domains were used in this analysis since the tandem repeat sequences may be phylogenetically misleading (ORTI et al. 1997 ). The third tree was found by a branch and bound search and included seven sequences from C. albicans, three from C. dubliniensis, and three from C. tropicalis, which spanned amino acids 231 to 357. The first and third trees were rooted with ALST1 because previous rDNA studies have placed C. tropicalis outside of C. albicans and C. dubliniensis (BARNS et al. 1991 ; GILFILLAN et al. 1998 ). The C. albicans-only tree was rooted with ALS7, as determined by the first tree. The maximum parsimony tree of SAP sequences included those studied by MONOD et al. 1998 and the three additional SAP sequences reported in GenBank (listed above). SAP sequences were aligned using PILEUP and a heuristic search conducted using S. cerevisiae YAP3 as the root (BARNS et al. 1991 ). In addition, the distance-based tree building method UPGMA was used to cross-validate inferred topology of the trees generated (HILLIS et al. 1996 ).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Detection of ALS gene sequences in non-albicans Candida species by Southern blotting:
Southern blots of genomic DNA from several non-albicans Candida species were hybridized with various ALS-specific fragments (Table 1). Because they hybridize to multiple ALS genes in C. albicans, fragments derived from the ALS1 and ALS5 tandem repeat domains were used first (HOYER et al. 1995 , HOYER et al. 1998A , HOYER et al. 1998B ; Table 1). At high stringency, the ALS1 repeats and ALS5 repeats probes largely differentiate between subfamilies of the ALS genes in C. albicans: the ALS1 repeats probe hybridizes to ALS1, ALS2, ALS3, ALS4, and ALS8 while the ALS5 repeats probe minimally recognizes ALS5, ALS6, and ALS7 (HOYER et al. 1998B ; HOYER and HECHT 2000 ). Each probe recognized multiple genomic fragments in strains of C. dubliniensis and C. tropicalis, although the hybridization signals in C. tropicalis were weaker (Fig 1). Decreasing the hybridization stringency increased the intensity of the C. tropicalis signals (data not shown); however, efforts were made to screen at higher stringencies to avoid potentially misleading nonspecific results. These initial results indicated the presence of ALS-like tandem repeat fragments in the genomes of C. dubliniensis and C. tropicalis.

View larger version (56K): In this window In a new window Download PPT slide   Figure 1. Southern blots of BglII-digested genomic DNA from C. albicans (Ca), C. dubliniensis (Cd), and C. tropicalis (Ct) strains. The Southern blot was probed with the indicated fragments derived from C. albicans ALS genes. The blots were hybridized at 50° and washed at 60° in 0.5x SSC/0.1% SDS. Molecular size markers (in kilobases) are indicated at the left of each blot.

The 5' domain of C. albicans ALS genes is conserved, showing 55–90% identity among known sequences (HOYER and HECHT 2000 ). To determine if ALS 5' domain sequences were also present in non-albicans Candida, the genomic Southern blot described above was stripped and reprobed with a KpnI-HpaI fragment derived from the 5' domain of ALS1 (Table 1). This probe recognizes multiple fragments in the C. albicans genome that are largely the same as the fragments that hybridize with the ALS1 tandem repeat probe (HOYER et al. 1998A ; Fig 1). These results suggested that the 5' domain and tandem repeat domain were found on the same genomic fragment in many cases. A similar result was achieved for C. dubliniensis DNA, but signals were not observed at higher stringency for C. tropicalis DNA (Fig 1).

In addition to a conserved 5' domain followed by a domain of tandem repeats, C. albicans ALS genes encode a 3' domain sequence that is variable in length and sequence among the known genes (HOYER et al. 1998B ; HOYER and HECHT 2000 ). To determine whether ALS genes in non-albicans species were homologous to certain C. albicans ALS genes, we hybridized blots of genomic DNA with fragments derived from the 3' end of ALS1, ALS3, ALS2/ALS4, ALS5/ALS6, and ALS7 (Table 1). Even at lowered stringencies, no 3'-domain-derived probes gave any signal on blots of C. dubliniensis or C. tropicalis DNA, with the exception of ALS7 (Fig 1). Results presented here suggest that there are a similar number of ALS genes in C. albicans and C. dubliniensis. While ALS genes are likely to be present in C. tropicalis, they are likely to be fewer in number and less related in sequence to C. albicans ALS genes. Finally, the juxtaposition of the 5' domain and tandem repeat domain of the ALS fragments in C. dubliniensis suggests that these genes have a similar three-domain structure as ALS genes in C. albicans, but have unique 3' sequences.

Northern blotting with ALS-specific probes:
Northern blot analysis was pursued to confirm that the ALS-hybridizing sequences detected on Southern blots encoded expressed genes. C. dubliniensis CD36 and C. tropicalis ATCC 13803 cells for RNA extraction were grown under a variety of conditions as described in MATERIALS AND METHODS above. For many growth conditions, C. dubliniensis showed multiple bands that cross-hybridized with the ALS1 tandem repeats probe (Fig 2). Hybridization of C. tropicalis RNA with the same probe failed to show strong signals with the exception of a high-molecular-weight band observed for RPMI-grown cells (data not shown). Lack of signals on C. tropicalis Northern blots may be due to difficulties in specific detection with C. albicans-derived sequences noted on Southern blots above and suggested that other means were needed to demonstrate a gene family in this organism.

View larger version (53K): In this window In a new window Download PPT slide   Figure 2. Northern blot of C. dubliniensis total RNA probed with the ALS1 tandem repeats fragment. RNA was isolated from strain CD36 grown overnight in YPD medium at 30° and 200 rpm shaking. Identical signals were also observed for CD36 grown in fresh YPD medium (30°) and RPMI medium (37°) for 2 and 5 hr and in Lee medium (37°) at four different pH values for 8 hr. Molecular size markers (in kilobases) are shown at the left of the blot.

Cell wall analysis of C. dubliniensis and C. tropicalis:
For cell wall analysis, C. dubliniensis CD36 and C. tropicalis 13803 were grown in YPD and RPMI media, respectively, since these conditions yielded the best signals on Northern blots described above. Western blot analysis of SDS-PAGE-separated, ß1,6-glucanase-released cell wall proteins with an antiserum raised against the N-terminal domain of Als5p revealed diffuse bands at 470 kD (Fig 3). These apparent molecular sizes were similar to those observed in the analysis of Als proteins in the C. albicans cell wall (KAPTEYN et al. 2000 ). These data suggested that, similar to C. albicans, both C. dubliniensis and C. tropicalis had Als proteins, which were incorporated in their cell wall through linkage to ß1,6-glucan.

View larger version (25K): In this window In a new window Download PPT slide   Figure 3. Western blot of C. dubliniensis CD36 and C. tropicalis 13803 cell wall extracts with an anti-Als serum.

Isolation of non-albicans Candida ALS sequences by PCR with consensus primers:
Southern and Western blot data presented above suggested that DNA encoding the 5' end of ALS genes was conserved across the three species studied. Alignment of the 5' domains of all known C. albicans ALS genes showed regions of sequence identity that could be used to design consensus PCR primers. Two forward primers and one reverse primer were selected from the aligned sequences. The combination of the first forward primer with the reverse primer predicted a PCR product of 1 kb; using the second forward primer predicted a 370-bp product. Amplification of genomic DNA from C. dubliniensis CD36 and C. tropicalis 13803 using the first primer set yielded PCR products of the expected size. Cloning of these products and DNA sequencing of selected clones revealed three distinct C. dubliniensis clones and one C. tropicalis clone with an open reading frame similar to the 5' end of C. albicans ALS genes. Because the newly isolated gene fragments did not directly correspond to known C. albicans ALS genes, nomenclature for the genes followed that in use for SAP genes in non-albicans Candida (MONOD et al. 1998 ): C. dubliniensis genes were designated ALSD1, ALSD2, and ALSD3; the C. tropicalis sequence was named ALST1. Despite sequencing many clones from C. tropicalis, no additional ALS-like coding regions were isolated. Cloning and sequencing of fragments isolated from amplification of C. tropicalis DNA with the second PCR primer pair revealed two new open reading frames (ALST2 and ALST3) that resembled ALS genes. Alignment of amino acid sequences from C. albicans, C. dubliniensis, and C. tropicalis Als proteins showed regions of conservation present in each; of particular note were the eight Cys residues, which were conserved in every sequence with the exception of the last Cys residue, which was missing from Alst2p (Fig 4). Comparison of the new ALS sequence fragments to corresponding regions of C. albicans ALS genes showed a lower degree of identity between C. tropicalis and C. albicans sequences (53 to 63% at the nucleotide level and 42 to 59% at the amino acid level), consistent with the weak hybridization signals observed in Northern and Southern blotting. Finding multiple ALS-like coding regions in C. dubliniensis and C. tropicalis suggested that ALS genes existed as a family in these organisms.

View larger version (98K): In this window In a new window Download PPT slide   Figure 4. Amino acid sequence alignment of predicted Als proteins corresponding to the PCR-amplified region. Amino acid sequences of Als proteins from C. albicans were aligned with those predicted from the PCR-amplified C. dubliniensis and C. tropicalis sequences. A consensus sequence is provided. The positions of conserved and semiconserved Cys residues are double-underlined in the consensus sequence. Amplification of the original clones with the Pfu proofreading polymerase and double-stranded sequencing of each fragment suggested that lack of the last Cys residue in Alst2p was not due to a PCR-induced or DNA sequencing error. Because their gene fragments were amplified with the second primer pair and yielded a shorter product, amino acid sequences of Alst2p and Alst3p do not begin until the third sequence block. Gaps in the alignment are denoted by periods; a tilde () is used to indicate sequence information that lies outside of the PCR-amplified region.

Chromosomal co-localization of ALS and SAP family sequences in C. dubliniensis and C. tropicalis:
Isolation of three C. dubliniensis and three C. tropicalis ALS sequences suggested the presence of an ALS gene family in each organism. The presence of SAP-like DNA sequences in C. dubliniensis was shown by cross-hybridization on genomic Southern blots and on CHEF gels (GILFILLAN et al. 1998 ). A SAP family in C. tropicalis was postulated from genomic Southern blots (MONOD et al. 1994 ) and substantiated by the presence of multiple C. tropicalis SAP gene sequences in the GenBank database. Previous work in C. albicans showed that ALS and SAP genes are located mainly on chromosomes 3, 6, and R (MONOD et al. 1994 , MONOD et al. 1998 ; HOYER et al. 1998A ; HOYER and HECHT 2000 ). To determine if this conservation of localization was also true for C. dubliniensis and C. tropicalis, Southern blots of CHEF-separated chromosomes were probed with SAP and ALS sequences.

Separation of C. dubliniensis chromosomes on a CHEF agarose gel showed a wide variability in karyotype between strains CD36 and CM1 (Fig 5, left). The presence of multiple C. dubliniensis strains with a karyotype similar to C. albicans has been demonstrated (GILFILLAN et al. 1998 ), although wide variation in karyotype for commonly studied C. dubliniensis isolates has also been shown (MAGEE et al. 1999 ). The ALS1 repeats and ALS5 repeats fragments hybridized to C. dubliniensis chromosomes the size of 3 and 6 and, in CM1, fragments that were likely derived from these chromosomes (Fig 5). These results matched data presented for C. dubliniensis where SAP probes hybridized to chromosomes the size of 3 and 6 and a fragment of similar size to R (GILFILLAN et al. 1998 ). Lack of signals for chromosome R with ALS probes indicated either that ALS sequences are not found on this chromosome or that ALS sequences present on chromosome R are sufficiently dissimilar in sequence that they cannot be detected with C. albicans-derived probes by high-stringency Southern hybridization.

View larger version (67K): In this window In a new window Download PPT slide   Figure 5. Chromosomal localization of ALS genes in C. albicans (Ca), C. dubliniensis (Cd), and C. tropicalis (Ct) strains as defined by hybridization with C. albicans ALS repeats probes. Chromosomes of two strains each of C. albicans, C. dubliniensis, and C. tropicalis were separated on a CHEF gel and stained with ethidium bromide (left). Subsequently, the gel was Southern blotted and probed with the ALS1 repeats (middle) and ALS5 repeats (right). Blots were hybridized at 65° and washed in 0.5x SSC/0.1% SDS at the same temperature. C. albicans chromosomes were numbered as previously indicated (WICKES et al. 1991 ). A numbering system has not been defined for C. dubliniensis chromosomes; C. tropicalis chromosomes were numbered from 1 to 7 (largest to smallest). The smudge at the upper left corner of the ALS5 repeats blot was not aligned with a single lane and is interpreted as a blotting artifact.

Limited references are available for the karyotype of C. tropicalis, but published information and experimentation with CHEF running conditions indicated that C. tropicalis chromosomes were separable with the same running conditions used for separating the largest C. albicans chromosomes (MAHROUS et al. 1992 ; Fig 5). Using these conditions, seven distinct chromosomal bands were separated and numbered from 1 to 7 (largest to smallest). Hybridization of C. tropicalis chromosomes with ALS sequences was done at high stringency to avoid potentially nonspecific hybridization and misleading signals. With this procedure the same two chromosomes were detected with each probe; however, hybridization signals were weak (Fig 5). The main question we sought to answer with this experiment was whether SAP and ALS sequences hybridized to the same chromosome. Since no SAP chromosomal localization data have been published for C. tropicalis, we constructed a consensus SAP probe by PCR using degenerate oligonucleotide primers designed from alignment of the C. tropicalis SAP sequences available in GenBank (see MATERIALS AND METHODS). Validation that this probe recognized C. tropicalis SAP genes was done by Southern blot of EcoRI-digested genomic DNA (Fig 6A). A similar blot probed with a consensus SAP oligo was reported by MONOD et al. 1994 . The fragments observed in our blot were all present in the published blot, demonstrating that our probe detected multiple C. tropicalis SAP genes. Hybridization of our SAP probe to CHEF-separated C. tropicalis chromosomes detected four chromosomal bands (Fig 6B). Hybridization of each C. tropicalis ALS gene to C. tropicalis chromosomes showed that ALS genes were located only on chromosomes where SAP genes are also found (Fig 6B). Interestingly, the ALST3 sequence hybridized to two chromosomes. Whether these chromosomes are homologous or whether there is more than one ALST3-like gene in C. tropicalis remains to be determined. Although characterization of the gene families in C. tropicalis is not complete, these initial data support the conclusion that ALS and SAP genes are found mainly on the same chromosomes in a variety of Candida species.

View larger version (65K): In this window In a new window Download PPT slide   Figure 6. Co-localization of ALS and SAP genes on C. tropicalis chromosomes. The PCR-amplified SAP consensus probe was hybridized to Southern-blotted, EcoRI-digested genomic DNA from two C. tropicalis strains (A) and also to a Southern blot of CHEF-separated chromosomes (B, right side). The chromosomes to which each C. tropicalis ALS gene fragment hybridized are indicated at the right of the blot. All blots were hybridized at 65° and washed in 0.5x SSC/0.1% SDS at the same temperature. Molecular size markers (in kilobases) are shown at the left of the genomic Southern blot in A.

Molecular evolution of the ALS family:
Phylogeny analysis of the ALS family was conducted to determine the oldest gene (most basal lineage) in the ALS family, to compare the rate of evolution of the ALS family to that of the SAP family, and to understand how the history of the ALS family compares to the history of C. albicans. To accomplish these goals, we constructed three maximum parsimony trees from amino acid sequences. UPGMA trees were also constructed with similar results (data not shown). The first tree was constructed to better resolve the C. albicans and C. dubliniensis family structure, specifying ALST1 as the root (Fig 7A). This phylogram was based on 336 characters of which 95 were constant and 156 were informative. The tree length was 708 with a consistency index of 0.8107 and a homoplasy index of 0.1893, including uninformative characters. C. tropicalis acted as an outgroup in the otherwise unrooted parsimony tree and placed ALS7 as the most basal lineage within C. albicans; ALS6 appeared to be the second most basal. The ALS4 and ALS6 ancestors existed before the C. albicans and C. dubliniensis split as each appeared to have a sister gene in C. dubliniensis. Because nearly all of the ALS genes from C. albicans have been characterized (HOYER and HECHT 2000 ), we concluded that ALSD2 probably arose within C. dubliniensis as it lacked a corresponding C. albicans gene. Concluding that ALS1, ALS2, ALS3, and ALS5 arose within C. albicans after the split from C. dubliniensis assumed that none of the other ALS sequences that were likely to exist in C. dubliniensis group with these genes.

View larger version (31K): In this window In a new window Download PPT slide   Figure 7. Phylogenetic trees. Maximum parsimony trees were constructed from Als and Sap amino acid sequences. Bars indicate the branch length that corresponds to 10 or 100 substitutions per 100 amino acids. Bootstrap values represent the percentage of times the topology was generated in 1000 replicates.

A second maximum parsimony tree was constructed solely from C. albicans sequences to determine the relative age of the genes in the family (Fig 7B). In this phylogram, there were 1016 characters of which 239 were constant and 410 were informative. The tree length was 1533, the consistency index was 0.9328, and the homoplasy index was 0.0672. This phylogram specified ALS7 as the root, based on results from the first tree. This analysis confirmed the conclusion that ALS6 was the second most-basal C. albicans lineage and that ALS2 and ALS4 were the youngest, most rapidly evolving genes in the family.

The third phylogram, with all sequences from C. albicans, C. dubliniensis, and C. tropicalis, was based on 127 total characters of which 33 were constant and 63 were informative (Fig 7C). Tree length was 340 with a consistency index of 0.76 and a homoplasy index of 0.24, including uninformative characters. Surprisingly, even though the SAP and ALS sequences examined co-localized in all three species, none of the known C. tropicalis ALS sequences grouped with C. albicans sequences as occurred for the SAP family (see below). This result was unexpected since genes located on the same chromosome in predominantly clonal organisms are expected to have the same evolutionary history (BARTON and WILSON 1996 ). As observed in the third tree (Fig 7C), the pairs of ALS6 with ALSD1 and ALS4 with ALSD3 indicated that family structure was detected in a close relative of C. albicans.

A maximum parsimony phylogram of all reported SAP sequences was found by a heuristic search in order to include additional C. tropicalis sequences not found in the most recently published SAP tree (MONOD et al. 1998 , Figure 7D). This new tree compared favorably with that reported by MONOD et al. 1998 . The SAP tree was based on 703 characters of which 140 were constant and 318 were informative. Tree length was 2408 with a consistency index of 0.75 and a homoplasy index of 0.25. The tree was rooted with the S. cerevisiae sequence YAP3. Unlike the ALS trees, the SAP sequences from C. tropicalis did not form a basal group separate from C. albicans. The C. tropicalis sequences instead grouped more distally, after branching off of the C. albicans SAP9 and SAP7 sequences.

ALS genes in other Candida species:
Because of the emerging similarities between the ALS and SAP gene families, species in which SAP genes have been documented are obvious ones to examine for the presence of ALS genes. An example of such an organism is C. parapsilosis (MONOD et al. 1994 ). Low-stringency hybridization of BglII-digested C. parapsilosis genomic DNA with the ALS1 repeats fragment showed a single cross-hybridizing fragment of 16 kb (data not shown). PCR amplification of C. parapsilosis DNA using both sets of ALS consensus primers yielded fragments of the predicted length; however, DNA sequencing of cloned fragments yielded sequences that did not resemble known ALS genes. Additional experimentation is required to define the nature of the C. parapsilosis cross-hybridizing fragment and to determine if this organism encodes ALS genes. Other species of Candida remain to be tested.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

DNA, RNA, and protein evidence reported here demonstrate that ALS gene families are found in C. dubliniensis and C. tropicalis. PCR screening procedures yielded the sequences of three ALS genes from each organism. These sequences revealed that the ALS family in the non-albicans species is not identical to that in C. albicans. Chromosomal analysis of each organism indicated that the ALS and SAP gene sequences are largely co-localized. Phylogenetic analysis of the ALS family suggests that the ALS family has a different evolutionary history from the SAP family as expected for an organism with an evolutionary history of mating (HULL et al. 2000 ; MAGEE and MAGEE 2000 ).

Comparative biology of Candida species:
Analysis of the ALS family in C. dubliniensis and C. tropicalis yielded new insights about two different biological processes in these organisms. First, release of Als proteins from the cell walls of C. dubliniensis and C. tropicalis with ß1,6-glucanase suggests that Als proteins in the non-albicans species encode the correct signals for cell wall localization and that the basic wall structure of these organisms is similar to that of S. cerevisiae and C. albicans (KAPTEYN et al. 1999 , KAPTEYN et al. 2000 ; SMITS et al. 1999 ). Second, studies of ALS genes in C. dubliniensis suggest differences in regulation of the gene family and in production of cell wall proteins. The multiplicity of similarly expressed ALS-hybridizing messages on Northerns of C. dubliniensis total RNA contrast sharply with the appearance of similar ALS Northerns for C. albicans in two ways: a seemingly increased number of expressed genes in C. dubliniensis and the apparent constitutive nature of the gene expression (HOYER et al. 1995 , HOYER et al. 1998A , HOYER et al. 1998B ). Understanding the phenotypic effect of this altered expression pattern requires additional analysis.

Phylogenetic relationship between ALS and SAP gene families:
The phylogenetic analysis of the ALS and SAP families focused on three main questions: (i) Which C. albicans ALS gene is the most basal and therefore likely to be the most ancestral form?, (ii) how does the rate of evolution compare between the ALS and SAP gene families?, and (iii) how does the evolutionary history of the ALS family compare with that of C. albicans? Phylogenetic reconstruction by maximum parsimony indicated that ALS7 is the most basal lineage of the C. albicans ALS family when the trees are rooted with C. tropicalis sequences; ALS6 is the second most-basal lineage in this analysis. Assuming a constant molecular clock for all ALS genes, this result implied that ALS7 is the oldest gene. However, ALS7 possesses a unique composition within the ALS family (HOYER and HECHT 2000 ), so it potentially evolved under a different clock. Since we believe that few C. albicans ALS sequences remain uncharacterized, we expect these results to be robust to the addition of sequences from other species.

Evidence for a younger ALS family was provided by examination of the ALS and SAP phylogenetic trees, which showed that all of the ALS sequences from C. tropicalis formed a basal group, while the SAP sequences from C. tropicalis branched off more distal nodes. One C. tropicalis SAP sequence even grouped as a sister with a C. albicans SAP sequence. If a family arose before the C. albicans/C. tropicalis split, the descendant species would each receive a copy of the gene and we would expect to see grouping by related family members across species as we do with SAP. If a family arose after the C. albicans/C. tropicalis split, we would see each species group separately, as occurs with ALS. Families that arise during a split exhibit a blend of these two patterns.

An alternative explanation for our molecular phylogeny data is concerted evolution between ALS genes, in which repeats have a homogenizing effect and all gene copies within a species become the same or very similar over time (ZIMMER et al. 1980 ; HUGHES 1999 ). Using this explanation, we would argue that the C. tropicalis sequences group together because of sequence homogenization within C. tropicalis, rather than because of ALS gene diversification after the C. tropicalis and C. dubliniensis/C. albicans split. Hence, the history of the gene family and the history of the species would not be the same. In this event, we would expect species to group separately and not in the pairs that occur for ALS4/ALSD3 and ALS6/ALSD1 (Fig 7). Because few additional ALS or SAP genes are likely to be found in C. albicans, it is unlikely the basal branches will change and we conclude that ALS is a younger family than SAP.

Given that ALS is a younger family, it is remarkable that the number of genes is comparable to that of the SAP family. It is possible that the repeats found in ALS genes provide a mechanism for the rearrangement and amplification of the family (BIERNE and MICHEL 1994 ; PARNISKE and JONES 1999 ). Homologous recombination between repeats could explain why the ALS and SAP families have such different histories.

Why are Candida gene families present?
The preservation and expansion of the Candida ALS and SAP families could suggest that gene families lend a selective advantage to the organism. Several possibilities exist to explain their presence. First, gene families may exist because of the need for multiple specificities of the same general function. For example, the various Sap proteins may digest specific proteins with varying degrees of efficiency while different Als proteins may allow Candida to adhere to a variety of host surfaces. Conversely, specific proteins in each family may have redundant function that provides backup function in case one protein in the family is compromised. One example of redundancy may be found in the hypha-specific genes in both the ALS and SAP families (HUBE et al. 1994 ; HOYER et al. 1999a; HOYER and HECHT 2000 ).

Other potential explanations for the presence of gene families include the possibility that a particular gene dosage is required to confer a specific phenotype on the cell. Although the necessary probes to detect all ALS and SAP genes in non-albicans Candida may not be defined, a positive correlation exists between the number of SAP and ALS genes and the frequency with which a given Candida species is isolated from clinical specimen: C. albicans has the most genes from each family and the gene number appears to decrease in C. tropicalis and further decrease in C. parapsilosis (DE VIRAGH et al. 1993 ). Perhaps the presence of additional proteins in each family signals greater colonization or pathogenic potential for that species.

Finally, the presence of two large gene families might suggest that products of each family have synergistic effects. These effects could occur between proteins of the same family or between proteins of the different families. Future research will clarify these possibilities and further define the role of gene families in Candida biology and pathogenesis.

	ACKNOWLEDGMENTS

We thank Patricia Kammeyer, David Coleman, and Richard Barton for Candida isolates. This work was supported by U.S. Public Health Service Grant AI39441, National Science Foundation Grant MCB-9630910, the Netherlands Technology foundation (STW) and the Earth Life Sciences Foundation (ALW). R.F. was supported in part by a Training Grant in Molecular and Cellular Mycology (T32-AI07373) from the National Institutes of Health.

Manuscript received September 4, 2000; Accepted for publication January 18, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BARNS, S. M., D. J. LANE, M. L. SOGIN, C. BIBEAU, and W. G. WEISBURG, 1991  Evolutionary relationships among pathogenic Candida species and relatives. J. Bacteriol. 173:2250-2255[Abstract/Free Full Text].

BARTON, N. H., and I. WILSON, 1996 Genealogies and geography, pp. 23–56 in New Uses for New Phylogenies, edited by P. H. HARVEY, A. J. L. BROWN, J. M. SMITH and S. NEE. Oxford University Press, New York.

BIERNE, H. and B. MICHEL, 1994  When replication forks stop. Mol. Microbiol. 13:17-23[Medline].

BORG-VON ZEPELIN, M., I. MEYER, R. THOMASSEN, R. WURZNER, and D. SANGLARD et al., 1999  HIV-protease inhibitors reduce cell adherence of Candida albicans strains by inhibition of yeast secreted aspartic proteases. J. Invest. Dermatol. 113:747-751[Medline].

CALDERONE, R. A. and P. C. BRAUN, 1991  Adherence and receptor relationships of Candida albicans.. Microbiol. Rev. 55:1-20[Abstract/Free Full Text].

CASSONE, A., F. DE BERNARDIS, A. TOROSANTUCCI, E. TACCONELLI, and M. TUMBARELLO et al., 1999  In vitro and in vivo anticandidal activity of human immunodeficiency virus protease inhibitors. J. Infect. Dis. 180:448-453[Medline].

DARWAZAH, A., G. BERG, and B. FARIS, 1999  Candida parapsilosis: an unusual organism causing prosthetic heart valve infective endocarditis. J. Infect. 38:130-131[Medline].

DE VIRAGH, P. A., D. SANGLARD, G. TOGNI, R. FALCHETTO, and M. MONOD, 1993  Cloning and sequencing of two Candida parapsilosis genes encoding acid proteases. J. Gen. Microbiol. 139:335-342[Medline].

DEVEREUX, J., P. HAEBERLI, and O. SMITHIES, 1984  A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12:387-395.

FRIDKIN, S. K. and W. R. JARVIS, 1996  Epidemiology of nosocomial fungal infections. Clin. Microbiol. Rev. 9:499-511[Abstract].

FU, Y., G. RIEG, W. A. FONZI, P. H. BELANDER, and J. E. EDWARDS, JR. et al., 1998  Expression of the Candida albicans gene ALS1 in Saccharomyces cerevisiae induces adherence to endothelial and epithelial cells. Infect. Immun. 66:1783-1786[Abstract/Free Full Text].

GAUR, N. K. and S. A. KLOTZ, 1997  Expression, cloning, and characterization of a Candida albicans gene, ALA1, that confers adherence properties upon Saccharomyces cerevisiae for extracellular matrix proteins. Infect. Immun. 65:5289-5294[Abstract].

GILFILLAN, G. D., D. J. SULLIVAN, K. HAYNES, T. PARKINSON, and D. C. COLEMAN et al., 1998  Candida dubliniensis: phylogeny and putative virulence factors. Microbiology 144:829-838[Abstract].

GRUBER, A., C. SPETH, E. LUKASSER-VOGL, R. ZANGERLE, and M. BORG-VON ZEPELIN et al., 1999  Human immunodeficiency virus type 1 protease inhibitor attenuates Candida albicans virulence properties in vitro. Immunopharmacology 41:227-234[Medline].

HILLIS, D. M., C. MORITZ and B. K. MABLE, 1996 Molecular Systematics, Ed. 2. Sinauer Associates, Sunderland, MA.

HOEGL, L., E. THOMA-GREBER, M. ROCKEN, and H. C. KORTING, 1998  HIV protease inhibitors influence the prevalence of oral candidosis in HIV-infected patients: a 2-year study. Mycoses 41:321-325[Medline].

HOPPE, J. E., M. KLAUSNER, T. KLINGEBIEL, and D. NIETHAMMER, 1997  Retrospective analysis of yeast colonization and infections in paediatric bone marrow transplant recipients. Mycoses 40:47-54[Medline].

HOYER, L. L. and J. E. HECHT, 2000  The ALS6 and ALS7 genes of Candida albicans.. Yeast 16:847-855[Medline].

HOYER, L. L. and J. E. HECHT, 2001  The ALS5 gene of Candida albicans and analysis of the Als5p N-terminal domain. Yeast 18:49-60[Medline].

HOYER, L. L., S. SCHERER, A. R. SHATZMAN, and G. P. LIVI, 1995  Candida albicans ALS1: domains related to a Saccharomyces cerevisiae sexual agglutinin separated by a repeating motif. Mol. Microbiol. 15:39-54[Medline].

HOYER, L. L., T. L. PAYNE, M. BELL, A. M. MYERS, and S. SCHERER, 1998a  Candida albicans ALS3 and insights into the nature of the ALS gene family. Curr. Genet. 33:451-459[Medline].

HOYER, L. L., T. L. PAYNE, and J. E. HECHT, 1998b  Identification of Candida albicans ALS2 and ALS4 and localization of Als proteins to the fungal cell surface. J. Bacteriol. 180:5334-5343[Abstract/Free Full Text].

HUBE, B., M. MONOD, D. A. SCHOFIELD, A. J. BROWN, and N. A. GOW, 1994  Expression of seven members of the gene family encoding secretory aspartyl proteinases in Candida albicans.. Mol. Microbiol. 14:87-99[Medline].

HUBE, B., R. RUCHEL, M. MONOD, D. SANGLARD, and F. C. ODDS, 1998  Functional aspects of secreted Candida proteinases. Adv. Exp. Med. Biol. 436:339-344[Medline].

HUGHES, A. L., 1999  Concerted evolution of exons and introns in the MHC-linked tenascin-X genes of mammals. Mol. Biol. Evol. 16:1558-1567[Abstract].

HULL, C. M., R. M. RAISNER, and A. D. JOHNSON, 2000  Evidence for mating of the "asexual" yeast Candida albicans in a mammalian host. Science 289:307-310[Abstract/Free Full Text].

KAPTEYN, J. C., H. VAN DEN ENDE, and F. M. KLIS, 1999  The contribution of cell wall proteins to the organization of the yeast cell wall. Biochim. Biophys. Acta 1426:373-383[Medline].

KAPTEYN, J. C., L. L. HOYER, J. E. HECHT, W. H. MULLER, and A. ANDEL et al., 2000  The cell wall architecture of Candida albicans wild-type cells and cell-wall-defective mutants. Mol. Microbiol. 35:601-611[Medline].

KUNOVA, A., J. TRUPL, A. DEMITROVICOVA, Z. JESENSKA, and S. GRAUSOVA et al., 1997  Eight-year surveillance of non-albicans Candida spp. in an oncology department prior to and after fluconazole had been introduced into antifungal prophylaxis. Microb. Drug Resist. 3:283-287[Medline].

LARONE, D. H., 1995 Medically Important Fungi. American Society for Microbiology, Washington, DC.

LEE, K. L., H. R. BUCKLEY, and C. C. CAMPBELL, 1975  An amino acid liquid synthetic medium for the development of mycelial and yeast forms of Candida albicans.. Sabouraudia 13:148-153[Medline].

LIPKE, P. N., D. WOJCIECHOWICZ, and J. KURJAN, 1989  AG1 is the structural gene for the Saccharomyces cerevisiae -agglutinin, a cell surface glycoprotein involved in cell-cell interactions during mating. Mol. Cell. Biol. 9:3155-3165[Abstract/Free Full Text].

MAGEE, B. B. and P. T. MAGEE, 2000  Induction of mating in Candida albicans by construction of MTLa and MTLalpha strains. Science 289:310-313[Abstract/Free Full Text].

MAGEE, B. B., M. SANCHEZ and P. T. MAGEE, 1999 Karyotypic analysis of the recently identified oral pathogen Candida dubliniensis, Abstract C52. ASM Conference on Candida and Candidiasis, Charleston, SC, American Society for Microbiology.

MAHROUS, M., A. D. SAWANT, W. R. PRUITT, T. LOTT, and S. A. MEYER et al., 1992  DNA relatedness, karyotyping and gene probing of Candida tropicalis, Candida albicans and its synonyms Candida stellatoidea and Candida claussenii.. Eur. J. Epidemiol. 8:444-451[Medline].

MONOD, M., G. TOGNI, B. HUBE, and D. SANGLARD, 1994  Multiplicity of genes encoding secreted aspartic proteinases in Candida species. Mol. Microbiol. 13:357-368[Medline].

MONOD, M., B. HUBE, D. HESS, and D. SANGLARD, 1998  Differential regulation of SAP8 and SAP9, which encode two new members of the secreted aspartic proteinase family in Candida albicans.. Microbiology 144:2731-2737[Abstract].

ORTI, G., D. E. PEARSE, and J. C. AVISE, 1997  Phylogenetic assessment of length variation at a microsatellite locus. Proc. Natl. Acad. Sci. USA 94:10745-10749[Abstract/Free Full Text].

PARNISKE, M. and J. D. JONES, 1999  Recombination between diverged clusters of the tomato Cf-9 plant disease resistance gene family. Proc. Natl. Acad. Sci. USA 96:5850-5855[Abstract/Free Full Text].

RANGEL-FRAUSTO, M. S., T. WIBLIN, H. M. BLUMBERG, L. SAIMAN, and J. PATTERSON et al., 1999  National epidemiology of mycoses survey (NEMIS): variations in rates of bloodstream infections due to Candida species in seven surgical intensive care units and six neonatal intensive care units. Clin. Infect. Dis. 29:253-258[Medline].

SANGLARD, D., B. HUBE, M. MONOD, F. C. ODDS, and N. A. GOW, 1997  A triple deletion of the secreted aspartyl proteinase genes SAP4, SAP5, and SAP6 of Candida albicans causes attenuated virulence. Infect. Immun. 65:3539-3546[Abstract].

SCHALLER, M., H. C. KORTING, W. SCHAFER, J. BASTERT, and W. C. CHEN et al., 1999  Secreted aspartic proteinase (Sap) activity contributes to tissue damage in a model of human oral candidosis. Mol. Microbiol. 34:169-180[Medline].

SMITS, G. J., J. C. KAPTEYN, H. VAN DEN ENDE, and F. M. KLIS, 1999  Cell wall dynamics in yeast. Curr. Opin. Microbiol. 2:348-352[Medline].

SUGITA, T. and T. NAKASE, 1999  Non-universal usage of the leucine CUG codon and the molecular phylogeny of the genus Candida.. Syst. Appl. Microbiol. 22:79-86[Medline].

SULLIVAN, D. J. and D. COLEMAN, 1997  Candida dubliniensis: an emerging opportunistic pathogen. Curr. Top. Med. Mycol. 8:15-25[Medline].

SULLIVAN, D. J., T. J. WESTERNENG, K. A. HAYNES, D. E. BENNETT, and D. C. COLEMAN, 1995  Candida dubliniensis sp. nov.: phenotypic and molecular characterization of a novel species associated with oral candidosis in HIV-infected individuals. Microbiology 141:1507-1521[Abstract].

WEINBERGER, M., T. SACKS, J. SULKES, M. SHAPIRO, and I. POLACHECK, 1997  Increasing fungal isolation from clinical specimens: experience in a university hospital over a decade. J. Hosp. Infect. 35:185-195[Medline].

WICKES, B., J. STAUDINGER, B. B. MAGEE, K. J. KWON-CHUNG, and P. T. MAGEE et al., 1991  Physical and genetic mapping of Candida albicans: several genes previously assigned to chromosome 1 map to chromosome R, the rDNA-containing linkage group. Infect. Immun. 59:2480-2484[Abstract/Free Full Text].

WINGARD, J. R., 1995  Importance of Candida species other than C. albicans as pathogens in oncology patients. Clin. Infect. Dis. 20:115-125[Medline].

WINGARD, J. R., W. G. MERZ, and R. SARAL, 1979  Candida tropicalis: a major pathogen in immunocompromised patients. Ann. Intern. Med. 91:539-543.

VAN'T WOUT, J. W., 1996  Fluconazole treatment of candidal infections caused by non-albicans Candida species. Eur. J. Clin. Microbiol. Infect. Dis. 15:238-242[Medline].

ZIMMER, E. A., S. L. MARTIN, S. M. BEVERLY, Y. W. KAN, and A. C. WILSON, 1980  Rapid duplication and loss of genes coding for the alpha chains of hemoglobin. Proc. Natl. Acad. Sci. USA 77:2158-