1 survival in the presence of antifungals

28 Pages·2007·0.94 MB·English

by  Vera Meyer

JBC Papers in Press. Published on September 5, 2007 as Manuscript M705856200
The latest version is at http://www.jbc.org/cgi/doi/10.1074/jbc.M705856200
1
SURVIVAL IN THE PRESENCE OF ANTIFUNGALS:
GENOME-WIDE EXPRESSION PROFILING OF ASPERGILLUS NIGER IN
RESPONSE TO SUB-LETHAL CONCENTRATIONS OF CASPOFUNGIN AND
FENPROPIMORPH
Vera Meyer1, 2, Robbert A. Damveld2†, Mark Arentshorst2, Ulf Stahl1, Cees A.M.J.J. van
den Hondel2, Arthur F.J. Ram2
(1)Berlin University of Technology, Institute of Biotechnology, Department Microbiology and
Genetics, Gustav-Meyer-Allee 25, D-13355 Berlin, Germany
(2)Leiden University, Institute of Biology, Molecular Microbiology
Wassenaarseweg 64, 2333 AL Leiden, The Netherlands
(†) Current address: DSM Food Specialties, A. Fleminglaan 1, 2600 MA Delft, The Netherlands
Running title: Global responses of A. niger to antifungals
Address correspondence to: Vera Meyer, Phone: [phone], Fax : [phone], Email:
[email protected]

How yeast cells respond to cell wall stress is The fungal cell wall is a dynamic structure
relatively well understood, however, how that is essential for sustaining cell morphology
filamentous fungi cope with cell wall damage and for protection against life-threatening
is largely unexplored. Here, we report the first environmental conditions. Morphological D
o
transcriptome analysis of Aspergillus niger characteristics during developmental processes in w
n
exposed to the antifungal compounds fungi depend upon the temporal regulation and lo
a
d
caspofungin, an inhibitor of β-1,3 glucan spatial localization of cell wall components and ed
synthesis, and fenpropimorph, which inhibits thereby ordered cell wall deposition (1-3). fro
m
ergosterol synthesis. The presence of sub- Moreover, cell wall rearrangements that h
laedthapalt dtor utgh ec ostnrceesns tcroantidointsio anlsl oawnedd tAo .c noingteinr utoe garuea ranotef e thvei tsatlr uctiumraplo irntatengcrei ty otof thew ciethlls twanaldl ttp://ww
growth by the establishment of new polarity environmental stress conditions such as osmotic w.jb
axes and formation of new germ tubes. By stress or the presence of antifungal substances c.o
comparing the expression profile between compromising cell wall and/or cell membrane rg
b/
caspofungin-exposed and non-exposed A. integrity (4-6). In order to prevent cell lysis and y
g
niger germlings, we identified a total of 172 to ensure cell survival, fungi have developed ue
s
responsive genes out of 14,509 open reading mechanisms to sense cell surface stress and to t o
n
frames present on the Affymetrix microarray respond to these stresses via a remodeling of the A
p
chips. Among 165 up-regulated genes, mainly cell wall (see also recent reviews by (1,7,8)). ril 1
genes predicted to function in (i) cell wall The composition of fungal cell walls and the 0, 2
assembly and remodeling, (ii) cytoskeletal mechanisms involved in assuring cell surface 01
9
organization, (iii) signaling and (iv) oxidative integrity have most intensively been studied in
stress response were affected. Fenpropimorph the model yeast Saccharomyces cerevisiae. The
modulated expression of 43 genes out of which cell wall of S. cerevisiae consists of a moderately
41 showed enhanced expression. Here, genes branched, flexible β-1,3 glucan network to which
predicted to function in (i) membrane to its external face β-1,6 glucan chains are bound
reconstruction, (ii) lipid signaling, (iii) cell which in turn are linked to GPI-mannoproteins.
wall remodeling and (iv) oxidative stress At the inner side of the β-1,3 glucan network,
response were identified. Northern analyses of chitin chains are attached (reviewed by (9)).
selected genes were used to confirm the Upon cell wall stress, the cell wall becomes
microarray analyses. The results further show reinforced among other things by a massive
that expression of the agsA gene encoding an increase of the chitin content in the lateral wall
α-1,3-glucan synthase is up-regulated by both (9-12) and by increased incorporation of certain
compounds. Using two PagsA-GFP reporter cell wall proteins in the cell wall (12-14). At
strains of A. niger and subjecting them to 16 least three signaling pathways, the Pkc1p-Slt2p
different antifungal compounds, including signaling pathway (also named cell wall integrity
caspofungin and fenpropimorph, we could ‘CWI’ pathway), the general stress response
show that agsA is specifically activated by pathway mediated by Msn2p/Msn4p and the
compounds interfering directly or indirectly Ca2+/calcineurin pathway have been shown to be
with cell wall biosynthesis. involved in the cell wall compensatory response
Copyright 2007 by The American Society for Biochemistry and Molecular Biology, Inc.
2
of S. cerevisiae (15). Moreover, genome-wide hyperbranched (23,24). An arrest in polarized
surveys and large-scale phenotypic analyses, growth and the induction of (sub)apical branches
aiming at an integrated view of pathways has been reported for A. niger when treated with
involved in cell wall assembly and integrity of S. the antifungal protein AFP, most recently shown
cerevisiae, have further contributed to the to be an inhibitor of chitin synthase activity in A.
understanding of its cell wall biology. As niger (25). Likewise, inhibition of β-glucan
summarized by Lesage and Bussey (1), five synthesis in A. fumigatus and A. oryzae by
levels of regulation contribute to a controlled cell pneumocandins or by a mutation in β-1,3-glucan
wall assembly and thereby coordinate cell synthase gene in N. crassa, causes considerable
morphogenesis in yeast: (i) the cell wall changes in morphology, such as swollen germ
synthetic machinery, (ii) surface signaling, (iii) tubes and highly branched hyphal tips (26,27).
cell cycle regulation, (iv) cell polarization and Finally, inhibition of the cross-linking of glycan
(v) the secretory machinery coupled with protein fibers by the antifungal agent calcofluor white
recycling through endocytosis. causes an arrest of polarized growth and swelling
In contrast to yeast, information about cell of hyphal tips in A. niger (21). Remarkably,
wall biology in filamentous fungi and the inhibition of polarized growth of filamentous
mechanisms important for maintaining cell fungi has not only been described as a
surface integrity is sparse. Although there are consequence of direct cell wall perturbations but
indications that architectural principles identified also for conditions that rather indirectly affect
D
in S. cerevisiae may also be valid for filamentous cell wall biosynthesis. For example, interference o
w
fungi (9,16), remarkable differences do exist with the assembly of the cytoskeleton (28), n
lo
both in the composition of the cell wall as well cAMP-PKA signaling (29-31), calcium signaling ad
e
as the relative amounts of the components. (32), plasma membrane integrity (33) and with d fro
Whereas the presence of β-1,6 glucan in S. the secretory machinery (34) caused apparent m
cerevisiae is undisputed, its presence in morphological changes (see also (27)). However, http
filamentous fungi is controversial discussed and the underlying molecular mechanisms and the ://w
if present, only in minor amount (17). The cell interconnections of the different pathways with w
w
wall of filamentous fungi also contains polymers cell wall assembly are far from being understood. .jb
c
that are not present in the S. cerevisiae cell wall The recent sequencing and annotation of the .o
rg
such as β-1,4-glucans, α-1,3-glucans and genome of A. niger (35) and the availability of b/
y
galactomannans (18,19). Moreover, the the Affymetrix microarray technology for A. g
u
distribution of polymers, such as chitin varies niger makes it now for the first time feasible to est o
markedly between yeast and filamentous fungi study the mechanisms involved in ensuring cell n A
(c2e0ll) .w Tahlle sctoremssp einns aftiolarmy erenatocutiso nfsu ning i rehsapvoen sbee etno spuorlfaarcizee d ingtergorwittyh ainn d thiitss bcioortereclhantioolno gicwaliltyh pril 1
0
first analyzed in Aspergillus niger. It has been important filamentous fungus. To get first , 2
0
shown that the cell wall stress response of A. insights into these processes, we screened in the 19
niger involves induced expression of agsA, present study for antifungal compounds that
encoding a putative α-glucan synthase (21). In affected the morphology of A. niger.
addition, the RlmA transcription factor is, Caspofungin, known as inhibitor of β-1,3 glucan
similarly to its S. cerevisiae homologue Rlm1p, synthesis in S. cerevisiae (36), and
required for the up-regulation of cell wall stress fenpropimorph, reported as inhibitor of S.
induced genes (4). Furthermore, the cell wall cerevisiae ergosterol biosynthesis (37) were
stress response of A. niger is, like in S. selected, as application of these compounds to A.
cerevisiae, also accompanied by increased chitin niger resulted in morphological alterations. Both
deposition, suggesting that part of the compounds were applied at sub-lethal
remodeling mechanism via the CWI pathway is concentrations to A. niger and global expression
conserved among fungi (22). profiling was performed aiming at (i)
Over the past years, evidence for a close identification of cellular responses involved in
correlation between cell wall assembly and cell cell integrity and adaptation to growth-inhibitory
morphology in filamentous fungi has been conditions, (ii) identification of drug-specific
accumulating. Several studies have shown the responses and thereby first insights into their
importance of chitin synthesis in determining mode of action in A. niger, (iii) identification of
hyphal morphology. For example, A. nidulans genes whose protein products are important for
and A. oryzae strains in which several chitin the establishment and maintenance of polarized
synthase genes have been disrupted are growth and (iv) prevention of secondary drug
3
effects or non-specific responses related to cell unbranched germlings)-1.
death. The experimental set-up of the study Construction of GFP reporter strains- The
involved the use of young, unbranched germlings reporter strain containing the PagsA-H2B-GFP-
as alteration of their morphology can easily be TtrpC reporter construct (RD6.47) has been
monitored and quantified by microscopic means described previously (38). The reporter strain
and because germlings represent a more with a cytoplasmatically expressed GFP under
homogenous cell population compared to control of the agsA promoter (PagsA-GFP-
mycelial hyphae. TtrpC) was constructed as follows: Plasmid
PagsA-GFP-TtrpC was constructed by ligation of
EXPERIMENTAL PROCEDURES a 2-kb SalI-NcoI fragment, containing PagsA
from PagsA-uidA-TtrpC (21), into an SalI-NcoI
Strains, growth conditions and antifungal opened PagsA-H2B::GFP-TtrpC vector, thereby
compounds- The A. niger strains N402 (wild removing the PagsA-H2B and replacing it with
type, lab collection), RD6.47 (38) and JvD1.1. PagsA to give pJD1. For the Agrobacterium
(this study) were used. The strains were grown at mediated transformation, the PagsA-GFP-TtrpC
37°C (unless otherwise stated) in minimal construct was inserted into a binary vector. The ~
medium (39) or complete medium (CM), 3-kb HindIII fragment containing PagsA-GFP-
consisting of minimal medium (MM) TtrpC from pJD1was cloned into a HindIII
supplemented with 1% yeast extract and 0.5% opened pTAS5 vector to give pTAS5-PagsA-
D
casamino acids. Fermentation medium (FM) was GFP-TtrpC. The pTAS5 vector consists of the o
w
composed of 0.75% glucose, 0.45% NH Cl, binary vector pSDM14 (41) with the hygromycin n
4 lo
0.15% KH2PO4, 0.05% KCl, 0.05% MgSO4, expression cassette from pAN7.1 (42) inserted ade
0.1% salt solution (39) and 0.003% yeast extract. between the borders. pTAS5-PagsA-GFP-TtrpC d fro
The pH of FM was adjusted to pH 3. was transformed to A. niger strain N402 using m
Caspofungin was purchased from (Merck, the Agrobacterium strain LBA1100. Hygromycin http
Cancidas®) and fenpropimorph from Sigma resistant transformants were subjected to ://w
Aldrich, respectively. AFP was isolated and Southern analysis to confirm complete w
w
purified from A. giganteus cultures as described integration. Genomic DNA was isolated .jb
c
by (40). All other antifungal compounds were according to (43) and digested with PstI or SstII .o
rg
made available by BASF (Germany). to determine the copy-number. Strain JvD1.1, b/
y
Screening for morphological changes induced by containing multiple copies (≥ 2) integrated in the g
u
antifungal compounds- 5 x 105 conidia of strain genome (data not shown), was selected as es
t o
N402 were inoculated in Petri dishes containing reporter strain. n
A
5 ml liquid MM supplemented with 0.003% Screening of antifungal compounds in glass p
yeast extract. Prior to inoculation, two cover bottom microtiter plates- The PagsA-GFP ril 1
0
slips were placed onto the bottom of the Petri (JvD1.1) and PagsA-H2B-GFP (RD6.47) , 2
0
dishes. Spores were allowed to germinate for 5 h reporter strains were used to screen antifungal 19
at 37°C and small germ tubes became visible in compounds for their ability to induce the cell
more than 90% of the spores. Compounds were wall integrity of A. niger. Conidia (2x104) from
added at various concentrations, whereas the the reporter strains were inoculated in each well
negative control was supplemented with the of 96-well optical glass bottom microtiter plates
same volume of H O. The following range of (Nunc art) in 100 μl 2x CM and grown for 6
2
concentrations were tested: caspofungin, 1 ng/ml hours at 37°C. After spore germination, 100 μl of
– 12.5 μg/ml; AFP, 0.2 – 0.9 μg/ml; a two-fold dilution series for each antifungal
fenpropimorph, 0.5 – 60 μg/ml; myriocin, 30 – compound was added to individual wells. The
200 μg/ml; 8-Br-cAMP, 1 – 10 mΜ; caffeine, 1 effect of each compound was tested for at least 7
different concentrations. After adding the
– 10 mΜ. After further cultivation for 1 hour at
antifungal solution, the microtiter plates were
37°C, germlings that were adherent to the cover
incubated for 3 more hours at 30 °C. After
slips were analyzed by microscopy (see below).
discarding the medium by inverting the
From at least 50 germlings per sample, the
microtiter plate, germlings that were adherent to
morphology was characterized as being either
the bottom of each well were analyzed by
unbranched (germlings with a single germ tube)
microscopy (see below). As a positive control,
or branched (germlings with apical and/or
strain MA26.1, containing PgpdA-H2B-GFP-
subapical branches). A “Branching Index (BI)”
TtrpC single copy at the pyrG locus (M.
was calculated that was defined as follows: BI =
Arentshorst, unpublished strain) was used. Strain
(Σ branched germlings) × (Σ branched +
4
N402 was used as a negative control. Acquired Technologies). For normalization, default
images were analyzed for both growth and GFP settings were used (50th percentile per chip,
levels. median per gene). Genes were defined as
Bioreactor cultivation- Freshly harvested conidia differentially expressed if their expression levels
(5x109) from strain N402 were used to inoculate varied at least 1.5 fold in the caspofungin- (or
5 l FM. Cultivations were performed in a fenpropimoph-) treated samples compared to the
BioFlo3000 bioreactor (New Brunswick control and if the difference was statistically
Scientific) using an agitation speed of 250 rpm. significant (Student’s t-test, P-value cut-off of
Temperature (37°C) and pH (set to 3) were 0.05).
controlled on-line using the program NBS Northern analyses using each 5 μg RNA from
Biocommand. Aeration was performed via the the six conditions were performed as described
headspace until the dissolved oxygen tension earlier (4). RNA samples were balanced
(DOT) dropped to 40% and was then switched to according to their content of the 18S mRNA
sparger aeration. After 5 h of cultivation, (data not shown). PCR amplicons obtained by
caspofungin or fenpropimorph (dissolved in 5 ml using different primer pairs as listed in Table 1
dest. H O) or 5 ml dest. H O (negative control) were labeled by random primer labeling using
2 2
were added. After one additional hour of 32P-labelled dATP (Amersham) and used as
cultivation, 400 ml of the culture broth were probes for Northern analysis. Hybridizations
quickly harvested via filtration and mycelial were carried out according to the manufacture´s
D
samples were immediately frozen using liquid instructions (Amersham). o
w
nitrogen. In addition, samples were taken for Microscopy- Pictures of A. niger germlings were n
lo
microscopic analysis (see below) and calculation captured using an Axioplan 2 (Zeiss) equipped ad
e
of the BI-value. with a DKC-5000 digital camera (Sony). Both d fro
RNA extraction, expression profiling and light (using DIC settings) and fluorescence m
Northern analysis- Total RNA was isolated from images (using GFP settings) were obtained with http
homogenized mycelial samples using TRIzol a 40x objective. For GFP images, a fixed ://w
reagent (Invitrogen). RNA quality control, exposure of two seconds was used. Images were w
w
labeling, microarray hybridization and scanning processed using Adobe Photoshop 6.0 (Adobe .jb
c
were performed at ServiceXS (Leiden, The Systems Inc.). .o
rg
Netherlands). Briefly, RNA quality was verified b/
y
using Agilent Bioanalyzer “Lab on chip” system RESULTS g
u
(Agilent Technologies, Palo Alto, CA). est o
Processing, labeling and hybridization of cRNA Screening for morphology-affecting n
A
to A. niger Affymetrix GeneChips were compounds. In order to select suitable p
performed according to Affymetrix protocols for compounds that affect the morphology of A. ril 1
0
“Eukaryotic Target Preparation” and “Eukaryotic niger, we have screened various substances , 2
0
Target hybridization”. For washing and staining, proposed to interfere with fungal cell wall 19
the protocol “Antibody Amplification for synthesis (caspofungin, AFP), ergosterol
Eukaryotic Targets” was followed. Hybridized synthesis (fenpropimorph), sphingolipid
probe array slides were scanned with Agilent synthesis (myriocin) and PKA signaling (8-Br-
technologies G2500A Gene Array Scanner at a 3 cAMP, caffeine). These compounds were applied
µm resolution and a wavelength of 570 nm. in different concentrations to 5 x 105 A. niger
Affymetrix Microarray Suite software MAS5.0 germlings grown on cover slips in 5 ml minimal
was used to calculate signals and P-values and to medium and their effect on germ tube elongation
set the algorithm’s absolute call flag, which and branching was microscopically followed.
indicates the reliability of the data points Morphological alterations were only observed in
according to P (present), M (marginal) and A response to caspofungin, AFP and
(absent). Microarray analyses for each condition fenpropimorph (Fig. 1), whereas no significant
(control, caspofungin-treated germlings, effect was observed when germlings were treated
fenpropimorph-treated germlings) were with a series of concentrations of myriocin, 8-Br-
performed on cells obtained from two cAMP or caffeine (data not shown). In Figure 1,
independent bioreactor cultivations (biological the morphological changes provoked by
duplicate). The complete set of transcriptional caspofungin, AFP and fenpropimorph are
raw data is available as Supplemental Material depicted. Tip swelling and an increase in
(Table S1). Expression data were analyzed using (sub)apical branching were observed after
the program GeneSpring 7.3. (Agilent treatment of the germlings with a minimal
5
concentration of 5 ng/ml, 5 μg/ml and 0.9 μg/ml disturbance of existing polarity growth sites by
for caspofungin, fenpropimorph and AFP, the formation of new polarity sites.
respectively. In order to judge and quantify the Global gene expression responses to
effect on morphology, a “Branching Index (BI)” caspofungin and fenpropimorph. Affymetrix
was determined that gives the percentage of microarray chips representing 14,509 ORFs of A.
germlings displaying (sub)apical branches niger were hybridized with RNA samples from
(n>50). As given in Figure 1, the strongest effect each two biological replicates of caspofungin-,
on branching were exerted by caspofungin and fenpropimorph- and non-treated samples,
fenpropimorph, where the BI-value was about 7- respectively, as described in Materials and
fold higher compared to the negative control. Methods. Following normalization to account for
Application of higher drug concentrations did deviations in hybridization intensity, genes
not significantly increase the BI, but resulted in showing at least 1.5 fold change (P-value cut off
the presence of dead germlings (data not shown). 0.05) in expression level were considered to be
Morphological responses to caspofungin and differentially expressed.
fenpropimorph. Caspofungin and fenpropimorph A total of 172 genes were differentially
were selected for further analysis as their effect expressed upon exposure to caspofungin, 165 of
on morphology was more prominent compared which showed increased expression and seven
to AFP. The screening assay described above genes decreased expression. In comparison, a
was repeated in large scale using cultivation of total of 43 genes were found to be responsive to
D
A. niger spores in a bioreactor (working volume treatment with fenpropimorph, 41 of these were o
w
5 l). Using such an experimental design, we up- and two were down-regulated (Supplemental n
lo
wanted to assure controlled and equal growth Material, Table S2). The modulated genes were ad
e
conditions between treated and non-treated functionally classified according to FunCat (44) d fro
germlings and thereby reliable expression data. as shown in Table 3. The category with the m
An increased starting inoculum (1 x 106 largest number of known genes that are http
spores/ml) and a slightly different minimal modulated by both antifungals is the category ://w
medium (FM) were used. During bioreactor runs, involved in metabolism. w
w
the dissolved oxygen tension (DOT) was Gene expression responses to caspofungin. .jb
c
followed and used as an indication for equal Caspofungin has been shown to be a potent .o
rg
growth behavior between the different inhibitor of β-1,3-glucan synthesis in S. b/
y
experiments (data not shown). After 5 h of total cerevisiae (36). As β-1,3-glucan is the central g
u
e
cultivation, caspofungin or fenpropimorph were cell wall polysaccharide to which other cell wall st o
added and the cultivations were continued for components of S. cerevisiae such as β-1,6- n A
otankee na dfdoirt idoentaelr mhoinuart,i oanf teorf wthhei cBhI -svaamlupel easn dw eforer glilnukceadn , (c1h)i,t int hea ndi nhmiabnitnioonp rotoefi nsβ -a1r,e3 -gclruocsasn- pril 10
transcriptomic analysis. Using this experimental synthesis by caspofungin causes cell wall , 20
1
set-up, we observed that a 10-fold increased disorganization and cell lysis in S. cerevisiae. 9
concentration of both caspofungin and One of the compensatory responses in yeast
fenpropimorph were necessary to significantly described to the presence of caspofungin is the
affect the morphology of A. niger germlings induction of the CWI pathway (Agarwal et al.
when compared to the screening experiment 2003; Lesage et al 2004), including induced
described above (Table 2 and data not shown). expression of cell wall protein encoding genes
On the one hand this can be explained by the and cell wall remodeling enzymes. In agreement
higher spore titer used for bioreactor inoculation with this, we observed an up-regulation of
and on the other hand by different cultivation several A. niger genes involved in cell wall
conditions used in both experiments. In addition assembly and remodeling (Table 4). Genes that
to inducing the formation of (sub)apical were up-regulated included genes coding for
branches, both caspofungin and fenpropimorph proteins involved in UDP-glucose synthesis
were observed to induce the establishment of (An02g07650/PgmB and An12g00820/UgpA),
new polarity axes that started from the spore and UDP-N-acetyl glucosamine synthesis
thus resulted in the formation of new germ tubes. (An18g06820/GfaA, An03g05940/GfaB and
Especially the amount of spores displaying three An12g07840/GnaA), chitin formation
or four germ tubes, which are usually rarely
(An09g04010/class III chitin synthase ChsB), α-
observed in A. niger, was significantly increased
glucan synthesis (An09g03100/AgtA), β-1,3-
by both antifungals (Fig. 2). This observation
glucan remodeling (An03g05290/BgtB,
indicated that the germlings may counteract the
An10g00400/GelA), cross-linkage of chitin to β-
6
1,6-glucan (An07g07530/CrhB, diacylglycerol pyrophosphate phosphatase Dpp1
An07g01160/ChrC), GPI anchor processing and An02g13220 predicted as lysophospholipase
(An14g03520/DfgC), protein mannosylation LplB. Moreover, a gene coding for a
(An03g01090/HocA, An18g06500/Sec53, geranylgeranyl transferase type II (An13g01040)
An16g04330/DpmA) and genes encoding involved in prenylation of proteins and therby in
putative cell wall proteins (An12g10200 and the membrane targeting and interaction of the
An14g01820). Genes involved in signaling modified proteins (48) showed increased
cascades ensuring cell wall integrity were also expression. Remarkably, a large number of
up-regulated such as An16g04200/RhoB (similar signaling proteins such as GTPases of the
to Rho2 GTPase of Schizosaccharomyces subfamilies Rho, Rac, Rab and Rap require this
pombe, regulator of α-1,3-glucan synthesis), modification for their cycling between
An10g00490/Rho-GAP and An18g03740/MkkA intracellular membrane compartments and hence
(similar to S. cerevisiae MAP kinase kinase 2 their activity (48,49). Increased expression of
involved in CWI signaling). An13g01040 might thus probably reflect a
The presence of sub-lethal concentrations of higher demand for relocalisation / recycling of
caspofungin allowed A. niger to adapt to its GTP binding proteins in response to caspofungin
inhibitory effect in such a way that the fungus treatment.
survived and continued growth by the Gene expression responses to
establishment of new polarity axes and formation fenpropimorph. Fenpropimorph belongs to the
D
of new germ tubes (Fig. 3). We thus expected, morpholine fungicides and interferes with o
w
beside identifying genes involved in cell wall ergosterol biosynthesis in S. cerevisiae by n
lo
maintenance, also to identify genes coding for inhibiting sterol C-14 reductase (ERG24 gene) ad
e
proteins having a function in cell growth and sterol C-8 isomerase (ERG2 gene) (37). We d fro
regulation and cell polarity. Indeed, three genes thus assumed that among the A. niger genes up- m
involved in growth control were up-regulated: regulated by fenpropimorph, a considerable http
An17g02350 showing strong similarity to the number of genes should be connected to lipid ://w
human Ras-related GTPase Rheb known to be a metabolism and especially to ergosterol w
w
key component of the TOR signaling pathway biosynthesis. Indeed, five genes encoding .jb
c
(45), An13g00100 (similar to the S. cerevisiae proteins with homology to the S. cerevisiae .o
rg
cell-cycle checkpoint protein kinase Dun1) and ergosterol pathway were responsive to b/
y
An14g00010 corresponding to the secretion fenpropimorph treatment (Table 5): An08g05400 g
u
related Rab-GTPase SrgA shown to be important (homologue of Erg10p which catalyses the first est o
for efficient secretion and maintenance of and rate-limiting step in ergosterol biosynthesis), n
A
polarity in A. niger (46). An03g06410 (Erg25p homologue; P-value 0.06), p
Furthermore, genes for which a function in An01g07000 (Erg24p homologue), An01g03350 ril 1
0
cytoskeleton organization and maintenance has (Erg2p homologue) and An15g00150 (Erg3p , 2
0
been established for eukaryotic organisms were homologue; P-value 0.06), indicating that 19
up-regulated such as: i) An08g06410 and fenpropimorph also targets ergosterol synthesis
An18g06590 showing homology to actin-binding in A. niger.
proteins important for the integrity of cortical Ergosterol is most abundant in the plasma
actin patches and actin-dependent endocytosis in membrane and secretory vesicles of S. cerevisiae
S. cerevisiae and S. pombe, ii) An01g03770 (50) and its metabolism has been shown to be
displaying homology to microtubule-based closely linked with the biosynthesis of sphingo-
motor proteins, iii) An16g03000 and and phospholipids (51-53). Further support for
An18g03900 with high homology to subunits of the link between ergosterol biosynthesis and
the prefoldin complex involved in the folding of other lipids and fatty acids biosynthesis was
tubulin and actin, iv) An05g00810 with provided by the observation of increased
homology to tubulin-specific chaperones and v) transcript levels of three genes predicted to
An01g13120 predicted to be a ADP ribosylation function in sphingolipid synthesis: An01g10030
factor-like 2 of the Ras superfamily of GTPases displaying homology to the S. cerevisiae
which has been shown to be important for sphinganine hydroxylase Sur2p as well as
tubulin stability and dynamics in human cells An01g14200 and An01g14190 coding for
(47). sphingolipid alpha-hydroxylase Scs7p.
Caspofungin also induced expression of Furthermore, a gene encoding an inositol-1-P
genes predicted to function in lipid metabolism synthase (An10g00530; homologous to S.
and signaling: An02g01180, coding for cerevisiae Ino1p) showed increased expression.
7
Perturbation of ergosterol biosynthesis in A. a crucial assembly factor of the spectrin-actin
niger was also accompanied by alteration of the membrane skeleton in higher eukaryotes (58).
cellular fatty acid metabolism. Six genes However, as ergosterol and sphingolipid
predicted to function in peroxisomal fatty acid β- metabolism have been shown to be important for
oxidation were up-regulated: An11g00400 and protein secretion and for the establishment of cell
An17g01150 both coding for an acyl-CoA polarity in yeast and filamentous fungi
dehydrogenase mediating the first committed (33,59,60) might suggest that some lipid genes
step of β-oxidation of fatty acids, two fatty acid mentioned above could also be involved in
dehydrogenases (An18g05210, An14g00990), polarity control of A. niger.
one fatty acid isomerase (An15g01280) and a Validation of transcriptome data by
sterol carrier protein 2 (An04g00740) which Northern analysis. In order to confirm the
functions in vitro as a chaperone for acyl-CoA changes in gene expression detected by the
oxidase (54) and was recently shown to bind to expression profiling, Northern analyses were
fatty acids (55). The main product of fatty acid performed using the same RNA samples as used
β-oxidation is acetyl-CoA that might serve as in the microarray experiments. For cells treated
precursor for de novo synthesis of ergosterol with caspofungin, four genes predicted to
(substrate for Erg10p) and fatty acids (substrate function in cell wall biosynthesis and integrity
for the fatty acid synthase multiprotein complex) were selected (An09g04010/chsC,
which in turn provides fatty acids for An03g05940/gfaA), An12g10200/hypothetical
D
sphingolipid and phospholipids biosynthesis. In cell wall protein and An18g03740/mkkA). In the o
w
agreement, three genes coding for proteins case of the fenpropimorph-treated samples, four nlo
involved in fatty acid synthase reactions showed genes coding for proteins putatively involved in ad
e
increased expression (An08g07520, lipid biosynthesis were selected d fro
An16g04520, An08g07520). Interestingly, a (An01g03350/ERG2 homologue), m
gene coding for hexose transporter (first and An01g07000/ERG24 homologue, http
rate-limiting step of glycolysis; An15g03940) An03g06410/ERG25 homologue and ://w
and citrate synthase (important for the shuttle of An01g14200/SCS7 homologue). As shown in ww
acetyl-CoA from mitochondria to the cytosol; Figure 3, the results of the Northern .jb
c
An15g01920) were also up-regulated, probably hybridizations are in good agreement with the .org
hinting at the possibility that the cytosolic microarray data. Genes that showed high/low b/
y
demand for acetyl-CoA necessary for fatty acid levels of induction in the expression profiling gu
biosynthesis is accommodated by using also showed signals of strong/moderate induction est o
glycolysis as alternative source for acetyl-CoA. in the Northern experiment (e.g. An03g06410 n A
Agepnaerst firnovmo lgveende si nc oennnseucrtiendg toc elilpl idw amlle tainbtoelgisrimty, and aAgns1A8 ge0x3p7r4es0s)i.o n is specifically induced by pril 10
were also up-regulated upon fenpropimorph compounds affecting cell wall integrity. The , 20
treatment: An01g1010 (CrhD) displaying strong expression profiling in this study revealed that 19
similarity to the S. cerevisiae cell wall protein the gene coding for the regulator of α-1,3-glucan
Crh1, most recently shown to function in the synthesis (Rho2-GTPase, An16g04200) and agtA
cross-linkage of chitin chains to β-1,6-glucan (GPI-anchored alpha-glucanotransferase,
(56) and An03g05000 predicted as ZIP family An09g03100) were up-regulated upon
zinc transporter with homology to the S. caspofungin treatment. We have previously
cerevisiae Yke4p. This protein is important for shown that the agsA gene coding for α-1,3
balancing zinc levels between the cytosol and the glucan synthase (An04g09890) is strongly
secretory pathway in yeast (57). induced in response to compounds that interfere
In response to fenpropimorph, new polar with cell wall or cell membrane integrity of A.
growth sites were developed by A. niger niger such as calcofluor white, SDS, caspofungin
germlings although to a lesser extent when and AFP (21,25), suggesting that α-1,3 glucan
compared with the caspofungin-treated samples synthesis might be generally involved in
(Fig. 3). This might be reflected by the lower securing cell surface integrity. In this study, the
number of up-regulated genes that putatively agsA gene was unexpectedly not found among
play a role in cell polarity of A. niger: the significantly up-regulated genes. A closer
An04g02340 (low homology to kinesin light look at the transcriptomic data revealed,
chain), An14g02370 (apyrase, required for Golgi however, that agsA was not expressed in the
N- and O-glycosylation in S. cerevisiae) and control experiment but strongly expressed when
An13g02780 exhibiting similarity to α-adducin, A. niger germlings were exposed to both
8
caspofungin and fenpropimorph (P-value > 0.05, leads to an up-regulation of the reporter,
see Table S2), implying that regulation of agsA suggesting that an efficient secretory pathway is
gene expression is actually under the control of required for proper cell wall biosynthesis.
stress conditions that affect the integrity of the Interfering with protein secretion might lead to
plasma membrane and/or the cell wall. cell wall weakening and subsequently to the
To further support this conclusion, we used activation of the cell wall salvage pathway.
two A. niger reporter strains, containing either a The second group of compounds comprises
cytoplasmatically (strain JvD1.1) or nuclear of chitosan and epoxiconazole. These
(strain RD6.47) targeted gfp gene under the compounds inhibited growth of A. niger and
control of the agsA promoter. Both strains were gave a moderate induction of GFP expression.
exposed to 16 antifungal compounds (including These compounds have been reported to have an
caspofungin and fenpropimorph) that target effect when the integrity of the cell wall was
different cellular processes and their effect on impaired, suggesting an indirect effect of these
growth and agsA expression was monitored by compounds on cell wall biosynthesis (65,66).
light and fluorescence microscopy (Table 6). The third group of compounds consisting of
Based on their effects, we have divided the hydrogen peroxide (Fig. 4D), pyraclostrobin,
compounds into four groups. The first group of benomyl and cycloheximide showed an
compounds includes calcofluor white, inhibition of growth at one ore more
caspofungin, tunicamycin, spiroxamine, concentrations used but showed no induction of
D
fenpropimorph, terbinafine, fludioxonil and GFP expression. To our knowledge, none of o
w
cyprodinil. These compounds inhibited growth these compounds were reported so far to affect n
lo
and provoked high expression of the GFP directly or indirectly cell wall biosynthesis. Into ad
e
reporter. In response to calcofluor white, fungal the last group, compounds were assorted that did d fro
growth became inhibited and aberrant hyphal not have any effect on either growth or GFP m
morphology such as tip swelling as well as a expression such as myriocin and nikkomycin. http
clear induction of GFP expression was observed Although nikkomycin was shown to interfere ://w
(Fig. 4A). Similarly, the induction of agsA in with chitin synthesis in S. cerevisiae (67), Li and w
w
response to the presence of caspofungin was also Rinaldi (68) showed that A. niger was not .jb
c
observed in both reporter strains, confirming sensitive to nikkomycin concentrations (MIC > .o
rg
previous results (21) and the results of the 64 µg/ml). These results are consistent with our b/
y
expression profiling in this study (Fig. 4B). finding that nikkomycin had no effect on either g
u
When the reporter strains were stressed with growth or GFP expression. es
t o
tunicamycin (inhibitor of protein N- Taken together, the data shows that the agsA n
A
glycosylation), swollen hyphae and high GFP promoter is specifically activated by compounds p
expression were visible (Fig. 4C). As N- interfering directly with cell wall biosynthesis or ril 1
0
glycosylation mutants in S. cerevisiae have been by compounds inhibiting plasma membrane , 2
0
shown to have defects in cell wall integrity function or the protein secretion machinery, 19
(15,61), it is very likely, that the addition of thereby disturbing cell wall biosynthesis more
tunicamycin to A. niger also results in weakening indirectly.
of the cell wall and activation of the cell wall
integrity pathway. The induction of the GFP DISCUSSION
reporter by the lipid synthesis disturbing
compounds spiroxamine, fenpropimorph and Growth and development of fungi as well as
terbinafine suggests that disturbance of the their ability to withstand internal turgor pressure
plasma membrane integrity negatively affects the and to survive environmental stress conditions
integrity of the cell wall and substantiates the depends on maintaining the integrity of their cell
expression data with respect to fenpropimorph. surface. As a first step towards a comprehensive
Activation of agsA::gfp expression by understanding of the regulatory network(s) of A.
fludioxonil (activator of the Hog1 osmotic signal niger involved in maintenance of cell surface
transduction pathway in S. cerevisiae) could hint integrity, we examined in this study the global
at the existence of a cross-talk between the cell gene expression profile of A. niger in response to
wall integrity pathway and the osmotic signal treatments with caspofungin and fenpropimorph.
transduction pathway in A. niger as shown Responsive genes to caspofungin. The
recently for S. cerevisiae (62,63). Interestingly, category with the highest number of genes
cyprodinil (interferes with methionine synthesis showing enhanced transcription in response to
and secretion of hydrolytic enzymes, (64)) also caspofungin is the group of genes required for
9
cell wall biogenesis and maintenance. About 12 site is strongly dependent on cortical actin patch
% of the up-regulated genes can be classified movement which oneself requires polymerization
into this category, suggesting that the primary (or on the Arp2/3 complex ((75) and references
an important) response to caspofungin is to therein). In this study, two components of the
counteract the inhibitory effect of caspofungin on Arp2/3 complex (An08g06410, An18g06590)
β-1,3-glucan synthesis by transcriptional showed increased expression upon caspofungin
activation of cell wall reinforcing genes. treatment. As also other proteins predicted to
Caspofungin inhibits β-1,3-glucan synthesis in S. function in actin and tubulin folding/stability
cerevisiae and several Aspergilli species (7,36) were up-regulated (An16g03000, An18g03900,
and has been shown to (mainly induce) up- An05g00810, An01g13120) suggests that
regulation of genes involved in the synthesis of inhibition of β-1,3-glucan synthesis may also
cell wall components and cell wall strengthening affect actin stability in A. niger which is
in the yeasts S. cerevisiae and Candida albicans counteracted by the induction of genes encoding
(69,70), implying that caspofungin triggers a for proteins that assist in actin stabilization and
similar response in A. niger as in yeast to cytoskeleton maintenance. The induced
reinforce the strength of the cell wall. expression of this class of genes has not been
One of the signal transduction pathways that observed in studies in yeasts (S. cerevisiae and
becomes activated in S. cerevisiae in response to C. albicans, (69,70)) and might therefore be
caspofungin is the CWI pathway (69). In brief, related to the filamentous growth of A. niger.
D
the CWI pathway of S. cerevisiae consists of the One pathway that has been described to be o
w
plasma membrane localized sensor proteins involved in actin polarization in S. cerevisiae is nlo
(Wsc1-4p and Mid2p) that mediate the cell wall the TOR signaling pathway. Basically, TOR ad
e
d
stress signal through the Rho1-GTPase and the signaling is conserved from yeast to humans and fro
Pkc1p kinase. Pkc1p initiates a phosphorylation consists of two signaling branches in S. m
h
cascade involving the MAP kinases Bck1p, cerevisiae (Tor1p-branch and Tor2p-branch) that ttp
Mkk2p and Slt2p. Slt2p finally phosphorylates couple nutrient signals to growth-related ://w
the transcription factor Rlm1p that induces processes such as protein synthesis, uptake of ww
expression of genes involved in cell wall amino acids, actin organization and endocytosis .jb
c
reinforcement (71). All components of the yeast (45). In A. niger, as in other filamentous fungi .org
CWI pathway are present in the genome of A. and higher eukaryotes, only a single Tor protein b/
y
niger (35), suggesting that this pathway is not is present (Meyer and Ram, unpublished data). g
u
e
only important for ensuring cell integrity in yeast The upstream activator of TOR signaling is the st o
but also in filamentous fungi. The results of this GTPase Rheb and one of the downstream n A
sptautdhyw apyro vbiedceo minedsi caactitoivnast ethda ti nt hAe . CnWigIe ri natengdr itiys eCfWfecI tpoarsth wofa yT o(4r25p,6 0is). tAhse iRn hthoi1s- GstTudPya,s ew eo fh atvhee pril 1
0
required for adaptation to caspofungin-mediated observed an up-regulation of a Rheb homologue , 2
0
1
inhibition of cell wall biogenesis (Fig. 5). First, (An17g02350), may suggest that Tor activates 9
An02g01180 displaying homology to the Rho-Pkc-MAP kinase cascade in A. niger
diacylglycerol pyrophosphate phosphatase is up- and thereby actin polarization. An additional hint
regulated. This enzyme generates diacylglycerol for involvement of the TOR pathway comes
which has been shown to be a physiological from the observation, that two putative targets of
activator of fungal Pkc1p homologues (72,73). TOR signaling (An09g03660, An04g09420
Second, a homologue to the S. pombe Rho2p coding for amino acid permeases; Table S2)
(An16g04200/RhoB) showed increased showed increased expression, as also observed
expression. In S. pombe, the Rho2-GTPase has for S. cerevisiae when subjected to caspofungin
been shown to stimulate α-1,3-glucan synthesis (69). Surprisingly, a homolog of the Rho1-
through activation of the Pkc1p homologue GTPase activator protein Sac7p showed also
Pck2p (74). Third, enhanced expression of increased expression (An10g00490). This GAP
An18g03740/MkkA (homologous to Mkk2p) is has been shown to be important for turning off
further indicative for an involvement of the CWI Rho1p activity (76), which contradicts the
pathway. And finally, targets of the A. niger conclusion that the A. niger Rho1p homologue
RlmA transcription factor such as gfaA (chitin becomes activated by Tor. However, in plants, it
synthesis) and agsA (α-1,3-glucan synthesis; (4)) has most recently been shown that the activity of
showed enhanced expression. GAPs is necessary to spatially restrict the action
of Rho-type GTPases to the tip of pollen tubes
The motility of the S. cerevisiae β-1,3-
and thereby maintains the subapical location of
glucan synthase Fks1p towards the polar growth
10
the GTPase and hence polarity of the cell (77). restructuring of the cell membrane as a
Thus, it might be conceivable that a similar compensatory response to fenpropimorph.
down-regulation of the A. niger Rho1p Lipids are essential components of
homologue at the flanks of the apex ensures that eukaryotic membranes affecting membrane
the active form of the GTPase is only present at permeability, fluidity, the activity of membrane-
the hyphal tip. associated proteins, vesicle targeting and also
It is interesting to note that the synthesis of participate in diverse signal transduction
two important lipid second messengers, pathways (50,83). Moreover, sterols and
diacylglycerol and phosphatidylcholine, seems to sphingolipids have been observed to form
be induced in response to caspofungin (enhanced segregated plasma membrane microdomains
expression of An02g01180 and An02g13220, (‘lipid rafts’) in organisms from yeast to human
respectively). Both lipids are involved in the (84). The asymmetric distribution of lipid rafts in
synthesis of the major phosphoinositide membranes is thought to provide a platform for
PtdIns(4,5)P , shown to be important for the signaling proteins such as GPI-anchored proteins
2
activity and localization of different GTPases and transporters (60) and contribute to
involved in actin assembly and membrane polarization events in different yeast such as S.
trafficking (78,79). Moreover, PtdIns(4,5)P cerevisiae and C. albicans (85,86). In A.
2
mediates the plasma membrane localization of nidulans, it has been reported that inhibition of
Rom2p that activates the Rho1/Rho2 GTPases, sphingolipid biosynthesis results in defects of
D
which in turn positively affects Pkc1p activity in actin polarization and thereby abolishes cell o
w
S. cerevisiae (71,80). Fadri et al. (81) have polarity (33,59). The parallel up-regulation of n
lo
shown that the two PtdIns(4,5)P2 interacting ergosterol and sphingolipid biosynthesis in A. ade
proteins Slm1p and Slm2p are essential for actin niger could thus point towards a reestablishment d fro
polarization and interact with the Tor2p of membrane polarization within the adaptation m
signaling complex, suggesting a link between process to fenpropimorph. http
PtdIns(4,5)P2 signaling, the TOR pathway and In this context, it is interesting to stress the ://w
organization of the actin cytoskeleton, which increased expression of An01g10030 (homolog w
w
might also be the case for A. niger. of the S. cerevisiae sphinganine hydroxylase .jb
c
Responsive genes to fenpropimorph. The Sur2p). The Sur2p product phytosphingosine is .o
rg
effect of the morpholine fenpropimorph on A. thought to stimulate Pkc1p phosphorylation and b/
y
niger or other filamentous fungi has never been thereby activation of the CWI pathway in S. g
u
studied so far. In yeast, the primary target of cerevisiae (60). A further hint for the es
t o
morpholines is inhibition of the ergostorol involvement of the CWI pathway in the adaptive n
A
biosynthetic enzymes Erg24p and Erg2p (37). response of A. niger to fenpropimorph comes p
Microarray analyses have shown that the from the observation that a ZIP family zinc ril 1
0
adaptation of S. cerevisiae and C. albicans to transporter (An03g05000 homologous to Yke4p) , 2
0
inhibition of ergosterol synthesis involves up- showed enhanced expression. Yke4p was shown 19
regulation of ERG24, ERG2 and other ERG to be strongly up-regulated during cell wall stress
genes such as ERG3 and ERG25 (51,82). via Pkc1p activity (6,69). Moreover, expression
Homologues to all of these genes showed of the known RlmA target agsA was also
enhanced expression in A. niger when treated modulated in response to fenpropimorph
with fenpropimorph, suggesting that treatment, suggesting that cell membrane
fenpropimorph also targets the ergosterol rearrangements are accompanied by remodeling
pathway in filamentous fungi and that the of the cell wall via the CWI pathway.
response mechanism to ergosterol biosynthesis Remarkably, An10g00530 (homologous to S.
inhibition is similar (Fig. 5). Perturbation of cerevisiae inositol-1-P synthase Ino1p) showed
ergosterol biogenesis in A. niger also affected the also an increased expression. Ino1p catalyses the
expression of genes belonging to related lipid conversion of glucose-6-phosphate to myo-
pathways (sphingolipid, phospholipid and fatty inositol phosphate, which is the first committed
acid metabolism). Overall, the group of up- step in the production of all inositol-containing
regulated genes involved in sterol, lipid and fatty compounds, including inositol phosphates and
acid metabolism represents the largest category phosphoinositides (87,88). Inositol phosphates
of genes responding to fenpropimorph (44%), and phosphoinositides are lipid second
suggesting a strong (inter)connection of the messengers necessary for diverse cellular
different metabolic pathways as also observed in functions and signaling processes in eukaryotes
yeast (51-53) and may further point towards a such as transcriptional regulation, mRNA