Monilinia fructicola Response to White Light

Light represents a powerful signal for the regulation of virulence in many microbial pathogens. Monilinia fructicola is the most virulent species causing brown rot in stone fruit crops. To understand the influence of light on M. fructicola, we measured the effect of white light and photoperiods on the colonial growth and sporulation of the model M. fructicola strain 38C on solid cultures. Searches in the M. fructicola 38C genome predicted a complete set of genes coding for photoreceptors possibly involved in the perception of all ranges of wavelengths. Since white light had an obvious negative effect on vegetative growth and the asexual development of M. fructicola 38C on potato dextrose agar, we studied how light influences photoresponse genes in M. fructicola during early peach infection and in liquid culture. The transcriptomes were analyzed in “Red Jim” nectarines infected by M. fructicola 38C and subjected to light pulses for 5 min and 14 h after 24 h of incubation in darkness. Specific light-induced genes were identified. Among these, we confirmed in samples from infected fruit or synthetic media that blue light photoreceptor vvd1 was among the highest expressed genes. An unknown gene, far1, coding for a small protein conserved in many families of Ascomycota phylum, was also highly induced by light. In contrast, a range of well-known photoreceptors displayed a low transcriptional response to light in M. fructicola from nectarines but not on the pathogen mycelium growing in liquid culture media for 6 days.


Introduction
Brown rot is a fungal disease affecting stone and pome fruit crops worldwide, caused mainly by three species of the genus Monilinia spp.[1], Monilinia fructicola (G.Winter), Monilinia fructigena (Honey) and Monilinia laxa (Aderhold and Ruhland).However, M. fructicola has a higher growth and aggressiveness than M. laxa and M. fructigena, which may be the cause of the displacement of both species on stone fruit orchards [2].In our laboratory strain, 38C has been largely used as a model to understand peach infection by M. fructicola [3].
M. fructicola, belonging to the Sclerotiniaceae family, has a necrotrophic lifestyle as it colonizes plant cell tissue and causes cell death in order to obtain nutrients [1].The M. fructicola infection process can be divided into three phases: pre-penetration, penetration and colonization.Infection begins when M. fructicola conidia germinate on the fruit surface to produce germ tubes and/or appressoria, which then penetrate through the fruit surface depending on the prevailing environmental conditions [3] and degree of maturity of the fruit [4].Temperature affects the production of germ tubes or appressoria by M. fructicola after conidial germination [4].M. fructicola produces the same number of germ tubes or appressoria at 4 • C but produces more germ tubes than appressoria at temperatures higher than 10 • C. Negligible or no germination of M. fructicola conidia occurs at 60% relative humidity (RH) [4].However, the effect of light on these early pre-penetration stages of M. fructicola is unknown.
Light has been identified as another important environmental factor in the development of plant diseases [5,6], affecting both the host response and the pathogen virulence.Light drives adaptive responses in fungi, which actively sense their environment in order to activate protective mechanisms against light-associated stress and to regulate many facets of their development, like carotenogenesis and conidiation [7], secondary metabolism [8] and the balance between sexual and asexual development [9].Behind this regulation of cell development, there is a sophisticated light signaling machinery which is composed of different apoproteins in association with chromophores [10].These complexes, named photoreceptors, react to light of a determined wavelength through absorption of a photon that promotes structural changes in the chromophore, which, in turn, induce conformational changes in the apoprotein [11].There are several families of photoreceptors depending on the chromophore they harbor and the wavelength they are sensitive to: near-UV/bluelight-sensing cryptochromes (CRYs, 350-500 nm), blue-light-sensing LOVs (light, oxygen, voltage) domain-containing proteins (450 nm), green light-sensing opsins (Ops, 540 nm), and phytochromes which detect red/far red ratio (650-780 nm) [11].
Recent genomic studies in M. laxa suggest that the photoreceptor's arsenal [12] is similar to those described in the well-known pathogen Botrytis cinerea [13], which also belongs to the Sclerotiniaceae family.A unique ortholog for each photoreceptor with a highly conserved domain architecture was identified in M. laxa [12].M. laxa has proven to be a light-responsive plant pathogen since different light conditions, intensities and photoperiods affect its transcriptional profile, phenotype, growth and conidiation rate [14,15].Light also affects the development of brown rot caused by M. laxa on nectarines and cherries [15].In this study, we focused on determining whether the model for M. fructicola infection in peaches, strain 38C, would be able to perceive the presence or absence of light by identifying the completeness of photoreceptors' machinery, which genes are under light regulation when the fungus is cultured on fruit in early pre-penetration stages or synthetic media, and how exposure to light influences the growth of this fungal plant pathogen.

Monilinia Fructicola Strains
Single-spore isolate from M. fructicola 38C is stored at the culture collection of the Plant Protection Department of INIA-CSIC (Madrid, Spain).From the same collection, we chose strain 1C as a control on solid cultures.The genome from M. fructicola 38C is sequenced and available at NCBI under the accession number GCA_016906325.1 (BioProject PRJNA503180) [16].Isolates were maintained as cultures on Potato Dextrose Agar (PDA Difco TM , Franklin Lakes, NJ, USA) in darkness at 4 • C for short-term storage and as a conidial suspension in 20% glycerol at −80 • C for long-term storage.For conidia production, isolates were cultured on PDA at 25 • C for 7 days in continuous darkness.

Effect of Different Lights on Mycelial Growth and Sporulation of M. fructicola on PDA
Effect of white light, generated using a set of four fluorescent bulbs of Osram36W/954 "daylight", on the growth rate and conidia production was evaluated for M. fructicola strains 1C and 38C.These bulbs produce an irradiance of 110.13 W m −2 , a luminous flux of 2850 lm, illuminance of 8700 lux and have a wavelength range between 300 and 700 nm, with a maximum around 550 nm (Supplementary Figure S1).Petri dishes filled with PDA were inoculated in the center with a 15 µL droplet of a suspension containing 10 6 conidia mL −1 and cultivated for 7 days in growth chambers at 22-25 • C under two different photoperiods: 8 h light/16 h darkness and 12 h light/12 h darkness.Distance between light bulbs and samples was set to 25 cm and control groups were set in continuous darkness.Two perpendicular diameters were measured starting on the third day of incubation until the end of the experiment, and daily growth speed (diameter in mm day −1 ) was calculated using regression analysis from the average measurements of each colony diameter.Total sporulation was calculated at the end of each assay and then referenced to each colony growth area.In order to do so, the surface of each plate was individually scraped after adding 5 mL of sterilized distilled water and filtered through glass wool to remove the mycelia.The number of conidia produced was counted using a hemocytometer.Data were expressed as total conidia divided by the colony area (cm 2 ).Five technical replicates were used for each condition and the whole experiment was repeated twice.

Effect of Different Lights on Gene Expression Levels of M. fructicola in Potato Dextrose Broth
Influence of white light generated using a set of four fluorescent bulbs of Osram 58W/840 "Cool White" was also tested.These bulbs produce an irradiance of 165.55 W m −2 , a luminous flux of 5200 lm, illuminance of 14,900 lux and have a wavelength range between 300 and 700 nm, with a maximum around 520 nm (Supplementary Figure S1).Flasks containing 100 mL of Potato Dextrose Broth (PDB, Difco TM , Franklin Lakes, NJ, USA) were inoculated with conidia from M. fructicola 38C and grown in darkness at 25 • C for 6 days to allow for mycelial growth.Lights were turned on and samples were taken at 5 min, 1 h, 2.5 h, 5 h and 14 h of uninterrupted illumination.Distance between light bulbs and samples was set to 25 cm and control groups were set in continuous darkness.Mycelium was then collected by filtering the culture through miracloth and immediately frozen in liquid nitrogen.Flasks maintained in darkness were used as control.Three biological replicates were used in each condition and the complete experiment was repeated twice.

Effect of Different Lights on Nectarine Brown Rot Disease
Nectarines (var.Red Jim) from Ebro valley (Lleida, Spain) harvested at optimal commercial maturity in the first weeks of August were used.Surface of nectarines were disinfected using the hypochlorite procedure [17].Afterwards, nectarines were prepared for infection experiments by being dried in a laminar flow cabinet and then each nectarine was inoculated with three drops of 15 µL of a suspension containing 10 6 conidia mL −1 from M. fructicola 38C.Inoculated fruit were placed on sterilized cellulose alveoli inside disinfected plastic containers lined up with soaked paper to maintain humidity.Each container was individually sealed with plastic film and incubated in the dark for 24 h at 22 • C and 100% RH to allow conidial germination and germ tube formation [4].Next, lights were turned on and fruit tissue samples with germinated conidia were taken at 5 min, 1 h, 2.5 h, 5 h and 14 h of continuous illumination.Inoculated fruit maintained in darkness were used as control.After each incubation time, samples of peel and pulp tissues of 1 cm diameter encompassing the inoculation sites were collected from three individual fruits (three biological replicates) per condition.All samples were frozen in liquid nitrogen before being kept at −80 • C until further transcriptional profiling analysis.

RNA Extraction
Total RNA of M. fructicola 38C was extracted from infected fruit tissues according to the rapid CTAB-based protocol [18] adapted to Monilinia spp.[12].Total RNA was resuspended in RNase-free Mili-Q water and frozen immediately at −80 • C until further use.
When using mycelium from PDB as source for RNA extraction, a TRIreagent based method was used (Sigma-Aldrich, St Louis, MO, USA), as previously described [19].Total RNA was resuspended in RNase-free MiliQ water and frozen immediately until further use.
In both cases, RNA concentration and purity were measured using a NanoDrop 2000 spectrophotometer and RNA integrity was checked by 1.2% agarose gel electrophoresis.

Identification of Photoreceptor Coding Genes in M. fructicola 38C Genome
The blast tool from NCBI Genomic workbench (https://www.ncbi.nlm.nih.gov/tools/gbench/ accessed on 13 December 2022) was used to identify a list of putative photoreceptor coding genes and light-response genes, described in M. laxa [12], in the available 38C genome and proteome prediction.

Transcriptomic Analysis
RNA samples from nectarines infected by M. fructicola 38C, from the same Osram 58W/840 "Cool White" as before and darkness at 5 min and 14 h post irradiance (pi), were sequenced using Illumina technology by Macrogen facilities.RawData files in .fastqformat were generated containing 150pb reads.These reads were trimmed and mapped against available Monilinia spp.genomes using CLC Genomics Workbench 22.0.3[20] software and differential expression analyses were carried out between the light and darkness in both time stamps.A transcript was considered to be differentially expressed when it had a logFC lower or higher than −2/2, respectively, and a Reads Per Kilobase Million (RPKM) value higher than 20 in at least one condition.
Gene Ontology (GO) and functional annotation analysis were performed using Omics-Box software 22.0.3[21].GO enrichment analyses were also carried out using the "Gene Set Test" tool in CLC Genomics Workbench (QIAGEN CLC Genomics Workbench 22.0.3[20].Filtering parameters to define a differentially expressed gene (DEG) were an RPKM value above 20 and a p-value < 0.05.

Gene Expression Analysis
Before cDNA synthesis, total RNA samples from 38C were treated with DNase I (Invitrogen, Carlsbad, CA, USA) following manufacturer's instructions.cDNA was synthesized from 1 µg of DNase-treated RNA using SuperScript First-Strand Synthesis System for RT-PCR and oligo(dT) primer (Invitrogen, Carlsbad, CA, USA) following manufacturer's indications.Real-Time PCR was performed with a 7500 Fast Real-Time PCR (Applied Biosystems, Foster City, CA, USA) using GoTaq qPCR Master Mix (Promega, Madison, WI, USA).Three technical replicates were analyzed for each biological replicate for both target and housekeeping genes, with a total volume of 20 µL, containing 10 µL of 2× GoTaq qPCR Master Mix, 7.8 µL of nuclease-free water, 300 nM of each primer and 100 ng of cDNA.The cycling program was 2 min at 95 • C, followed by 40 cycles of 15 s at 95 • C and 1 min at 60 • C.After the amplification reaction, a melt curve analysis was performed to check the specificity.Primers (oligonucleotides listed in Supplementary Table S1) for histone H3 gene (Ml_histoneH3) (BK012065) were used from [15] Primers for vivid1 (vvd1), far1, phytochrome 1 and 2 (phy1 and phy2), opsine1 and 2 (op1 and op2), cryptochrome 1 and 2 (cry1and cry2) and velvet 2 and 4 (vel2 and vel4) genes (gene codes are in Table 1) were designed using Vector NTI (Thermo Fisher, Waltham, MA, USA) and MEGA X [22].

Statistical Analysis
Data were analyzed by ANOVA analysis.When F-test was significant at p ≤ 0.05, the means were compared using Tukey s multiple test range.

Effect of Light on Growth Rate and Sporulation of M. fructicola
Strain 38C has previously been used to study the infection process on peaches by M. fructicola.Strain 38C was subjected to different conditions of light exposure to determine the possible effects of light on its colony phenotype.As a control, M. fructicola strain 1C was used for comparison.Strains 1C and 38C showed differences in colony development when exposed to light.Strain 38C grew faster than strain 1C in every condition tested (Figure 1).

Statistical Analysis
Data were analyzed by ANOVA analysis.When F-test was significant at p ≤ 0.05, the means were compared using Tukey´s multiple test range.

Effect of Light on Growth Rate and Sporulation of M. fructicola
Strain 38C has previously been used to study the infection process on peaches by M. fructicola.Strain 38C was subjected to different conditions of light exposure to determine the possible effects of light on its colony phenotype.As a control, M. fructicola strain 1C was used for comparison.Strains 1C and 38C showed differences in colony development when exposed to light.Strain 38C grew faster than strain 1C in every condition tested (Figure 1).Since photoperiods are important for fungal growth [12], in order to quantify the effect of white light or darkness on the growth and conidiation rates of M. fructicola, we tested two different photoperiods with 8 and 12 h of light each.The growth speed of both M. fructicola isolates was affected by the 12 h light photoperiod.The 8 h light photoperiod only had a significant effect on the growth rate of isolate 38C (Figure 2A,B).However, the effect of light on the sporulation of both isolates was different (Figure 2C,D).The sporulation of 38C was reduced by two photoperiods, with a further reduction at 12 h of light photoperiod.Meanwhile, light did not have any effect on the sporulation of 1C.These data indicated that our model strain 38C might differentially respond to light.Since photoperiods are important for fungal growth [12], in order to quantify the effect of white light or darkness on the growth and conidiation rates of M. fructicola, we tested two different photoperiods with 8 and 12 h of light each.The growth speed of both M. fructicola isolates was affected by the 12 h light photoperiod.The 8 h light photoperiod only had a significant effect on the growth rate of isolate 38C (Figure 2A,B).However, the effect of light on the sporulation of both isolates was different (Figure 2C,D).The sporulation of 38C was reduced by two photoperiods, with a further reduction at 12 h of light photoperiod.Meanwhile, light did not have any effect on the sporulation of 1C.These data indicated that our model strain 38C might differentially respond to light.

Light Sensing Machinery in M. fructicola 38C
Fourteen genes encoding putative light sensing and photoresponse related (photoreceptors) proteins (listed in Table 1) were identified in the M. fructicola 38C genome after BLASTP searches using the NCBI Genomic Workbench and available genomic data from B. cinerea strain B05.10 and M. laxa strain 8L, together with the predicted proteome of M. fructicola 38C.Deduced protein sequences showed shared and well-conserved domains with those of B. cinerea and M. laxa (Tables 1 and 2).
The predicted domain architecture of photoreceptors agreed with previous studies.Three red/far red ratio light-sensing photoreceptors known as phytochromes (mfcphy1, mfcphy2 and mfcphy3) were found.These proteins contain a light-sensing PAS-GAF-PHY module, in addition to domains commonly found in hybrid histidine kinases (HKs).Rhodopsins are green-light-driven proton pumps integrated in the cell membrane and carrying a covalently bound all-trans-retinal as the chromophore, locating two of this type of photoreceptors in M. fructicola strain 38C, mfcop1 and mfcop2.For blue-light-sensing, there are many unrelated proteins, commonly referred as LOVs, with an LOV (Light, Voltage, Oxygen) domain bounded to a flavin adenine dinucleotide (FAD/FMN).M. fructicola strain 38C had three putative photoreceptors from two distinct LOVs groups, namely L1, which are GATA-type zinc transcription factors; and L2, a group of proteins with a conserved LOV domain but lacking any effector domain.These putative photoreceptors were mfcwcl1, mfcwcl2 and mfcvvd1, respectively.The last photoreceptor group found in the predicted proteome of strain 38C were 2 cryptochromes, mfccry1 and mfccry2.These proteins were very similar to photolyases that have gained a signaling function.They sense near-UV/blue light with methenyltetrahydrofolate (MTHF) as the antenna pigment and FAD as the photocatalytical chromophore, both covalently bound to PHR-and FAD-binding domains.Another important component of the light sensing and signaling machinery is the velvet protein family, which links development and secondary metabolism with light.mfcvel1, mfcvel2, mfcvel3 and mfcvel4 were found in 38C (Figure 3).The Gene Ontology and Functional Annotation analysis confirmed the putative photoresponse function of the genes identified in the 38C predicted proteome (Table 2).

Light Sensing Machinery in M. fructicola 38C
Fourteen genes encoding putative light sensing and photoresponse related (photoreceptors) proteins (listed in Table 1) were identified in the M. fructicola 38C genome after BLASTP searches using the NCBI Genomic Workbench and available genomic data from B. cinerea strain B05.10 and M. laxa strain 8L, together with the predicted proteome of M. fructicola 38C.Deduced protein sequences showed shared and well-conserved domains with those of B. cinerea and M. laxa (Tables 1 and 2).Disease symptoms were not visible on fruit inoculated with 38C conidial suspension after the incubation and irradiation period, nor on the control group in continuous darkness.A comparative analysis was performed among RNAseq samples of infected fruit tissue by 38C after 5 min and 14 h of daylight irradiation and continuous darkness.The majority of transcripts that M. fructicola 38C expressed during the infection process, 6309 at 5 min of illumination and 4800 after 14 h, did not show any significant differences in expression levels regarding light conditions.There was another group of transcripts, 3612 at 5 min of illumination and 4927 after 14 h, which were not expressed by the fungus or had expression values under the cut off.However, there was a set of transcripts that changed their expression profile which was affected by light conditions, 165 at 5 min of illumination and 359 after 14 h (Table 3).Among the 165 genes from 38C that showed a modified expression after 5 min of 58 w "daylight" exposure, 126 of which were upregulated while the remaining 39 were downregulated.At 14 h of illumination, the number of upregulated transcripts was 188, a 49.2% increase, and the amount also grew to 171 for the downregulated gene class, which meant a 47.0% increase.Samples at 5 min and 14 h of illumination shared 34 upregulated and 9 downregulated transcripts (Supplementary Table S2).Raw reads corresponding to this RNAseq are available in PRJNA1011491.

Transcriptional Profiles in the Light and Darkness in M. fructicola
Disease symptoms were not visible on fruit inoculated with 38C conidial suspension after the incubation and irradiation period, nor on the control group in continuous darkness.A comparative analysis was performed among RNAseq samples of infected fruit tissue by 38C after 5 min and 14 h of daylight irradiation and continuous darkness.The majority of transcripts that M. fructicola 38C expressed during the infection process, 6309 at 5 min of illumination and 4800 after 14 h, did not show any significant differences in expression levels regarding light conditions.There was another group of transcripts, 3612 at 5 min of illumination and 4927 after 14 h, which were not expressed by the fungus or Using CLC Genome Workbench (Qiagen), we performed a GO enrichment analysis among the DEGs in our RNAseq results for biological process, cell wall component and molecular function categories.In total, we found 165 different GO terms for the 5 min comparison and 125 for the 14 h comparison (Figure 4).The most overrepresented GO terms at 5 min and 14 h post irradiation were peptide metabolic process (14 hpi) and metabolic process (5 min pi) for the biological process category, cytoplasmic part (14 hpi) and membrane part (5 min pi) for the cellular component category, and oxidoreductase activity (14 hpi) and catalytic activity (5 min pi) for the molecular function category.A summary of the GO terms search in DEGs is shown in Supplementary Material (Gene set test.xlsx).

Photoregulation of Photoresponse Genes in M. fructicola
Among DEGs in RNAseq, only one of the putative photoreceptor coding ge present in 38C, vvd1.The remaining predicted photoreceptor genes were unaffe light (Table 4) or showed very low expression values (measured as transcript per m TPMs).Searching for upregulated transcripts that could be related to photoresp 38C, we found the gene MFRU_021g00620.1,named from now on far1, which was cantly upregulated in light over darkness in both time stamps (Table 4).The Gene ogy and functional annotation did not assign any function nor GO terms for this g  We detected a shift in the M. fructicola 38C strain's transcriptional activity due to light exposure in the number of DEGs and which genes were transcribed.The GO terms dealing with the carbohydrate metabolic process, polysaccharide catabolic process, extracellular region and hydrolase activity of different substrates were also enriched among DEGs.This suggests that light might have a regulatory function in the production of carbohydrate active enzymes (CAZymes), some of which are related to the host cell wall degrading process acting, thus, as pathogenicity factors for M. fructicola.

Photoregulation of Photoresponse Genes in M. fructicola
Among DEGs in RNAseq, only one of the putative photoreceptor coding genes was present in 38C, vvd1.The remaining predicted photoreceptor genes were unaffected by light (Table 4) or showed very low expression values (measured as transcript per millions, TPMs).Searching for upregulated transcripts that could be related to photoresponse in 38C, we found the gene MFRU_021g00620.1,named from now on far1, which was significantly upregulated in light over darkness in both time stamps (Table 4).The Gene Ontology and functional annotation did not assign any function nor GO terms for this gene.FAR1 is a short protein (82aa) with an undefined function, lacking any known conserved domains or associated GO Terms.Putative homologous genes to far1 were found in related fungal genomes.The blastp tool in UniProt [24] returned 250 hits, of which 248 belonged to the Ascomycota phylum and all of them belonging to the Pezizomycotina subdivision, with hits in six of twelve classes within this clade.The most abundant classes were Eurotiomycetes, with 91 hits including genus Aspergillus and Penicilium spp.; Sordariomycetes, with 70 hits including genus like Colletotrichum, Fusarium, Verticilium, Neurospora, Magnaporthe and Rosellinia spp.; and Leotiomycetes, with 58 hits including genus like Botrytis, Monilinia and Sclerotinia spp.Thus, FAR1 is a widely distributed protein in the phylum Ascomycota, present in some important species of phytopathogenic fungi and well-studied model organisms (Figure S1).
When reconstructing the evolutionary relations using the amino acidic sequence of the protein, we found that all Monilinia spp.Sequences formed a clade together, with Sclerotinia spp.And Botrytis spp.As close relatives.The rest of the clades in the tree are much more uncertain, suggesting conserved functions in those orthologs belonging to the Leotiomycetes class (Figure 5).

Differential Expression Analysis
Using RT-qPCR in M. fructicola far1 and only vvd1 among the putative photoreceptors and putative light-sensing protein coding genes were differentially expressed under light in our transcriptomic analysis.We confirmed these transcriptional profiles using RT-qPCR.We used the same infected plant tissue samples by 38C that were the subjects of RNAseq analyses.Genes vvd1 and far1 showed statistically significant upregulation when exposed to 58 w daylight by RT-qPCR.vvd1 greatly increased its expression levels after a short light pulse of 5 min, decreasing with time but still being significantly different than continuous darkness, and rising up again at the 14 h post irradiation time stamp.The expression levels of far1 were upregulated upon light exposure, reaching its peak at 14 h post irradiance when 38C grew over plant tissue (Figure 6).3.6.Differential Expression Analysis Using RT-qPCR in M. fructicola far1 and only vvd1 among the putative photoreceptors and putative light-sens tein coding genes were differentially expressed under light in our transcriptomic a We confirmed these transcriptional profiles using RT-qPCR.We used the same The tree with the highest log likelihood (−2766.09) is shown.The percentage of trees in which the associated taxa clustered together is shown next to the branches.Initial tree(s) for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the JTT model, and then selecting the topology with superior log likelihood value.The tree is drawn to scale, with branch lengths measured in the number of substitutions per site.This analysis involved 46 amino acid sequences.There was a total of 221 positions in the final dataset.Evolutionary analyses were conducted in MEGA X [22].upregulated upon light exposure, reaching its peak at 14 h post irradiance when 38C grew over plant tissue (Figure 6).We also questioned whether the presence of fruit would modify the effect of light exposure on the expression levels of photoreceptors.Only 5 out of 8 putative photoreceptor genes and the unknown function gene far1 significantly increased their expression levels over the time M. fructicola 38C was exposed to light when growing in PDB (Figure 7).Blue-light-sensing vvd1 was the most upregulated gene, increasing its expression levels sixfold at 1 h post irradiation and maintaining these expression levels throughout the whole experiment.Genes encoding near-UV/blue-light-sensing proteins cry1 and cry2 both showed the same pattern, continuously increasing their expression levels but never reaching those of vvd1.phy2, encoding a red light photoreceptor, shares the same expression profile as cry1 and cry2 but with a lesser fold change overall.In contrast, green light photoreceptor op2 activates its expression at five minutes of light exposure and reaches its highest at 1 h post illumination, at 3.5 fold, and then decreases its expression levels to not be significantly different from continuous darkness.Not all of these genes were modified We also questioned whether the presence of fruit would modify the effect of light exposure on the expression levels of photoreceptors.Only 5 out of 8 putative photoreceptor genes and the unknown function gene far1 significantly increased their expression levels over the time M. fructicola 38C was exposed to light when growing in PDB (Figure 7).Blue-light-sensing vvd1 was the most upregulated gene, increasing its expression levels sixfold at 1 h post irradiation and maintaining these expression levels throughout the whole experiment.Genes encoding near-UV/blue-light-sensing proteins cry1 and cry2 both showed the same pattern, continuously increasing their expression levels but never reaching those of vvd1.phy2, encoding a red light photoreceptor, shares the same expression profile as cry1 and cry2 but with a lesser fold change overall.In contrast, green light photoreceptor op2 activates its expression at five minutes of light exposure and reaches its highest at 1 h post illumination, at 3.5 fold, and then decreases its expression levels to not be significantly different from continuous darkness.Not all of these genes were modified in a continuous darkness condition (Figure 7).The remaining photoreceptor coding genes that were tested did not show any significant changes in gene expression under these conditions. in a continuous darkness condition (Figure 7).The remaining photoreceptor coding genes that were tested did not show any significant changes in gene expression under these conditions.

Discussion
White light emitted by the fluorescent bulbs used in this work has a significant effect on the growth and sporulation of M. fructicola and on some of its photoresponse genes in pre-penetration stages.Light is an essential source of abiotic environmental information for many fungi, regulating key elements of their behavior.Aspergillus nidulans is regulated in a light-dependent manner [26], with red light playing a dominant role, while B. cinerea needs light-dark cycles for the production of conidia [27].The growth and conidiation of M. laxa on PDA were significantly affected by light and photoperiod [12].However, M. fructicola responds to light stimulus differently than M. laxa, especially because the sporulation of M. laxa was always favored with light and photoperiod [12,28], an effect not observed in M. fructicola.
The light effect on fungal growth and sporulation might be specific to fungal genus, species of the same genus, strains or different light sources [26,29].The same isolate of M. fructicola 38C showed no difference in growth and sporulation under LEDs and darkness [30], which could be explained by the different qualities of light emitted by LEDs vs. fluorescent tubes.Also, there are differences in light wave composition between different models and brands of fluorescent light bulbs.For example, our light sources had less quantity of light emitted in the 400-450 nm range than the 500-550 nm range, which could affect the expression of genes controlled by photoreceptors sensing light of either of those

Discussion
White light emitted by the fluorescent bulbs used in this work has a significant effect on the growth and sporulation of M. fructicola and on some of its photoresponse genes in pre-penetration stages.Light is an essential source of abiotic environmental information for many fungi, regulating key elements of their behavior.Aspergillus nidulans is regulated in a light-dependent manner [26], with red light playing a dominant role, while B. cinerea needs light-dark cycles for the production of conidia [27].The growth and conidiation of M. laxa on PDA were significantly affected by light and photoperiod [12].However, M. fructicola responds to light stimulus differently than M. laxa, especially because the sporulation of M. laxa was always favored with light and photoperiod [12,28], an effect not observed in M. fructicola.
The light effect on fungal growth and sporulation might be specific to fungal genus, species of the same genus, strains or different light sources [26,29].The same isolate of M. fructicola 38C showed no difference in growth and sporulation under LEDs and darkness [30], which could be explained by the different qualities of light emitted by LEDs vs. fluorescent tubes.Also, there are differences in light wave composition between different models and brands of fluorescent light bulbs.For example, our light sources had less quantity of light emitted in the 400-450 nm range than the 500-550 nm range, which could affect the expression of genes controlled by photoreceptors sensing light of either of those wavelengths.This could explain why we were unable to find photoactivation of the white collar complex (WCC) as this photoreceptor senses light with a wavelength of 450 nm.
Furthermore, growth and sporulation differences between the two M. fructicola strains were observed in the present study.While strain 1C, less virulent than 38C, did not show any statistically significant difference in conidia production between light and continuous darkness, strain 38C was unable to produce conidia under the two photoperiods tested and only produced conidia under continuous darkness.There are reported cases in the literature of "blind" strains resulting in anomalous regulation of conidia production in which the mutation resulting in such phenotype has been identified in a photoreceptor or a light-response-related protein.cinerea T4 is a wild isolate with a mutation in the Velvet protein BcVEL1.It produces a truncated protein resulting in a loss of function.This strain is considered as "blind" and shows an "always conidia" phenotype, producing conidia regardless of illumination conditions.This phenotype is also associated with a reduction in virulence on several plant hosts [31].This shows that vel1 might be implicated in the regulation mechanisms for conidia production.The "always conidia" phenotype is also found in A. nidulans carrying the veA1 mutation [26].veA is a homologous gene to vel1 from B. cinerea.The A. nidulans VeA1 mutant protein lacks the first 36 amino acids at the N-terminus and is not able to migrate from the cytoplasm to the nucleus, failing to respond to light, thus explaining the light independent conidiation phenotype displayed by those strains carrying the veA1 allele [32].In B. cinerea, a mutation in the BcveA gene and loss-of-function mutants of the photoreceptor coding gene bcphy3 were identified, all causing evident morphological defects and reduced virulence in a light-independent way [33].However, even though M. fructicola is more virulent than M. laxa, no differences were found between its photoreceptors that could explain the light effect on 38C conidia production.
Differential transcriptional regulation mediated by light of photoreceptors and lightresponse-related proteins could also explain the anomalous conidiation phenotype of strain 38C.Some transcripts expressed by M. fructicola 38C during the fruit infection process and measured by RNAseq showed significant differences between light and dark infection conditions in the present study.Among photoreceptors, only vvd1 showed significant differences in its transcripts between light and dark infection conditions in infecting plant tissue.However, these results could be affected by growth culture media and, as we discussed before, the light wave composition of light sources.Light had a significant effect on the M. fructicola 38C strain's transcriptional profile, affecting the genetic expression of key photoreceptors, namely vvd1, cry1, cry2, op2 and phy2, when M. fructicola was cultured in liquid medium.All the above photoreceptors were also found to be light inducible in M. laxa strain 8L [12].They also found that light produces differential expressions on vel1 and vel4, which are involved in conidiation regulation, but we found no light effect on 38C for those genes.However, M. fructicola 38C showed serious difficulties to sporulate on PDA after incubation for 6 days under white light.
While we did not find any evidence of photo regulation for wcl1 and wcl2 among the DEGs in our RNAseq on infected nectarine tissue, our study shows that vvd1, a blue light photoreceptor with an antagonistic regulatory function of the WCC by physical interaction [34], had an early, high and consistent photoinduction in every condition we tested.Another photoreceptor that could be involved in explaining our results was cry2, which had a role in the negative regulation of conidia formation [13] and was found to be photoinducible for M. fructicola strain 38C in PDB.Additional studies are required to determine whether vvd1 or cry2 have a role in regulating conidiation in M. fructicola and if it could explain 38C's anomalous conidiation behavior.Wcl1 and wcl2 join together to form the White Collar Complex (WCC), which is a primary regulator of light-responsive genes [7,35,36].These photoreceptors did not show any light regulation in M. laxa [12] and some unpublished results suggest the same for M. fructigena.
On the other hand, gene MFRU_021g00620.1,from now on named far1, was significantly upregulated in light against darkness on our RNAseq on infected plant tissue.And, RT-qPCR confirmed these results using total RNA from mycelial samples grown in liquid culture.far1, which encodes for a small protein and which may serve as a light response signal, is present in some Monilinia species and other related fungi.The Far1 protein does not have any known function and our GO annotation did not assign any GO terms; however, a putative homologue of Far1 in B. cinerea is already labeled as "conidiation protein" in the Uniprot data base.Far1 is highly induced by light and might also play a role in explaining the abnormal conidiation behavior of 38C.
Further investigation is required to unravel the function of far1 and its role in photoreception.Genetic transformation is a powerful tool for functional characterization of genes; however, M. fructicola strain 38C is recalcitrant to transformation and, due to the lack of successful and transformation protocols, we have not been able to perform gene knockout or overexpression experiments to confirm a possible direct function of Far1 in photoreception.Our laboratory has made some efforts in order to overcome this situation but success in obtaining mutant strains is daunting [37].Both vvd1 and far1 showed early, high and consistent light upregulation in all the conditions tested in this study and are present in all Monilinia species.Thus, we think they are perfect candidates for light response signaling genes.

Supplementary Materials:
The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/jof9100988/s1, Figure S1: Light wave composition of used light sources, Figure S2: Relative abundance of species encoding a Far1-like protein within Pezizomycotina classes; Table S1: qPCR primers used in this study, Table S2: Differentially expressed genes shared between samples from 5 min and 14 h of illumination.;Archive S1: Gene set test.xlsx.

Figure 1 .
Figure 1.Effect of white light on two Monilinia fructicola strains after seven days of incubation on Potato Dextrose Agar (PDA): (A) strain 38C exposed to a 12 h white light photoperiod; (B) strain 1C exposed to a 12 h white light photoperiod; (C) strain 38C exposed to continuous darkness; (D) strain 1C exposed to continuous darkness.

Figure 1 .
Figure 1.Effect of white light on two Monilinia fructicola strains after seven days of incubation on Potato Dextrose Agar (PDA): (A) strain 38C exposed to a 12 h white light photoperiod; (B) strain 1C exposed to a 12 h white light photoperiod; (C) strain 38C exposed to continuous darkness; (D) strain 1C exposed to continuous darkness.

Figure 2 .
Figure 2. Effect of two different white light photoperiods on mycelial growth rate and sporulation of two strains of M. fructicola, 38C and 1C, after seven days of incubation on Potato Dextrose Agar (PDA): (A) mycelial growth rate under 8 h light photoperiod; (B) mycelial growth rate under 12 h light photoperiod; (C) sporulation under 8 h light photoperiod; (D) sporulation under 12 h light photoperiod.Data are the average and standard deviation from two completed assays with five technical replicates each one.Bars with same letter in each graph were not significantly different by Tukey´s multiple test range.

Figure 2 .
Figure 2. Effect of two different white light photoperiods on mycelial growth rate and sporulation of two strains of M. fructicola, 38C and 1C, after seven days of incubation on Potato Dextrose Agar (PDA): (A) mycelial growth rate under 8 h light photoperiod; (B) mycelial growth rate under 12 h light photoperiod; (C) sporulation under 8 h light photoperiod; (D) sporulation under 12 h light photoperiod.Data are the average and standard deviation from two completed assays with five technical replicates each one.Bars with same letter in each graph were not significantly different by Tukey's multiple test range.

JFigure 4 .
Figure 4. Distribution of GO terms by category on the Differentially Expressed Genes (DE min and 14 h post irradiation.

Figure 4 .
Figure 4. Distribution of GO terms by category on the Differentially Expressed Genes (DEGs) at 5 min and 14 h post irradiation.

Figure 5 .
Figure 5. Evolutionary analysis of far1 orthologs by Maximum Likelihood method The evo history was inferred by using the Maximum Likelihood method and JTT matrix-based m The tree with the highest log likelihood (−2766.09) is shown.The percentage of trees in w associated taxa clustered together is shown next to the branches.Initial tree(s) for the search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a pairwise distances estimated using the JTT model, and then selecting the topology with sup likelihood value.The tree is drawn to scale, with branch lengths measured in the number tutions per site.This analysis involved 46 amino acid sequences.There was a total of 221 in the final dataset.Evolutionary analyses were conducted in MEGA X [22].

Figure 5 .
Figure 5. Evolutionary analysis of far1 orthologs by Maximum Likelihood method The evolutionary history was inferred by using the Maximum Likelihood method and JTT matrix-based model[25].The tree with the highest log likelihood (−2766.09) is shown.The percentage of trees in which the associated taxa clustered together is shown next to the branches.Initial tree(s) for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the JTT model, and then selecting the topology with superior log likelihood value.The tree is drawn to scale, with branch lengths measured in the number of substitutions per site.This analysis involved 46 amino acid sequences.There was a total of 221 positions in the final dataset.Evolutionary analyses were conducted in MEGA X[22].

Figure 6 .
Figure 6.Changes in relative gene expression of two putative light-sensing protein coding genes (vvd1 and far1) after daylight exposure and continuous darkness at different time points of M. fructicola strain 38C growing on fruit tissue.The scale is Log2 2 −∆∆Ct of mean fold change values, including standard deviation, from three biological replicates, with three technical replicates each, after normalization against 0 h post illumination using the 2 −∆∆Ct method [23], with histone H3 as endogenous control gene.

Figure 6 .
Figure 6.Changes in relative gene expression of two putative light-sensing protein coding genes (vvd1 and far1) after daylight exposure and continuous darkness at different time points of M. fructicola strain 38C growing on fruit tissue.The scale is Log 2 2 −∆∆Ct of mean fold change values, including standard deviation, from three biological replicates, with three technical replicates each, after normalization against 0 h post illumination using the 2 −∆∆Ct method [23], with histone H3 as endogenous control gene.

Figure 7 .
Figure 7. Changes in relative gene expression of six putative light-sensing protein coding genes after daylight exposure and continuous darkness on different time stamps when M. fructicola strain 38C grew in Potato Dextrose Broth (PDB).The scale is Log2 2 −∆∆Ct of mean fold change values from two separate experiments, including standard deviation, from three biological replicates, with three technical replicates each, after normalization against 0 h post illumination using the 2 −∆∆Ct method [23], with histone H3 as endogenous control gene.

Figure 7 .
Figure 7. Changes in relative gene expression of six putative light-sensing protein coding genes after daylight exposure and continuous darkness on different time stamps when M. fructicola strain 38C grew in Potato Dextrose Broth (PDB).The scale is Log 2 2 −∆∆Ct of mean fold change values from two separate experiments, including standard deviation, from three biological replicates, with three technical replicates each, after normalization against 0 h post illumination using the 2 −∆∆Ct method [23], with histone H3 as endogenous control gene.

Table 1 .
Photoreception related proteins presented in M. fructicola 38C and homology of their conserved domains in Botrytis cinerea and Monilinia laxa.a Protein unique identifier in 38C predicted proteome.b Putative protein function based on functional annotation.c Predicted protein length in 38C proteome d Gene unique identifier in M. laxa 8L genome; BLASTP identity and coverage.e Gene unique identifier in B. cinerea B05.10 genome; BLASTP coverage and BLASTP identity.

Table 2 .
Gene ontology (GO) and functional annotation for putative photoreception related genes identified in the M. fructicola 38C predicted proteome.
3.3.Transcriptional Profiles in the Light and Darkness in M. fructicola

Table 4 .
Expression values measured in Transcripts Per Million (TPMs) for putative photor related genes in M. fructicola 38C.

Table 4 .
Expression measured in Transcripts Per Million (TPMs) for putative photoreception related genes in M. fructicola 38C.