Aspergillus nidulans CkiA is an essential casein kinase I required for delivery of amino acid transporters to the plasma membrane

Summary Type I casein kinases are highly conserved among Eukaryotes. Of the two Aspergillus nidulans casein kinases I, CkiA is related to the δ/ε mammalian kinases and to Saccharomyces cerevisiæ Hrr25p. CkiA is essential. Three recessive ckiA mutations leading to single residue substitutions, and downregulation using a repressible promoter, result in partial loss-of-function, which leads to a pleiotropic defect in amino acid utilization and resistance to toxic amino acid analogues. These phenotypes correlate with miss-routing of the YAT plasma membrane transporters AgtA (glutamate) and PrnB (proline) to the vacuole under conditions that, in the wild type, result in their delivery to the plasma membrane. Miss-routing to the vacuole and subsequent transporter degradation results in a major deficiency in the uptake of the corresponding amino acids that underlies the inability of the mutant strains to catabolize them. Our findings may have important implications for understanding how CkiA, Hrr25p and other fungal orthologues regulate the directionality of transport at the ER-Golgi interface.

: ckiA is an essential gene: inactivation in a diploid strain.
Panel A shows a scheme of the chromosome III homologues of strain LH61. In red ckiA, in green argB.
The strain is ckiA + /ckiA + homozygous and argB + /argB2 heterozygous. It is also homozygous for pantoB100 in chromosome VII and thus a pantothenate auxotroph. If a ckiA deletion is a recessive lethal, as indicated by the heterocaryon experiments, the pantoB + gene (in blue) could substitute one of the ckiA genes, resulting in a ckiAΔ hemizygote which is pantoB + (shown in blue). This can occur in coupling (as it occurred in T12.) or in repulsion with argB2 (as it occurred in T15). Panel B, Sothern blot of representative sectors originating form the T15 diploid transformant. (XbaI digest) showing the 6.5 kB band of ckiA present in the recipient strain; and in the diploid transformant, the two bands of 6.5 kB and 2.8 kB representing the intact ckiA gene and the restriction fragment predicted to result from the pantoB + substitution, respectively. All haploid (argB + ) strains show only the wt ckiA restriction fragment. Panel C, representative sectors originating from the T15 diploid, analysed as above. Three haploid strains (argB2) behave as predicted, while the two putative haploid sectors (to the left of the panel), which show both bands were shown on further testing to be prototrophic for all markers and actually diploid. A 1140 bp PCR fragment corresponding to the region downstream the deletion cassette (from position +1392 to position +2530 relative to the ckiA translation start) was used as a probe.

Supplementary Experimental procedures
Supplements: Ammonium L (+)-tartrate, and urea were used at 5 mM, acetamide, nitrate (NaNO 3 ), and nitrite (NaNO 2 ) at 10 mM. Uric acid, xanthine, hypoxanthine, adenine and allantoin at 0.1 mg/ml, as sole nitrogen sources. Amino acids were generally used at 5 mM as carbon or nitrogen sources in solid media (unless differently described in the text) and at 10 mM as nitrogen sources in liquid media. Aspartate and glutamate were used as monosodium salts (SIGMA). In MM, glucose was used as a carbon source at 1 % w/v. Acetate was used at 50 mM as a carbon source. The amino acid toxic analogues D-serine and DLparafluorophenylalanine (FPA) were used at 5 mM and 55 μM respectively. Uracil and uridine were used at 5 mM and 10 mM respectively. ckiAmutant selection and characterisation: ckiA2 was selected after N-methyl-N'-nitro-Nnitrosoguanidine mutagenesis of a strain of genotype biA1;puA2;fwA1 as a faster growing, nitrogenstarved sector (Herman and Clutterbuck, 1966) on biotin-supplemented glucose minimal medium containing limiting (10 nM) putrescine and 5 mM delta-aminovaleric acid as nitrogen source. ckiA102 was isolated as described by Arst et al. (1981) as a spontaneous mutation conferring resistance to 50 mM proline to a strain carrying proA6 (conferring proline auxotrophy) and sasA60 (conferring toxicity to proline) (Sharma, 1984). The selection of novel mutants resistant to D-serine or FPA (including ckiA1919) was carried out during this work by inoculating conidia of CS2290 on solid supplemented MM medium containing D-serine with uric acid as a sole nitrogen source. After 4-5 days of incubation at 37 o C, mutant sectors were recovered and purified on supplemented MM containing ammonium as a nitrogen source.
Putative ckiA fbaA1013 double mutants were checked by outcrossing to a wild type, with recovery of both parental classes and for the strains tested in Fig. 1B, also by confirming the presence of the relevant ckiA mutation after PCR amplification of the ckiA ORF (see text for the nature of each mutation).

Southern blot analyses:
The cross hybridisation of pRG3C to cosmid L12G07 was tested in HindIII and EcoRI digests with the purified SmaI-NotI fragment of plasmid pRG3C. The restriction enzymes used to monitor the copy number of ckiA in the genome of CS2498, CS1901 and CS1902 were NcoI, which does not have a recognition site within the genome sequence included in ckiA probe employed (a PCR product using primers pRG3F and prG3R and plasmid pRG3C as a template) and EcoRI that cuts twice within this sequence. Southern blot analysis of the transformants obtained with pAA21 and pAA22 plasmids was carried out using HindIII and the latter PCR product as a probe. To characterise the transformants obtained with pAA61, pAA62 and pAA63 (see below), a Southern blot was carried out using BglII and the purified BglII-SalI fragment from plasmid pAA51 as a probe.

Identification of ckiA using the AMA-NotI genomic library: this gene library was constructed in the self-
replicating plasmid pRG3-Not1 (Osherov and May, 2000) carrying the pyr4 gene of Neurospora crassa as a selection marker, which complements pyrG89 mutation of A. nidulans. CS1901 carrying ckiA102 and pyrG89 (resulting in a uracil/uridine requirement) was transformed with 10 μg of the amplified library.
Selection of transformants was carried out on supplemented MM containing 10 mM glycine as a sole nitrogen source in the absence of uracil and uridine. The transformants were tested for their ability to grow on a number of amino acids (proline, tryptophane, arginine, glutamate, β-alanine, leucine and serine) as sole nitrogen sources and by their sensitivity to the toxic analogues D-serine and FPA (on uric acid as a nitrogen source). The free plasmids harboured by nineteen independently obtained transformants, which were sensitive to both toxic analogues and able to grow on amino acids as nitrogen sources, were recovered by electroporating the E. coli strain DH10B with dialysed genomic DNA (extracted from mycelia grown in the presence of glycine as nitrogen source). The recovered plasmids were characterized by restriction (using BamHI, SphI and KpnI that cut once in pRG3-NotI polylinker) and one representative from each restriction profile was used to re-transform CS1901. Two plasmids, pRG3B and pRG3C, were able to fully complement ckiA102. Sequencing showed that pARG3C carries a 4301 bp insert comprising the whole putative open reading frame of ANID_04563.1 (accession number of the last release of the Broad Institute database, http://www.fgsc.net/aspergenome.htm), together with 322 bp of its upstream region and a partial sequence of ANID_04562.1. pARG3B carries an approximately 14 kb insert, which fully comprises the insert of plasmid pARG3C. In order to check whether the whole insert carried in plasmid pARG3C was able to complement ckiA102 after single ectopic integration, the latter was cloned in the integrative vector pBKS+ to give plasmid pAA2 (Table S3). pAA2 was used to transform CS1901 in the same conditions as previously mentioned. The insert carried by plasmid pARG3C only affords full complementation of the ckiA102 mutation when present in multiple copies, but only complements partially ckiA102 by single copy ectopic insertion at a number of different positions in the genome when cloned in the integrative vector pAA2. A full wild type phenotype was obtained only after single in locus insertion of the pAA2 plasmid (resulting presumably in correction of the mutation), which was visualized by Southern blot analysis as a locus duplication. Eventually, only one cosmid L12G07 was able to complement the mutation. L12G07 showed crosshybridisation with the insert contained in pRG3C (see above). Selection of transformants was carried out on supplemented MM containing 10 mM glycine as a sole nitrogen source. In order to isolate the whole ckiA gene in an integrative vector, first pAA5 was constructed (Table S3) by ligating the integrative pGEM TM -T easy vector to a 6.1 kb insert, which was PCR amplified using cosmid L12G07 as a template.

Identification of ckiA using the cosmid library
This was 1918bp longer than the pRG3C insert upstream the ANID_04563.1 gene sequences and shorter by 207 bp in the adjacent ANID_04562.1 sequences. pAA5 was used to transform CS1901 in the same conditions as previously. Subsequently, the BglII-SalI fragment (3099 bp) of pAA5 insert was subcloned into pBKS+ to construct plasmid pAA6 (Table S3), which was also used to transform CS1901 as previously. This fragment contains the complete ANID_04563.1 extended by 1053 bp upstream and by 595 bp downstream, and is able to complement ckiA102 in a single copy, as differently from pARG3C and derived plasmids (see above) it presumably carries a complete 5' upstream region. 1.78) the ckiA ORF, to be called respectively region A and C, were amplified using the primer pair Delk1-Delk2 (for region A) and Delk5-Delk6 (for region C) and genomic DNA from strain CS2498 as template.

Construction of ckiA deletions
The pantoB marker gene, to be called region B, was amplified from genomic DNA of CS2498 using the primer pair Delk3-Delk4. Regions A and C were independently fused by DJ-PCR with region B. Then, a nested fragment of each of these fused DNA segments was amplified using the primer sets ABFW-ABRV (for the region A+B) and BCFW-BCRV (for the region B+C), giving fragments D and E respectively. homologues. Haploidisation of diploid transformants was carried out on CM in the presence of benlate at 5 mg/ml (Hastie, 1970). Southern blot analysis (as above) was used to check the status of the ckiA gene in the issuing haploids. Figure S2 illustrates the results obtained.
Additionally, a second deletion cassette containing the 5'UTR-AFriboB-3'UTR fusion was constructed by sequential cloning of the corresponding PCR-amplified fragments in the ApaI, SphI, and NdeI sites of the polylinker of the pGEM-T Easy vector (Promega) using primer pairs carrying restriction enzyme adaptors (Table S4). The ckiA 1256 bp upstream region (starting at 285961 of contig 1.78 and ending 670 bp before the translation initiation codon), and the ckiA 1261 bp downstream region (starting at 289074 of contig 1.78, immediately after the translation termination codon), were, first, amplified using as template genomic DNA of TNO2A7 and the primer pairs CkiA 5' ApaI F-CkiA 5' SphI R, CkiA 3'SphI F-CkiA 3'NdeI R respectively. Subsequently, they were digested and cloned into the pGEM-T Easy vector, resulting in plasmid pCkiAKO. The AFriboB amplified from plasmid pnirAΔ using the primer pair AFriboB SphI F-AFriboB SphI R was cloned at the SphI site of the above plasmid. The resulting transformation cassette was amplified by the primer pair CkiA 5' ApaI F-CkiA 3' NdeI R and used to complement the riboB2 auxotrophic mutation of strain TNO2A7 by selecting on MM media with urea as sole nitrogen source, in the absence of riboflavin.

Construction of strains containing in-locus ckiA transcriptional and translational fusions:
The cassette containing the ckiA::sgfp::AFpyrG sequences was constructed by joining three different PCR fragments, as described by Szewczyk et al. (2006). DNA fragments corresponding to the central part of the construction were amplified from the p1439 plasmid using the primer pair CkiAp1439 F-CkiAp1439 R (Table S4). The upstream flanking sequence of the sgfp ORF is a 1109 bp fragment (starting at 287993 of contig 1.78) followed by the 5GA linker, while the downstream flanking sequence is a fragment containing a 1239 bp fragment corresponding to the 3' end of the ckiA gene (starting at 289115 of contig 1.78), 2bp just after the chain termination codon. Both flanking fragments were amplified from genomic DNA of a TNO2A7 strain using the CkiA P1-CkiAP3 and CkiA P4-CkiA P6 primer pairs respectively.
The entire fusion cassette was amplified using the CkiA P2-CkiA P5 nested primer pair and was used to transform the TNO2A25 strain. Transformants were selected on MM containing urea as nitrogen source in the absence of uracil and uridine.
The thiamine repressible promoter was originally described in A. oryzae, (Shoji et al., 2005)  Membrane protein extracts were prepared from strains LH121, LH127, CAM13 and AMC264 (Calcagno-Pizarelli et al., 2007). Conidiospores were inoculated in an appropriately supplemented MM containing ammonium L (+) tartrate as nitrogen source and incubated for 14 h at 30° C. Mycelia were harvested and transferred to supplemented MM containing 10 mM GABA (inducing conditions) or 10 mM ammonium (repressing conditions) and incubated for an additional 3 h at 37°C. Total membrane protein was determined using a bicinchoninic acid protein assay kit (Sigma). Samples were loaded onto 12% SDSpolyacrylamide gels (Bio-Rad) adjacent to Rainbow molecular mass markers (10,000 to 250,000 Da; Amersham Biosciences). Proteins were transferred onto nitrocellulose membranes that were reacted with rat anti-HA primary antibodies (catalog no. 3F10; Roche). The secondary antibody was a peroxidasecoupled goat anti-rat immunoglobulin G (Southern Biotechnology). Peroxidase activity was revealed using an ECL Western blotting detection system (Amersham Pharmacia).
Amino acid uptake assays: Four strains (CS2498, CAM45, CS1947 and CS1903) were assayed for proline uptake and of two strains (CS2498 and CS1947) for glutamate uptake. 3 H-labeled amino acid uptake was measured in germinating conidia as described in Robinson et al. (1973) and Tazebay et al.
(1995), modified as follows. Conidiospores were inoculated in supplemented liquid MM using urea as nitrogen source and incubated under agitation at 37 o C. At the onset of germ tube appearance (4 h for ckiA+ strains and 4 h 40 min for ckiAstrains), conidia were collected by filtration on a Millipore filter (d=2.5 μm) using a vacuum pump and re-suspended in 1 ml supplemented MM with urea as nitrogen source and then transferred on ice to prevent further growth. After pre-incubation of 90 μl of a germinating conidia suspension for 10 min at 37 o C, 10 μl "Hot Solution" was added and the samples were  an Axioplan Zeiss phase-contrast epifluorescence microscope with appropriate filters was used and the resulting images acquired with a Zeiss MRC5 driven by AxioVs40 (version 4.40.0) software were then processed with Adobe Photoshop CS2 (version 9.0.2) software. Observation of asexual compartments was according to Pantazopoulou et al. (2007). Sample preparation and vacuolar staining with CMAC (7amino-4-chloromethyl coumarin) (Molecular Probes) was according to Gournas et al. (2010).