Intraspecific functional and genetic diversity of Petriella setifera

The aim of the study was an analysis of the intraspecific genetic and functional diversity of the new isolated fungal strains of P. setifera. This is the first report concerning the genetic and metabolic diversity of Petriella setifera strains isolated from industrial compost and the first description of a protocol for AFLP fingerprinting analysis optimised for these fungal species. The results showed a significant degree of variability among the isolates, which was demonstrated by the clearly subdivision of all the isolates into two clusters with 51% and 62% similarity, respectively. For the metabolic diversity, the BIOLOG system was used and this analysis revealed clearly different patterns of carbon substrates utilization between the isolates resulting in a clear separation of the five isolates into three clusters with 0%, 42% and 54% of similarity, respectively. These results suggest that genetic diversity does not always match the level of functional diversity, which may be useful in discovering the importance of this fungus to ecosystem functioning. The results indicated that P. setifera strains were able to degrade substrates produced in the degradation of hemicellulose (D-Arabinose, L-Arabinose, D-Glucuronic Acid, Xylitol, γ-Amino-Butyric Acid, D-Mannose, D-Xylose and L-Rhamnose), cellulose (α-D-Glucose and D-Cellobiose) and the synthesis of lignin (Quinic Acid) at a high level, showing their importance in ecosystem services as a decomposer of carbon compounds and as organisms, which make a significant contribution to carbon cycling in the ecosystem.The results showed for the first time that the use of molecular biology techniques (such as AFLP and BIOLOG analyses) may allow for the identification of intraspecific diversity of as yet poorly investigated fungal species with favourable consequences for our understanding their ecosystem function.


Introduction
To identify the isolated fungal species, we have used the Large Subunit Ribosomal (LSU) 46 sequencing. According to many authors (Schoch et al., 2012;Pawlik et al., 2015a,b), to identify a 92 potential. In this paper, we have demonstrated for first time a combination of genomic and 93 functional diversity assays in P. setifera and the development of the first protocol on the AFLP 94 fingerprinting analysis applied to this species. The results showed for the first time that the use of 95 molecular biology techniques (such as AFLP and BIOLOG analyses) can allow the identification 96 of intraspecific diversity without knowing a lot of information on the analysed fungal species.  The sequencing of the D2 LSU region was performed with the use of universal primers (Table  208 of fungal species belonging or not to the same Petriella setifera family. This process has been done 209 to have a certain identification of the five fungal strains isolated from industrial compost and both 210 to compare the P. setifera strains with others published fungal genome.

211
To illustrate the BIOLOG results, the similarity of the carbon utilization patterns between the 212 strains, was presented by heatmaps graph and the percent of total carbon source utilization. For 213 the substrate richness (R) and AWDD indices were assessed, by two-way ANOVA analysis, the 214 effect of the incubation hours and the strain on them. Successively, the significant differences were 215 calculated by a post hoc analysis using the Tukey test. In function of the carbon utilization, we 216 drew a cluster analysis using a dendrogram calculated with the Ward method and Sneath's 217 dissimilarity criterion which was calculated in function of the dissimilarity of fungal groups on the 218 basis of their response to standard tests (Sneath & Sokal, 1973).

219
On the other hand, for the AFLP results, we have considered only the peaks of amplified 220 fragments are longer than 200 bp. The results obtained were shown using dendrograms calculated 221 with the Ward method and cluster analysis with Sneath's dissimilarity criterion (Sneath & Sokal, 222 1973). 223 All the statistical analyses, which are described above, were performed with the use of 224 STATISTICA 10.0 software (StatSoft, Inc., USA).  The genetic relationship between the isolates was presented by the dendrogram (Figure 4). The 251 subdivision of all isolates is in accordance with the less restrictive Sneath criterion (66%). The 252 isolates exhibited the following percentage of similarity: isolates G11/16 and G16/16 51% DNA 253 profile similarity; isolates G17/16, G14/16, and G18/16 62% DNA profile similarity. In turn, at 254 33% of Sneath`s restrictive criterion, we noted separation between all the tested isolates. 255 Moreover, through this analysis, we saw that four monomorphic peaks were present only in one 256 strains. The utilization profiles of carbon sources by these isolates revealed a broad intraspecific 261 variability ( Figure 5). Significant differences (approximately up to 6 times) were demonstrated in 262 the substrate richness (R) index and especially, we saw that the strains G16/16, G11/16 and G17/16 263 presented a significant different substrate richness between them and between the two remaining 264 strains (G14/16 and G18/16) ( Figure 6). These findings were supported by the ANOVA analysis 265 and the post hoc Tukey test. Through the ANOVA analysis, we found that the strain, the incubation 266 time and the interaction between these two factors had significant effect (p < 0.05) on the substrate 267 richness (Table 3). All the five strains used an average of 92% of the 95 available carbon substrates; 268 especially, they used more carbohydrate sources (average of 95.45% of the total 44 analysed 269 substrates). In total, each strain utilised more amino acid, carbohydrate and polymer; but for the 270 total utilization of carboxylic acid and miscellaneous, we saw a different utilization between the 271 strains ( Figure 7). 272 We found that all the P. setifera strains were extensively capable of metabolising the carbon The dendrogram showed that the strains were separated into three clusters, in accordance with 284 Sneath's dissimilarity criterion (66%) (Figure 8). The first group included isolate G18/16 with 285 metabolic profile similarity of 0%, the second one consisted of isolates G16/16 and G11/16, and 286 the third included G17/16 and G14/16 isolates with metabolic profile similarity of 42% and 54%, 287 respectively.  All the analysed strains can be regarded as Petriella setifera (Figure 1), as revealed in the 300 phylogenetic tree and especially it confirmed that the five fungal isolates were not know and there 301 is published partial genome for these strains. This analysis explained the good separation between 302 the other genera belonging to the Microascaceae family, but this approach did not show any 303 significant differences within the Petriella sp.. The lack of the intraspecific variability may be 304 related to the use of sequencing of the LSU region and not of the ITS region. In fact, to reveal the 305 separation of strains at the family level in the fungal domain, the sequencing of LSU region should 306 be carried out. Christ et al. (2011) revealed that to view the differences within a family, the best 307 attempt is to sequence the ITS region because of its high variability and resolution at the species 308 level. This was also confirmed by the phylogenetic study on the Microascaceae family performed   The analysis of the metabolic potential has revealed the presence of intraspecific variability 322 within the P. setifera strains and the differences were found in the affinity and modality to use 323 these carbon substrates. When we analysed the dendrogram of the patterns of carbon sources 324 utilization (Figure 8), we noted that the subdivision into the three clusters was a function of the 325 utilization of these substrates. Strains G16/16, G11/16 and G17/16 metabolised more substrates 326 than the others, and this was confirmed by the high substrate richness index (R index, Figure 6). 327 Another aspect that distinguishes the P. setifera strains in the functional diversity was the different 328 pattern of substrates utilization between the isolates. Figure 7 showed clearly that cluster G11/16 329 and G16/16 used the five principal carbon source groups in the same way, which was completely 330 different from cluster G14/16 and G17/16; in fact, these clusters exhibited metabolic profile 331 similarity of 42% and 54%, respectively (Figure 8). We observed that strain G18/16 utilized these 332 carbon substances in a different way than the other two groups, especially we saw this different 333 utilization for carboxylic acid and miscellaneous (Figure 7). Moreover, Figure 7 demonstrated that 334 all the strains were characterized by a different C-substrate utilization ratio, especially for 335 carboxylic acids, polymers, and miscellaneous substrates, whereas the patterns for the other three 336 groups (i.e. amines/amides, amino acids, and carbohydrates) were the same for all the strains. The 337 results of the BIOLOG FF Plates ™ analysis indicated intraspecific differences in the phenotypic 338 profiles. This means that these isolates have different metabolic abilities to degrade the analysed 339 carbon sources. These findings were confirmed by the analysis of the density of each isolate. The 340 AWDD showed that this measure for all the analysed isolates increased after the 24 incubation 341 hours and it remained higher throughout the time of incubation. At the beginning of the experiment 342 (until the 24 incubation hours) all the five strains had the same lower fungal activity and after this 343 point, we saw a bigger increase (an exponential phase) of the activity for all strains from 48 to 72 344 incubation hours. From 72 to 120 incubation hours, the analysed strains had an equal activity 345 (similar a plateau situation). After this moment, only for 144 and 192 hours of the incubation, we 346 saw that four analysed strains had a same activity and only the G16/16 strain presented an 347 increment of the activity until the end of the experiment. These modifications on the fungal activity 348 mean that in the moment when the strains come into contact with the carbon substrates, they 349 present a lower fungal activity followed by an exponential phase. In this last phase, it could be the 350 phase in which the substrates were more degraded. In the last 120 incubation hours, we saw a 351 plateau phase due to the possible limitation of the substrate amount or the excessive presence of 352 the inhibitor products. The significant different behaviour in the fungal activity between the five 353 strains, were seen at 144 and 192 incubation hours (Figure 9).

354
The Petriella setifera, which can be found in decaying wood, belongs to soft rot fungi that 355 degrade cellulose and hemicellulose. We found that all the isolates degraded at high level the 356 substances that can be produced during hemicellulose degradation (i.e. D-Arabinose, L-Arabinose,  Figure 4 that the 369 dendrogram based on cluster analysis divides the analysed strains into two groups (in accordance 370 with Sneath's dissimilarity criteria of 66%). However, at a 33% dissimilarity coefficient, the 371 analysed strains are not related to each other. This differentiation was made in function of the 372 number of detected polymorphisms. The cluster with G11/16 and G16/16 had a 52% of AFLP 373 profile similarity, since these two isolates exhibited in total an average of 24 common peaks of a 374 total 27 polymorphic peaks and five polymorphic peaks were not observed in the other strains. The 375 cluster with G14/16, G18/16, and G17/16 had a 62% of AFLP profile similarity with an average 376 of 17 common peaks of a total 19 polymorphic peaks, there was only one common peak, which 377 was not detected in the previous cluster. This means that more polymorphism peaks were detected 378 in the cluster with G11/16 and G16/16 than in the other strains. The results of the AFLP analysis 379 confirm that this new protocol has successfully differentiated the isolated P. setifera strains.

380
In general, the results of grouping obtained in the BIOLOG FF Plates™ and AFLP analyses 381 revealed differences in the graphs (Figures 4 and 8). To evaluate intraspecific variability among 382 isolates, the BIOLOG and AFLP are proper tools, as proved in our experiments, which is also  Fig 6). For the DNA 393 profile (Figure 4), strain G18/16 had profile similarity of 66% (it was clustered with strain G14/16), 394 as suggested by the detection of only 16 polymorphism peaks for this strain (16/28, 57.14%). For 395 this reason, strain G18/16 displays lower variability in the genetic and metabolic profiles. Finally, 396 isolate G17/16 had metabolic profile similarity of 54%, which was similar to strain G14/16, given 397 their similar pattern of carbon substances utilization (Figure 8). Regarding the DNA profile, 398 G17/16 exhibited similarity of 62%, which separated it from the cluster of isolates G14/16 and 399 G18/16. We found that this separation between G17/16 and the latter cluster was revealed by the 400 number of polymorphic peaks in common (14 out of a total of 19); additionally, a peak that was 401 not present in the others two strains (G14/16 and G18/16) was detected for isolate G17/16.