qPCR-based quantification reveals high plant host-specificity of endophytic colonization levels in leaves

Abstract

Premise: Despite the high functional importance of endophytes, we still have limited understanding of the biotic and abiotic factors that influence colonization of plant hosts along major ecological gradients and lack quantitative estimates of their colonization extent. In this study, we hypothesized that the developmental stage of the ecosystem will affect the levels of bacterial and fungal endophytic assemblages in the foliar endosphere.

Methods: We quantified levels of bacterial and fungal endophytes in leaves of four plant hosts at four stages of vegetation succession using an optimized qPCR protocol with bacteria-specific 16S and fungi-targeting primers.

Results: (1) The ecosystem developmental stage did not have a significant effect on the colonization levels of bacterial or fungal endophytes. (2) Colonization levels by bacterial and fungal endophytes were governed by different mechanisms. (3) Endophytic colonization levels and their relationship to foliar tissue stoichiometry were highly host specific.

Conclusions: Quantifying colonization levels is important in the study of endophytic ecology, and the fast, relatively low-cost qPCR-based method can supply useful ecological information, which can significantly enhance the interpretation potential of descriptive data generated, for example, by next-generation sequencing.

Keywords: cell counts; ecological succession; foliar endophyte; fungi‐bacteria ratios; qPCR; soil chronosequence.

Figure 1

Bacterial quantification (no. 16S rDNA copies ng DNA^–1) using qPCR with primer set 515 F/806 R (yellow) or 335 F/769 R (gray). Nine sample compositions in two dilutions each were tested: mock community sample of both bacterial and fungal genomic DNA (MC; 0.1 and 0.01 ng DNA µL^–1), Chlorella vulgaris genomic DNA from an axenic culture (Chl; 0.1 and 0.01 ng DNA µL^–1), and mixed samples with 100:1, 10:1, 1:1, 1:10, or 1:100 of the mock community and algal genomic DNA (MC:Chl). Values were log‐transformed (base 10), shaded area represents the 95.0% confidence interval.

Figure 2

Accuracy of quantifying individual cells in a standardized mock community consisting of three yeast strains, a bacterium, and a green alga, using direct counts with fluorescence‐microscopy and qPCR with three target fungal genes (18S, ITS, and β‐actin). Samples of mock communities with different proportions of the five organisms were quantified using both approaches. Microscopy counts (brown line) were based on DAPI staining of nuclei. Primers for qPCR quantification nu_SSu_1196/nu_SSu_0817 for 18S (blue line), ITS1/ITS2 for ITS (pink line), and ACT‐512F/ACT‐783R for β‐actin (green line). Values were log‐transformed (base 10), and the shaded area represents the 95.0% confidence interval.

Figure 3

Violin plots showing the level of bacterial endophytes (as 16S rDNA gene copies ng DNA^–1) in leaves from four plant host species (Calamagrostis epigejos, Picea abies, Salix caprea, and Tussilago farfara) from the experimental locations along a chronosequence gradient, Sokolov, Czech Republic, at three sampling times in the growing season (spring, summer, and autumn). Values were log‐transformed (base 10); box plots display the median (central constriction), interquartile range (upper and lower box borders); lower/upper adjacent values (black lines), and outliers (open circles and stars).

Figure 4

Violin plots showing the level of fungal endophytes (as β‐actin gene copies ng DNA^–1) in leaves from four plant host species (Calamagrostis epigejos, Picea abies, Salix caprea, and Tussilago farfara) from the experimental locations along a chronosequence gradient, Sokolov, Czech Republic, at three sampling times in the growing season (spring, summer, autumn). Values were log‐transformed (base 10); box plots display the median (central constriction), interquartile range (upper and lower box borders), lower/upper adjacent values (black lines), and outliers (open circles and stars).

Figure 5

Violin plots showing ratio of fungi to bacteria in the leaves of four plant host species (Calamagrostis epigejos, Picea abies, Salix caprea, and Tussilago farfara) from the experimental locations along a chronosequence gradient, Sokolov, Czech Republic, at three sampling times in a growing season (spring, summer, and autumn). Values were log‐transformed (base 10), box plots display the median (central constriction), interquartile range (upper and lower box borders), lower/upper adjacent values (black lines), and outliers (open circles and stars).

Figure 6

Scatter plots of Pearson correlation between the qPCR‐based estimates of fungal and bacterial levels (β‐actin gene copies ng DNA^–1 and 16S rDNA gene copies ng DNA^–1, respectively) in the foliar endosphere of four plant hosts (Calamagrostis epigejos, Picea abies, Salix caprea, and Tussilago farfara), from Sokolov experimental locations, Czech Republic. Values were log‐transformed (base 10); plots, Pearson's r, P values, and regression equations correspond in color to the plant host.