MFS transportome of the human pathogenic yeast Candida albicans
- First Online:
- Cite this article as:
- Gaur, M., Puri, N., Manoharlal, R. et al. BMC Genomics (2008) 9: 579. doi:10.1186/1471-2164-9-579
The major facilitator superfamily (MFS) is one of the two largest superfamilies of membrane transporters present ubiquitously in bacteria, archaea, and eukarya and includes members that function as uniporters, symporters or antiporters. We report here the complete transportome of MFS proteins of a human pathogenic yeast Candida albicans.
Computational analysis of C. albicans genome enabled us to identify 95 potential MFS proteins which clustered into 17 families using Saier's Transport Commission (TC) system. Among these SP, DHA1, DHA2 and ACS represented major families consisting of 22, 22, 9 and 16 members, respectively. Family designations in C. albicans were validated by subjecting Saccharomyces cerevisiae genome to TC system. Based on the published available genomics/proteomics data, 87 of the putative MFS genes of C. albicans were found to express either at mRNA or protein levels. We checked the expression of the remaining 8 genes by using RT-PCR and observed that they are not expressed under basal growth conditions implying that either these 8 genes are expressed under specific growth conditions or they may be candidates for pseudogenes.
The in silico characterisation of MFS transporters in Candida albicans genome revealed a large complement of MFS transporters with most of them showing expression. Considering the clinical relevance of C. albicans and role of MFS members in antifungal resistance and nutrient transport, this analysis would pave way for identifying their physiological relevance.
Current evidence suggests that Candida albicans acquires azole resistance by employing multiple mechanisms that include (a) alterations in the azole-target protein Erg11p (b) upregulation of the ERG11 gene [1, 2, 3, 4] as well as (c) failure of drug accumulation mediated by efflux pumps. Most commonly, genes encoding drug efflux pumps belonging to ATP binding cassette (ABC) and Major facilitator (MFS) superfamilies of proteins are overexpressed in azole resistant Candida isolates [5, 6, 7, 8, 9]. ABC family permeases are in general multicomponent primary active transporters, capable of transporting both small molecules and macromolecules which is coupled to ATP hydrolysis while the MFS transporters are single-polypeptide secondary carriers capable only of transporting small solutes in response to chemiosmotic ion gradients. We have earlier annotated and classified ABC transporters in C. albicans , however, clinically relevant MFS superfamily in C. albicans largely remains uncharacterized.
MFS superfamily is ubiquitously present in all kingdoms of life and includes members of direct medical and pharmaceutical significance. They are involved in the symport, antiport or uniport of various substrates [11, 12] and are known to exhibit specificity for sugars, polyols, drugs, neurotransmitters, Krebs cycle metabolites, phosphorylated glycolytic intermediates, amino acids, peptides, osmolites, siderophores (efflux), iron-siderophores (uptake), nucleosides, organic and inorganic anions, etc [11, 12]. Most MFS proteins vary between 400 and 600 amino acid residues in length and possess either 12 or 14 putative transmembrane segments (TMS). The MFS superfamily consists of 61 families according to the Transport Commission (TC) system given by Saier and group http://www.tcdb.org/. TC is a comprehensive classification system for membrane transport proteins and is analogous to the Enzyme Commission (EC) system, except that it incorporates both functional and phylogenetic information [13, 14, 15]. This system allocates five digits to each phylogenetic cluster of transporters. The first two digits ("class" and "subclass") identify the transport mode and energy-coupling mechanism. The third digit characterizes phylogenetic "families" or "superfamilies." The fourth digit identifies phylogenetic "subfamilies." The fifth digit ("clusters") corresponds to the substrate specificity, as presumed by experimental data or stringent sequence identity . In TC, the designation 2.A.1 represents MFS and the next digit denotes the family, for instance, 2.A.1.1 represents sugar transporters and so on. Any two transport systems in the same subfamily of a transporter family that transport the same substrate(s) are given the same TC number, regardless of whether they are orthologues or paralogues.
Till date only a few MFS transporters namely MDR1, FLU1, NAG3, NAG4, JEN1, ARN1 and NGT1 [16, 17, 18, 19, 20, 21, 22], have been identified and characterized in C. albicans. Additionally, over 20 hexose transporters and glucose sensors are known to exist in C. albicans that reflect the varied niches in which this pathogen thrives . However, very limited knowledge about other MFS transporters is available in C. albicans. Out of all the known MFS in C. albicans MDR1, its alleles and FLU1 are shown to be the only drug efflux pumps transporters. MDR1 was initially identified as a gene, which conferred resistance to the tubulin binding agent benomyl and tetrahydrofolate reductase inhibitor methotrexate [24, 25]. MDR1 expression in S. cerevisiae confers resistance to several unrelated drugs and its overexpression has been linked to azole resistance in C. albicans. The expression of MDR1 in C. albicans cells is enhanced by benomyl, methotrexate and several other unrelated drugs, and is found to be more pronounced in some of the azole resistant clinical isolates [26, 27]. Keeping in view, the relevance of the MFS multidrug transporters in general and in multidrug resistance (MDR) in particular, in the present study, we have examined MFS superfamily of proteins in C. albicans. Although, earlier annotation of the previous Candida genome assembly (version 19) predicted 71 MFS genes, no systematic classification was given . To address this question, we have applied a comprehensive bioinformatics approach to identify and annotate all MFS transporter genes in the Candida genome from assembly version 21 and systematically searched for evidence of their expression. It is hoped that these findings will provide the scientific community with the necessary framework needed for the functional characterization of the MFS proteins and a better understanding of this medically and pharmaceutically significant superfamily.
Identification and sequence-based functional grouping of C. albicansputative MFS genes
The Sugar Porter (SP) Family (TC # 2.A.1.1)
The SP family is widespread and have members from all of the major groups of living organisms: bacteria, archaea, eukaryotic protists, fungi, animals and plants . It forms one of the largest families with 22 members in C. albicans. We have previously characterized HGT1  and many studies since then have also reported other SP members namely HGT2-HGT20 [23, 30]. Apart from identifying HGT1-HGT20, in the present study, we have also identified two previously unidentified members, namely MAL31 (orf19.3981) and orf19.4923 (Additional file 1) both having homologues in S. cerevisiae, MAL31 and YFL040W, respectively. Further, based on the homology, the 22 SP members show significant similarity with various sugars, namely arabinose, quinate, myoinositol, maltose, fructose, glycerol, monosaccharide, glucose, hexoseand xylose (Additional file 1).
The Drug: H+Antiporter-1 (DHA1) Family (TC # 2.A.1.2)
DHA1 family like the SP is widely distributed and the members include both drug-specific and MDR efflux pumps. Like SP it also forms one of the largest families having 22 members in Candida (Additional file 1). MDR1 gene in C. albicans is one of the best characterized members, originally known to confer resistance to benomyl and methotrexate [19, 20, 24, 25]. Subsequent studies indicated that MDR1 also encodes resistance to cycloheximide, benztriazoles, 4-nitroquinolone-N-oxide and sulfometuron methyl [24, 31]. Disruption of the MDR1 gene reduced the virulence of C. albicans . Other characterized members of this family include FLU1, NAG3 and NAG4. Disruption of FLU1 in C. albicans hyper-susceptibility to mycophenolic acid thus suggesting that it could be a preferred substrate for the transporter . On the other hand, NAG3 (TMP1) and NAG4 (TMP2) show susceptibility to a number of unrelated compounds such as cycloheximide, 4-nitroquinoline-N-oxide and 1,10-phenanthroline and are upregulated in response to these drugs, suggesting that they function as multiple drug efflux pumps . Apart from MDR1, which is known as a clinically relevant efflux pump protein, none of the other characterized members have been directly linked to MDR of C. albicans.
The Drug: H+Antiporter-2 (DHA2) Family (TC # 2.A.1.3)
The DHA2 family of drug:H+ antiporters with 14 predicted transmembrane-spanning segments, consists of nine members in C. albicans which show significant similarity to transporters, namely aminotriazole, 4-nitroquinoline-N-oxide, Me2+-tetracycline antiporter, vacuolar basic amino acid (Arg, Lys, His) transporter and metal:tetracycline/oxytetracycline efflux pump (Additional file 1). In C. albicans no member of this family has yet been characterized whereas in S. cerevisiae two DHA2 proteins, SGE1 and ATR1 are well studied. ATR1 has been shown to confer resistance to the structurally unrelated compounds aminotriazole and 4-nitroquinolone-N-oxide and expression of ATR1 is inducible by the former but not the latter [32, 33]. SGE1 appears to confer resistance to crystal violet  and ethidium bromide [35, 36].
The Fucose: H+Symporter (FHS) Family (TC # 2.A.1.7)
FHS is a small family with two ORFs identified in Candida namely: orf19.4090 and orf19.7490 (Additional file 1). They are homologous to S. cerevisiae BSC6, which encodes a protein of unknown function exhibiting genomic organization compatible with a translational read through-dependent mode of expression .
The Phosphate: H+Symporter (PHS) Family (TC # 2.A.1.9)
PHS family is unusual in that it has representatives only in yeast, fungi and plants but none in bacteria, animals and other eukaryotes . The occurrence of distant homologues in both the plant and fungal kingdoms suggests that they possess isoforms that diverged from each other well before plants diverged from fungi . Two well characterized members of the PHS family are the Pho84 inorganic phosphate transporter of S. cerevisiae  and the GvPT phosphate transporter of Glomus versiforme . In this study, we have identified five ORFs belonging to PHS family in C. albicans with homology to phosphate: H+ symporters (Additional file 1). However, none of the members identified in Candida has yet been characterized.
The Oxalate: Formate Antiporter (OFA) Family (TC # 2.A.1.11)
OFA family members are widely distributed in nature, being present in the bacterial, archaeal and eukaryotic kingdoms . In C. albicans, our searches revealed two members (Additional file 1). OxlT, the oxalate:formate antiporter from Oxalobacter formigenes, is the hallmark protein and provides the basis for naming the OFA family [40, 41]. This protein has been purified, reconstituted in an artificial membrane system as well as structurally and functionally characterized [42, 43].
The Sialate: H+Symporter (SHS) Family (TC # 2.A.1.12)
SHS family, like the PHS family, is very small with only two members namely JEN1 and JEN2, identified in the present as well as a previous study . JEN1 has been described as the first monocarboxylate transporter of C. albicans showing loss of all measurable lactate permease activity upon its disruption. Further, lactate uptake by JEN1 was competitively inhibited by pyruvic and propionic acids while acetic acid behaved as a non-competitive substrate .
The Monocarboxylate Porter (MCP) Family (TC # 2.A.1.13)
MCP family is exclusively present in yeasts and animals. In mammals, these permeases are known to transport monocarboxylates, namely pyruvate, lactate and mevalonate with inwardly-directed polarity and presumably function as proton symporter  while it is reported that the yeast monocarboxylate transporter proteins perform functions other than their mammalian counterparts . This family has six members identified in Candida.
The Anion: Cation Symporter (ACS) Family (TC # 2.A.1.14)
ACS is a relatively large family having representation in bacteria, yeasts and animals comprising mainly of symporters that are known to accumulate their substrates in symport with either Na+ or H+, depending on the system. They may transport either inorganic (e.g. phosphate) or organic anions (e.g. glucarate, hexuronate, tartrate, allantoate or 4-hydroxylphenyl acetate) . In Candida, we have identified 16 members showing significant similarity to transporters having prefered substrates, namely tartrate, allantoate, nicotinate, biotin and pantothenate (Additional file 1).
The Aromatic Acid: H+Symporter (AAHS) Family (TC # 2.A.1.15)
The members of AAHS family occur exclusively in gram-negative bacteria where they are known to transport a variety of aromatic acids like benzoate, 4-hydroxybenzoate, 3-hydroxyphenylpropionate, 2,4-dichlorophenoxyacetate as well as niacin and cis, cis-muconate . In Candida, a single member of AAHS family has been identified, namely orf19.6952 showing significant similarity to putative niacin uptake porter (Additional file 1). This family has no representation in S. cerevisiae and thus unique to Candida.
The Siderophore-Iron Transporter Family (TC # 2.A.1.16)
All the known members of this family are from yeast species. In C. albicans this family is represented by a single protein known as siderophore transporter, SIT1/ARN1 (orf19.2179) which is required in ferrichrome-iron uptake. Previous reports suggest that deletion of ARN1 leads to reduced ability of C. albicans to use iron bound to the hydroxamate-type siderophore ferrichrome and upon deletion of the two high-affinity iron permease C. albicans genes (FTR1 and FTR2), the activity was completely abolished [18, 46]. According to another study, siderophore uptake by Sit1p/Arn1p is required in a specific process of C. albicans infection, namely epithelial invasion and penetration, while in the blood or within organs other sources of iron, including heme, may be used .
The Organic Cation Transporter (OCT) Family (TC # 2.A.1.19)
In C. albicans this family is represented by a single uncharacterized member FGR2 (orf19.7071) showing similarity to organic anion: dicarboxylate transporter. These proteins are known to transport organic cations and/or anions and catalyze uptake of cationic drugs such as tetramethyl ammonium, cimetidine, procainamide, quinidine and some endogenous metabolites such as N-methyl-nicotinamide [48, 49, 50, 51].
The Vesicular Neurotransmitter Transporter (VNT) Family (TC # 2.A.1.22)
These proteins are more closely related to SP family than to other MFS families. The better characterized members of the VNT family are synaptic vesicle proteins from mammals, the electric eel and insects [52, 53, 54, 55]. In C. albicans this family is represented by a single member orf19.6578 with significant similarity to dopamine transporter.
The Peptide-Acetyl-Coenzyme A (PAT) Transporter Family (TC # 2.A.1.25)
Members of the PAT family are present across bacteria, yeast and animals . Amongst the well characterized proteins of this family include acetyl-CoA transporter localized in the endoplasmic reticulum and Golgi membranes of humans . AmpG protein of E. coli belonging to PAT family, brings into the cell peptides, including cell wall degradative peptides and glycopeptides, to act as inducers of β-lactamase synthesis . The acetyl-CoA transporter is expected to function by acetyl-CoA:CoA antiport while the AmpG protein is most likely energized by substrate:H+ symport. In C. albicans this family is represented by a single member orf19.3782 with significant homology to acetyl-CoA:CoA antiporter.
The L-Amino Acid Transporter-3 (LAT3) Family (TC # 2.A.1.44) (also called the SLC43 family)
LAT3 transports neutral amino acids such as L-leucine, L-isoleucine, L-valine and L-phenylalanine by a Na+-independent, electroneutral, facilitated diffusion process and also transports amino acid alcohols. In C. albicans, this family is represented by two ORFs: orf19.6654 and orf19.6316.
The Proton Coupled Folate Transporter/Heme Carrier Protein (PCFT/HCP) Family (TC # 2.A.1.50)
In C. albicans, this family is represented by a single member orf19.6976 showing homology to high-affinity folate transporter. PCFT from human has been shown to act both as an intestinal proton-coupled high-affinity folate transporter and as an intestinal heme transporter which mediates heme uptake from the gut lumen into duodenal epithelial cells. The iron is then released from heme and may be transported into the bloodstream [58, 59].
The N-Acetylglucosamine Transporter Family (TC # 2.A.1.58)
NGT1 from C. albicans represents the first eukaryotic N-acetylglucosamine (GlcNAc) transporter and is the only known member of this family. It is required for efficient GlcNAc uptake and for inducing hyphae development at low GlcNAc concentrations . High concentrations of GlcNAc could bypass the need for NGT1 to induce hyphae, indicating that elevated intracellular levels of GlcNAc induce hyphal formation. Expression of NGT1 in S. cerevisiae promoted GlcNAc uptake, indicating that NGT1 acts directly as a GlcNAc transporter . No homologue of NGT1 was detected in S.cerevisiae.
Most of the identified members of MFS superfamily are expressed
Oligomers used for RT-PCR
Expected amplicon length (bp)
Discussion and conclusion
In this study, we report the complete transportome of MFS superfamily of C. albicans. Computational analysis of the C. albicans genome assembly (version 21) from CGD enabled us to identify 95 potential MFS permeases. The latter were classified according to both phylogeny and function based TC system earlier developed by Saier [13, 15]. This approach enabled us to cluster these 95 MFS proteins into 17 distinct families. Indeed each of the predictions must be tested experimentally before final conclusions are reached with reference to the expression and function of the proteins analyzed.
Our expression analysis of the published work revealed that out of 95 MFS genes, 87 are expressed under either basal (uninduced) or in different specific experimental conditions (Additional file 1). Most of the genes identified in the present investigation are expressed either at mRNA or protein level thus validating our analysis. The expression of the remaining 8 genes was not detected. This would imply that either these genes are expressed under specific growth condition or that they may be candidate pseudogenes. To dissect the role of each putative member of MFS superfamily, an essential part of the process will now involve construction of multiple knockout mutants, which will enable to unravel their role in drug or nutrient transport.
Media chemicals were obtained from HiMedia (Mumbai, India). Luria Bertani broth and agar media was purchased from Difco, BD Biosciences, NJ, USA. Taq DNA polymerase, ultra pure deoxyribonucleotides (dATP, dGTP, dCTP and dTTP) were obtained from New England Biolabs (NEB Inc.), USA. Moloney murine leukemia virus (M-MuLV) reverse transcriptases (RT) and RNase inhibitor were obtained from MBI Fermentas. Oligotex mRNA Mini Kit was purchased from Qiagen. Oligonucleotides used were commercially synthesized from Sigma-Aldrich. All Molecular Biology (MB) grade chemicals used in this study were obtained from Sigma Chemical Co. (St. Louis, USA).
Identification of C. albicansMFS transporter genes
C. albicans genome assembly version 21 http://www.candidagenome.org was searched for MFS genes using well known MFS proteins from Swiss-Prot database http://www.expasy.ch/sprot as queries in TBLASTN searches . Our initial query dataset had 230 MFS proteins which were used individually to BLAST Candida genome. Out of 230 sequences, 38 were chosen which gave significant E-values and were maximally dissimilar among themselves covering diversity of MFS from plants, fungi and mammals. It should also be noted that although a rather relaxed E-value (0.0001) cut-off was used, the observed E-values between the test sequence and the closest query sequence were much below this threshold indicating that the hits obtained were highly significant. The high-scoring segment pairs (HSPs) returned from TBLASTN searches were checked for duplications using an in-house written Perl script and only those that gave the lowest E-value with one or the other sequences from the query set were kept for further analysis. The overlapping HSPs were merged so as to obtain the largest contiguous stretch of nucleotides in the C. albicans genome, which had strong sequence homology with the MFS proteins in the query dataset. Since the test sequences always gave significant hits with a number of query sequences the best individual alignments were merged using overlaps. The availability of multiple TBLASTN matches (on account of the large number of query sequences used) made the merging step relatively easy and unambiguous. It also greatly increased the reliability of identifying a true hit and distinguishing it from false positives. The protein sequences were obtained by a six frame translation of the HSPs, using the tool "transeq" from the EMBOSS package http://www.ebi.ac.uk/emboss/transeq and taking the largest open reading frame (ORF). 95 ORFs identified from C. albicans after the initial TBLASTN searches were then pooled with the query dataset of 38 sequences to form a new query dataset and used iteratively for subsequent searches until no new ORFs were obtained. Subsequently, all potential genes were analyzed for MFS domains using the programs ExPASY PROSITE , InterPro  and Conserved Domain Database at NCBI . Transmembrane domains were predicted using TMHMM http://www.cbs.dtu.dk/services/TMHMM/.
Sequence-based functional grouping of C. albicansMFS genes
C. albicans MFS genes, as identified above, were further subjected to sequence-based classification according to TC system which is based on both functional and phylogenetic information [13, 14, 64]. Each putative MFS was individually searched against the TCDB. For this purpose the BLAST server at the transporter database http://www.tcdb.org was used with the default settings and E-value cut-off of 1.0 from the given choices of E-values (1000 to 0.0001). It should be noted that here also we chose a rather relaxed E-value cut-off and the potential MFS identified in Candida returned much lower E-values with the MFS sequences in the transporter database. To validate the family designations obtained for C. albicans using TC system all the known S. cerevisiae MFS proteins were also searched against the TC database using the same method as described for C. albicans.
Phylogenetic relationship with S. cerevisiae
A systematic search for S. cerevisiae homologues of the proteins was done with each C. albicans MFS gene by using SGD BLASTP tool http://www.yeastgenome.org/.
Expression analysis of the putative MFS genes
In order to validate the existence of the putative MFS genes in C. albicans, expression analysis was done by extensive mining of the data available from the previous genome and proteome-wide studies (Additional file 1) as well as experimentally by RT-PCR.
Total RNA isolation
Total RNA from C. albicans isolate SC5314 was prepared from mid-logarithmically grown phase cells. In a standard preparation, 10 ml of cells, optical density at 600 nm (OD600) of 1.0, were pelleted and washed with 10 ml of ice-cold H2O and spun at 5000 rpm. The pellet was resuspended in 1.0 ml of TRI® Reagent (Sigma) and 0.3 ml of ice-cold, acid-washed 0.4–0.6 mm diameter glass beads (Sigma, St. Louis, MO, USA) were added and vortexed for 5 min. Chloroform (0.2 ml) without isoamyl alcohol was added and the tubes were shaken vigorously for 15 s. The samples were incubated at room temperature for 15 min, centrifuged at 12,000 × g for 15 min at 4°C. The upper colourless aqueous phase was transferred to a new tube and 0.5 ml of isopropanol was added. The tubes were incubated at room temperature for 10 min, centrifuged at 12,000 × g for 10 min and the pellet washed with 75% ethanol and recentrifuged. The pellet was air dried and resuspended in 100 μl of H2O. All the experiments were done with diethyl pyrocarbonate (DEPC) treated H2O. DNA free RNA was prepared by treating total RNA with DNase RQ1 (Promega). The OD260 and OD280 were measured and the integrity of the total RNA was visualized by subjecting 2–5 μl of the sample to electrophoresis through a denaturing 1% agarose/2.2 M formaldehyde gel. The total RNA preparation isolated was stored at -80°C till further use.
Reverse transcription PCR (RT-PCR)
The nucleotide sequence of the oligonucleotide primers used for the RT-PCR was taken from CGD http://www.candidagenome.org. Total RNA isolated from SC5314 (as described above) was enriched with poly(A)+ (polyadenylated) mRNA using the Oligotex mRNA Mini Kit protocol (Qiagen) and used subsequently for performing the reverse transcription reaction as described elsewhere . To synthesize cDNA, ca. 0.1 μg of poly(A)+ RNA was placed in a 0.5 ml reaction tube with 1 μM of oligo(dT)18 anchor primer stock and the volume was adjusted to 11 μl with DEPC treated water. The mixture was incubated for 10 min at 70°C and chilled on ice for 1 min, after which the remainder of the reaction mixture was added from a master mix to the reaction tube in order for each reaction to contain a 1 mM concentration each of dATP, dCTP, dGTP and dTTP; 40 U of RNase inhibitor in a buffer consisting of 50 mM Tris-HCl (pH 8.3), 50 mM KCl, 4 mM MgCl2 and 10 mM DTT. After brief mixing, the reaction was incubated for 10 min at 37°C followed by addition of 40U of M-MuLV reverse transcriptase. Finally, the reaction was incubated at 37°C for 60 min and then stopped by heating at 70°C for 10 min followed by chilling it on ice for 1 min. The synthesized cDNA was purified from unincorporated dNTPs, oligo(dT)18 anchor primer and proteins by using Oligotex mRNA Mini Kit. Amplification of specific mRNA of each gene was performed using corresponding appropriate dilution of cDNA as template (generally 1:4) and 1 μM of each specific forward and reverse PCR primer as mentioned in Table 1. (parameters: initial denaturation of 95°C for 5 min followed by 35 cycles denaturation at 95°C for 15 s, annealing at 55°C for 30 s, elongation at 72°C for 30 s and final extension at 72°C for 10 min). As a positive control, MDR1 specific forward MDR1-F and reverse MDR1-R primer (corresponding to positions 1038–1396 in the MDR1 genomic sequence) was also used. The negative control (without RT) established that the PCR products generated in the RT-PCR were not due to genomic DNA contamination (data not shown). Resulting RT-PCR products were electrophoresed on a 1.2% agarose gel in 1× TAE.
The work presented in this paper has been supported in parts to RP by grants from Department of Biotechnology, (DBT/PR4862/BRB/10/360/2004), Council of Scientific and Industrial Research (38(1122)/06/EMR-II, 22/3/2006), Department of Science and Technology (SR/SO/BB-12/2004, 5/9/2005) and Indo-French Centre for the Promotion of Advanced Research (IFC/A/3403-2/2006). MG and NP acknowledge the University Grants Commission, India and Indian Council for Medical Research, respectively, for the support in the form of junior and senior research fellowships.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.