Fungal nucleobase transporters

REVIEW ARTICLE

Fungal nucleobase transporters Areti Pantazopoulou & George Diallinas Faculty of Biology, Department of Botany, University of Athens, Panepistimioupolis, Athens, Greece

Correspondence: George Diallinas, Faculty of Biology, Department of Botany, University of Athens, Panepistimioupolis, Athens 15781, Greece. Tel.: 3 0210 7274649; fax: 3 0210 7274702; e-mail: [email protected] Present address: Areti Pantazopoulou, Centro de Investigaciones Biologicas CSIC, Ramiro de Maeztu 9, Madrid 28040, Spain. Received 30 April 2007; revised 5 July 2007; accepted 5 July 2007. First published online 3 September 2007. DOI:10.1111/j.1574-6976.2007.00083.x ´ D´ıaz Orejas Editor: Ramon

Abstract Early genetic and physiological work in bacteria and fungi has suggested the presence of highly specific nucleobase transport systems. Similar transport systems are now known to exist in algae, plants, protozoa and metazoa. Within the last 15 years, a small number of microbial genes encoding nucleobase transporters have been cloned and studied in great detail. The sequences of several other putative proteins submitted to databases are homologous to the microbial nucleobase transporters but their physiological functions remain largely undetermined. In this review, genetic, biochemical and molecular data are described concerning mostly the nucleobase transporters of Aspergillus nidulans and Saccharomyces cerevisiae, the two model ascomycetes from which the great majority of data come from. It is also discussed as to what is known on the nucleobase transporters of the two most significant pathogenic fungi: Candida albicans and Aspergillus fumigatus. Apart from highlighting how a basic process such as nucleobase recognition and transport operates, this review intends to highlight features that might be applicable to antifungal pharmacology.

Keywords purine; pyrimidine; transcriptional regulation; trafficking; structure–function relationships; kinetic modelling.

Introduction Bacteria, fungi, protozoa, plants, insects and mammalian tissues, all have the ability to take up purines and pyrimidines. The universality of specific purine–pyrimidine transport systems is reflected in the importance of bases in nucleotide and nucleic acid biosynthesis, but also to several other roles related to cellular nutrition, cell communication and defence mechanisms. Most fungi can use purines, but not pyrimidines, as fairly good nitrogen sources (Scazzocchio, 1982). This is due to the degradation of purines, first to ureides (allantoin, allantoic acid) and eventually to urea, via several enzyme-catalysed oxidations. The complete purine catabolic pathway (Fig. 1) is present in most filamentous fungi and is the same as that of most bacteria and plants. In contrast, most yeasts have degenerate variations of the purine degradation pathway. Interestingly, most lack xanthine dehydrogenase (HxA, also known as purine hydroxylase I), the major enzyme-oxidizing hypoxanthine to xanthine and xanthine to uric acid, and urease (UreA), the last enzyme-oxidizing allantoic acid to urea (Scazzocchio, 1982). Some yeasts, like Candida albicans and Schizosaccharomyces pombe, can use purines as nitrogen sources through FEMS Microbiol Rev 31 (2007) 657–675

XanA (xanthine a-ketoglutarate-dependent dioxygenase; Goudela et al., 2005; see Fig. 1). Saccharomyces cerevisiae lacks all the enzymes necessary for purine oxidation, while conserving the ability to break down ureides to urea, the last step taking place via the allophanate pathway, and not through a typical urease. Thus, Saccharomyces cerevisiae can grow on allantoin or allantoic acid, but not on purines as sole nitrogen sources. Early genetic and biochemical studies established the presence of highly specific nucleobase transporters in fungi. The lack of growth on purines or the use of purine or pyrimidine toxicity, caused either by an excess of a base (e.g. uracil, uric acid), or by a cytotoxic analogue (e.g. allopurinol, 5-fluorouracil, 8-azaguanine), provided a powerful tool to select mutants and identify the corresponding genes (Darlington & Scazzocchio, 1967; see Fig. 1b). Out of the work performed mostly in Saccharomyces cerevisiae and Aspergillus nidulans, but also with Neurospora crassa, Candida albicans and Schizosaccharomyces pombe, it became evident that fungi possess at least three distinct nucleobase uptake systems: one for adenine– guanine–hypoxanthine–cytosine, one for uric acid–xanthine and one for uracil. 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

658

(a)

A. Pantazopoulou & G. Diallinas

transporters Guanine

Adenine

adenine deaminase nadA

guanine deaminase

AzgA & FcyB in A. nidulans/A. fumigatus Fcy2p/Fcy21p in S. cerevisiae/C. albicans

Hypoxanthine purine hydroxylase II hxnS

purine hydroxylase I hxA

Xanthine xanthine dioxygenase xanA

UapA/UapC in A. nidulans/A. fumigatus

hxA

Xut1p in C. albicans

Uric Acid uricase uaZ

(b)

A. nidulans mutant strains on xanthine

Allantoin allantoicase alX

Allantoic acid

wt

(c)

allantoicase aaX

Ureidoglycolic acid

uapA uapC

uaZ

Urea

ureidoglycolase

Urease ureB,C,D

Ammonia

uric acid

xanthine

Families and genomics Cloning and genome sequencing showed that the fungal nucleobase-specific transporters belong to three evolutionary distinct proteins families (De Koning & Diallinas, 2000; http://www.membranetransport.org): the nucleobase cation symporter family 1 (NCS1), also known as the purinerelated transporter family (PRT), the nucleobase-ascorbate transporter family (NAT or NCS2) and the AzgA-like family. All these families are classified as secondary active transporters catalysing the symport of purines with protons (these two chemical species are transported together in the same direction in a tightly coupled process, not coupled to a direct form of energy other than chemiosmotic energy; http:// www.tcdb.org/tcdb/). Tables 1 and 2 summarizes all recognized fungal nucleobase-specific transporters that are discussed in more detail below.

The NAT/NCS2 family This family consists of hundreds of currently sequenced proteins derived from Gram-negative and Gram-positive bacteria, archaea, fungi, diatoms, plants and animals. Proteins of the NAT family are 414–650 amino acid residues in length, probably possess 12 transmembrane a-helical spanners (TMSs) and cytoplasmic N- and C-termini, characteristics that resemble the structure of major facilitator superfamily (MFS) members. The MFS (http:// www.tcdb.org/tcdb/2.A.1) is a very old, large, diverse and extensively studied superfamily that includes members that catalyse uniport, solute : cation (H1 or Na1) symport and/ or solute : H1 or solute : solute antiport. Most are 400–600 amino acid residues in length and possess 12 putative 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

hxA

C. albicans on purines

hypoxanthine

Fig. 1. (a) Purine catabolism in fungi. Purine transporters (UapA, UapC, AzgA, FcyB, Xut1p, Fcy2p, Fcy21p) and key catabolic enzymes are discussed in the text. (b) Growth phenotypes of Aspergillus nidulans mutants on purine (uric acid) as the sole nitrogen source: wild-type (wt) growth, lack of nitrogen (uric acid) utilization exemplified by an hxA mutant; lack of nitrogen (uric acid) uptake exemplified by a uapA uapC double mutant; toxicity of uric acid due to increased accumulation because of a metabolic block caused by a uaZ (uricase) null mutation. (c) Differential growth of Candida albicans on purines (uric acid 4 xanthine 4 hypoxanthine). Saccharomyces cerevisiae does not grow on purines as the sole nitrogen source due to lack of several enzymes and transporters (see text).

transmembrane a-helical spanners. Evidence exists that the 12 TMS arose from a three TMS element by two successive duplication events, and this is reflected in a twofold symmetry (616 TMS separated by a long connecting loop) in the high-resolution three-dimensional structures of the glycerol-3-P : P antiporter (GlpT; TC #2.A.1.4.3) and the lactose : H1 symporter (LacY; TC #2.A.1.5.1). The substrate pathway is predicted to exist between the two halves of the permeases using an alternating access mechanism with a single substrate binding site. However, there is no significant primary amino acid sequence or common motives between the NAT/NCS2 and the MFS. Moreover, the predicted secondary structure of NAT/NCS2 is not compatible with the twofold symmetry (616 TMS) of MFS. Most functionally characterized NAT/NCS2 members, arising from bacteria (Escherichia coli, Bacillus subtilis), fungi (A. nidulans, Aspergillus fumigatus, Candida albicans) and plants (maize), are specific for either oxidized purines, such as xanthine and/or uric acid, or uracil (only in bacteria). Fungal uracil transporters do not belong to NAT/ NCS2 but to NCS1/PRT. Several of the microbial NAT/ NCS2 proteins have been shown to be nucleobase: H1 symporters. However, the two closely related mammalian members of the family, SVCT1 and SVCT2, cotransport Lascorbate and Na1 with a high degree of specificity and high affinity for the vitamin (Liang et al., 2001). Aspergillus nidulans has two NAT/NCS2 members, called UapA and UapC, that are extensively characterized with respect to transcriptional regulation, down-regulation by endocytosis and sorting in the vacuoles in the presence of ammonium, expression during asexual and sexual development and structure–function relationships. UapA is a 574 FEMS Microbiol Rev 31 (2007) 657–675

659

Fungal nucleobase transporters

Table 1. Classification, specificity and kinetics of fungal nucleobase-specific transporters Fungus/ transporter

Family

Physiological substrates [Km/i (mM)]

A. nidulans UapA UapC

NAT/NCS2 NAT/NCS2

X [7], UA [8] X[4], UA[136]

AzgA FcyB FurD

AZGA-like NCS1/PRT NCS1/PRT

HX[1.5], AD[3], GU[3] AD[7], HX[20], CY[20], GU[12] UR [0.4]

NAT/NCS2 AZGA-like NCS1/PRT NCS1/PRT

X [6], UA [171] AD[3.5], HX[6], GU[8] ND ND

8MX [50], 1MX [60], 2TX [65], 3MX [100], OX [103], 2,6 DAPU [3], PU [40], 6TPU [48], 9MGU [69] ND ND

H M L H

NCS1/PRT NCS1/PRT

AD[1.8], HX[2.5], CY [1.8], GU[ND] UR [2.5], URD [ND]

ND 5FUR, 5FURDz

H H

NAT/NCS2

X [4], UA [50]

H

Fcy21p

NCS1/PRT

H

Fur4p‰

NCS1/PRT

HX[4], AD[16], GU[53], CY [4] ND

2TX [30], 2TUA [97], 3MX [22], 6TUA [84], 8MX [80], 8AZX [55], OX [4], 3MAD [22], 3DAZG [72], 5MCY [20], 5FCY [35] ND

H

A. fumigatus UapC AzgA FcyBw FurDw S. cerevisiae Fcy2p Fur4p C. albicans Xut1p

Other ligands o 100 mM [Km/i (mM)] 2TX [63], 3MX [28], 8MX [100], OX [103], ALL 1MX [o 50], 2TX [o 50], 3MX [100], 8MX [100], OX [38], ALL PU [99], 6TPU [78], 2,6DAPU [2], 8AZX [11] ND THY [3.3], 1MUR [4], 2TUR [8], 4TUR [14], 5FUR [0.5], 6AZUR [2], 6MUR [15], X [94], UA [99]

Transport capacity H M H L H

Allopurinol affinity cannot be determined due to non-Michaelis–Menten kinetics (M. Koukaki, G. Diallinas, unpublished data). Based on genetics and

growth tests allopurinol is formally shown to be a substrate of UapA and UapC and that concentrations as low as 2 mM lead to allopurinol sensitivity. FcyB and FurD orthologues of Aspergillus fumigatus have only been recognized in silico based on the Aspergillus nidulans proteins. z Based on growth tests. ‰ A Fur4p orthologue of Candida albicans has only been recognized in silico based on the Saccharomyces cerevisiae protein. X, xanthine; UA, uric acid; AD, adenine, HX, hypoxanthine; GU, guanine; CY, cytosine; UR, uracil; 1MX, 1-methylxanthine; 2TX, 2-thioxanthine; 3MX, 3-methylxanthine; 8MX, 8-methylxanthine; OX, oxypurinol; ALL, allopurinol; PU, purine; 6TPU, 6-thiopurine; 2,6DAPU, 2,6-diaminopurine; 8AZX, 8-azaxanthine; THY, thymine; 1MUR, 1-methyluracil; 2TUR, 2-thiouracil; 4TUR, 4-thiouracil; 5FUR, 5-fluorouracil; 6AZUR, 6-azauracil; 6MUR, 6methyluracil; 9MGU, 9-methylguanine; 2TUA, 2-thiouric acid; 6TUA, 6-thiouric acid, 3MAD, 3-methyladenine; 3DAZG, 3-deazaguanine; 5MCY, 5-methylcytosine; 5FCY, 5-fluorocytosine; URD, uridine; FURD, 5-fluorouridine. H, M and L stand for high, moderate or low capacity of transport, respectively, as this is judged by calculated Vm or/and growth tests. ND, not determined. w

amino acid, high-affinity, high-capacity, transporter responsible for the uptake of uric acid and xanthine, also able to transport 2-thiouric acid, 2-thioxanthine, 3-methyloxanthine, allopurinol and oxypurinol (Koukaki et al., 2005; Table 1). UapC is a 580 amino acid protein, a very similar paralogue of UapA (62% identity), which has a high affinity for xanthine and a moderate affinity for uric acid and other xanthine analogues (Table 1). UapC also seems to have a very low affinity (Km 4 500 mM) and very low capacity for transport of other purines, not recognized by UapA, such as adenine or hypoxanthine (E. Tsilivi & G. Diallinas, unpublished data). A single A. fumigatus UapA/UapC homologue (61–63% identity with UapA or UapC) has been characterized kinetically by expression in an A. nidulans strain carrying deletions of its endogenous purine transporter genes (S. Goudela, U. Reichard, G. Diallinas & S. Amillis, unpublished data). This carrier resembles UapC with respect to its substrate affinity and specificity (see Table 1) but has a FEMS Microbiol Rev 31 (2007) 657–675

high transport capacity, similar to UapA. A C. albicans homologue (47% identity to the Aspergilli proteins) is a high-capacity transporter with a high affinity for xanthine and a moderate affinity for uric acid (Goudela et al., 2005; Table 1). UapA/C homologues, usually one in each species, are encoded by genes in most fungi [Neurospora crassa, Gibberella zeae, Magnaporthe grisea, Fusarium gramineum, Coprinus cinereus, Phanerochaete chrysosporium, Ustilago maydis (66–69% identities), Schizosaccharomyces pombe, Cryptococcus neoformans (56%), Kluyveromyces lactis, Yarrowia lipolytica and Debaryomyces hansenii (44–47%)]. Saccharomyces cerevisiae does not have any NAT/NCS2 protein, reflecting its lack of enzymes involved in xanthine or uric acid utilization. In fact, very few organisms lack NAT/NCS2 transporters, the most prominent examples being protozoa. Plants have numerous NAT/NCS2 transporters (23–25% identical to UapA), the only one with known function being Lpe1, a high-affinity, high-capacity, uric acid–xanthine/H1 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

660

A. Pantazopoulou & G. Diallinas

Table 2. Overview of fungal nucleobase transporter regulation Regulation of expression Developmental

Physiological Transcriptional

Posttranscriptional Posttranslational

Nitrogen Nitrogen Fungus/ Vegetative Asexual Sexual Nucleobase Nitrogen starvation or substrate transporter Germination form differentiation differentiation induction repression induction mRNA decay A. nidulans UapA 1 UapC 1 AzgA 1 FcyB 1 FurD 1 S. cerevisiae Fcy2p NR Fur4p NR

Nitrogen Substrate inactivation inhibition

1 1 1 1 1

1 1w  1 ND

1 ND ND 1 ND

1 1 1  

1 1 1 1 

1 1 1 1 

 ND ND ND 1z

1 1 1 1 

1 ND ND ND 1

1 1

NR NR

ND ND

 

ND 

ND 

ND 1

 

 1

Expressed in the metulae. w

Expressed in the metulae in the presence of uric acid. Uracil addition to growth medium leads to down-regulation of FurD mRNA steady-state levels but there is no formal proof as to whether this is due to repression of de novo synthesis or increased mRNA turnover. ND, not determined; NR, not relevant. z

symporter, necessary for chloroplast development in maize (Argyrou et al., 2001). Recently, several Arabidopsis NAT proteins have been cloned and studied with respect to differential transcription in various plant tissues. Despite analysis of double and triple knock-out mutants, no phenotype was assigned to any of them (Maurino et al., 2006). Drosophila melanogaster has a single NAT homologue (G. Diallinas, C. Delidakis & C. Gournas, unpublished observations), Caenorhabditis elegans six (G. Diallinas, M. Koukaki & N. Tavernarakis, unpublished observations) and mammals two (SVCT1 and SVCT2), the latter having diverged to become ascorbate/Na1 symporters (Liang et al., 2001). All these metazoan proteins are 23–24% identical to UapA and 35–40% identical among themselves. The insect and nematode NAT proteins are of unknown specificities and despite an overall greater similarity to the mammalian vitamin C transporters, functional motifs are more similar to the Aspergillus proteins. Interestingly, despite biochemical evidence for the existence of specific nucleobase uptake activities in mammals, no genes similar to the known fungal nucleobase transporters have been identified (De Koning & Diallinas, 2000).

The NCS1/PRT family Members of this family are, in general, 419–635 amino acid residues long, and possess, most probably, 12 putative TMSs. N- and C-termini are predicted to be cytoplasmic and there is no long hydrophilic loop or repeated motifs, as in MFS members. At least some of them have been shown to 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

function in uptake by substrate: H1 symport. The NCS1/ PRT family is restricted to prokaryotes, fungi and plants and includes transporters for purines, cytosine, uridine, allantoin, pyridoxine or thiamine (http://www.tcdb.org/). Based on the fact that some bacterial NAT/NCS2 transporters (e.g. PbuX) have similarities to NCS1/PRT members, the two families were proposed to be distantly related (http:// www.tcdb.org/) (De Koning & Diallinas, 2000). Purine and pyrimidine transporters of the NCS1/PRT family were characterized more than 15 years ago in Saccharomyces cerevisiae. Two of them, the Fcy2p adenine– guanine–hypoxanthine–cytosine permease (Weber et al., 1990) and the Fur4p uracil permease (Jund et al., 1988), have become paradigms for studies concerning transporter function and cellular regulation. Fcy2p and Fur4p share limited overall sequence identity (19% identity) but show significant local similarity and share common motifs (De Koning & Diallinas, 2000). The FCY2 gene of Saccharomyces cerevisiae encodes a 533 amino acid-long protein compatible with 9–12 TMS (Weber et al., 1990; Andr´e, 1995). Biochemical evidence for a nine TMS topology has been reported (Andre´, 1995; Ferreira et al., 1997). Fcy2p has been shown to be a high-affinity (see Table 1), high-capacity, adenine– guanine–hypoxanthine–cytosine/H1 symporter (Chevallier et al., 1975; Brethes et al., 1992). Genetic and molecular studies have also addressed posttranslational regulation (Weber et al., 1990; Bloch et al., 1992; Ferreira et al., 1997, 1999) and structure–function relationships of Fcy2p. A Saccharomyces cerevisiae protein similar to Fcy2p has been shown to be involved in pyridoxine uptake (Tpn1p) (Andre´, FEMS Microbiol Rev 31 (2007) 657–675

661

Fungal nucleobase transporters

1995). Two very close homologues (85–89% identity) of Fcy2p of unknown physiological function, called Fcy21p and Fcy22p, and Tpn1p and Fur4p contribute to 5-fluorocytosine toxicity in Saccharomyces cerevisiae (Paluszynski et al., 2007). The FUR4 gene of Saccharomyces cerevisiae encodes a protein of 633 amino acid residues (Jund et al., 1988). The Fur4p protein is a high-affinity, high-capacity, uracil/H1symporter and might also be involved in minor uridine transport (Wagner et al., 1998; Table 1). Other Fur4p-like proteins in Saccharomyces cerevisiae have been shown to be involved in the bulk uptake of uridine (Fui1p; 41% identical to Fur4p), allantoin (Dal4p; 76% identical to Fur4p) or thiamine (Thi7p; 30% identical to Fur4p) (Andre´, 1995). Fur4p topology has been studied experimentally (Garnier et al., 1996). The N- and C-tails of the protein were shown to be cytoplasmic using several biochemical approaches. Particularly, the N- and C-tails were accessible to trypsin degradation, when appropriate epitopes were added to the termini, or to immunofluorescent antibodies against the last 10 amino acids, only in permeabilized protoplasts. In addition, carboxypeptidase digested the C-terminus of uracil permease inserted into the sealed dog-pancreas microsomes. The orientation of several hydrophilic loops with respect to the membrane has also been investigated by introducing potential glycosylation sites into these regions. The resulting mutant proteins were analysed for glycosylation when expressed in the presence of dog-pancreas microsomes. Two loops of the protein were found to be lumenal. Together with the in silico predictions, these result indicated that uracil permease is a 10 TMS transporter, with rather small external loops and three main cytoplasmic regions (the N- and C-termini and a central 60-residue loop). Structure–function relationships in Fur4p have also been approached using mutational analysis that will be discussed later. The most important contribution of studies on Fur4p, however, concerns the discovery of mechanisms that control the dynamic trafficking of this transporter in response to various physiological signals, which is also discussed later. Candida albicans has transport systems with kinetic and specificity properties very similar to Fur4p and Fcy2p (Rao et al., 1983; Goudela et al., 2006). In contrast, Candida glabrata has a transport system for adenine–guanine (Km values 0.6–5.7 mM), a different transport system for cytosine (Km 4.0 mM) and no hypoxanthine transporter (Gupta et al., 1995). Several Fcy2p and Fur4p homologues of unknown function exist in all other yeasts. Recently, FcyB and FurD, two A. nidulans homologues (45–55% identities) of Fcy2p and Fur4p, have been functionally characterized (Amillis et al., 2007; A. Vlanti & G. Diallinas, unpublished data). The FurD protein is able to recognize with high affinity not only uracil but also thymine and several 5-substituted analogues of uracil, and with moderate-affinity uric acid and xanthine (Table 1). Kinetic evidence supports a model on how FurD FEMS Microbiol Rev 31 (2007) 657–675

recognizes uracil. Several other, in silico identified, Fur4p/ FurD and Fcy2p/FcyB homologues in A. nidulans and other fungi might be specific, among others, for uridine, pyridoxine, thiamine or allantoin (Z. Hamari, S. Amillis, G. Diallinas & C. Scazzocchio, unpublished data).

The AzgA-like family This family includes homologues in bacteria, archaea, fungi and plants, but only the A. nidulans and A. fumigatus proteins have been characterized as hypoxanthine–adenine– guanine/H1 symporters (Cecchetto et al., 2004; Goudela et al., 2006; S. Goudela, U. Reichard, G. Diallinas & S. Amillis, unpublished data). Fungal proteins share up to 35% and 44% identity with bacterial and plant homologues, respectively, while identities among fungi vary from 45% to 75%. AzgA-like proteins are 423–594 amino acids long and predicted to possess 10–12 TMSs. Based on overall primary and secondary sequence comparisons, AzgA-like proteins are grouped as a separate subfamily within the MFS (http:// www.tcdb.org/). However, AzgA also shares some common features with NAT/NCS2, a family clearly grouped outside the MFS. Morover, we could not detect the twofold symmetry present in MFS proteins. The AzgA protein (580 amino acids, 12 TMS) of A. nidulans is a high-affinity, highcapacity, transporter specific for adenine, guanine, hypoxanthine, 8-azaxanthine and 2,6-diaminopurine. It also transports efficiently the toxic analogues purine and 8-azaguanine (Cecchetto et al., 2004; Goudela et al., 2006; Table 1). A very similar A. fumigatus protein (75% identity) has been expressed in A. nidulans and shown to be very similar in function and specificity with AzgA (S. Goudela, U. Reichard, G. Diallinas & S. Amillis, unpublished data). A kinetic model was proposed on how the AzgA carriers of A. nidulans and A. fumigatus interact with their substrates (Goudela et al., 2006; S. Goudela, U. Reichard, G. Diallinas and S. Amillis, unpublished data). No homologues are present either in Saccharomyces cerevisiae, Candida albicans or any known hemiascomycete yeasts. Schizosaccharomyces pombe has a very close homologue (59%). Among filamentous fungi, no homologues were found in Cryptococcus neoformans, but several in most other fungi: one in Aspergillus oryzae, Coprinus cinereus, Phanerochaete chrysosporium and U. maydis, two in N. crassa, G. zeae and M. grisea and three in F. gramineum.

Regulation of purine uptake at the level of transcription Most fungal catabolic pathways are inducible and repressible at the level of transcription. Expression of transporters taking-up solutes that can be used as nitrogen or carbon sources is a primary target of transcriptional control circuits. This is exemplified by the three major purine transporters of 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

662

A. nidulans, UapA, UapC and AzgA. Transporters are also very efficient scavengers for supplying energy-cheap metabolites, thus avoiding the cost of biosynthesis, or other necessary metabolites that are not synthesized by a fungus. Scavenging transporters are, in principle, expressed constitutively, and this was shown for the FUR4 (uracil) and FCY2 (cytosine, adenine, guanine, hypoxanthine) genes in Saccharomyces cerevisiae, a yeast that only takes up purines for anabolic purposes. Table 2 summarizes the regulation of transcriptional activation of all known fungal transporters.

Regulation of A. nidulans purine transporters in response to nitrogen source and purine availability The regulation of expression of the uapA, uapC and azgA genes is mediated, at the level of mRNA transcription, in response to purine induction and nitrogen metabolite repression. The same mechanism regulates the expression of most genes encoding the enzymes involved in purine catabolism (Scazzocchio, 1982, 1994). In vegetative mycelia or germinated germlings growing in the absence of an inducer (uric acid, other purines or the gratuitous inducer 2-thiouric acid), uapA, uapC and azgA are transcribed at low basal levels, when a nonrepressing nitrogen source (urea, proline) is present in the medium. Upon purine induction, uapA, uapC and azgA message accumulation increases eight-, three- and ninefold, respectively (average increases in several experiments appearing in a number of publications; Amillis et al., 2004; Cecchetto et al., 2004). The presence of ammonium or glutamine (4 5 mM) drastically represses the expression of uapA and uapC, and partially that of azgA, irrespective of the presence of an inducer (uric acid or any other purine). As the physiological inducer of the purine catabolic pathway is uric acid, the tight repression of uapA and uapC, genes that encode for the two sole ‘gates’ for uric acid uptake, results in inducer exclusion, and thus provides a ‘double-lock’ mechanism for repression of purine utilization. Uric acid induction is mediated by the positive-acting regulatory protein UaY. This pathway-specific regulator is a constitutively expressed transcriptional activator, which, in the presence of uric acid, mediates the induction of most genes involved in purine utilization in A. nidulans (Scazzocchio, 1994). UaY contains a typical zinc binuclear cluster domain through which it binds to the promoter regions of, among others, the transporter genes uapA and uapC (Suarez et al., 1995). A fine detailed analysis including gel shifts, DNAse I footprinting and interference assays showed that the binding site of UaY is 5 0 -TCGG-6X-CCGA, but nevertheless, the identity of the base immediately 3 0 of the 5 0 TCGG sequence affects the affinity of UaY for this site. Ammonia repression is mediated by the inactivation of the general transcription factor AreA, which is necessary for the 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

A. Pantazopoulou & G. Diallinas

transcription of more than 100 genes encoding products implicated in the utilization of nitrogen sources in A. nidulans (Kudla et al., 1990). Gel shifts, DNAse I footprinting and interference assays have shown that the binding site of AreA is 5 0 -HGATAR (Muro-Pastor et al., 1999). The single DNA-binding domains of AreA and the homologous fungal proteins show striking similarities to the zinc fingers and adjacent basic regions of the metazoan GATA factors (Kudla et al., 1990). AreA mediates nitrogen metabolite repression by responding to the glutamine or ammonia concentration in the cell, while glutamine is thought to destabilize the areA mRNA (Platt et al., 1996; Morozov et al., 2001). In fact, the specificity of the transcript degradation response was qualitatively assessed with respect to a variety of nitrogen sources. The down-regulation response to glutamine, asparagine and NH41 required the same discrete region of the areA 3 0 -UTR, but both NH41 and asparagine needed to be metabolized to glutamine before they were effective as a signal. uapA and uapC maximum message accumulation is absolutely dependent on both functional UaY and AreA products, while the expression of the azgA gene is partially dependent, showing lower but still detectable levels in uaY and areA backgrounds (Diallinas & Scazzocchio, 1989; Gorfinkiel et al., 1993; Diallinas et al., 1995). The molecular and functional analysis of a number of areA DNA-binding specificity mutations and cis-acting regulatory mutations in the promoter regions of uapA and uapC have proved to be valuable tools in identifying the physiological binding sites, among several other putative sequences conforming to the consensus target sites, of both UaY and AreA (Diallinas & Scazzocchio, 1989). A change of a universally conserved leucine residue (L683) to valine in the DNA-binding domain of AreA results in the inability to activate the promoters of uapA and uapC, while some other promoters become able to function more efficiently than in the wild-type context. A methionine in the same position (M683) results in a less extreme but mirror-image effect. The uapA and uapC cis-acting mutations have been selected as specific suppressors of the AreA V683 mutation and established the identity of the physiological AreA- and UaY-binding sites. They further showed that interactions between the conserved leucine at the seventh position of the zinc finger of AreA and the first base upstream of the physiological GATAbinding sequences define the efficiency of transcriptional activation (Ravagnani et al., 1997). This work provided a rationale for the almost universal evolutionary conservation of leucine at the seventh position of the Zn finger of GATA factors. Molecular models of the wild-type and mutant AreA–DNA complexes account both for the phenotypes observed in vivo and the binding differences observed in vitro (Ravagnani et al., 1997) Finally, a specific uapC promoter mutation, which creates a second UaY-binding site very close FEMS Microbiol Rev 31 (2007) 657–675

663

Fungal nucleobase transporters

to the physiological site, suppresses the inefficient binding of a mutant form of AreA (AreA V683), suggesting that UaY binding in the promoter region of purine transporter genes should play a critical role in directing the binding of the general factor AreA, and thus initiating chromatin remodelling and gene expression activation. (G. Diallinas, N. Oestreicher & C. Scazzocchio, unpublished data).

Regulation of A. nidulans purine transporters during conidiospore germination Conidiospores of Aspergillus are asexually produced dormant, mononucleate cells that are characterized by very low rates of metabolism. In a suitable environment, dormancy breaks, conidiospores swell rapidly, the nucleus reorganizes and a germ tube emerges to eventually develop into a growing hypha (Momany, 2002). The term ‘germination’ is usually defined as the set of very early events required for sensing and transmitting the signal to germinate. A recent model for conidial germination is emerging, in which the first essential step is the uptake of a carbon source, and the subsequent activation of the rasA and cAMP signalling pathways (Osherov & May, 2000; Fillinger et al., 2002). Conidial germination seems to specifically involve the transcriptional activation of genes encoding transporters. A detailed analysis has been performed for the purine transporters UapA, UapC and AzgA (Amillis et al., 2004). Transcriptional activation of the uapA, uapC and azgA genes occurs during the isotropic growth phase (Momany, 2002), before the first nuclear division, and leads to the appearance of the corresponding purine transport activities within a short time delay (30–60 min). uapA, uapC and azgA transcriptional activation is independent of UaY, the pathwayspecific transcription factor, as both loss-of-function (uaY) and constitutive (uaYc) mutations have no effect on transcription during germination. In fact, the only requirement for this de novo transcriptional activation is hydration of dormant conidiospores (Amillis et al., 2004). The presence of a carbon source (glucose or fructose) in the germination medium is, however, necessary for translation of uapA, uapC or azgA mRNA messages to purine uptake activities, possibly by providing a necessary energy source. The temperature, pH, the quality of the carbon source (glucose vs. fructose) or the presence of different nitrogen sources (urea, proline, nitrate), including ammonium or glutamine, which in mycelia repress all genes involved in nitrogen utilization, were shown not to affect the transient activation of purine transporter transcription. The lack of repression in the presence of ammonium or glutamine was in line with the observation that, for at least uapC and azgA, this novel transcriptional activation mechanism is also independent of AreA, the general GATA factor, mediating nitrogen catabolite repression (Amillis et al., 2004). A component of uapA FEMS Microbiol Rev 31 (2007) 657–675

expression might also be independent of AreA, but this could not be definitively proved due to the very low levels of uapA transcript in germinating conidia. The expression observed early during germination for uapA, uapC and azgA is transporter-specific. uaZ and hxA (Fig. 1), two genes coding for enzymes necessary for purine catabolism, are not transcriptionally activated during the isotropic growth phase of germination. Their transcription becomes evident at later stages and is dependent on UaY and AreA. Thus, the two major regulatory systems known to control uapA, uapC and azgA transcription in mycelia, that is, nitrogen repression and substrate induction, become operational after the germination-specific transcriptional activation of transporter genes and after the first nuclear division, at stages coincident with polarity establishment and maintenance. This novel mechanism regulating the expression of purine transporters is also valid for several other solute transporters. Direct evidence exists that all fur-like genes (NCS1/ PRT), including those encoding uracil (furD) and allantoate (furA) transporters (Amillis et al., 2007, Z. Hamari, S. Amillis, G. Diallinas, C. Scazzocchio, unpublished data), and prnB, encoding the highly specific L-proline transporter (Tazebay et al., 1997), are also transcriptionally activated during the isotropic phase of germination. Their expression is also independent of known transcription regulatory circuits operating at later stages of A. nidulans vegetative development. Indirect evidence, through uptake measurements, also strongly suggests that nitrate, ammonium aspartate/glutamate and lactate transporters are induced/ activated early during germination (Magill & Magill, 1975; A. Apostolaki, C. Scazzocchio, V. Sophianopoulou, G. Diallinas, unpublished data). A similar developmental control of purine transporters has been detected by early studies in N. crassa (Pendyala & Wellman, 1977). Thus, unlike Saccharomyces cerevisiae, where highly specific sensor proteins can activate true transporters (Forsberg & Ljungdahl, 2001), A. nidulans and possibly other filamentous fungi might use their transporters per se for both sensing the environment and for the bulk transport of solutes. The molecular protagonists controlling this novel regulation system remain unknown.

Regulation of nucleobase transporters at a posttranslational level: dynamic endocytosis and trafficking Changes in fungal growth conditions, apart from affecting the transcriptional regulation of several transporter genes (mainly those encoding ‘catabolic’ transporters), might also lead to down-regulation of specific transporters by Rsp5dependent ubiquitylation, endocytosis and targeting of these transporters to the endosome/vacuolar pathway for degradation. Rsp5p/Npi1p is an HECT-type ubiquitin 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

664

ligase, homologous to proteins of the Nedd4 ligase family in higher eukaryotes, which is involved in regulated endocytosis of transmembrane proteins in yeast. RSP5 is an essential gene (Dupre et al., 2004). Studies on the Rsp5-dependent trafficking fate of the Fur4p uracil transporter of Saccharomyces cerevisiae have become a fine paradigm of the regulation of membrane protein expression at a posttranslational level (for a review see Dupre et al., 2004). Here, only the basic aspects of this regulation mechanism will be outlined and will be compared with analogous phenomena concerning the nucleobase transporters of A. nidulans. When no or very low levels of uracil exist in the growth medium, the FUR4 gene is transcribed at relatively low levels constitutively. This leads to Fur4p translation and translocation in the plasma membrane. Fur4p-GFP chimeras mark the plasma membrane of yeast cells but some fluorescence can also be detected in the vacuoles. This is a typical picture of transporters tagged with GFP. No ER, golgi, endosomal or vesicle fluorescence staining is apparent, suggesting that the flow to the plasma membrane and apparent endocytosis is continuous, not allowing concentrative accumulation of transporter molecules in any of these compartments. Endocytosis and subsequent vacuolar degradation reflects the natural turnover of the transporter. Down-regulation by endocytosis and vacuolar degradation is enhanced by various adverse conditions, such as nitrogen, carbon or phosphate starvation, inhibition of protein synthesis by cycloheximide or entrance to the stationary phase of growth (Volland et al., 1994; Galan et al., 1996). The addition of high concentrations of uracil to the medium also decreases uracil transporter levels, partly by destabilizing FUR4 mRNA in the absence of the endogenous FUR4 promoter (Se`ron et al., 1999), but mostly, by triggering the rapid degradation of the existing transporter through accelerated Rsp5-dependent ubiquitylation and endocytosis (Se`ron et al., 1999). As the adverse conditions that accelerate transporter turnover also lead to ribosome degradation, they are likely to trigger an increase in the internal pool of uracil of catabolic origin, which may be the main signal for Fur4p down-regulation. Interestingly, the uracil-induced control of Fur4p trafficking requires direct binding of uracil to the transporter, as a mutant transporter with an abnormally low affinity for uracil displays impaired transport activity (Urban-Grimal et al., 1995) and is not internalized following the addition of uracil to the medium (Se`ron et al., 1999). Several aspects concerning the mechanism of Fur4p endocytosis have become clear using GFP-tagged wild-type or variant versions of the transporter (Volland et al., 1994; Galan et al., 1996; Galan & Haguenauer-Tsapis, 1997; Dupre et al., 2004). Fur4p has an N-terminal PEST-like sequence that was shown to be modified by Ser phosphorylation (S42, S43, S45, S55, S56) and this phosphorylation is necessary for subsequent ubiquitylation of two nearby lysines (K38, K41). 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

A. Pantazopoulou & G. Diallinas

Fur4p phosphorylation at this site is partly dependent on Yck1p/Yck2p casein kinase I homologues but the kinase that phosphorylates Fur4p remains unknown. For Fur4p endocytosis and degradation, mono- or oligo-ubiquitylation (chain of ubiquitin molecules through K63 bonds) was shown to be necessary. Consistent with this, a single ubiquitin molecule fused N-terminally to Fur4p causes internalization of a transporter version lacking its physiological target Lys. However, this chimeric transporter is internalized at a fifth of the rate of a wild-type fully ubiquitylated Fur4p, while some of the chimeric protein is retained in the vacuolar membranes, suggesting that monoubiquitylation is not sufficient for efficient sorting into multivesicular bodies (MVB), the vesicular compartments that eventually fuse with vacuoles where degradation of transporters takes place (Blondel et al., 2004). Fur4p was also shown to display direct vacuolar routing if synthesized de novo in the presence of uracil (Se`ron et al., 1999). A mutant version of Fur4p with a very low affinity for uracil could not be diverted to the vacuolar pathway when synthesized de novo in the presence of uracil. Instead, the transporter was targeted to the cell surface. These experiments were conducted on cells producing a GFP-tagged version of the mutant transporter, together with a wild-type untagged transporter to permit uracil uptake. In contrast to the fate of this mutant transporter, in cases of defective ubiquitylation (rsp5 mutants in particular), Fur4p synthesized de novo in the presence of uracil was able to leave the Golgi apparatus and reach the vacuole, but remained missorted in the vacuolar membrane. Correct luminal delivery was restored by the in-frame N-terminal fusion of ubiquitin, which also resulted in the partial diversion of Fur4p to the vacuolar pathway (Blondel et al., 2004). These data suggest that the binding of intracellular uracil to the transporter promotes a conformational change preventing translocation of the transporter to the plasma membrane. Missorted transporter then undergoes Rsp5-dependent ubiquitylation, leading to its delivery to the vacuolar lumen (Blondel et al., 2004). It should be noted that defective ubiquitylation (rsp5 cells) does not seem to prevent Fur4p from leaving the Golgi apparatus and reaching the late endosome. It is the subsequent MVB sorting of the transporter that is specifically impaired. This mechanism is similar to the model proposed for the control by amino acids of the Gap1p transporter, a general amino acid permease (Andre´, 1995), except that this protein is mainly retargeted to the cell surface if not correctly ubiquitylated. The requirements for Fur4p ubiquitylation at the plasma membrane and during Golgi-to-MVB sorting have been compared. Both events are Rsp5p-dependent. The ubiquitylation required for MVB sorting did not require phosphorylation of the PEST sequence of the transporter (Blondel et al., 2004), whereas plasma membrane ubiquitylation did (Marchal et al., 1998). Fur4p K38 FEMS Microbiol Rev 31 (2007) 657–675

665

Fungal nucleobase transporters

and K41, the only target Lys for plasma membrane ubiquitylation (Marchal et al., 2000), were also sites of ubiquitylation for MVB sorting, together with other Lys (Blondel et al., 2004). The type of ubiquitylation (mono-, poly-, multi-) occurring during Fur4p Golgi-to-vacuole targeting has not yet been determined. Interestingly, in E vps (vacuolar protein sorting) mutants, Fur4p seems to be stabilized in the plasma membrane under conditions that normally lead to endocytosis and degradation. E VPS proteins are involved in MVB sorting. Loss of class E VPS function rapidly leads to the accumulation of large multilamellar cisternal compartments near the vacuole and lack of internal vacuolar vesicles, without affecting internalization from the plasma membrane. A consequence of defective MVB sorting is the stabilization of ubiquitinated proteins on the limiting membrane of endosomes, preventing the entry of polyubiquitin chains into the degradative pathway (Dupre et al., 2004). Unlike other cases, the recycling of Fur4p does not take place through the Golgi (Bugnicourt et al., 2004). Whether this dynamic recycling detected in different vps mutants also occurs in wild-type cells in response to different metabolic needs remains unknown (Bugnicourt et al., 2004; Dupre et al., 2004). Very little is known on how general or specific transacting trafficking factors cross-talk with cis-active elements on Fur4p. Unlike amino acid transporters, hexose transporters or phosphate transporters (Kuehn et al., 1996; Kota & Ljungdahl, 2005), no trafficking chaperones have been identified specifically for Fur4p. Despite several studies concerning ubiquitin-induced degradation of transporters and receptors, one still has no clear idea as to how Rsp5 or other ubiquitin ligases and their partners interact with transporters, such as Fur4p, at which point such an interaction take place or what defines specificity in interactions. This last aspect of specificity in transporter-ubiquitin ligase interactions is highlighted by the observation that candidate adaptors for Rsp5p, such as Bul1p and Bul2p, are needed for the down-regulation of several transporters (e.g. Gap1p, Fui1p, Tat2p) but not of others (e.g. Fur4p, Tat1p) (Dupre et al., 2004), while specific vps or other genes needed for efficient Gap1p endosomal trafficking are irrelevant to Fur4p trafficking (Rubio-Texeira & Kaiser, 2006). Recently, Fur4p was found to accumulate in late endosomal compartments in specific COPIb mutants, a phenotype similar to E vps mutants, suggesting that COPI subunits play a role in vacuolar protein sorting to the MVB compartment (Gabriely et al., 2007). COPI multisubunit coat components confer the retrograde trafficking of cargo molecules from the Golgi apparatus to the endoplasmic reticulum. The effect of COPIb mutants on Fur4p trafficking is consistent with the clathrin-like characteristics of COPIb activity. Clathrin is another multisubunit coat complex involved in Golgi apparatus-to-vacuole/lysosome transport, endocytosis and endosomal protein sorting (Gabriely et al., 2007). FEMS Microbiol Rev 31 (2007) 657–675

Another interesting aspect of fungal transporter topology, in general, comes from work showing that Fur4p, regardless of its ubiquitination status and similar to other yeast secondary transporters, behaves like a raft-associated protein (Dupre & Haguenauer-Tsapis, 2003; Hearn et al., 2003). Lipid rafts, formed by the lateral association of sphingolipids and cholesterol or ergosterol in the external membrane leaflet, have been implicated in membrane traffic and cell signalling in mammalian and yeast cells. Raft association may be essential for regulating the breakdown of Fur4p in response to stresses and other factors that govern uracil transport activity. Recently, it was suggested that the yeast plasma membrane has two nonoverlapping subcompartments: one, represented by a network-like structure, is occupied by the proton ATPase, Pma1, and the second one, houses a number of proton symporters, including Fur4p. In this later compartment membrane depolarization governs the lateral segregations of H1 symporters (Grossmann et al., 2007). The cellular and subcellular expression of the UapA, UapC, AzgA and FcyB nucleobase transporters has been examined in A. nidulans using GFP tags (Valdez-Taubas et al., 2000, 2004; Pantazopoulou et al., 2007; A. Vlanti, G. Diallinas, unpublished data). Similar to yeast, in vegetative Aspergillus mycelium functional GFP-tagged transporters appear in the plasma membrane and in the vacuoles. Expression of UapA and AzgA was also examined during conidiospore germination (Pantazopoulou et al., 2007). Resting conidiospores do not express these nucleobase transporters. UapA-GFP expression becomes evident only after 4 h of germination at 25 1C (equivalent to 2 h germination at 37 1C). Interestingly, after 6 h of germination at 251, UapA-GFP was localized in cytoplasmic rings corresponding to the ER membrane (Vlanti et al., 2006). At later stages of germination and in mycelia, these ER rings are not apparent anymore, suggesting that the rate of UapA exit from the ER membrane is rapid. The expression of AzgAGFP is more rapid than that of UapA-GFP, being detectable in the ER membrane after 4 h of germination and clearly visible in the periphery of conidiospores after 6 h. An interesting suggestion arising from these observations is that ER exit of purine and possibly other type of transporters is efficient only after the first nuclear division and establishment of polarity. Recently, preliminary evidence was obtained for substrate-induced nucleobase transporter inactivation and/or endocytosis in A. nidulans. FurD-mediated H3-uracil uptake or UapA-mediated H3-xanthine uptake was shown to be abolished or reduced when cells are preincubated with unlabelled uracil or xanthine or nonmetabolizable analogues of these nucleobases (Amillis et al., 2007; G. Diallinas, unpublished data). In the case of UapA, there was direct evidence that substrates induce very rapid endocytosis and sorting in the vacuole (G. Diallinas, unpublished data). In 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

666

A. Pantazopoulou & G. Diallinas

addition to substrate-induced inactivation of nucleobase transporters, UapA and UapC are also down-regulated by ammonium through endocytosis and degradation in the vacuole (Valdez-Taubas et al., 2000, 2004; Pantazopoulou et al., 2007). This was clearly shown by fluorescent microscopy of GFP-tagged versions of UapA and UapC expressed from the alcA glucose-repressible, ethanol-inducible, promoter, which is not regulated by ammonium. In the case of UapA, Western blotting of membrane-associated GFP- or His-tagged transporter has confirmed the negative effect of ammonium on plasma membrane expression (Pantazopoulou et al., 2007). Interestingly, AzgA, the other major purine transporter of A. nidulans, is not regulated by ammonium-induced endocytosis (Pantazopoulou et al., 2007). Recently, evidence was obtained that UapA endocytosis by ammonium is dependent on its C-terminal cytoplasmic tail (A. Vlanti, S. Amillis, G. Diallinas, unpublished data). Whether the mechanism of ammoniuminduced endocytosis involves posttranslational modifications such as phosphorylation and ubiquitination, like the case of the general amino acid permease (Gap1p) in yeast (Dupre et al., 2004), is under investigation. A question to be answered is whether substrate-induced, ammonium-induced or other conditions leading to endocytosis and vacuolar degradation share similar mechanisms and genetics. The case of the UapA purine transporter of A. nidulans provides an ideal tool to answer several of these questions, not only because it is down-regulated by both ammonium and substrates but also because of a wealth of a genetic analysis that has provided mutant versions of altered substrate-binding affinities and specificities, as well as mutants with altered trafficking and endocytosis (Koukaki et al., 2005; Pantazopoulou & Diallinas, 2006; Vlanti et al., 2006). In addition, the availability of the genome sequence and the ease of designing genetic screens to identify the molecules involved in both endocytosis and exocytosis of UapA, or other A. nidulans nucleobase transporters, makes this fungus not only an alternative model organism to study membrane protein trafficking but also a unique system to study this phenomenon in polarized cells with large sizes faced with the need to transmit and co-ordinate metabolic signals over longer distances and via multiple nuclei.

inducing (uric acid) or noninducing (urea) conditions. UapC was conditionally expressed in metulae, in samples grown only in the presence of uric acid. AzgA-sGFP was not expressed in any of the asexual structures of A. nidulans, under any condition used (urea or hypoxanthine). In a similar study, the PrnB-GFP proline transporter was shown to be expressed specifically and intensively in conidiospores and much less in phialides, but not at all in the metulae, the vesicle or the conidiophore (Pantazopoulou et al., 2007). The expression of UapA or UapC ‘in the air’ seems paradoxical. A speculation might be that low levels of uric acid are continuously synthesized by oxidation of purines in the mycelium and that UapA or UapC expression in the metulae mediates uric acid transport towards asexual compartments. However, it is totally unknown whether transporters in general are mechanistically needed for metabolite distribution in different fungal compartments or whether metabolites simply diffuse freely through septal pores. Moreover, strains lacking UapA and UapC activities do not seem to have any phenotype with respect to asexual reproduction. Purine transporter expression was also examined in sexual compartments of A. nidulans, such as cleistothecia, ascospores, ascogenous hyphae and h¨ulle cells. UapA and AzgA are expressed in ascogenous hyphae, in h¨ulle cells and interconnecting hyphae of the latter (see Fig. 2). Both UapA and AzgA fluoresce strongly in the hyphae interconnecting h¨ulle cells and surrounding the cleistothecia. Very recently, evidence was obtained that the poorly expressed purinecytosine transporter FcyB is also expressed at a significant level in ascogenous hyphae, in h¨ulle cells and interconnecting hyphae, but also in immature asci (A. Vlanti & G. Diallinas, in preparation). A role, if any, of purine transporters during sexual development, apart from nitrogen source supply, is doubtful because deletion of uapA, azgA, uapC or fcyB does not affect the development of cleistothecia or the production of viable ascospores.

Purine transporters in fungal sexual and asexual development: new roles?

Work from the author’s laboratory has led to several conclusions concerning structure–function relationships in the UapA uric acid–xanthine transporter. An approach using chimeric UapA–UapC transporters has initially identified a relatively short segment, including two putative transmembrane domains and their connecting loops, that determines the kinetics and, possibly, the specificity of these purine transporters (Diallinas et al., 1998). In the original article, UapA was proposed to have 14 TMS but today, after a revision of the translation start of UapA and using improved topological algorithms, it is now believed that it

Strains expressing functional versions of UapA-GFP, UapCGFP and AzgA-GFP were also used for studying purine transporter expression in the asexual and sexual structures of A. nidulans (Pantazopoulou et al., 2007). Figure 2 shows that UapA-GFP is not expressed in the conidiophore stalk and the vesicle, but is highly expressed in the periphery of the metulae, and may be at low levels in phialides. The same result was obtained when the strain was grown under 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

Structure--function relationships Towards an understanding of the molecular determinants underlying UapA function

FEMS Microbiol Rev 31 (2007) 657–675

667

Fungal nucleobase transporters

Germination of conidiospores 2h

4h

6h

8h

PM

ER

cs

Asexual cycle

96 h

16 h

S V

PM

PM

S

S

M Ph

cs

Vegetative Hyphae

Hypha and Conidiophore 240 h Fig. 2. Expression of UapA-GFP in Aspergillus nidulans. The timing of conidiospore germination at 37 1C in minimal media is shown. Diiferent Aspergillus structures and cellular structures: V, vacuoles; M, metulae; Ph, phialidae, Sc, conidiospores; S, septa; H, hulle ¨ cells; Asc, ascospores; Cl, cleistothecium; Ah, ascogenous interconnecting hypha.

H

asc

most probably has 12 TMS, with an extra, topologically ambiguous, amphipathic a-helix between TMS8 and TMS9 (Koukaki et al., 2005; Fig. 3). In that model, the results from chimeric analysis suggest that the critical region determining UapA (or UapC) kinetics corresponds to a sequence starting in the loop downstream from TMS8 and including the following topologically ambiguous amphipathic a-helix, the next hydrophilic loop and TMS9 (see Fig. 3). Interestingly, the short loop connecting the ambiguous amphipathic a-helix and TMS9 includes a highly conserved sequence ([Q/E/P]408-N-X-G-X-X-X-X-T-[R/K/G])417, called the FEMS Microbiol Rev 31 (2007) 657–675

H

IH

asc

H

CL

Sexual cycle reproductive structures

NAT signature motif (Diallinas et al., 1998; Koukaki et al., 2005;Fig. 3). Using an in vitro directed mutagenesis approach and extensive kinetic analysis of mutants with a large collection of purine analogues as competitive inhibitors, this motif was shown to be part of the substrate translocation pathway interacting with the imidazol part of purines (Meintanis et al., 2000; Amillis et al., 2001; Goudela et al., 2005; Koukaki et al., 2005). Residues Q408, N409 and G411 were shown to modify the kinetics and specificity of UapA, without affecting targeting in the plasma membrane. Q408 was proposed to be able to interact with position N9 of the 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

668

A. Pantazopoulou & G. Diallinas

(b)

70

80

90

-+---------+---------+L

L

L

(a)

L

OUT 99 L Q L

77

H

2

3

4 5 6 7

8

L V P I K L NE

R T L A I V G

A E

R

?

10 11

F 528 IN

N N Q 408

Selectivity filter

(c)

(d)

H86N

H86A

H86K

H86D

NAT signature motif

SVCT1 PyrP UraA Lpe1 UapA UapC Xut1 YgfO YicE PbuX PucJ Consensus

(e)

L-ascorbate uracil uracil uric acid / xanthine uric acid / xanthine uric acid / xanthine uric acid / xanthine xanthine xanthine xanthine uric acid

Fig. 3. (a) Speculative topology of UapA showing the location of various mutations affecting expression in the plasma membrane, transport catalysis or specificity. Twelve a-helical TMS are shown. An a-helical segment of ambiguous topology is indicated with a ‘?’. Absolutely conserved amino acids in the NAT/NCS2 family are highlighted by dark grey circles. The UapA-specific Leu Repeat and residue F528 acting as a selectivity filter are highlighted by lighter grey circles. (b) Alignment of the sequence corresponding to TMS1 of NAT/NCS2 transporters. The ubiquitous HQ motif (in bold/light grey squares) and the UapA-specific Leu Repeat (in dark grey squares) are highlighted. (c) H86 mutations affecting trafficking of UapA. Notice ER retention in H86D and numerous vacuoles and other punctuate bodies (probably endosomes or Golgi) in other mutants. (d) Alignment of the sequence corresponding to the NAT/NCS2 signature motif considered to be part of the purine-binding site (absolutely conserved amino acids in bold). The substrate specificity of each transporter shown is indicated on the right. (e) Physiological (wild-type) expression of UapA in the plasma membrane exhibited by all mutants affecting the NAT/NCS2 signature motif and residue F528 (selectivity filter). The amino acid sequences shown in the alignments (b and d) belong to members of the NAT/NCS2 family with known function, discussed in the text, or to homologues of unknown function from bacteria (Q9RYX7), plants (Q6H5A2, Q8RWE9), Dictyostelium discoideum (Q54ZLO), Schizosaccharomyces pombe (Q9HE12), insects (Q8MU86, Q9VH02) and Caenorhabditis elegans (Q18771).

imidazol ring of purines while N409 was shown to be absolutely necessary for transport catalysis. Residue G411 seems to affect the functional flexibility of this motif. In silico predictions and a search in structural databases strongly suggested that these residues (Q408XN410G411) should form a loop, probably including a b-turn. Recently, cysteine-scanning analysis of the NAT signature motif in the YgfO permease of E. coli has further confirmed these conclusions (Karatza et al., 2006). A random genetic approach, used to select second-site suppressors of mutation 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

Q408E, has repeatedly led to the suppressor mutation F528S, located within TMS12 (Amillis et al., 2001). This mutation, by itself, was sufficient to convert UapA into a general, low-affinity, high-capacity, purine transporter. By systematically mutating residue F528, it was subsequently shown that replacement of an aromatic amino acid (Phe or Tyr) by small residues (Ala, Ser, Thr) at this position allowed very low-affinity, but high-capacity, H1 symport of several novel purine and pyrimidine substrates, without affecting significantly the kinetics of UapA transport for its FEMS Microbiol Rev 31 (2007) 657–675

669

Fungal nucleobase transporters

physiological substrates, uric acid and xanthine (Vlanti et al., 2006). It seems that the presence of an aromatic amino acid residue plays the role of an independent selectivity filter, excluding nonsubstrate purines, even at mM concentration, from leaking in through the UapA-binding site (Vlanti et al., 2006; Fig. 3). Allele-specific combinations of F528 mutations with substitutions of Q408 (Vlanti et al., 2006) or N409 (I. Papageorgiou, G. Diallinas, unpublished data), residues proposed to be involved in purine binding and transport, led to an array of UapA molecules with different kinetic and specificity profiles, suggesting that a molecular cross-talk of part of the purine-binding site (corresponding to the NAT signature motif) with an aromatic residue at position 528 (TMS12) determines substrate translocation. Using similar mutational analyses, the role of TMS1 in UapA function was also highlightened (Pantazopoulou & Diallinas, 2006; Fig. 3). The function of a short motif (Q85H86) conserved in all NATs was investigated. All Q85 mutants were cryosensitive, decreasing (Q85L, Q85N, Q85E) or abolishing (Q85T) the capacity for purine transport, without affecting physiological substrate binding or expression in the plasma membrane. All H86 mutants showed nearly normal substrate-binding affinities but most (H86A, H86K, H86D) were cryosensitive, a phenotype associated with partial ER retention and/or targeting of UapA in small vacuoles (see Fig. 3). Thus, residues Q85 and H86 modify the flexibility of UapA, in a way that affects either transport catalysis per se (Q85) or the expression in the plasma membrane (H86). In addition to the above motif, the role of a transmembrane Leu Repeat (LR) motif present in TMS1 of UapA, but not in other NATs, was also investigated (Pantazopoulou & Diallinas, 2006; Fig. 3). Mutations replacing Leu with Ala residues alter differentially the binding affinities of xanthine and uric acid (the affinity for xanthine was 55-fold higher than the affinity for uric acid) in a temperature-sensitive manner. This result strongly suggests that the presence of L77, L84 and L91 modifies the flexibility of the UapA substrate-binding site in a way that is necessary specifically for high-affinity uric acid transport. This verifies the previous results, where the two basic substrates of UapA, uric acid and xanthine, were found to make different hydrogen bond contacts with the active site of the transporter (Goudela et al., 2005; see Fig. 4).

Possible residues of the Fcy2p substrate-binding site and analogies with UapA A mutagenic analysis has been performed for the Fcy2p transporter in Saccharomyces cerevisiae. Originally, various FCY2 mutant strains were genetically selected and shown to have altered Km of uptake for one or several nucleobases (Chevallier et al., 1975). Three of the mutations are subFEMS Microbiol Rev 31 (2007) 657–675

stitutions in the segment I371-A-N-N-I-P-N377, within a loop (L7) downstream from TMS7 (Bloch et al., 1992; Brethes et al., 1992). Subsequent oligonucleotide-directed mutagenesis and functional analysis of several single mutations emphasized the role of the two Asn residues, at positions 374 and 377, in the binding of Fcy2p substrates (Ferreira et al., 1997). Changes in Kms were directly correlated with changes in substrate-binding constants (Kd) and suggest that residues N374 and N377 should play a crucial role in the local conformation of the Fcy2p active-binding site. Interestingly, the effects of mutations at these residues were partially due to a shift of the pKa of an ionizable amino acid residue of the unliganded transporter (Ferreira et al., 1997). Mutations of residue P376 have also been analysed. The mutant protein Fcy2p-P376G was able to transport nucleobases with increased affinity and decreased turnover, while Fcy2p-P376R was normally expressed in the plasma membrane, binds nucleobases, but it is completely unable to transport any of its ligands. In addition, the Kd for hypoxanthine binding was independent of the pH within the range 3.5–6.0, showing that no conformational change was induced by ligand binding. It has been proposed that P376 is part of a b-turn motif, which plays a dynamic role in the translocation process (Ferreira et al., 1997). Thus, the analogy with what was proposed for the NAT signature motif of UapA (Q408NNG411), which also possibly forms a loop with a b-turn, is impressive. In both cases, the most important amino acid residue for function is an Asn residue (N409 in UapA or N374 and N377 in Fcy2p). These residues are absolutely conserved in all NAT/NCS2 or PRT/NCS1 sequences, respectively. Immediately upstream from residue N409 in UapA or N374 in Fcy2p, there are polar residues (Q408 or N373, respectively) critical for transport catalysis. Furthermore, two amino acids downstream from N409 in UapA or N374 in Fcy2p, there are two highly conserved structural residues (G411 in UapA, P376 in Fcy2p) that seem critical for forming a b-turn within a flexible loop. If not fortuitous, this implies a common architecture and mechanism in active-site interactions of nucleobase transporters belonging to evolutionarily distinct families. In an attempt to identify domains that structurally and/or functionally interact with the hydrophilic segment 371–377 in Fcy2p, Ferreira et al. (1999) isolated and studied a secondsite suppressor, S272L, of the mutant Fcy2p-N374I. This suppressor mutation partially restores the binding of hypoxanthine and cytosine. This effect is not specific for the Fcy2p-N374I allele as it also suppresses mutations N377G and T213I. Kinetic and biochemical analysis of a strain overexpressing Fcy2p-S272L by itself showed that this substitution increases significantly the binding capacity of Fcy2p for its ligands and reduces the turnover of the transporter. S272 is probably located in TMS5, but more experiments are needed to test whether this residue is part of 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

670

A. Pantazopoulou & G. Diallinas

N N

H

≥17.8 5.6

N

N

=34.6 kJ mol G ( (G ))= ≥ 33.2 kJ mol

Fcy21p

(C. albigans)

R

H

6.6

N

R O

R

≥7.6

N R H ≥7.6

∆G = ≥ 32 kJ mol ( (G ))= ≥ 30.4 kJ mol

Hypoxanthine

H

NH

≥17.8

O

N

N N H

R

N

≥7.6

∆G = ≥ 25.4 kJ mol ( (G ))= ≥ 23.8 kJ mol

H

Adenine

N

N N

N

R 9.8

R ≥7.6

R

N H

∆G = ≥ 32 kJ mol (∆(G ))= ≥ 30 kJ mol

Cytosine

O N

N

H

H

R R 8.8

8.8 kJ 12.3 kJ

NH N

H N

N

Uric acid

≥15.2

O

O

O

= 30.5 kJ mol G ( (G )) = ~ 34.6 kJ mol

R

≥15.2

∆G = ≥ 28 kJ mol ( (G ))= ≥ 26.4 kJ mol

Guanine

H

Xanthine

R ≥7.6

N

H

R R R H 5.2-5.4 ~14-16 G = 30.3 –32.0 kJ mol (G )) = ~ 30.5 –31.8 kJ mol

≥11.2

NH

R ≥7.6

~16 R

R O H H

Adenine R

N N

N

R

N H G = 33.3 kJ mol ( (G ))= ≥ 31.9 kJ mol

8.6

N

9.3 – 10.4

9.6

N

Guanine

8.6 N

N

N H

=33.3 kJ mol G ( (G ))= ≥ 31.9 kJ mol

Hypoxanthine R O

N

NH

H

R 4.5 NH R R N ≥17.8 N

FurD

H N 6.9

R

R

8.5

(A. nidulans)

R

O

UapA

R

8.5

(A. nidulans)

AzgA

(A. nidulans)

R O

R

7.6 kJ R

R O H O

N N H 5.5 kJ R

DG DG

= 37.7 kJ mol = 34.2kJ mol

Uracil

Fig. 4. Kinetic models of transporter–nucleobase interactions in fungi. For details of the energetic contribution of different positions, see the text.

the hydrophilic pore involved in the translocation of nucleobases and/or the proton (Ferreira et al., 1999).

Structure--function relationships in Fur4p The presence of 10 TMS and long hydrophilic N- and Ctermini has been supported by biochemical experiments in Fur4p. Fur4p N-terminal segments of increasing lengths have been fused to a reporter glycoprotein, acid phosphatase and the in vitro and in vivo fates of the resulting hybrid proteins were analysed to identify the first signal anchor sequence of the transporter and demonstrate the cytosolic orientation of its N-terminal hydrophilic segment (Silve et al., 1991). Other experimental approaches (trypsin and carboxypeptidase degradation and immunofluorescent analysis) have been used to confirm this orientation, and to establish that both the N- and C-termini of Fur4p are cytoplasmic, which in turn demonstrates that the transporter polypeptide spans the membrane an even number of times (Garnier et al., 1996). The orientation of several hydrophilic loops with respect to the membrane has also been investigated by introducing potential glycosylation sites into these regions. The resulting mutant proteins have been analysed with respect to whether they are glycosylated when expressed in the presence of dog-pancreas microsomes (Garnier et al., 1996). Mutagenic analysis of Fur4p has placed emphasis on the role of a number of charged amino acid residues that are located within membrane-spanning segments. These are E243 in TMS3, K272 in TMS4 and E539 in TMS9 (Pinson et al., 1999). All mutant transporters were correctly ex2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

pressed and targeted to the plasma membrane of a fur4 null mutant. No evidence was found for ionic interactions between either of the glutamic acid residues and the lysine residue. Of the three charged residues, only K272 was critical for the transport activity of Fur4p. All three substitutions K272E, K272A and K272R strongly impair both the binding and transport of uracil. The large increases in Km seem to be attributable to a large increase in Kd. Together with the fact that all K272 mutants are still functional, the results from these studies suggest that K272 is not an absolutely essential residue for active transport but that it is critical for the proper functioning of the binding site (Pinson et al., 1999). Interestingly, most of the Fcy2p-like sequences have a positively charged residue (H or R) at the same position.

Kinetic modelling of transporter--substrate interactions: purine transporters as specific gateways for drug delivery? Analogues of purine and pyrimidine nucleobases are being widely used as antimetabolites against a host of different diseases and infections, ranging from antitumour and leukaemia chemotherapy (5-fluorouracil, 6-mercaptopurine, thioguanine) and antiviral compounds (acyclovir, ganciclovir, carbovir) to antibiotics and drugs against parasitic disease (e.g. allopurinol, pyrimethamine) and even for the prevention of organ transplant rejection (azathioprine) and the treatment of gout (allopurinol) (Elion, 1989). Kinetic and genetic analyses of nucleobase transporters in fungal model systems, where a given protein can be studied in a FEMS Microbiol Rev 31 (2007) 657–675

671

Fungal nucleobase transporters

genetic background lacking all other related transporters, can lead to modelling of transporter–substrate interactions. Such models can be tested by reverse genetics and be used to understand the functioning of nucleobase transporters of pathogenic fungi and bacteria. This knowledge can, in the long run, constitute an essential step in predicting the use or/ and design of purine-based antifungals that will not be taken up by the analogous transporters of the infected host tissues. In the author’s lab, using appropriate genetic backgrounds, detailed kinetic analyses were performed of substrate binding by the UapA (Goudela et al., 2005), AzgA (Goudela et al., 2006) and FurD (Amillis et al., 2007) transporters of A. nidulans, and by the Xut1p (a true orthologue of UapC; see Table 1; Goudela et al., 2005) and Fcy21p (a true orthologue of Fcy2p or FcyB; see Table 1; Goudela et al., 2006) transporters of C. albicans. Analogous models for UapA and AzgA have also been constructed for A. fumigatus (S. Goudela, U. Reichard, G. Diallinas, S. Amillis, unpublished data; see Table 1), and for YgfO, an E. coli NAT homologue (Goudela et al., 2005; Karatza & Frillingos, 2005). The rationale of this approach has also been used successfully to define transporter–substrate interactions for protozoan and mammalian nucleobase and nucleoside transporters (De Koning & Diallinas, 2000; Wallace et al., 2002). In brief, using a rich collection of purines and purine analogues, Ki values for compounds inhibiting the uptake of radiolabelled xanthine (for UapA, Xut1 or YgfO), hypoxanthine (for AzgA or Fcy21p) or uracil (for FurD) were determined from full dose–response curves. Gibbs free energies DG0 were then calculated from DG0 =  RTln(Ki), where R is the ideal gas constant and T the absolute temperature (1K). By comparing the difference in DG0s in the interaction of two putative substrates differing at a single position each time (e.g. xanthine vs. 2-thioxanthine or adenine vs. 3-deazadenine), the contribution of this position in binding was estimated. Taking into account losses due to possible steric hindrance or due to loss of planarity, and the structure of preferred tautomeric forms of substrates, models describing the interactions of nucleobase transporters with possible substrates could be developed. A summary of all these models is given in Fig. 4. From these models, some predictions could be drawn with respect to possible nucleobase-related analogues that can be used to target fungi. For example, AzgA has 10- and 26-fold higher affinities for adenine or guanine and 2-amino purine, respectively, compared with the human nucleobase transporter hFNT1 (Wallace et al., 2002). Even more prominently, Fcy21p has a more than 1000-fold higher affinity for 3-methyl-adenine compared with hFNT1 (Wallace et al., 2002). Even higher levels of selectivity could be achieved, or at lest tested, with rationally designed purine analogues, based on the models proposed. For example, it seems that a major difference between fungal transporters analysed hereFEMS Microbiol Rev 31 (2007) 657–675

in and analogous human transporters (e.g. hFNT1) is the involvement of N3 in substrate binding (Kraupp & Marz, 1995; Wallace et al., 2002). AzgA, Fcy21p and UapA do not use N3 to bind purines. Bulky groups at this position have a severe steric effect on AzgA substrate recognition, but Fcy21p and UapA are tolerant in similar substitutions. This leads to the prediction that purine analogues with N3 substitutions will be good substrates specifically for Fcy21p and UapA, but also for AzgA, given that in the latter case they will not increase the volume at this position, e.g. 3deazapurines. It is known that N3-substituted analogues are channelled to nucleic acid synthesis and can be used at low concentrations as specific antiviral inhibitors (Shigeta et al., 1988; Andrei & De Clercq, 1990). Similarly, 3-deazapurines and their analogues might also be efficient drugs against pathogenic fungi. Based on a similar logic, purine drugs targeted to filamentous fungi, via AzgA-like transporters, might also be based on nonbulky N9 substitutions, which should be extremely nontolerant to recognition by the human purine transporter hFNT1 (Wallace et al., 2002).

Epilogue This review has as a goal to highlight recent advances in fungal purine-pyrimidine transporters and show how can these molecules become tools to investigate several aspects of the biology of transporter proteins. Knowledge on fungal nucleobase transporter regulation of expression, the physiological and developmental control of their targeting to the plasma membrane and on structure–function relationships determining their specificities and kinetic properties, apart from satisfying one’s basic scientific curiosity, might also provide ideas and means for the use of these transporters as specific gateways for the systematic development of targeted pharmacological therapies against pathogenic fungi. In addition, the homology between fungal and bacterial nucleobase transporters is expected to channel knowledge obtained from fungal proteins to bacterial proteins. This in turn leads to the question as to whether existing or novel purine analogues can also be used as antibacterial drugs. A prerequisite of all the above ideas concerning the use of microbial purine transporters for targeted delivery of antimicrobials is the integration of knowledge obtained on analogous nucleobase transporters of the human body or any other host infected by fungi or bacteria. Saccharomyces cerevisiae fcy2 or fur4 and A. nidulans uapA, uapC and azgA null mutants have already served as recipient strains for the cloning or/and functional analysis of protozoan purine transporters and plant purine or uracil transporters. In general, plant or microbial transporters are much more easily expressed in fungi than metazoan transporters. Often, aberrant transporter trafficking, mostly ER retention, constitutes a major problem for expressing 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

672

heterologous transporters in yeast (Wieczorke et al., 2003; Flagelova et al., 2006) or in A. nidulans (C. Gournas, M. Billini, V. Sophianopoulou, P. Kafasla, A. Vlanti, S. Amillis, G. Diallinas, unpublished observations). However, yeast mutants deficient in the ubiquitin ligase Rsp5p, in which endocytosis or MVB sorting are impaired and recycling of transporters towards the plasma membrane is promoted, proved to be powerful tools to solve the many problems inherent to heterologous expression of membrane proteins in yeast, including ER retention (Flagelova et al., 2006; Froissard et al., 2006). An interesting aspect of the work carried out with fungal nucleobase transporters is their dynamic regulation at the level of exocytosis–endocytosis. The work of R. Haguenauer-Tsapis and colleagues on Fur4p contributed significantly in revealing several mechanisms controlling transporter topogenesis in respect of variable physiological signals. Work with filamentous fungi has provided evidence that purine transporter (UapA, UapC) endocytosis operates as a mechanism for rapid turnover in the presence of preferred nitrogen sources or purine excess. Compared with yeasts, filamentation is a different life style associated with polar growth, cell size in the range of tens of micrometer and polarized targeting of plasma membrane proteins. The fact that TMS-containing proteins, typical of transporters, channels and receptors, diffuse very slowly within the plasma membrane (Valdez-Taubas & Pelham, 2003), and the involvement of endosomes, which were shown to move bidirectionally at a speed of 2–3 mm s1 (Penˇalva 2005), may provide a means by which metabolic signals travel in a biologically meaningful speed and elicit changes in the multiple nuclei in an orchestrated manner. Thus, both yeasts and filamentous fungi provide excellent alternative and complementary model systems to study endocytosis–exocytosis of transmembrane proteins, such as nucleobase transporters, using genetics, genomics and molecular biology. Last but not the least is that fungal nucleobase transporters belong to three families of transmembrane proteins that are evolutionarily distinct from any membrane protein of known structure. The difficulty of obtaining solved structures of polytopic transmembrane proteins through crystallography or nuclear magnetic resonance is well known and any powerful genetic system, such as that of Saccharomyces cerevisiae and A. nidulans, provides attractive alternative methodologies to approach structure through function.

Acknowledgements The authors would like to thank all past and present members of the lab, and especially Sotiris Amillis, Anna Vlanti, Christos Gournas, Ioannis Papageorgiou, Sofia Goudela and Utz Reichard (G¨ottingen), Claudio Scazzocchio (Orsay) for sharing unpublished results and for critical 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

A. Pantazopoulou & G. Diallinas

discussions. A. P. has been financed by PENED 2001 (GSRT, Ministry of Development) and the Archimedes EU award. Work in the lab of G.D. has been financed from the GSRT (Ministry of Development), the Ministry of Education and the University of Athens.

References Amillis S, Koukaki M & Diallinas G (2001) Substitution F569S converts UapA, a specific uric acid-xanthine transporter, into a broad specificity transporter for purine-related solutes. J Mol Biol 313: 765–774. Amillis S, Cecchetto G, Sophianopoulou V, Koukaki M, Scazzocchio C & Diallinas G (2004) Transcription of purine transporter genes is activated during the isotropic growth phase of Aspergillus nidulans conidia. Mol Microbiol 52: 205–216. Amillis S, Hamari Z, Roumelioti K, Scazzocchio C & Diallinas G (2007) Regulation of expression and kinetic modelling of substrate interactions of a uracil transporter in Aspergillus nidulans. Mol Membr Biol 24: 206–214. Andr´e B (1995) An overview of membrane transport proteins in Saccharomyces cerevisiae. Yeast 11: 1575–1611. Andrei G & De Clercq E (1990) Inhibitory effect of selected antiviral compounds on arenavirus replication in vitro. Antiviral Res 14: 287–299. Argyrou E, Sophianopoulou V, Schultes N & Diallinas G (2001) Functional characterization of a maize purine transporter by expression in Aspergillus nidulans. Plant Cell 13: 953–964. Bloch JC, Sychrova H, Souciet JL, Jund R & Chevallier MR (1992) Determination of a specific region of the purine-cytosine permease involved in the recognition of its substrates. Mol Microbiol 6: 2989–2997. Blondel MO, Morvan J, Dupre S, Urban-Grimal D, HaguenauerTsapis R & Volland C (2004) Direct sorting of the yeast uracil permease to the endosomal system is controlled by uracil binding and Rsp5p-dependent ubiquitylation. Mol Biol Cell 15: 883–895. Brethes D, Chirio MC, Napias C, Chevallier MR, Lavie JL & Chevallier J (1992) In vivo and in vitro studies of the purinecytosine permease of Saccharomyces cerevisiae. Functional analysis of a mutant with an altered apparent transport constant of uptake. Eur J Biochem 204: 699–704. Bugnicourt A, Froissard M, Sereti K, Ulrich HD, HaguenauerTsapis R & Galan JM (2004) Antagonistic roles of ESCRT and Vps class C/HOPS complexes in the recycling of yeast membrane proteins. Mol Biol Cell 15: 4203–4214. Cecchetto G, Amillis S, Diallinas G, Scazzocchio C & Drevet C (2004) The AzgA purine transporter of Aspergillus nidulans: characterization of a protein belonging to a new phylogenetic cluster. J Biol Chem 279: 3132–3141. Chevallier MR, Jund R & Lacroute F (1975) Characterization of cytosine permeation in Saccharomyces cerevisiae. J Bacteriol 122: 629–641.

FEMS Microbiol Rev 31 (2007) 657–675

673

Fungal nucleobase transporters

Darlington AJ & Scazzocchio C (1967) Use of analogues and the substrate-sensitivity of mutants in analysis of purine uptake and breakdown in Aspergillus nidulans. J Bacteriol 93: 937–940. De Koning HP & Diallinas G (2000) Nucleobase transporters. Mol Membr Biol 17: 75–94. Diallinas G & Scazzocchio C (1989) A gene coding for the uric acid-xanthine permease of Aspergillus nidulans: inactivational cloning, characterization and sequence of a cis-acting mutation. Genetics 122: 341–350. Diallinas G, Gorfinkiel L, Arst HN Jr, Cecchetto G & Scazzocchio C (1995) Genetic and molecular characterization of a gene encoding a wide specificity purine permease of Aspergillus nidulans reveals a novel family of transporters conserved in prokaryotes and eukaryotes. J Biol Chem 270: 8610–8622. Diallinas G, Valdez J, Sophianopoulou V, Rosa A & Scazzocchio C (1998) Chimeric purine transporters of Aspergillus nidulans define a domain critical for function and specificity conserved in bacterial, plant and metazoan homologues. EMBO J 17: 3827–3837. Dupre S & Haguenauer-Tsapis R (2003) Raft partitioning of the yeast uracil permease during trafficking along the endocytic pathway. Traffic 4: 83–96. Dupre S, Urban-Grimal D & Haguenauer-Tsapis R (2004) Ubiquitin and endocytic internalization in yeast and animal cells. Biochim Biophys Acta 1695: 89–111. Elion GB (1989) The purine path to chemotherapy. Science 244: 41–47. Ferreira T, Brethes D, Napias C & Chevallier J (1997) Functional analyses of mutated purine-cytosine permease from Saccharomyces cerevisiae. A possible role of the hydrophilic segment 371–377 in the active carrier conformation. J Biol Chem 272: 9697–9702. Ferreira T, Chevallier J, Paumard P, Napias C & Brethes D (1999) Screening of intragenic second-site suppressor of purinecytosine permease from Saccharomyces cerevisiae. Possible role of Ser272 in the base translocation process. Eur J Biochem 260: 22–30. Fillinger S, Chaveroche MK, Shimizu K, Keller N & d’Enfert C (2002) cAMP and ras signaling independently control spore germination in the filamentous fungus Aspergillus nidulans. Mol Microbiol 44: 1001–1016. Flagelova H, Haguenauer-Tsapis R & Sychrova H (2006) Heterologous expression of mammalian Na/H antiporters in Saccharomyces cerevisiae. Biochim Biophys Acta 1760: 504–516. Forsberg H & Ljungdahl PO (2001) Sensors of extracellular nutrients in Saccharomyces cerevisiae. Curr Genet 40: 91–109. Froissard M, Belgareh-Touze N, Buisson N, Desimone M, Frommer WB & Haguenauer-Tsapis R (2006) Heterologous expression of a plant uracil transporter in yeast: improvement of plasma membrane targeting in mutants of the Rsp5p ubiquitin protein ligase. Biotechnol J 1: 308–320. Gabriely G, Kama R & Gerst JE (2007) Involvement of specific COPI subunits in protein sorting from the late endosome to the vacuole in yeast. Mol Cell Biol 27: 526–540.

FEMS Microbiol Rev 31 (2007) 657–675

Galan JM & Haguenauer-Tsapis R (1997) Ubiquitin lys63 is involved in ubiquitination of a yeast plasma membrane protein. EMBO J 16: 5847–5854. Galan JM, Moreau V, Andre B, Volland C & Haguenauer-Tsapis R (1996) Ubiquitination mediated by the Npi1p/Rsp5p ubiquitin-protein ligase is required for endocytosis of the yeast uracil permease. J Biol Chem 271: 10946–10952. Garnier C, Blondel MO & Haguenauer-Tsapis R (1996) Membrane topology of the yeast uracil permease. Mol Microbiol 21: 1061–1073. Gorfinkiel L, Diallinas G & Scazzocchio C (1993) Sequence and regulation of the uapA gene encoding a uric acid-xanthine permease in the fungus Aspergillus nidulans. J Biol Chem 268: 23376–23381. Goudela S, Karatza P, Koukaki M, Frillingos S & Diallinas G (2005) Comparative substrate recognition by bacterial and fungal purine transporters of the NAT/NCS2 family. Mol Membr Biol 22: 263–275. Goudela S, Tsilivi H, Diallinas G, Goudela S, Tsilivi H & Diallinas G (2006) Comparative kinetic analysis of AzgA and Fcy21p, prototypes of the two major fungal hypoxanthine–adenine– guanine transporter families. Mol Membr Biol 23: 291–303. Grossmann G, Opekarova M, Malinsky J, Weig-Meckl I & Tanner W (2007) Membrane potential governs lateral segregation of plasma membrane proteins and lipids in yeast. EMBO J 26: 1–8. Gupta S, Kerridge D & Chevallier MR (1995) A purine permease in Candida glabrata. FEMS Microbiol Lett 126: 93–96. Hearn JD, Lester RL & Dickson RC (2003) The uracil transporter Fur4p associates with lipid rafts. J Biol Chem 278: 3679–3686. Jund R, Weber E & Chevallier MR (1988) Primary structure of the uracil transport protein of Saccharomyces cerevisiae. Eur J Biochem 171: 417–424. Karatza P & Frillingos S (2005) Cloning and functional characterisation of two bacterial members of the NAT/NCS2 family in Escherichia coli. Mol Membr Biol 22: 251–261. Karatza P, Panos P, Georgopoulou E & Frillingos S (2006) Cysteine-scanning analysis of the nucleobase-ascorbate transporter signature motif in YgfO permease of Escherichia coli: Gln-324 and Asn-325 are essential, and Ile-329-Val-339 form an alpha-helix. J Biol Chem 281: 39881–39890. Kota J & Ljungdahl PO (2005) Specialized membrane-localized chaperones prevent aggregation of polytopic proteins in the ER. J Cell Biol 168: 79–88. Koukaki M, Vlanti A, Goudela S, Pantazopoulou A, Gioule H, Tournaviti S & Diallinas G (2005) The nucleobase-ascorbate transporter (NAT) signature motif in UapA defines the function of the purine translocation pathway. J Mol Biol 350: 499–513. Kraupp M & Marz R (1995) Membrane transport of nucleobases: interaction with inhibitors. Gen Pharmacol 26: 1185–1190. Kudla B, Caddick MX, Langdon T, Martinez-Rossi NM, Bennett CF, Sibley S, Davies RW & Arst HN Jr (1990) The regulatory gene areA mediating nitrogen metabolite repression in

2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

674

Aspergillus nidulans. Mutations affecting specificity of gene activation alter a loop residue of a putative zinc finger. EMBO J 9: 1355–1364. Kuehn MJ, Schekman R & Ljungdahl PO (1996) Amino acid permeases require COPII components and the ER resident membrane protein Shr3p for packaging into transport vesicles in vitro. J Cell Biol 135: 585–595. Liang WJ, Johnson D & Jarvis SM (2001) Vitamin C transport systems of mammalian cells. Mol Membr Biol 18: 87–95. Magill JM & Magill CW (1975) Purine base transport in Neurospora crassa. J Bacteriol 124: 149–154. Marchal C, Haguenauer-Tsapis R & Urban-Grimal D (1998) A PEST-like sequence mediates phosphorylation and efficient ubiquitination of yeast uracil permease. Mol Cell Biol 18: 314–321. Marchal C, Haguenauer-Tsapis R & Urban-Grimal D (2000) Casein kinase I-dependent phosphorylation within a PEST sequence and ubiquitination at nearby lysines signal endocytosis of yeast uracil permease. J Biol chem 275: 23608–23614. Maurino VG, Grube E, Zielinski J, Schild A, Fischer K & Flugge UI (2006) Identification and expression analysis of twelve members of the nucleobase-ascorbate transporter (NAT) gene family in Arabidopsis thaliana. Plant Cell Physiol 47: 1381–1393. Meintanis C, Karagouni AD & Diallinas G (2000) Amino acid residues N450 and Q449 are critical for the uptake capacity and specificity of UapA, a prototype of a nucleobase-ascorbate transporter family. Mol Membr Biol 17: 47–57. Momany M (2002) Polarity in filamentous fungi: establishment, maintenance and new axes. Curr Opin Microbiol 5: 580–585. Morozov IY, Galbis-Martinez M, Jones MG & Caddick MX (2001) Characterization of nitrogen metabolite signaling in Aspergillus via the regulated degradation of areA mRNA. Mol Microbiol 42: 269–277. Muro-Pastor MI, Gonzalez R, Strauss J, Narendja F & Scazzocchio C (1999) The GATA factor AreA is essential for chromatin remodelling in a eukaryotic bidirectional promoter. EMBO J 18: 1584–1597. Osherov N & May G (2000) Conidial germination in Aspergillus nidulans requires RAS signalling and protein synthesis. Genetics 155: 647–656. Paluszynski JP, Klassen R, Rohe M & Meinhardt F (2007) Various cytosine/adenine permease homologues are involved in the toxicity of 5-fluorocytosine in Saccharomyces cerevisiae. Yeast 23: 707–715. Pantazopoulou A & Diallinas G (2006) The first transmembrane segment (TMS1) of UapA contains determinants necessary for expression in the plasma membrane and purine transport. Mol Membr Biol 23: 337–348. Pantazopoulou A, Lemuh ND, Hatzinikolaou DG, Drevet C, Cecchetto G, Scazzocchio C & Diallinas G (2007) Differential physiological and developmental expression of the UapA and AzgA purine transporters in Aspergillus nidulans. Fungal Genet Biol 44: 627–640.

2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c

A. Pantazopoulou & G. Diallinas

Penˇalva AM (2005) Tracing the endocytic pathway of Aspergillus nidulans with FM4-64. Fung Genet Biol 42: 963–975. Pendyala L & Wellman AM (1977) Developmental-stagedependent adenine transport in Neurospora crassa. J Bacteriol 131: 453–462. Pinson B, Chevallier J & Urban-Grimal D (1999) Only one of the charged amino acids located in membrane-spanning regions is important for the function of the Saccharomyces cerevisiae uracil permease. Biochem J 339: 37–42. Platt A, Langdon T, Arst HN Jr, Kirk D, Tollervey D, Sanchez JM & Caddick MX (1996) Nitrogen metabolite signalling involves the C-terminus and the GATA domain of the Aspergillus transcription factor AREA and the 3 0 untranslated region of its mRNA. EMBO J 15: 2791–2801. Rao TV, Verma RS & Prasad R (1983) Transport of purine, pyrimidine bases and nucleosides in Candida albicans, a pathogenic yeast. Biochem Int 6: 409–417. Ravagnani A, Gorfinkiel L, Langdon T, Diallinas G, Adjadj E, Demais S, Gorton D, Arst HN Jr & Scazzocchio C (1997) Subtle hydrophobic interactions between the seventh residue of the zinc finger loop and the first base of an HGATAR sequence determine promoter-specific recognition by the Aspergillus nidulans GATA factor AreA. EMBO J 16: 3974–3986. Rubio-Texeira M & Kaiser CA (2006) Amino acids regulate retrieval of the yeast general amino acid permease from the vacuolar targeting pathway. Mol Biol Cell 17: 3031–3050. Scazzocchio C (1982) Positive regulation in a eukaryote, a study of the uaY gene of Aspergillus nidulans: I. Characterization of alleles, dominance and complementation studies, and a fine structure map of the uaY-oxpA cluster. Genetics 100: 185–208. Scazzocchio C (1994) The purine degradation pathway, genetics, biochemistry and regulation. Prog Ind Microbiol 29: 221–257. Se`ron K, Blondel MO, Haguenauer-Tsapis R & Volland C (1999) Uracil-induced down-regulation of the yeast uracil permease. J Bacteriol 181: 1793–1800. Shigeta S, Konno K, Yokota T, Nakamura K & De Clercq E (1988) Comparative activities of several nucleoside analogs against influenza A, B, and C viruses in vitro. Antimicrob Agents Chemother 32: 906–911. Silve S, Volland C, Garnier C, Jund R, Chevallier MR & Haguenauer-Tsapis R (1991) Membrane insertion of uracil permease, a polytopic yeast plasma membrane protein. Mol Cell Biol 11: 1114–1124. Suarez T, de Queiroz MV, Oestreicher N & Scazzocchio C (1995) The sequence and binding specificity of UaY, the specific regulator of the purine utilization pathway in Aspergillus nidulans, suggest an evolutionary relationship with the PPR1 protein of Saccharomyces cerevisiae. EMBO J 14: 1453–1467. Tazebay UH, Sophianopoulou V, Scazzocchio C & Diallinas G (1997) The gene encoding the major proline transporter of Aspergillus nidulans is upregulated during conidiospore germination and in response to proline induction and amino acid starvation. Mol Microbiol 24: 105–117.

FEMS Microbiol Rev 31 (2007) 657–675

675

Fungal nucleobase transporters

Urban-Grimal D, Pinson B, Chevallier J & Haguenauer-Tsapis R (1995) Replacement of Lys by Glu in a transmembrane segment strongly impairs the function of the uracil permease from Saccharomyces cerevisiae. Biochem J 308: 847–851. Valdez-Taubas J & Pelham HR (2003) Slow diffusion of proteins in the yeast plasma membrane allows polarity to be maintained by endocytic cycling. Curr Biol 1318: 1636–1640. Valdez-Taubas J, Diallinas G, Scazzocchio C & Rosa AL (2000) Protein expression and subcellular localization of the general purine transporter UapC from Aspergillus nidulans. Fungal Genet Biol 30: 105–113. Valdez-Taubas J, Harispe L, Scazzocchio C, Gorfinkiel L & Rosa AL (2004) Ammonium-induced internalisation of UapC, the general purine permease from Aspergillus nidulans. Fungal Genet Biol 41: 42–51. Vlanti A, Amillis S, Koukaki M & Diallinas G (2006) A novel-type substrate-selectivity filter and ER-exit determinants in the UapA purine transporter. J Mol Biol 357: 808–819.

FEMS Microbiol Rev 31 (2007) 657–675

Volland C, Urban-Grimal D, Geraud G & Haguenauer-Tsapis R (1994) Endocytosis and degradation of the yeast uracil permease under adverse conditions. J Biol Chem 269: 9833–9841. Wagner R, de Montigny J, de Wergifosse P, Souciet JL & Potier S (1998) The ORF YBL04 of Saccharomyces cerevisiae encodes a uridine permease. FEMS Microbiol Lett 159: 69–75. Wallace LJ, Candlish D & De Koning HP (2002) Different substrate recognition motifs of human and trypanosome nucleobase transporters. Selective uptake of purine antimetabolites. J Biol Chem 27729: 26149–26156. Weber E, Rodriguez C, Chevallier MR & Jund R (1990) The purine-cytosine permease gene of Saccharomyces cerevisiae: primary structure and deduced protein sequence of the FCY2 gene product. Mol Microbiol 4: 585–596. Wieczorke R, Dlugai S, Krampe S & Boles E (2003) Characterisation of mammalian GLUT glucose transporters in a heterologous yeast expression system. Cell Physiol Biochem 13: 123–134.

2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved


c