Epichloid endophytes are well known symbionts of many cool-season grasses that may alleviate environmental stresses for their hosts. For example, endophytes produce alkaloid compounds that may be toxic to invertebrate or vertebrate herbivores. Achnatherum robustum, commonly called sleepygrass, was aptly named due to the presence of an endophyte that causes toxic effects to livestock and wildlife. Variation in alkaloid production observed in two A. robustum populations located near Weed and Cloudcroft in the Lincoln National Forest, New Mexico, suggests two different endophyte species are present in these populations. Genetic analyses of endophyte-infected samples revealed major differences in the endophyte alkaloid genetic profiles from the two populations, which were supported with chemical analyses. The endophyte present in the Weed population was shown to produce chanoclavine I, paspaline, and terpendoles, so thus resembles the previously described Epichloë funkii. The endophyte present in the Cloudcroft population produces chanoclavineI, ergonovine, lysergic acid amide, and paspaline, and is an undescribed endophyte species. We observed very low survival rates for aphids feeding on plants infected with the Cloudcroft endophyte, while aphid survival was better on endophyte infected plants in the Weed population. This observation led to the hypothesis that the alkaloid ergonovine is responsible for aphid mortality. Direct testing of aphid survival on oat leaves supplemented with ergonovine provided supporting evidence for this hypothesis. The results of this study suggest that alkaloids produced by the Cloudcroft endophyte, specifically ergonovine, have insecticidal properties.

The endophyte presence and diversity in natural populations of Poa alsodes were evaluated along a latitudinal transect from the southern distribution range in North Carolina to New York. Two distinct Epichloë hybrid taxa were identified from 23 populations. Each taxon could easily be distinguished by polymerase chain reaction (PCR) genotyping with primers designed to mating type genes and alkaloid biosynthesis genes that encode key pathway steps for ergot alkaloids, indole-diterpenes, lolines, and peramine. The most commonly found Epichloë taxon, Poa alsodes Taxonomic Group-1 (PalTG-1), was detected in 22 populations at high infection frequencies (72-100%), with the exception of one population at high elevation (26% infection). The second taxon, PalTG-2, was observed only in five populations in Pennsylvania constituting 12% of infected samples. Phylogenetic analyses placed PalTG-1 as an interspecific hybrid of E. amarillans and E. typhina subsp. poae ancestors, and it is considered a new hybrid species, which the authors name Epichloë alsodes. PalTG-2 is an intraspecific hybrid of two E. typhina subsp. poae ancestors, similar to E. schardlii from the host Cinna arundinacea, which the authors propose as a new variety, Epichloë schardlii var. pennsylvanica. Epichloë alsodes isolates were all mating type MTA MTB and tested positive for dmaW, easC, perA, and some LOL genes, but only the alkaloid N-acetylnorloline was detected in E. alsodes-infected plant material. Epichloë schardlii var. pennsylvanica isolates were all mating type MTB MTB and tested positive for perA, but peramine was not produced. Both E. alsodes and E. schardlii var. pennsylvanica appeared to have complete perA genes, but point mutations were identified in E. alsodes that would render the encoded perA gene nonfunctional.

Ergot alkaloids are mycotoxins with an array of biological effects. With this study, we investigated for the first time the application of atmospheric pressure photoionization (APPI) as an ionization method for LC-MS analysis of ergot alkaloids, and compared its performance to that of the more established technique of electrospray ionization (ESI). Samples of the grass Achnatherum robustum infected with the ergot producing Epichloë fungus were extracted using cold methanol and subjected to reserved-phase HPLC-ESI-MS and HPLC-APPI-MS analysis. The ergot alkaloids ergonovine and lysergic acid amide were detected in these samples, and quantified via external calibration. Validation parameters were recorded in accordance with ICH guidelines. A triple quadrupole MS operated in multiple reaction monitoring yielded the lowest detection limits. The performance of APPI and ESI methods was comparable. Both methods were subject to very little matrix interference, with percent recoveries ranging from 82% to 100%. As determined with HPLC-APPI-MS quantification, lysergic acid amide and ergonovine were extracted from an A. robustum sample infected with the Epichloë fungus at concentrations of 1.143±0.051 ppm and 0.2822±0.0071 ppm, respectively. There was no statistically significant difference between these concentrations and those determined using ESI for the same samples.

Some cool-season pooid grasses partner with symbiotic fungal endophytes in the Epichloë genus for defense against insect herbivores via fungal alkaloids. Poa alsodes, North American woodland grass, independently hosts two species of Epichloë that vary by produced alkaloids. E. alsodes produces insecticidal N-acetylnorloline. E. schardlii var. pennsylvanica (E. schardlii hereafter) has the gene for peramine, an insect-deterring alkaloid, production, but peramine was not detected. We tested the effects of the two endophytes on survival, feeding preference, and plant damage by the generalist herbivore, Spodoptera frugiperda. No larvae survived when feeding on plants harboring E. alsodes. In contrast, survival was only slightly reduced by plants harboring E. schardlii. However, larvae that fed on E. schardlii infected plants experienced delayed development and reduced pupal mass. Uninfected plants and plants infected with E. schardlii were damaged severely when single larvae fed upon them, whereas larvae fed negligibly on plants infected with E. alsodes. Preference did not match performance. Larvae strongly avoided feeding on E. schardlii but not E. alsodes-infected leaves where survival was zero. When E. schardlii was experimentally removed, larval leaf choices suggested that this endophyte is responsible for deterrence. High levels of N-acetylnorloline were detected from E. alsodes infected plants. Peramine was not detected in the experimental plants harboring E. schardlii, so it remains unclear what mechanisms caused avoidance and developmental delays. The two endophytes may protect their common host in different ways: E. alsodes by larval mortality and E. schardlii by deterring feeding and negative effects on development.