Genetic studies reveal that myristoylCoA:protein N-myristoyltransferase is an essential enzyme in Candida albicans

Molecular Microbiology (1995) 16(2), 241-250

Genetic studies reveal that myristoylCoA:protein N-myristoyltransferase is an essential enzyme in Candida albicans Robin A. Weinberg,^" Charles A. McWherter,^ Sandra K. Freeman,^ David C. Wood,^ Jeffrey I. Gordon^ and Stephen C. Lee^ ^ Department of Research and Development. G. D. Searle and Company, St. Louis, Missouri 63198, USA. ^Department of Molecular Biology and Pharmacology, Washington University School of Medicine, St. Louis, Missouri 63110. USA. Summary MyristoylCoA:protein N-myristoyltransferase (Nmt) catalyses the co-transiational, covalent attachment of myristate (014:0) to the amino-terminal glycine residue of a number of eukaryotic proteins involved in cellular growth and signal transduction. The NMT1 gene is essential for vegetative growth of Saccharomyces cerevisiae. Studies were carried out to determine if Nmt is also essential for vegetative growth of the pathogenic fungus Candida albicans. A strain of C. albicans was constructed in which one copy of NMT was partially deleted and disrupted. A Gly-447 - Asp mutation was Introduced into the second JVAfr allele. This mutation produced marked reductions in catalytic efficiency at 24 and 37 C, as judged by in vitro kinetic studies of the wild-type and mutant enzymes which had been expressed in, and purified from, Escherichia coli. The growth characteristics of isogenic NMT/NMT, WAfr/Anmt, and nnitA/nmtG447D C. albicans strains were assessed under a variety of conditions. Only the nmtA/nmtG447D strain required myristate for growth. This was true at both 24 and 37 C. Palmitate could not substitute tor myristate. Incubation of nmt\/nmtG447D cells at 37 C in the absence of myristate resulted in cell death as observed by the inability to form colonies on media supplemented with 500 MM myristate. Studies in an immunosuppressed-mouse model of C. albicans infection revealed that the NMllSnmt strain produced 100% lethality within 7d after intravenous administration while the isogenic nmtA/nmtG447G strain produced Received 5 September. 1994; revised 16 December, 1994; accepted 20 December, 1994. 'For corespondence. Tel (314) 537 7053; Fax

{314)537 6480. 1995 Blackwel) Science Ltd

no deaths even after 21 d. These observations establish that Nmt is essential for vegetative growth of C. afbicans and suggest that Inhibitors of this acyttransferase may be therapeutical Iy useful fungicidal agents. Introduction N-myristoylation of proteins is catalysed by myristoylCoA:protein N-myristoyttransterase (Nmt). Nmt transfers myristate {C14;0) from myristoylCoA to the amino-terminal Gly residue of a variety of eukaryotic cellular and viral proteins (reviewed in Rudnick et ai, 1993a). Metabolic labelling studies mdicate that Saccharomyces cerevisiae produces at least 12 N-myristoylproteins during exponential growth (Duronio etai, 1991). Seven of these cellular N-myristoylproteins have been identified to date (Johnson et ai, 1994a) and three — ADP ribosylation factor-1 (Arf 1 p), Arf2p, and the alpha subunit of the heterotrimeric G protein involved in the mating response (Gpa1 p) — are essential proteins which need a covalently bound myristoyi group for full expression of their physiological functions (Stone et ai., 1991; Dohlman et al.. 1993; Kahn et al.. 1995). Genetic studies have established that the NMT1 gene is essential for the viability of S. cerevisiae (Duronio etai, 1989; 1991; Johnson etai, 1994a). Candida albicans is a dimorphic, asexual fungus. A laboratory strain of C. albicans (B311) synthesizes a small number of cellular N-myristoylproteins during exponential growth (Langner e^ ai, 1992; Wiegand et ai, 1992), The C. albicans NMT gene (encoding Nmt) has been isolated. Its 451 amino acid protein product shares 55% identity with the S. cerevisiae acyltransferase (Wiegand et al., 1992). C. albicans contains at least two ARF genes (Denich etai, 1992; Langner etai, 1992). One of the Arts is known to be a substrate for Nmt (Lodge et ai. 1994). C. albicans CAG1 can complement the gro\wth arrest and mating defects found in strains of S. cerevisiae with gpal null alleles (Sadhu et ai, 1992). The aminoterminal sequence of Cag1 (GCGASVPVDD) makes it a likely Nmt substrate (Wiegand et ai.. 1992). The peptide substrate specificities of C. albicans and S. cerevisiae Nmts differ from that of humah Nmt (Rudnick etai, 1992a; Wiegand etai, 1992; Rocque etai., 1993). In contrast, surveys of a large panel of myristic acid

242 fl. A. Weinberg eX 3\. analogues indicate that the acylCoA-binding sites of the orthologous enzymes are quite similar (Kishore et ai, 1993; N. Kishore and J.I. Gordon, in preparation). This apparent divergence in the enzymes' peptide but not acylCoA-binding sites probably reflects their similar requirements for myristoylCoA and differences in the numbers and types of protein substrates they acylate in vivo. These observations raise the possibility that it may be possible to design species-specific inhibitors of C. albicans Nmt which could function as fungistatic or fungicidat agents. C. albicans is by far the most common cause of fungal infections in immunocompromised patients. For example, 90% of all patients with acquired immune deficiency syndrome develop C. albicans infections at some point during the course of their disease (Dupont ef ai, 1992). Currently, there are few fungicidal drugs available and those that exist have side-effects which often limit their therapeutic usefulness. However, before an Nmtinhibitor approach can be used to treat systemic C. albicans infections, further information is required to determine whether this enzyme is essential for vegetative growth of this organism. One way of addressing this

10-fo!d reduction in affinity for myristoylCoA (Duronio etai, 1991; Rudnick etai, 1992b). S. cerevisiae strains with nmti-i81 exhibit growth arrest at 37 C owing to reductions in the level of acylation of several cellular Nmyristoylproteins (Duronio et ai. 1991; Johnson et al., 1994a). The growth arrest observed at the non-permissive temperatures can be relieved by addition of myristate to the media. Other fatty acids such as C12;0, C15:O or C16:0 cannot be substituted for C14:O. Based on these findings, we hypothesized fhat a Gly-447 - Asp mutation in C. albicans NMT would yield an allele that produces temperature-sensitive growth arrest and myristic acid auxotrophy in this fungus. Therefore, isogenic strains containing single copies of a wild-type or mutant NMT allele have been generated and characterized in culture and in an immunosuppressed-mouse model of systemic C. albicans infection.

Results and Discussfon 2.0

Construction of an nmtA/nmtG447D heterozygote

-

I

2

3 4 5 6 7 B

Fig. 1. Southern biot analysis of genomic DNA prepared from isogenic C. albicans strains containing wild-lype or mutant alleles of NMT. Genomic DNA (lOng) was digested with EcoRI. Blots were probed with a digoxigenin-labelled C- aibicans WMf fragment encompassing ils open reading (rame. The following strains were used for this analysis: CAI-4 (assigned genotype = A//Vfr/WMr),

MONCA101 {Anmt::hisG-URA3-hisG/NMTi. MONCA102 {.\nmt::hisG-URA3-hisG/NMT). MONCA103 {Anml::hisG/NMT), MONCA104 {Anmt::hisG/NMT). MONCA106 iAnmt::hisG/ nmtG447D::hisG-URA3-hisG). MONCA107 {Anmt::hisG/ nmlG447D::hisG-URA3-hisG), and MONCA105 (Anmtv.hisG/ NMT::tiisG-URA3~hisG). See text for further discussion.

Various protocols have been developed for generating specific mutant aileles in C. albicans (e.g. see Kurtz et ai, 1986; Kelly etai. 1987). Production of strains homozygous for a given mutation requires fhe use of two selectable marker genes or sequential transformations with the same marker (Kurtz and Man-inan, 1989; Gorman et ai, 1991; Fonzi and Irwin, 1993). Fonzi and Invin (1993) have described a very useful strategy that uses (i) a Ura strain of C. albicans (CAI-4) which has never been subjected to chemical mutagenesis, thereby providing a wild-type genetic background, and (ii) a URA3 selectable marker that can be repeatedly employed in successive transformations because the URA3 gene is flanked by direct repeats of the Salmonella typhimurium hisG gene. Spontaneous recombination between the direct repeats occurs at high frequency (Aiani etai, 1987), resulting in excision of the URA3 locus and restoration of a Ura phenotype. Strains with a Ura " phenotype can be selected based on their resistance to 5-f!uoro-orotic acid (5-FOA) and used for another round of targeted mutagenesis. 1995 Blackwell Science Ltd, Molecular Microbiology, 16, 241-250

Mutagenesis of the NMT locus in Candida albicans Table i . Candida aibicans stTains used in ihis sluby. Strain

Genotype

Source

CAW

Aura3::imnT434/Aura3::imrrH34

WIONCA101

Fonzi and Irwin (1993) Triis work

Anmti.hisG-URAS-hisG/NMT Aura3:: imm434/A ura3:: imm434 Anmt::hisG-URA3-hisG/NMT This work A ura3:: imm434/A ura3:: imm434 Anmt::hisG/NMT Aura3::imm434/ This work Aura3::imin434 Anmt::hisG/NMT Aura3::imm434/ This work Aura3::imm434 This work Anmt::hisG/NMT:.hisG-URA3hisG Aura3 ::imm434/A ura3:: imm434 Anml::hisG/nmtG447D::hisGThis work URAS-hisG A ura3:: imm434/A ura3:: imm434 Anmt::hisG/nmlG447D::hisGThis work URA3-hisG A ura3 ::imm434/A ura3:: imm434

MONCA102 MONCA103 MONCA104 MONCA105 MONCA106 MONCA107

We adopted this method to generate a strain of C albicans with one copy of a disrupted nmt aliete and one copy of an nmt allele containing a Gly-447 — Asp mutation. C. albicans NMT\s located on chromosome 4, Sfi\ fragment 4N or 4B, in strain 1006 (B. Magee, personal communication). NMT is also a single-copy gene in strain CAl-4 (data not shown). We disrupted one allele of this strain's NMTgene with hisG-URA3-hisG (Fig. IB). Southern blotting established that several of the uracil prototrophs had the expected NMT/Anmt..hisG-URA3~hisG genotype (MONCA101 and MONCA102 in Table 1; see Fig. 1). The URA3 marker was removed from NMT/ Anmt::hisG-URA30'hisG heterozygotes by selecting for 5-FOA-resistant colonies. Two isolates with a NMT/ Anmt:.his genotype were retained for further study (MONCA103 derived from MONCA101 and MONCA104 derived from MONCA102; see Table 1 and Fig. 1). A recombinant DNA fragment which specified the carboxyterminal half of the mutant nmt (lle-214 —Leu-451) and included the URA3 marker as well as downstream sequences which flank the NMT gene was used to replace the remaining copy of NMT in MONCA103 and MONCA104 with nmtG447D (Fig. 2D). Transformanfs were selected on ura" media containing 500MM myristate. The resulting uracil prototrophs were then examined for their ability to grow in the absence of myristate at both 37 C and 24 C. Of the 20 colonies tested, eight were unable to grow in the absence of myristate even at 24 C. Southern blot analysis indicated that all eight strains had the pattern expected if the original nmth..hisG disruption had been maintained and the hisG-URA3-hisG cassette, linked to the carboxy-terminal coding portion of nmtG447D. had recombined with the wild-type WMTallele (see MONCA106and MONCA107 in Fig. 1). Interestingly, 1995 Blackweli Science Ltd, Molecular Microbiology. 16. 241-250

243

one myristic acid prototroph (MONCA105) had a pattern of reactive EcoRI fragments in Southern blots that was identical to the pattern found in DNA prepared from myristic acid auxotrophs (Fig. 1). DNA sequence analysis confirmed the presence of the Gly-447 — Asp mutation in MONCA106 and MONCA107. The one anomalous myrisfic acid prototroph, MONCA105, did not contain this mutation in the NMT open reading frame, indicating thaf Gly-447 - ' Asp causes the myristic acid requirement of MONCA106 and MONCA107 at 24-37'C. MONCA105 was used as an isogenic NMTI/nmtA control in subsequent experiments.

Growth requirements o^Anmt;:hisG/nmtG447D;.hisGURA3-hisG strains The need for exogenous myristate. MONCA102 has a doubling time of 2.2 h at 24 C and 1.5 h on YPD at 37 C. These rates are similar to those of the isogenic strain CAI4 which is homozygous for NMT. As noted above, strains with an nm/null allele and an nm/allele containing

NMT genomic clone 2SO0

ATG BclJV Clal Seal

H—H-

Nsil TAABdl • . EcoRl\ Psil

B UfiA3 disruption

URA3 resolution 166

I

863 803 (Sail)

16531903

H

(SamHI \

1

EcoRI

2511 EcoRV

2900

1

Construct (or introducing Gty*'*7—• Asp mutation

JBglll

Fig. 2. Strategy used tor targeted mulagenesis of the C. albicans WfWgene, See the text for further discussion. The bold line indicates coding region of NMT.

244

R. A. Weinberg et al. AnmtrhisG/NMT

Anmt. hisG/NMT

Anmf \h)sG/nmtG447D.:hisG-URA3-hisG

Anml. hisG/NMT. hi5G-URA3-f\,sG

24-0

30'C

YPD

37°C

Fig. 3. Growth characteristics of isogenic C. albicans strains that produce wild-type or mutant Nmts. Strains were grown on YPD alone or YPD suppiemented with 500 jiM myristate (MYR) or palmitate (PAL) for 3d at 24, 30, or 37 C. The genotypes of the strains used for this study are indicated at the top of the Figure.

YPO + MYR

YPD + PALM

the Gly-447 -- Asp substitution are unable to grow at either 24 or 37 C unless media are supplemented with 500 nM myristic acid (MYR). This myristic acid auxotrophy was evident when the two Anmt::hisG/nmtG447D ..hisGURA3-hisG strains. MONCA106 and MONCA107, were grown on rich medium (YPD) as well as the defined minimal medium used for screening Jransformants. Palmitate (C16:0) cannot support growth of either MONCA106 or M0NCA107 at 24. 30. or 37 C (Fig. 3). Growth curves were calculated from cultures grown on YPD containing 100|.iM myristate to determine if any temperature effect was evident in liquid culture. Under these conditions MONCA105 and !VIONCA107 had similiar doubling times at 24 C (2h and 1.8h, respectively). However, when the cells were grown at 37 C there was a slight temperature effect: MONCA107 had a doubling time of 2.1 h whereas the doubling time of MONCA105 was only 1.5h. Phase-contrast microscopy of strains grown at 24 C on YPD/MYR revealed that cells which only produce nmtG447D (MONCA107) are larger, more rounded. and have larger vacuoles than the heterozygote parent MONCA102 or the isogenic prototroph MONCA105 (data not shown). The effects of perturbing the de novo pathway for fatty acid biosynthesis. S. cerevisiae fatty acid synthetase (Fas) produces long-chain acylCoAs from acetylCoA and malonylCoA. PalmitoylCoA and stearoylCoA are its principal products while myristoylCoA represents 3-5% of the total acylCoAs produced (Schweizer and Boiling, 1970; Singh et ai. 1985). The de novo pathway for acylCoA biosynthesis can be blocked by cenjlenin (CER; (2f?,3S)2,3-epoxy-4-oxo-7,10-frans.frans-dodecanienamide). The

metabolic effect is quite specific: cerulenin inhibits Fas (Funabashi et ai, 1989) but does not appear to inhibit the yeast's fatty-acid elongation or desaturation systems (Awaya etai. 1975). S. cerev/s/aestrains containing NMT1 cannot grow on YPD/CER at 24 or 37 C unless supplemented with long-chain fatty acids (e.g. C14:0, C16:0, C18:r^^) (Awaya et al., 1975; Johnson et al., 1994b). Isogenic strains containing nmt1~181 can grow in YPD/CER at 37 C if the media are supplemented with myristate but not palmitate (Duronio etai, 1991; Rudnick etai. 1992b; Johnson etai, 1994b). This latter result suggests that the mutant acyltransferase can utilize myristoylCoA derived from 014:0 which has been imported and then activated to its CoA derivative by cellular acylCoA synthetases (Johnson et ai, 1994b). This pathway maintains myristoylCoA pools that would otherwise be depleted by CER-mediated inactivation of the de novo pathway. The ability of isogenic Nf\/IT/NMT, nmtA/NMT, and nmtM nmtG447D strains of C. albicans to grow on YPD/CER was assessed at 24. 30, and 37 C. Cerulenin inhibited growth of CAI-4 (NMT/NMT). MONCA104 {nmtA/NMT). and f^ONCA107 {nmtMnmtG447D) on YPD at 24 and 37 C. The growth inhibition was slightly more severe with the nmtG447D-producing strain (Fig. 4). Palmitate could rescue the cerulenin-dependent growth inhibition of cells which contained one wild-type copy of NMT (MONCA104). Palmitate could not rescue cerulenin-treated cells that only synthesized nmtG447D (MONCA107). Myristate overcame the cerulenin inhibition of both strains and allowed better growth of MONCA104 than did C16:O (Fig. 4). These results suggest that both wild-type C. albicans Nmt and nmtG447D can access myristoylCoA generated through activation of exogenousiy derived myristate. CO 1995 Blackweil Science Ltd. Molecular Microbiology. 16. 241-250

Mutagenesis of the NMT locus in Candida albicans 245 Anmf Anml :hisG/nmtG447D :his6-URA3-hisG 24'C

CER

CER + MYR

CER+PALM

3T'C

I I I

Fig. 4. The effect ol cerutenin on growth of isogenic C. albicans strains containing wild-type or mutant A/M7"alieies, Isogenic strains MONCA104 [Anmtv.hisG/NMT) and MONCA107 (Anmtv.hisG/ nmtG447D::hisG-URA3~hisG) were grown for 3d at the indicated temperature on YPD containing cenjlenin {CER. final concentration 25f^M). or YPD + CER plus the indicated fatty acid (final concentration 500 fiM).

Similar results were obtained when C albicans A/MTand nmtG447D were used to complement the Ifethal phenotype of an S. cerevisiae nmtiA allele (Lodge et ai, 1994).

etai, 1991; Rocque etai, 1993; Rudnick etai, 1993b; Bhatnagarefa/.. 1994). The myristoylCoA and peptide kinetic parameters of purified E. co//-derived, wild-type C albicans Nmt and the G447D mutant did indeed provide an explanation for the observed myristic acid auxotrophy of the C. albicans nmtG447D strain at 24 and 37 C. The Kmopp) for myristoylCoA is essentially identical (0.3-0.4 ^M) for C. albicans Nmi and nmtG447D at 24 C (Table 2). In contrast. myristoylCoA Vmeu is sharply reduced leading to a greater than fivefold decrease in myristoyiCoA catalytic efficiency (Wnax^/C) for nmtG447D relative to Nmt at 24 C. Raising the temperature to 37 C leads to a small decrease in the myristoylCoA K^ of the wild-type C. albicans Nmt. and to a prodigious increase in myristoylCoA catalytic efficiency due to the almost 10-fold increase in V^ax (Table 2). In contrast, the myristoylCoA K^japp) of the mutant enzyme increases sevenfold when the temperature is raised from 24 to 37 C while the V^ax increases by a factor of only 2.5 (Table 2). While nmt181p and nmtG447D both exhibit profound defects in myristoylCoA catalytic efficiencies at 37 C. only the mutant C. albicans acyltransferase manifests marked reductions in myristoylCoA V^ax^Km at 24 C (see Table 2 and Duronio etai, 1991).

Steady-state levels of Nmt and nmtG447D The steady-state levels of wild-type and mutant Nmt were compared in isogenic C. albicans strains grown in YPD/ MYR at 24 and 37 C. Western blots revealed that both MONCA107 and its parent MONCA104 contain equivalent steady-state levels of the enzyme at 24 or at 37 C. For both strains, these levels are slightly higher at 37 C than at 24 C (Fig. 5). These observations indicate that introduction of a Gly-447-> Asp mutation in C. albicans Nmt does not produce a thermoiabile enzyme and suggest that the specific activity of the mutant acyltransferase is lower than the wild-type enzyme at both temperatures.

V

V

24 37 24 37'C

— ^

—52 kDa

Comparative kinetic studies of C. albicans Nmt &nd nmtG447D To clarify the basis of the myristic acid auxotrophy seen at 24 and 37 C, in vitro kinetic studies were performed. Nmt has an ordered Bi Bi reaction mechanism: myristoylCoA binds to apo-Nmt forming a high-affinity myristoylCoAiNmt complex. Binding of myristoylCoA is required for formation of a functional peptide-binding site. Once a ternary myristoylCoA:Nmt:peptide complex has formed, myristate is transferred from CoA to the peptide; CoA is then released, followed by myristoytpepfide (Rudnick 1995 Bladtweli Science Ltd. Molecular Microbiology, 16, 241-250

Fig. 5. Western blot analysis of the levels of wild-type Nmt and nmtG447D in isogenic MONCA104 and MONCA107 strains grown on YPD media supplemented with myristate. Twenty micrograms ot protein were reduced, denatured, and fractionated by SDS-PAGE and transferred to nylon membranes. The resulting blots were probed with monospecific, rabbit anti-C. albicans Nmt sera. Two species of C albicans Nmt and nmtG447D are seen, probably reflecting proteolytic processing at near its amino-terminus (cf Wiegand etai. 1992).

246

R. A. Weinberg et al.

Table 2. Kinetic studies of C. albicans Nmt and nmtG447D*.

MyristoylCoA

Peptide Temp Enzyme

Nmt nmtG447D Nmt nmtG447D

24 24 37 37

&r

{% of Nmt)

Sr

100 18 141 10

0.38 ±0.09 0.54 ±0.23 0.41 ±0.02 2.8*0.2

^TWKtapp)

(pmolmin

42±6 60±4 210±17 96 ±20

mg )

1.1 ±0.4x10^

(pmol min" ^ mg" ^)

(% of Nmt)

7.3±2.3x 10^

100 13 533 4

a. Peptide (GNAASARR-NH2} kinetic parameters were determined using a myristoylGoA concentration of 1 [iM. MyristoyiCoA kinetic parameters were determined using GNAASARR-NH3 at approximately its Kr,,(apparent)- This ensured that the fractional saturation of the peptide site was approximately equal in Ihe determination oi all myristoylCoA kinetic parameters. The GNAASARR-NH2 concentrations used were: Nmt. peptide = 40MM at 24 C and 200 |iM at 37 C; nmtG447D. peptide = 60MM at 24 C and 100 nM at 37 C.

The three-dimensional structure of C albicans Nmt has not been solved, although we have identified conditions for crystallizing the apo-enzyme purified from E. co//(Shieh ef ai, 1995). Given our current knowledge, it is not possible to predict what conformational perturbations are produced as a consequence of the mutation of the Gly residue, whioh. as noted above, is absolutely conserved in all six Nmts whose primary structures are known (Lodge et ai. 1994; this includes the 450-residue Nmt produced by C. elegans — E. Anderson, J- Lodge and J- I- Gordon, manuscript in preparation).

washed twice in H2O to remove residual MYR. and then introduoed into YPD at a density of -10'' cells ml \ The culture was inoubated at 37 C and aliquots were removed at various time points to test for colony-forming units (c.f.u.) both in the presence and absenoe of myristate (Table 3). MONCA105 continued to grow in the absence of MYR. However, MONCA107 exhibited an eightfold reduction in viability 6h after withdrawal from MYRcontaining media and a >100-fold reduction in c.f.u, ml'"^ by 24 h. No viable cells were detectable after 48 h (Table 3).

Withdrawal ofC. albicans nmtAytimtG447D strains from myristate-containing media produces cell death

The myristic acid auxotrophy of MONCA107 was quite stable. These oells were grown at 28 C in YPD/MYR to stationary phase and then plated on YPD to see if they would revert to myristic aoid prototrophy. The reversion rate was found to be 2.5 x 10 ®.

An experiment performed using the isogenic C. albicans nmtA/NMT and nmtA/nmtG447D strains established that Nmt is essential for vegetative growth. Cells were grown to early to mid-log phase at 28 C in YPD/MYR, pelleted.

Table 3. Effects of withdrawing myristate from the media on a nmtAl nmt447D strain of C. albicans. MONCA^05 iAnml::hisG/NMT::hisG-URA3-hisG) Time after inoculation c.f.u,ml ' (h) on YPD

0 1 2 6

19600 30700 44000 609000

c,l,u.mr^on YPD + MYR 19600 30700 ND^ ND

MONCA\07 (Anml::hisG/nmtG447D::hisG-URA3~hisG) Time after inoculation c.f.u.ml"'' c.f.u.ml"' on (h) on YPD YPD + MYR

0 1 4 5 6 24 48 a. Not determined.

0

0 0 0 0 0 0

16700 16200 6730 4070 2110 90 0

The nmtA/timtG447D strain ofC. albicans does not produce the rapid lethality seen with the NMT/nmtA strain in an immunocompromised mouse model of systemic fungal Infection Cyclophosphamide-treated mice have been used as a model of systemic C. albicans infection in immunocompromised hosts (Kirsch and Whitney. 1991). We used these animals to determine whether reductions in Nmt activity would reduce the lethality produced by systemic candidiasis. Differently size inocula of the wild-type or mutant strain were injected intravenously into groups of adult Swiss Webster mice (10^-10^c.f.u./animal; n=10 animals strain ^ dose ^). The wild-type strain (MONCA105) produced a dose-dependent lethality: a single intravenous infusion of lO^o.f.u. resulted in the death of all members of the group within 7d; 10''c.f.u. produced 50% lethality by 13d while all mice that received lO^c.f.u. survived for at least 21 d (Fig. 6). In contrast, injection of up to lO^c.f.u. of the nmtAlnmtG447D strain produced no deaths even after 21 d (Fig. 6). Cultures of kidney homogenates prepared at this 21 d timepoint revealed no viable C. albicans. © 1995 Blackwell Science Ltd, Molecular Microbiology, 16, 241-250

Mutagenesis of the NMT locus in Candida albicans

MONCA105 10^-CFU * UONCAI05 10*-CFU • MONCA107 10*-CFU X PBS Control —1—I—1—m—I—P—I—I—I—I—I—r—

7

9

11 13 15 17 19 Days

21

Fig. 6. Sun/ival of immunosuppressed adult Swiss Webster mice infected with MONCA105 or MONCA107. 10^-10^c.f.u. of either strain were given as a single dose in the tail vein of adult mice (n = 10 animals group '' dose ^) and the number of sun/iving mice was monitored over a 21 d period- Mote lbai nerther 1O*^ c.i.u. ol MONCA107 {nmtA/nmlG447D), 10^c.f.u. of MONCA105 {NMT/ nmtA]. nor the PBS vehicle alone, produced any deaths during the course of ^ e experiment.

Prospectus The results of genetic and biochemical studies now lend proof to the concept that Nmt is an attractive target for development of drugs which are active against C. albicans. The genetic analyses described above indicate that Nmt is required for the viabilty of this organism, Biochemioal analyses have indicated that selective inhibitors oould be developed by exploiting differences in the peptide substrate specificities of human and fungal Nmls. Biological experiments have confirmed that Nmt inhibition wiil allow an immunocompromised host to rid itself of a systemic C albicans infection. The isogenic NMT/nmt and nmt/Gnmt447D C. albicans strains represent a starting point for developing assays to quantify the levels of acylation of specific cellular N-myristoylproteins in vivo at various times after withdrawal of myristate. This will need to be done in both exponentially growing and stationary-phase cells. The results would allow a prediction to be made about the extent to which a drug would have to inhibit Nmt in vivo to produce lethality. Experimental procedures Strains and media The strains of C. albicans used in this study are listed in Table 1 and were routinely maintained in YPD medium (Sherman ef ai, 1986). Defined media contained glucose (2% w/v), yeast nitrogen base (0.17% w/v) and complete supplement mix powder minus uracil (Biol 01), Media were (D 1995 Blackwell Science Ltd. Molecular Microbiology. 16, 241-250

247

supplemented with uridine (25Mgml \ Difco) as required. Sorbitoi was added to the appropriate media (final concentration IM) to regenerate sphaeroplasts. Solid media was made by adding agar (Difco; 2% w/v). Ura auxotrophs were selected on medium containing 5FOA (Sigma) (Boeke etai, 1984), Seiection medium was prepared as described by Ausubel etai (1989) except that uracil was replaced with uridine (25i4gml"^), Prior to 5-FOA selection, cells were grown in YPD medium supplemented with uridine at 30 C for 48 h. Cells were then plated on 5-FOAcontaining media and subsequently restreaked on defined media lacking uridine to confirm their Ura phenotype. The fatty acid requirements of various strains were determined by supplementing media with myristate (MYR, obtained from NuChek Prep; final concentration 500 (iM) or palmitate (500MM). Note that all media supplemented with these fatty acids also contained 0,5% Brij58. Cerulenin {CER, Sigma) was dissolved in ethanol and added to media to a final concentration of Construction of DNAs used for transformation of C. albicans Flanking regions of C albicans NMTwere needed for replacement of portions of this gene with a selectable marker by homologous recombination. The sequences were obtained by the polymerasa chain reaction (PCR) using C. albicans genomic DNA from strain CAI4 (Fonzi and Inwin, 1993) as a template. The PCR reactions were run using Amplitaq polymerase (Perkin Elmer) and the following cycling conditions: denaturation at 94 C for 1 min; annealing at 50 C for 2 min; and extension at 72 C for 3 min for a total of 30 cycles. The primers 5-TGGTGGTACCCGCA11 I I I IGTTG-3' and 5'ATGCAATGTCGACCAATTGATTG-3 were used to isolate an 880 bp fragment that started 150 bp upstream of the initiator ATG of C. a/d;cans W/W7"(Fig. 2A). The primers introduced a Kpn\ site at the 5' end and a Sal\ site at the 3' end of the fragment. Two other oligonucleotides. 5'-TATAATATAAAGGATCCCATGTnTG-3' and 5-GAACAAACAGCTCTCTAGAACTGTT-3\ were used to isolate a 1250bp fragment that began 42 bp downstream of the NhAT TAA termination codon (Fig, 2A). These primers introduced a BamHI site at the 5' end and an Xba\ site at fhe 3' end of the fragment. Both fragments were sequentially cloned into pBluescript SK+ (Stratagene). The hisG-URA3-hisG cassette from pMB7 (Fonzi and Imin, 1993) was placed at the BamHl and Sa/I sites, between the two flanking regions of C. albicans NMT {Fig. 2B), yielding pMON25111, This construct contained a 660 bp deletion of the 1360bp coding region of NMT (Fig. 2B). The entire insert of pMON25111 was removed as a Kpn\~Not\ fragment and used for transformation of C. albicans. The method used to introduce a Gly-447 — Asp mutation into the C. aibicans NMT gene (yielding nmtG447D) is described in Lodge et ai (1994). The recombinant pBtuescript SK piasmid containing nmtG447D was digested with Cla\-Kpn\ to release a fragment that began 718 bp upstream and ended 1280 bp downstream of the terminator TAA of NMT (Fig. 2). This DNA fragment was subcloned into pBluescript. The cloned fragment contained a Bcl\ site 148 bp downstream of the mutant Asp-447 codon. A hisG-URA3-hisG

248

R. A. Weinberg ef al.

selection cassette from p5921 (Fonzi and Irwin, 1993) was subsequently introduced into the Bcl\ site as a BamH\-Bgl\\ fragment (Fig. 2D) yielding pMON25112, The entire insert in pMON25112 was excised with Cla\-Kpn\ and the fragment used to transform C. albicans to uracil prototrophy (see below). Note that there are only 70 bp and 100bp upstream and downstream, respectively, of the hisG cassette present in the pMON25112 nmfG'^47D fragment which are available for recombination with the disrupted nmtAv.hisG ailele (Figs 2C and 2D). This feature shoutd favour recombination with the intact NMT allele when the DNA shown in Fig, 2D is used for transformation of an NMT/nmtA heterozygote.

Transformation ofC. albicans C. albicans strains were transformed using the yeast sphaeroplast transformation kit (BioiOI) and the manufacturer's protocol. C. albicans CAI-4 (Table 1) was used as the starting strain for generation of the A/MTmutant strains, DNA was linearized as described above prior to transformation.

Genomic DNA isolation and analysis C. albicans genomic DNA was isolated according to the method described by Ausubel et al. (1989) for S. cerevisiae. Genomic DNA was analysed using Southern blot hybridizations (Ausubei ef a/., 1989). All DNA probes were labelled with digoxigenin using the Genius" kit (Boehringer Mannheim). The NMT prQhe consisted of the 880 bp Scal-Afe/( fragment from the gene's coding region. The URA3 probe consisted of a 1350bp BamH\-Sal\ fragment from pMON25113 and encompassed the entire open reading frame of the C albicans gene. The Salmonella typhimurium hIsG probe was the 950 bp BamH\-Nco\ fragment recovered from pNKY51 (Alani et ai. 1987), The stringencies selected for hybridization and washing of Southern blots were identical to those recommended for aqueous solutions by the manufacturer of the Genius" kit. Blots were developed with LumiPhos (Boehringer Mannheim) and exposed to XAR film (Eastman Kodak). When we wanted to sequentially probe a blot with two or more probes, the first probe was removed from the filter by incubation for 1 h at 75 C in a solution containing 60% formamide, 50 mM Tristhydroxymethyl]-aminomethane hydrochloride (Tris-HCI), pH 8.0, and 1% SDS (w/v). The membrane was then rinsed thoroughly in H2O prior to reprobing.

DNA sequencing A 571 bp fragment containing the G447D mutation was recovered from genomic DNA using PCR, the cycling conditions described above, and the following oligonucleotides: 5-TAGAACATTCTGCAGCCCTATGGATGG-3' which lies within NMT and 5-CATGATGCAGGATCCCGCGCCAACTG-3' which lies at the 5' end of hisG. Three independent amplifications were performed with DNA prepared from each of three transformants, MONCA105, MONCA106 and MONCAl 07 (Table 1), Each of the three independently amplified DNA segments from each of the three strains was sequenced.

Western blot studies C. albicans strains were grown overnight at either 28 C or 37 C in 40 ml YPD supplemented with 500 nM MYR, Cells were pelleted at 2500 x g. washed twice with H2O, and then resuspended in 2 ml of lOmM Tris 1 mM EDTA (pH7,5). The cells were lysed using a Mini-Bead beater (Biospec) and acid-washed 500|.im glass beads (Sigma). The lysates were spun in an Eppendorf microfage for 2 min at room temperature. The protein concentration of the resulting supernatant fraction was determined using the BCA kit (Pierce) according to the manufacturer's instructions. Samples (20j.ig total protein) were reduced, denatured, and fractionated by eiectrophoresis through 12% polyacrylamide gels containing SDS (0.1%; Laemmli, 1970). The fractionated proteins were then transferred to Immobilon (Millipore) by electroblotting using a buffer consisting of lOmM 3-(cyclohexylamino)-1propanesulphonic acid (CAPS, Sigma). p H I I , 10% MeOH. and 0.005% SDS. The blot was incubated with an affinity purified anti-C. albicans Nmt polyclonal antibody (496-2) generated in rabbits using enzyme purifed from E. coli (Wiegand et al., 1992). The rabbit antibody was used al a dilution of 1:5000. An anti-rabbit Ig horseradish peroxidase-linked F(ab')2 fragment from donkey (Amersham NA9340) was used as the secondary antibody at a dilution of 1:10000. Antigen-antibody complexes were visualized using an enhanced chemiluminesence (ECL) kit and a protocol developed by the manufacturer (Amersham).

Enzyrrw purification Wild-type C. albicans Nmt and its mutant derivative, nmtG447Dj were expressed in E. co//strain JM101 using vectors described in Wiegand et ai (1992) and Lodge et al. (1994). The theoretical mass of nmfG447D Is 51 876 amu as compared to 51 818 amu for wild-type C. albicans Nmt. Electrospray mass spectrometry of the purified mutant enzyme yielded a value of 51 870 amu (data not shown).

Kinetic studies An in vitro Nmt assay system described in Rocque et ai (1993) was used to determine the apparent K^ and V^ax values for myristoylCoA and the substrate peptide GNAASARR-NH2. Peptide kinetic parameters were determined for Nmt and nmtG447D at 24 and 37 C using purified pH]-myristoylCoA at a concentration of 1 ^M (specific activity = 9.09 Cimmol"'). GNAASARR-NH2 concentrations were varied between 5 and i60|.iM at 24 C and between 20 and 320 ^M at 37 C for both enzymes. MyristoylCoA kinetic parameters were determined by using GNAASARR-NH2 at its apparent Krr, (see iegend to Table 2) in order to achieve approximately equal fractional saturation of the peptide site independent of assay conditions of temperature or enzyme. The total concentration of myristoylCoA was varied by adding aliquots of a stock of unlabelled myristoylCoA, whose concentration was determined by absorbance at 260 nM (r.= 1.54 x 1O''M ' cm '), to purified [^H]-myristoylCoA (specific activity = 55.1 Ci mmol"''). The total concentration of myristoylCoA was varied between 0.165 and 1.6 MM for both Nmt and nmtG447D at 24 C, between 0.165 and 1,5 nM for Nmt at 37 C and between 1995 Blackwell Science Ltd. Molecular Microbiology, 16. 241-250

Mutagenesis of the NMT locus in Candida albicans 249 0.165 and 2.64 ^lM for nmtG447D at 37 C. Kinetic parameters were determined using non-linear regression analysis of the initial velocities with the program k.cat (version 1.31, Biometallics). All assays were performed in triplicate. Assays were repeated on at least two separate occasions.

Studies in an immunocompromised mouse modet of systemic C. albicans infection This mouse model has been described by Kirsch and Whitney, (1991). Strains MONCA105 and MONCAl 07 were grown to early stationary phase in YPD and YPD/MYR, respectively. Cells were pelleted and resuspended in phosphate-buffered saline (PBS). An aliquot of the mutant strain (MONCAl 07) was plated directly onto YPD and YPD/MYR to establish that the culture contained no myristic acid prototrophs. Aliquofs (100 fil) containing 10^, 10'' or 10^ c.f.u. of MONCAl 05 or MONCAl 07 were injected into the tail vein of adult (20-25 g) Swiss Webster mice (n = 10 animals for each dose of each strain; control group = PBS vehicle atone). Mice were treated withcyclophosphamide(100mgkg " \ intraperitoneally) 4d and 1 d prior to infection and every 3-4 d thereafter. At 21 d, any surviving animals were sacrificed, their kidneys removed and homogenized in PBS. Aliquots of the suspension were plated on YPD or YPD/MYR to define the number of viable organisms per kidney. All animals were treated in accordance with guidelines established by the American Association for the Accreditation of Laboratory Animal Care. Acknowledgements We thank Jennifer Lodge, Roger Wiegand. and Nandini Kishore for many helpful discussions concerning this work, Stan Dotson for his assistance with the animal studies, William Fonzi for supplying C. albicans strain CAI-4 and plasmids. B. Magee for sharing information about the chromosomal location of C. albicans NMT. and Kevin Duffin for mass spectral analysis of purified E co//-derived C. albicans nmtG447D. This work was supported in part by a grant from the National institutes of Health (AI30188), References Atani, E., Cao, L., and Kleckner, N. (1987) A method for gene disruption that allows repeated use of URA3se\Bcl\on in the construction of multiply disrupted yeast strains. Genetics 116: 541-545. Ausubel, F.M., Brent R., Kingston, R.E., Moore, D.D., Seidman, J.E., Smith, J.A., and Struhl, K. (1989) Current Protocols in Molecular Biology. New York: Wiley. Awaya, J., Ohno, T.. Ohno, H., and Omura, S. (1975) Substitution of cellular fatty acids in yeast cells by the antibiotic cerulenin and exogenous fatty acids. Biochim Biophys Acta 409: 267-273. Bhatnagar, R.S., Jackson-Machelski, E., McWherter, C.A., and Gordon, J.I. (1994) Isothermal titration calorimetric studies of S. cerevisiae myristoylCoA:protein N-myristoyltransferase: determinants of binding energy and catalytic discrimination among acylCoA and peptide ligands. J Biol Chem 269: 11045-11053. 1995 Blackwell Science Ltd, Molecular Microbiology. 16, 241-250

Boeke, J.D., LaCroute, F., and Fink, G.R. (1984) A positive selection for mutants lacking orotidine-5'-phosphafe decarboxylase activity in yeast: 5-fiuoro-orotic acid resistance. Mol Gar) Genet 197; 345-346. Denich, K.T., Malloy, P.J., and Feldman, D. (1992) Cloning and characterization of the gene encoding the ADPribosylation factor in Candida albicans. Gene 110:123-128. Dohlman, H.G., Goldsmith, P.. Spiegel, A.M., and Thorner. J. (1993) Pheromone action regulates G-protein alphasubunit myristoylation in the yeast Saccharomyces cerevisiae. Proc Natl Acad Sci USA 90: 9688-9692 Dupont, J., Graybill, J.R.. Armstrong, D,, Laroche, R.. Touze, J.E., and Wheat. L J . (1992) Fungal infections in AIDS patients. J Med Vet Mycol 30: 19-28. Duronio, R.J.. Towler, D.A,, Heuckeroth, R.O., and Gordon, J.I. (1989) Disruption of the yeast N-myristoyltransferase gene causes recessive lethality, Science 243: 796-800. Duronic. R.J., Rudnick. D.A.. Johnson. R.J.. Johnson. D.R.. and Gordon, J.I. (1991) Myristic acid auxotrophy caused by mutation of S. cerevisiae myristoyl-CoA:protein N-myristoyltransferase. J Cell Biot 113: 1313-1330. Duronio, R.J., Reed, S.I., and Gordon, J.I. (1992) Mutations of human myristoyl-CoA:protein N-myristoyltransferase cause temperature-sensitive myristic acid auxotrophy in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 89: 4129-4133. Fonzi. W.A., and Irwin, M.Y. (1993) Isogenic strain construction and gene mapping in Candida albicans. Genetics 134: 717-728. Funabashi, H., Kawaguchi. A., Tomoda, H., Omura, S., Okuda, S., and Iwasaki, S, (1989) Binding site of cerulenin in fatty acid synthetase. J Biochem 105: 751-755. Gorman. J.A., Chan, W., and Gorman, J.W. (1991) Repeated use of GAL1 for gene disnjption in Candida albicans. Genetics 129: 19-24. Johnson, DR., Cok, SJ., Feldmann, H., and Gordon, J.I. (1994a) Suppressors of a conditional lethal allele of the S. cerevisiae myristoylCoA: protein N-myristoyttransferase gene (nmti-161) reveal proteins involved in regulating protein N-myristoylation. Proc Natl Acad Set USA 91: 10158-10162. Johnson, DR., Knoll, L.J.. Levin, D.E., and Gordon, J.I. (1994b) Saccharomyces cerevisiae contains four fatty acid activation (FAA) genes: an assessment of their role in regulating cellutar lipid metabolism and protein N-myristoylation. J Cell Biol 127: 751-762, Kahn. R.A., Clark, J., Rulka. C , Stearns, T.. Zhang, C-J.. Randazzo, P.A., Terui, T., and Cavenagh, M. (1995) Mutational analysis of Saccharomyces cerevisiae ARF1. J Biol Chem 270: 143-150. Kelly, R.S., Miller, S.M., Kurtz, M.B., and Kirsch, D.R. (1987) Directed mutagenesis in Candida albicans: one-step gene disruption to isolate ura3 mutants. Mol Celt Biol 7:199-207. Kirsch, D.R.. and Whitney, R.R. (1991) Pathogenicity of C. albicans auxotrophic mutants in experimental infections. Infect Immun 59: 3297-3300. Kishore, N.S., Wood, D.C.. Mehta, P.P., Wade, A.C., Lu, T., Gokel, G.W., and Gordon, J.I. (1993) A comparison of the acyi chain specificities of human myristoylCoA synthetase and human myristoyolCoA:protein N-myristoyltransferase. J Biol Chem 268: 4889-4902.

250

R. A. Weinberg et al.

Kurtz. M.B., Cortelyou. M.W., and Kirsch, D.R. (1986) Integrative transformation of Candida albicans, using a cloned Candida ADE2 gene. Mol Cell Biol 6: 142-149. Kurtz, MB., and Marrinan, J. (1989) Isolation of Hem3 mutants from Candida albicans by sequential gene disruption. Mol Gen Genet 2^7: 47-52. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685. Langner, C.A., Lodge, J.K., Travis. S.J., Caldwell, J.E., Lu, T., Li, O.. Bryant, M.L., Devadas, B., Gokel, G.W.. Kobayashi, G.S., and Gordon. J,l. (1992) 4-Oxatetradecanoic acid is fungicidal tor C. neoformans and inhibits repiication of human immunodeficiency virus I. J Biol Chem 267:1715917169. Lodge, J.K., Johnson, R.L., Weinberg. R.A.. and Gordon, J.I. (1994) Comparison of myristoylCoA:protein N-myristoyltransferases from three pathogenic fungi—Cryptococcus neoformans. Histoplasma capsulatum, and Candida albicans. J Biol Chem 269: 2996-3009. Rocque, W.J., McWherter, C.A., Wood, D.C.. and Gordon, J.t. (1993) A comparative analysis of the kinetic mechanism and peptide substrate specificity of human and S. cerevisiae myristoylCoA:protein N-myrisoyltransferase. J Biol Chem 268: 9964-9971. Rudnick, D.A., McWherter, C.A.. Rocque, W.J., Lennon, P.J., Getman. D.P., and Gordon, J.l, (1991) Kinetic and structural evidence for a sequential ordered bi bi mechanism of catalysis by S. cerevisiae myristoyi-CoA: protein Wmyristoyltransferase. J Biol Chem 266: 9732-9739. Rudnick, D.A., Johnson. R.L, and Gordon, J.I. (1992a) Studies of the catalytic activities and substrate specificities of S. cerevisiae myristoyl-coenzymeA:protein W-myristoyltransferase deletion mutants and human/yeast Nmt chimeras in Escherichia coli and S. cerevisiae. J Biot Chem 267:23852-23861. Rudnick, D.A., Lu, T., Jackson-Machelski, E.. Hernandez, J.C, Li, Q,, Gokel, G.W.. and Gordon, J.I. (1992b) Analogs of palmitoylCoA that are substrates for myrisoyl-

CoA:protein A/-myristoyltransferase. Proc Natl Acad Sci USA 89: 10507-10511. Rudnick, D.A., McWherter, C.A., Gokel, G.W.. and Gordon, J.I. (1993a) MyristoylCoA:protein W-myristoyltransferase. Adv Enzym 67: 375-430. Rudnick, D.A., Rocque, W.J., McWherter. C.A., Toth. M.V., Jackson-Machelski, E., and Gordon, J.I. (1993b) Use of photoactivatable peptide substrates of S. cerevisiae myristoy!CoA:protein N-myristoyltransferase (Nmtip) to characterize a myristoylCoA:Nmt1p:peptide ternary complex and to provide evidence for an ordered reaction mechanism. Proc Natl Acad Sci USA 90: 1087-^1091. Sadhu, C, Hoekstra, D., McEachern, M.J., Reed, S.I.. and Hicks, J.B. (1992) A G-protein a subunit from asexual Candida albicans functions in the mating signal transduction pathway of Saccharomyces cerevisiae and is regulated by fhe 7i-Tf2 repressor. Mol Cell Bion2: 1977-1985. Schweizer, E., and Boiling, H. (1970) A Saccharomyces cerevisiae mutant defective in saturated fatty acid biosynthesis. Proc Natl Acad Sci USA 67: 660-666. Sherman, F., Fink. G.R., and Hicks. J.B. (1986) Laboratory Course Manual for Methods in Yeast Genetics. Cold Spring Harbor. New York: Cold Spring Harbor Laboratory Press. Shieh. H.-S,, Stallings, W.C, Stevens. A.M., and Stegeman, R.A. (1995) Using sampling techniques in protein crystallization. Acta Cryst Section D. in press. Singh, N., Wakil, S.J.. and Stoops, J.K. (1985) Yeast fatty acid synthase: structure to function relationship. Biochemistry 24: 6598-6602. Stone. D.E., Cole, G.M., Lopes, M.B., Goebl. M., and Reed, S.I. (1991) W-myristoylation is required for function of the pheromone-responsive G^ protein of yeast: conditional activation of the pheromone response by a temperaturesensitive /V-myristoyttransferase. Genes Dev5:1969-1961. Wiegand, R.C, Carr. C, Minnerly, J.C, Pauley, A.M.. Carron, C.P., Langner, C.A., Duronio, R.J., and Gordon. J.). (1992) The Candida atbicans myristoylCoA:profein /V-myristoyltransferase gene: Isolation and expression in S. cerevisiae and E. coli. J Biol Chem 267: 8591 -8598.

1995 Blackwell Science Lid, Molecular Microbiology. 16. 241-250