Determining The Genetic Basis Of Glyphosate Resistance In Wild Populations Of Erigeron Bonariensis

Herbicide resistance is the heritable ability of weeds to survive and reproduce in the presence of herbicide doses that are lethal to the wild type of the species. Erigeron bonariensis (hairy fleabane) is an agricultural weed that infests orchards and crop fields in California’s Central Valley, and has become resistant to the herbicide chemical glyphosate (RoundUp®), through an unknown genetic mechanism. One mechanism of glyphosate resistance demonstrated in E. canadensis, a close relative of E.

bonariensis, is non-target site reduced translocation of the herbicide, in which vacuolar sequestration prevents the chemical from spreading around the plant. Resistance of Erigeron canadensis to glyphosate is believed to involve upregulation of the target gene EPSPS in combination with the ABC transporter genes M10 and M11. This study aims to determining through quantitative PCR (qPCR) if these candidate genes are involved in glyphosate resistance in wild populations of Erigeron bonariensis.

 Sample leaves of the weed were collected before and after glyphosate spraying in plants from 10 different populations wild-collected from the Central Valley and two control populations of Erigeron bonariensis. Response to glyphosate was used to characterize percent resistance for each wild-collected population. RNA was extracted from the leaves of glyphosate-treated and untreated individuals, and used for cDNA synthesis. Quantitative PCR primers were designed for the E.

bonariensis orthologues of the E. canadensis genes EPSPS, ABC M10, and ABC M11, and pre- and post-spraying expression levels of each gene (relative to the housekeeping gene actin) were analyzed through qPCR. This experiment will examine gene expression patterns in ABC transporter genes (M10 and M11) to determine if the respective genes are upregulated in response to Glyphosate application and therefore establish their role in glyphosate resistance.   Future RNA-Seq analysis via Illumina HiSeq may reveal other genes that are differentially up- or down-regulated in resistant populations of E.

bonariensis after glyphosate exposure.  Determination of the genetic basis of herbicide resistance will provide fundamental data about parallel evolution in response to strong selection pressures, and help suggest alternative mechanisms for field control of this weed.

Herbicide Resistance

Moss (2002) defines herbicide resistance as the heritable ability of weeds to survive and reproduce in the presence of herbicide doses that are lethal to the wild type of the species. Herbicide resistance was first observed by an ornamental nursery owner in 1968 (Jasieniuk et al. 1996). The first resistance to herbicide was recorded in Senecio vulgaris; seeds from the resistant biotypes were found to be insensitive to the chemicals simazine and atrazine (Pieterse 2010). In 1974, resistance to glyphosate herbicide became a problem for corn growers (Gressel et al. 1982). Since then, more than 187 species of weeds throughout the globe have developed resistance against various herbicides that target a broad range of biochemical processes (Pieterse 2010). The first case of herbicide resistance in California was reported in 1981 at UC Riverside (Holt at al. 1981) and recently, more species have also evolved resistance to various other herbicide chemicals employed by farmers in California (see Table 1; Malone 2014).

Table 1. Weed species resistant to herbicide in California

Herbicide(s)	Species	Target site
Triazine	Senecio vulgaris	D1 protein
Sulfonylurea	Lolium perenne, Sagittaria montevidensis, Scirpus mucronatus, Salsola tragus, Ammania auriculata, Scirpus mucronatus	Acetaloacetate synthase (ALS)
Thiocarbamate	Echinochloa phyllopogon	Lipis synthesis
Aryloxyphenixy propionic acid	Echinochloa phyllopogon	Acetyl-CoA carboxylase (ACCase)

Glyphosate herbicide (marketed by Monsanto as RoundUp®) contains N-phosphonomethyl glycine that acts against plants by hindering aromatic amino acid synthesis (Bridges 2003). Upon application on leaves, the plant takes in glyphosate through the stomata and is transported through the phloem alongside products of photosynthesis to all parts of the plant. In susceptible plants, glyphosate hinders the role of the plastid-expressed enzyme 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS), important in the prechorismate step of the shikimate pathway. In the absence of glyphosate, this enzyme works by condensing shikimate-3-phosphate and phosphoenolpyruvate into 5-enolpyruvylshikimate-3-phosphate (EPSP) with inorganic phosphate, initiating the anabolism of aromatic amino acids (Ferreira 2008). Disruption of the synthesis of aromatic amino acids (phenylalanine, tryptophan, and tyrosine) eventually kills the plant (Herman and Weaver 1999).

Glyphosate has become the world’s most commonly used herbicide since its market introduction in 1974 (Baylis 2000). Its use around the globe results from several factors including potency against an extensive variety of species (monocots and dicots), less harmful activity against animals than other herbicides (the enzyme EPSPS is not present in animals), rapid microbial degradation in the soil, and low cost (Duke and Powles 2008). Also, it has been commonly used in recent years as an alternative to manual weeding (tillage) as a method of weed control. In this low or no-tillage method, there is huge dependence on herbicides to control weeds, and these have numerous environmental benefits and economic importance (Owen 2008; Powles 2008; Shaner 2000). With the adoption of genetically modified crops with glyphosate resistance in 1996, the already high levels of glyphosate application increased (Powles & Preston 2006). The combined effects of excessive over-application against already glyphosate resistant crops and completely reduced tillage has created an agricultural environment that has an increased risk of evolution of glyphosate resistance (Neve at al. 2003). Herbicide resistance is stimulated by the re-current use of herbicides with the same active chemical ingredients (LeBaron 1991), and so inevitably, glyphosate resistance has evolved in many weeds (Powles et al. 1998). However, because it still has the ability to control numerous weed species, has adaptability to low tillage systems, and low animal toxicity, high glyphosate use remains a key factor in modern systems because growers are unwilling to return to greater tillage systems or older, more toxic herbicides (Beckie 2012).

Glyphosate as a Herbicide

Two mechanisms that confer resistance and have contributed to glyphosate resistance in weeds are target-site and non-target site resistance. Target site-based resistance is a condition where resistance evolves due to a gene mutation resulting in a structural or chemical change to the target enzyme, so that the herbicide fails to effectively inhibit the normal enzyme function (Powles and Preston 2006). A missense mutation may involve a specific nucleotide substitution in the coding region producing a different amino acid that results in structural, hydrophobicity or charge change in the active site of the target enzyme making it less sensitive to inhibition by the herbicide. A few weeds such as goosegrass (Eleusine indica (L.) Gaertn.) have undergone target site mutagenesis (Lee and Ngim, 2000; Dinelli et al. 2006) of the enzyme EPSPS, via a substitution mutation that replaced the amino acid proline with serine at position 106 (Pro106-Ser). Ng et al. (2004 & 2005) demonstrated that substitution of proline by threonine (Pro106-Thr) also confers resistance glyphosate resistance to goosegrass. The mutated EPSPS enzyme has a low affinity for glyphosate but almost normal affinity for phosphoenol pyruvate (the enzyme’s usual substrate), permitting the shikimate pathway to proceed normally even in the presence of glyphosate (Gaines et al. 2010). 

Non-target site mutations induce sequestration, detoxification, and reduced absorption function by limiting the translocation of glyphosate to target sites (Wakelin et al. 2004). According to Claus and Brehrens’ (1976) study, rapid and widespread glyphosate translocation is necessary to achieve high herbicide efficacy. There is therefore a possibility that changes in its translocation may confer resistance. This is an intriguing observation that would be helpful in the understanding of non-target-site-based resistance (NTSR). The study that unraveled this phenomenon was carried out in rigid ryegrass (Lolium rigidum Gaud; Lorraine-Colwill et al. 2002), and indicated that resistance in at least one glyphosate-resistant biotype was not due to EPSPS enzyme target mutagenesis or degradation. In the same study, it was shown that there was no significant difference between glyphosate resistant and susceptible biotypes in EPSPS sensitivity or expression level or in glyphosate absorption. However, the patterns of glyphosate translocation differed. The researchers observed accumulation of glyphosate at lower parts of the plant and to some extent in the roots in susceptible plants, whereas in resistant plants, it accumulated in the tip of the leaves with a negligible amount transported to the roots (Lorraine-Colwill et al. 2002). Wakelin et al. (2004) found the same pattern of reduced glyphosate translocation when working with four glyphosate resistant ryegrass populations in Australia.

Researchers investigating mechanisms of glyphosate resistance in other Lolium rigidum populations have not found large differences in glyphosate translocation. Perez et al. (2004) found no significant difference in glyphosate absorption and translocation between susceptible and resistant Chilean Lolium plants. In an investigation of glyphosate resistance in Californian Lolium, Simarmata et al. (2003) found significantly higher glyphosate concentration in treated leaves of glyphosate resistant plants 2-3 days after treatment. Interestingly, these authors observed no other significant differences in glyphosate absorption or translocation between resistant and susceptible plants. These varying results suggest potentially different non-target resistance mechanisms to glyphosate may occur in different Lolium populations. In general, most studies observed no differences in glyphosate absorption, but phloem translocation is sometimes reduced greatly (Feng et al. 2004; Koger and Reddy 2005) in resistant populations. Failure of glyphosate translocation from leaves to the roots seems to be an important mechanism that leads to resistance in certain Lolium biotypes (Preston 2002).  

Mechanisms of Glyphosate Resistance

Species in the genus Erigeron are annual or short-lived perennial plants native to the Americas that have in the recent past become cosmopolitan and invasive weeds of many crops and arable lands (Prieur-Richard et al. 2000). The genus is in the sunflower family (Asteraceae) and the weed species were formerly placed in the genus Conyza before taxonomic revision (Baldwin 2012). Erigeron spp. are prolific seed producers: a single plant is capable of producing thousands of viable wind dispersed seeds (Weaver 2001; Shields et al. 2006).

Erigeron spp. have become very common and problematic weedy plants in agronomic crops around the world (Weaver 2001). The opportunistic nature of Erigeron in undisturbed areas makes them well-suited for becoming established in agricultural fields and surrounding areas, particularly in no-tillage or low tillage systems, e.g., conservation tillage (Bruce and Kells 1990; Buhler et al. 1992). Considering the fact that over the last two decades the amount of crop hectares in conservation tillage has increased by a margin of about 50% in Central California (Young 2006), weeds such as Erigeron spp. are becoming a great concern.

The two main weed species of Erigeron in California are hairy fleabane (Erigeron bonariensis L.) and horseweed (Erigeron canadensis L.), both summer annuals. Unlike horseweed, which is weedy across the U.S., hairy fleabane is an agricultural problem specific to California (Shrestha 2008). The optimal temperature for germination of hairy fleabane ranges between 20.3?C and 23.4 ?C and the seeds usually germinate under moderate water availability conditions (Karlsson and Milberg 2007). Based on these characteristics, conditions are ideal for fleabane germination in the cooler seasons of California’s Central Valley, October-March. Hairy fleabane has adapted to a wide range of conditions ranging from irrigated vineyard and orchard systems to dry non-crop areas (Shrestha et al. 2014).

Erigeron bonariensis has been shown to exhibit high resistance to glyphosate in the southern Central Valley of California (Okada et al. 2013), but the actual resistance mechanism is still obscure, especially the dynamics that occur at the genetic level. A number of studies have been performed on fleabane’s close relative, E. canadensis, in an attempt to establish the presence or absence of target site mutations at codon 106 of the EPSPS gene, as well the synchronization mechanism of EPSPS enzyme overexpression levels (Tani et al. 2015). However, the presence and number of these genes and their specific roles in glyphosate resistance in fleabane have not been established. The relationship between ABC transporter gene expression and glyphosate resistance needs to be examined in hairy fleabane as it has been examined in E. canadensis (Tani et al. 2015).  It is possible that reduced glyphosate translocation is the suspected mechanism of resistance in hairy fleabane from other parts of the world (Ferreira et al. 2008). It is also imperative to examine target mutations in the EPSPS gene (Gaines et al. 2010), to evaluate the potential involvement in fleabane glyphosate resistance. In addition to sequencing the gene region containing the known target site mutation in EPSPS (Pro106 in Lolium rigidum), it is important to determine if there are other mutations that confer target site resistance via changes in the EPSPS enzyme itself.

Target-Site Resistance

Candidate ABC transporter genes for our study were selected based on previous research on horseweed (Nol et al. 2012; Tani et al. 2015).  In these studies it was demonstrated that application of glyphosate to either resistant or susceptible biotypes may or may not influence EPSPS gene expression levels, but ABC transporter genes M10 and M11 were significantly upregulated in the resistant plants, and not in the non-resistant biotypes (Nol et al. 2012; Tani et al. 2015). This suggests a possible role of M11 and M10 ABC transporter genes in resistance to glyphosate in horseweed. These ABC transporters may function to sequester glyphosate into the vacuole resulting in a non-target site mechanism of glyphosate resistance in horseweed (Peng et al. 2010).

Weeds are very competitive for resources, both above and below ground, and therefore contribute to reduction in crop yield. As herbicides are the most commonly used weed control method, weed populations are subject to significant evolutionary pressure for mechanisms to survive herbicide exposure. This herbicide selection pressure has been present for decades in weed populations, resulting in reduced herbicide efficacy and rising crop losses (Green et al. 2008). Modern molecular biology techniques permit the determination of the genes that are upregulated in response to herbicide exposure in particular weed species, to begin to decipher the mechanisms behind non-target site resistance (NTSR). The aim of this study is to elaborate on the specific role of the EPSPS and ABC transporter genes in herbicide resistance in fleabane. This study will reaffirm the presence of glyphosate-resistance in hairy fleabane in the Central Valley, and will aid in identifying if resistance has evolved once or multiple times. This information is critical to the agricultural industry to ensure that effective and sustainable weed management and control strategies are promoted. Information about the genetic basis of herbicide resistance is key in effective weed management and has the potential to bolster agricultural productivity. Elucidating the role of both target site and non-target mutations in resistance to various chemicals is also important in the genetic engineering of crop plant resistance, especially by plant transformation techniques.

Broad Objective: To identify the genetic basis of glyphosate (RoundUp®) resistance in Central Valley populations of the agricultural weed hairy fleabane (Erigeron bonariensis L.).

To examine gene expression patterns in known populations of glyphosate-resistant vs. sensitive fleabane before and after different glyphosate application rates.

To establish if the candidate genes EPSPS and ABC transporter genes M10 and M11 are significantly associated with glyphosate herbicide resistance in fleabane populations.

To determine whether different glyphosate-resistant populations of fleabane from the Central Valley have different genetic bases for resistance.

Hypothesis: Glyphosate resistance in fleabane is due to non-target site resistance, facilitated by transcriptional upregulation of ABC transporter genes.

Glyphosate application will produce significant upregulation of ABC transporter genes in glyphosate-resistant hairy fleabane but not the glyphosate-sensitive populations.

Target site resistance in the EPSPS gene is not significantly associated with glyphosate resistance but M10 and M11 (and possibly other ABC transporter genes) underlie non-target site herbicide resistance in fleabane.

Different glyphosate-resistant Central Valley fleabane populations have different expression levels for M10 and M11 (and possibly other ABC transporter genes), due to different origins of resistance.

Non-Target Site Resistance

Sample preparation

Ten different populations of Erigeron bonariensis were selected for germination: these seeds were collected from the San Joaquin Valley in 2016 based on previous fleabane collection sites of known glyphosate resistance or sensitivity (Okada et al. 2014, Shrestha et al. 2014). In each Valley population (labeled “1v” to “10v”), seeds collected from 20 individual plants were bulked. A sensitive control population (HFS) and a resistant control population (HFR) were used as control populations; these populations (from seed collected in 2007 by Shrestha et al.) are known to be glyphosate susceptible and glyphosate resistant, respectively. Initial germination was performed in Petri dishes in growth chambers, with subsequent growth in the greenhouse.  Four Petri dishes were labeled for each population, except for the HFR population, which has very low germination success, for which 8 Petri plates were used. Twenty-five of the bulked seeds were added to filter paper in each Petri plate, for 100 seeds total per population (200 seeds for HFR). After adding the seeds, 10 ml of distilled water was added to each plate, and they were sealed with Parafilm. A growth chamber was used to germinate seeds, set at 20/15 C day/night temperature and 13 hr day length (Fig 1). There is no control over humidity in these chambers (as per Nandula et al. 2006).

Figure 1: Growth chamber with Petri plates used to germinate seeds  

At the 2-3 true leaf stage, 18 plants/population were transplanted into 4-inch pots. After transplanting into small pots, plants were grown to the 5-8 leaf stage, as per Okada et al. 2014. Each treatment group contained 6 randomly selected plants per population. One treatment group was sprayed with the label rate (1x) of glyphosate and another with twice the label rate (2x). The third treatment group was left unsprayed and used as the control. All individuals were randomized in a complete block design after spraying to eliminate effects of greenhouse position.  35 days after spraying, above-ground biomass was harvested for all living individuals, and dead individuals were recorded.  Biomass was dried at 60?C for 21 days to constant weight, and then weighed on a scale with an accuracy of 0.001 grams. ANOVA was used to calculate statistical differences in biomass after spraying between treatments, for each population.  If control and sprayed treatments showed a significant difference at the p<0.01 level, the population was considered glyphosate sensitive (Table 3).

Leaves were harvested at two time points in the experiment: prior to the glyphosate spraying time point, and twenty-four hours after glyphosate spraying occurred (as per Tani et al. 2015). The leaves were selected from among the medium-aged leaves on each plant (1-2 leaves per plant for about 100mg of tissue), stored at -80°C in Eppendorf 1.7 mL tubes, and ground in liquid nitrogen. Total RNA was isolated from the leaves using the Qiagen Plant RNeasy Mini Kit, using the protocol included with the kit, and stored at -20°C until downstream gene-expression analysis. Each individual sample was not extracted alone for most of the RNA extractions: instead, sample leaves from 2-3 plants that were from the same population and received the same treatment were bulked together at this stage, to provide more RNA for cDNA synthesis and to reduce expense. Potential genomic DNA was removed using DNase (FisherSci) and RNA was quantified using Qubit^® dsDNA HS Assay Kits. Denaturing gel electrophoresis was also performed to assess RNA quality. 

Study on Glyphosate Resistance

The qPCRs targeted the genes EPSPS, ABC M10, ABC M11, and actin as a control, “housekeeping” gene. These genes (other than actin) were upregulated in glyphosate-resistant E. canadensis (Nol et al. 2012; Tani et al. 2015), which is closely related to E. bonariensis. First-strand cDNA was reverse-transcribed from 1µg of total RNA. The starting reaction mixture (15 µL in volume) contained 2 µl oligo (dT) 18 mer and 1 µg of DNase treated RNA, and 4 µL of RNase-free water. The mixture was denatured at 70⁰C for 5 min followed by quick incubation on ice for 5 mins. After the addition of 5 µL M-MLV 5X Reaction Buffer, 1.25 µL 10mM dNTPs (RNase-free), 1ul M-MLV Reverse Transcriptase (company?), and RNase-free water to a 20 µL final volume, the reaction mixture was then incubated at 42⁰C for 60 min, followed by quick chilling on ice. Target cDNAs (50-150bp) were PCR-amplified using gene-specific primers shown in Table 2 (designed specifically for E. bonariensis using Primer3 software and vetted with standard PCR).

Quantitative PCR was performed using Thermo Scientific ABsolute SYBR® Green Master Mix (company?) on an Eppendorf^® Mastercycler^® RealPlex thermocycler (Applied Biosystems). The reaction mixture (20 µL) contained gene-specific oligonucleotides at a final concentration of 0.2 µM each and 6.8 µL of the cDNA as a template. PCR cycling started with the initial polymerase activation at 50⁰C for 5 min and 95⁰C for 10 min, followed by 40 cycles of 95⁰C for 15 s, 58⁰C for 15 s. The primer specificity and the formation of primer dimers was monitored by dissociation curve analysis. The expression level of the E. bonariensis actin gene was used as an internal standard to normalize small differences in cDNA template amounts. Relative transcript levels of the genes of interest were calculated as a ratio to the actin gene transcripts, and relative expression of each gene of interest before and after glyphosate treatment was calculated using the ΔΔ Ct method (implemented in Microsoft Excel). PCR efficiency for each amplicon was calculated by employing the linear regression method on the log (fluorescence) per cycle number data, using the LinRegPCR software (Ramakers et al., 2003). All real-time qPCR was performed on three biological replicates for population 3, 4, 6, 7, 8, 10, HFS and HFR (e.g., three plants per population which received the same treatment) and two biological replicates for population 1, 2, 5, and 9 (due to lack of RNA) for appropriate statistical power.

Table 2. Primers used for real-time quantitative PCR

Target gene	Forward primer	Reverse primer
EPSPS	5’-TTACTTCTTAGCTGGTGCTG-3’	5’-GGCATTTTGTTCATGTTCACATC-3’
M10	5’-TTGGCTCAACTTCGTGGTATTGGG-3’	5’-CCAAGAAATTCCAAGCGGAACCCT-3’
M11	5’-ATGCTGTCTTCTTTTACCTTTGC-3’	5’-CGACTTCCCACTACCAGTTCTTC-3’
Actin	5’-GTGGTTCAACTATGTTTCCCTG-3’	5’-CTTAGAAGCATTTCCTGTGG-3’

Glyphosate Resistance of Hairy Fleabane (Erigeron bonariensis), Based on Biomass Harvesting.

Thirty-five days after glyphosate spraying, the majority of the fleabane plants were still alive, although showing signs of chlorosis, stunted growth, and wilting. This made a visual survival estimate an inaccurate way to judge which plants showed glyphosate resistance (GR) and which showed glyphosate sensitivity (GS). Instead, their dry biomass was recorded and an analysis of variance was run. The control populations did not behave as expected.  More than 65% of the resistant control population (HFR) died after the glyphosate spraying, whereas only 11% died from the sensitive control population (HFS). Our suspicion is that the seeds’ labels were switched.  This did not affect our results because we did not base their resistance or sensitivity on their label but on our spraying results.

Based on the ANOVA used to calculate statistical differences in biomass after spraying between treatments, out of 12 populations, only four showed promise of glyphosate resistance. If control and sprayed treatments (either 1X or 2X) showed a significant difference at the p<0.001 level, the population was considered very glyphosate sensitive; somewhat sensitive populations had 0.001 < p < 0.05; somewhat resistant populations had 0.05 < p < 0.1; and very resistant populations had p > 0.1 (Table 3).  Table 3 shows that populations 1v, 2v, 4v, and 8v show evidence of glyphosate resistance. In contrast, populations 3v, 6v, 7v, 9v, 10v, HFS, and HFR appear to be at least somewhat glyphosate sensitive.  Populations 5v had such low germination that only 10 plants could be obtained for the experiment, and thus this population’s resistance could not be accurately determined.

Table 3. Summary table of ANOVA results from glyphosate spraying experiment.

Population	ANOVA Results	Resistance Inference
1v	F(2,15) = 3.17368, p = 0.07089	Somewhat resistant
2v	F(2,15) = 1.70497, p = 0.21518	Very resistant
3v	F(2,15) = 10.34117, p = 0.00150	Somewhat sensitive
4v	F(2,15) = 1.29119, p = 0.30381	Very resistant
5v	F(2,7) = 6.85536, p = 0.02245	Undetermined
6v	F(2,15) = 5.05745, p = 0.02095	Somewhat sensitive
7v	F(2,15) = 33.50911, p = 2.93E-06	Very sensitive
8v	F(2,14) = 2.79437, p = 0.09525	Somewhat resistant
9v	F(2,11) = 6.38642, p = 0.01443	Somewhat sensitive
10v	F(2,15) = 6.77714, p = 0.00800	Somewhat sensitive
HFS	F(2,15) = 9.26914, p = 0.00239	Somewhat sensitive
HFR	F(2,9) = 68.25843, p = 3.64E-06	Very sensitive

After isolation of RNA from collected leaves from each population, all RNA samples were quantified using Qubit^® dsDNA HS Assay Kits for their concentration range per ul. I selected only samples with RNA of concentration range ~40ng/uL or above for further analysis. Then RNA denaturing gel electrophoresis was performed to observe the quality of the chosen RNA samples. Agarose gel electrophoresis was run on 1 µg of the RNA sample. In the gel image (Fig. 3), each RNA sample shows two consecutive sharp and clear 28S rRNA and 18S rRNA bands in a 2:1 ratio. This band ratio indicates that RNA is completely intact. Degraded RNA lacks this band ratio and has a smeared appearance.

Figure 3: Depicts denaturing gel electrophoresis to assess RNA quality after measuring it with Qubit^® dsDNA HS assay kit. SHFS5, SHFS12, and SHFS14 represent different samples of the sensitive control population after spraying with glyphosate, and HFS5, HFS12, and HFS14 represents samples of the sensitive control population prior to spraying with glyphosate. SHFR3, SHFR5, and SHFR10 are different samples of the resistant control population after spraying with glyphosate, and HFR3, HFR5, and HFR10 are different samples of the resistant control population prior spraying with glyphosate. RFU values are the sample concentration range measured using Qubit quantitation.

RNA was selected for cDNA synthesis based on the results obtained from Qubit quantitation and RNA denaturing gel electrophoresis. To confirm cDNA synthesis, I ran a general PCR reaction, with primers for the target gene (i.e. actin, M10, M11, or EPSPS). After visualizing the gel with the UV light box, gel images similar to the following two example figures (Figs. 4 and 5) were obtained from gel electrophoresis for each population.  Thereafter, these cDNA samples were used for qPCR.

Figure 4: Depicted on top - Target cDNAs (50-150bp) for sprayed samples in different treatments (i.e. 1X, 2X, 0X) and unsprayed samples of the HFS control population were PCR amplified using gene-specific primers, i.e for the actin gene (housekeeping gene; first six lanes after 1kb ladder) and for the M10 gene (gene of interest).

Depicted on bottom - Target cDNAs (50-150bp) for sprayed samples in different treatments (i.e. 1X, 2X, 0X) and unsprayed samples of the HFS control population were PCR amplified using gene-specific primers, i.e the M11 gene (gene of interest; first six lanes after ladder) and the EPSPS gene (gene of interest). NC = negative control.

Figure 5: Depicted on top - Target cDNAs (50-150bp) for sprayed samples in different treatments (i.e. 1X, 2X, 0X) and unsprayed samples of the HFR control population were PCR amplified using gene-specific primers, i.e for the actin gene (housekeeping gene; first six lanes after 1kb ladder) and for the M10 gene (gene of interest).

Depicted on bottom - Target cDNAs (50-150bp) for sprayed samples in different treatments (i.e. 1X, 2X, 0X) and unsprayed samples of the HFR control population were PCR amplified using gene-specific primers, i.e the M11 gene (gene of interest; first six lanes after ladder) and the EPSPS gene (gene of interest). NC = negative control.

Quantitative PCR experiments on cDNA derived from leaves of all treated plants (i.e. control plants and plants treated at 1X and 2X concentrations of glyphosate, revealed several overall results (Table 5).  The ABC-transporter genes seemed to respond to glyphosate spraying in hairy fleabane regardless of glyphosate resistance or sensitivity. M10 and M11 expression was upregulated in both resistant and sensitive populations. Population 9 is probably not significantly upregulated for M10 simply because of low sample size. M10 showed very high upregulation in populations HFR, 1v, 2v and 5v with 1X treatment with glyphosate. Also, M10 was very highly upregulated in HFR and HFS with 2X treatment with glyphosate. M11 showed very high upregulation in HFR and 2v with 1x treatment. Whether EPSPS expression is upregulated or downregulated is unpredictable based on inferred glyphosate resistance; in half of the 12 populations, EPSPS is not significantly differently expressed in the 1X or 1X treatments vs. the 0X control (Table 5).

Table 5: Combined qPCR and glyphosate spraying results.  The first three rows show average normalized fold change in gene expression for the three treatment groups (0X, 1X and 2X glyphosate) for all genes of interest (M10, M11, and EPSPS).  ANOVA was used to calculate statistical differences in gene expression between treatments at the time point after spraying, for 12 populations. Degrees of freedome, F-values, and p-values are reported for the ANOVAs in rows 4-6. If the gene of interest showed greater than 1 fold change after spraying, the gene was considered up-regulated; <1 fold change, down-regulated (row 7). Finally, inferences about glyphosate resistance from the biomass results are summarized in row 8: R = very resistant, S = very sensitive, ~R = somewhat resistant, and ~S = somewhat sensitive.

M10
Population	HFS	HFR	1v	2v	3v	4v	5v	6v	7v	8v	9v	10v
1X: Fold change	3.87	39.42	38.99	77.86	35.51	2.73	17.70	4.32	2.39	2.62	4.75	2.21
2X: Fold change	13.54	81.40	3.28	4.77	2.61	2.96		4.56	3.47	2.89	4.49	6.00
0X: Fold change	0.99	1.12	1.04	0.85	1.65	1.21	1.65	0.99	0.85	1.87	0.72	1.21
Degrees of freedom	2, 27	2, 21	2, 13	2, 11	2, 15	2, 15		2, 9	2, 15	2, 15	2, 9	2, 15
F-value	238.42	10.70	52.97	20.48	36.89	9.82	28.6503	21.02	5.05	3.35	4.23	70.08
p-value	< .00001	0.00	< .00001	0.00	< .00001	0.00	< .00001	0.00	0.02	0.06	0.05	< .00001
Up or down-regulation: 1x/2x	up/up	up/up	up/up	up/up	up/up	up/up	up	up/up	up/up	up/up	up/up	up/up
R/S (spraying)	~S	S	~R	R	~S	R	N/A	~S	S	~R	~S	~S
M11
Population	HFS	HFR	1v	2v	3v	4v	5v	6v	7v	8v	9v	10v
1X: Fold change	4.68	43.13	1.33	30.90	2.62	2.70	0.86	4.19	2.68	2.28	3.64	1.69
2X: Fold change	9.14	2.55	2.10	3.45	0.81	2.89		4.29	4.60	2.47	3.22	2.33
0X: Fold change	0.89	1.06	0.98	1.27	1.18	1.01	0.81	0.86	1.16	1.03	1.16	1.24
Degrees of freedom	2, 27	2, 21	2, 11	2, 11	2, 15	2, 15		2, 9	2, 15	2, 15	2, 9	2, 15
F-value	20.55	27.45	0.73	5.86	4.66	5.92	4.5894	4.45	8.33	5.92	0.64	4.53
p-value	< .00001	< .00001	0.51	0.02	0.03	0.01	0.0101	0.05	0.00	0.01	0.55	0.03
Up or down-regulation: 1x/2x	up/up	up/up	up/up	up/up	up/down	up/up	down	up/up	up/up	up/up	up/up	up/up
R/S (spraying)	~S	S	~R	R	~S	R	N/A	~S	S	~R	~S	~S
EPSPS
Population	HFS	HFR	1v	2v	3v	4v	5v	6v	7v	8v	9v	10v
1X: Fold change	0.39	0.70	4.27	9.25	0.49	2.19	0.05	3.43	1.98	2.02	2.45	1.22
2X: Fold change	0.53	1.70	0.24	3.58	0.89	2.05		2.32	1.32	1.17	2.90	1.50
0X: Fold change	0.77	0.95	1.85	0.75	1.28	0.99	0.89	1.01	0.99	0.93	0.72	1.22
Degrees of freedom	2, 27	2, 21	2, 11	2, 11	2, 15	2, 15		2, 9	2, 15	2, 15	2, 9	2, 15
F-value	32.40	6.58	8.29	3.26	0.69	2.90	3.6606	14.09	2.38	2.90	15.65	0.14
p-value	< .00001	0.01	0.01	0.08	0.51	0.09	0.0216	0.00	0.13	0.09	0.00	0.87
Up or down-regulation: 1x/2x	down/down	down/up	up/down	up/up	down/down	up/up	down	up/up	up/up	up/up	up/up	up/up
R/S (spraying)	~S	S	~R	R	~S	R	N/A	~S	S	~R	~S	~S

The majority of the fleabane plants were still alive after thirty-five days of spraying showing signs of stunted growth, chlorosis and wilting. This experiment was inaccurate as the plants showed glyphosate resistance.  The control of the experiment did not behave as projected.  The suspicion was that the labels were interchanged but it did not affect the results as the experiment did not base the sensitivity and resistance on the label but on spraying result. The quantitative PCR experiment on cDNA derived from the leaves of all treated plants. The ABC- transporter genes looked to react to glyphosate spraying in hairy fleabane regardless of glyphosate sensitivity or resistance. M11 displayed a very high up-regulation in HFR and 2V with 1x treatment. Also, M10 showed a highly unregulated in HFR and HFS with 2x treatment with the glyphosate. Both M10 and M11 expression was up-regulated in both sensitive and resistance populace.

 ANOVA was utilized to calculate statistical variances in biomass after spraying for each populace between treatments. If the control and sprayed treatment showed a considerable variance at the P<0.01 level, the populace was deliberated glyphosate sensitive.  From the table above, it can be seen that the 1v, 2v, 4v, and 8v displays evidence of glyphosate resistance.  Populace 5v had low germination that merely 10 plants could be got for the trial, and thus this populace’s resistance could not be precisely determined.  On contrarily, the populace 3v, 6v, 7v, 9v, 10v, HFR and HFS display evidence of glyphosate resistance. The six distinctive per populations is a very trivial sample dimension, and thus, these preliminary results ought to be tested with more replicates.  This is particularly true because the HFR and HFS populace did not behave as projected and because the literature reports about this populace, it disagree with experiment results in numerous instances.  All the subsequent interpretations about qPCR are contingent upon this spraying replication backing experiment preliminary conclusion about the resistance.  If the future biomass outcomes are varying, the experiment interpretations of sensitivity and resistance will differ, and so will the conclusion about the gene resistance and expression (Gaines et al. 2010).

Our experiment partially agreed with the hypothesis that glyphosate application will produce significant up-regulation of ABC transporter genes in glyphosate-resistant hairy fleabane but not the glyphosate-sensitive populations (Peng et al. 2010). However, the ABC-transporter genes seemed to respond to the glyphosate spraying regardless of glyphosate sensitivity or resistance.  The data in the experiment also partially support the hypothesis that the target site resistance in the EPSPS gene is not significantly associated with glyphosate resistance but M10 and M11 (and possibly other ABC transporter genes) underlie non-target site herbicide resistance in fleabane (Peng et al. 2010).  From the data above,  different plants show different levels of sensitivity or resistance and therefore, relate to our hypothesis that different glyphosate-resistant Central Valley fleabane populations have different expression levels for M10 and M11 (and possibly other ABC transporter genes), due to different origins of resistance.

In the first hypothesis, it does not correspond to the experiment outcomes. Erigeron bonariensis has been shown to exhibit high resistance to glyphosate in the southern Central Valley of California (Okada et al. 2013), but the actual resistance mechanism is still obscure, particularly the dynamics that happen at the genetic level. A number of studies have been performed on fleabane’s close relative, E. canadensis, in an attempt to establish the presence or absence of target site mutations at codon 106 of the EPSPS gene, as well the synchronization mechanism of EPSPS enzyme overexpression levels (Tani et al. 2015). However, the presence and number of these genes and their specific roles in glyphosate resistance in fleabane have not been established. The relationship between ABC transporter gene expression and glyphosate resistance needs to be examined in hairy fleabane as it has been examined in E. canadensis (Tani et al. 2015).   In the second hypothesis, it is not supported by the experiment results. The studies demonstrates that application of glyphosate to either resistant or susceptible biotypes may or may not influence EPSPS gene expression levels, but ABC transporter genes M10 and M11 are significantly upregulated in the resistant plants, and not in the non-resistant biotypes (Nol et al. 2012; Tani et al. 2015). This suggests a possible role of M11 and M10 ABC transporter genes in resistance to glyphosate in horseweed. These ABC transporters may function to sequester glyphosate into the vacuole resulting in a non-target site mechanism of glyphosate resistance in horseweed (Peng et al. 2010). The above reaffirms the hypothesis of different glyphosate-resistant Central Valley fleabane populations have different expression levels for M10 and M11 (and possibly other ABC transporter genes), due to different origins of resistance.  The EPSPS was not associated with resistance at all as the quantitative PCR primers were planned for the E. bonariensis and E. Canadensis showed different levels of resistance contrary to what was expected.  The above statement restates the reason why third hypothesis correlate to the outcomes of the trial.

Future RNA-Seq evaluation through the Illumina HiSeq may show other genes that are differentially up or down-regulated in resistance populace of E. borariensis after glyphosate contact. Determination of the genetic foundation of herbicide will offer fundamental info about the parallel evolution in reaction to robust selection pressure and assist propose substitute mechanisms for field control of this weed.

The objective of this paper was to identify the genetic foundation of glyphosate resistance in central valley populaces of the agricultural weed hairy fleabane. Glyphosate is an herbicide that comprises N-phosponomethyl glycine that performs against vegetation by inhibiting aromatic amino acid synthesis. It inhibits the roles of the Plastid-expressed enzyme 5-enolpyruvylshikimate-phosphate synthase (EPSPS) (Ferreira 2008). The control populace did not behave as anticipated. More than 60% of the resistance control populace died after the glyphosate spray whereas merely 11% died from the sensitive control populace (HFS). The main suspicion of the deviation of the above finding from the literature is due to the seeds labels switch. We matched our resistance outcomes with the results from the corresponding literature and established that our results sometimes contradicted the literature outcomes (table 4). These variations to glyphosate sensitivity or resistance could have been caused by a number of ecological aspects in the field, in the years amidst the lastly collections and our collections.

Table 4: Glyphosate spraying results from our experiment, compared with those from the corresponding literature (Okada et al. 2014; Shrestha et al. 2008; Shrestha et al. 2014). 

Population	In literature	Literature results	Our results
1v	Okada et al. 2014 H3	GR	Somewhat resistant
2v	Shrestha et al. 2014 #7	GS	Very resistant
3v	Shrestha et al. 2014 #4	GR	Somewhat sensitive
4v	Shrestha et al. 2014 #5	GR	Very resistant
5v	Shrestha et al. 2014 #6	GS	Undetermined
6v	Shrestha et al. 2014 #3	GR	Somewhat sensitive
7v	Shrestha et al. 2014 #9	GS	Very sensitive
8v	Okada et al. 2014 H1	GS at high rates (60%)	Somewhat resistant
9v	N/A	N/A	Somewhat sensitive
10v	Okada et al. 2014 K1	Possible GS (70% resistance)	Somewhat sensitive
HFR	Shrestha et al. 2008	Suspected GR	Very sensitive
HFS	Shrestha et al. 2008	Suspected GS	Somewhat sensitive

References

Anders, S., & Huber, W. (2010). Differential expression analysis for sequence count data. Genome biology, 11(10), R106.

Anders, S., Pyl, P. T., & Huber, W. (2015). HTSeq—a Python framework to work with high-throughput sequencing data. Bioinformatics, 31(2), 166-169.

Baldwin, B. G., & Goldman, D. H. (2012). The Jepson manual: vascular plants of California. University of California Press.

Baylis, A. D. (2000). Why glyphosate is a global herbicide: strengths, weaknesses and prospects. Pest Management Science, 56(4), 299-308.

Beckie, H. J., & Tardif, F. J. (2012). Herbicide cross resistance in weeds. Crop Protection, 35, 15-28.

Bruce, J. A., & Kells, J. J. (1990). Horseweed (Conyza canadensis) control in no-tillage soybeans (Glycine max) with preplant and preemergence herbicides. Weed Technology, 642-647.

Buhler, D. D. (1992). Population dynamics and control of annual weeds in corn (Zea mays) as influenced by tillage systems. Weed Science, 241-248.

Claus, J. S., & Behrens, R. (1976). Glyphosate translocation and quackgrass rhizome bud kill. Weed Science, 149-152.

Conesa et al. (2016). A survey of best practices for RNA-seq data analysis; Genome biology. https://doi.org/10.1186/s13059-016-0881-8

Dinelli, G., Marotti, I., Bonetti, A., Minelli, M., Catizone, P., & Barnes, J. (2006). Physiological and molecular insight on the mechanisms of resistance to glyphosate in Conyza canadensis (L.) Cronq. biotypes. Pesticide Biochemistry and Physiology, 86(1), 30-41.

Duke, S. O., & Powles, S. B. (2008). Glyphosate: a once-in-a-century herbicide. Pest management science, 64(4), 319-325. 

Feng, P. C., Tran, M., Chiu, T., Sammons, R. D., Heck, G. R., & CaJacob, C. A. (2004). Investigations into glyphosate-resistant horseweed (Conyza canadensis): retention, uptake, translocation, and metabolism. Weed Science, 52(4), 498-505.

Ferreira, E. A., Galon, L., Aspiazú, I., Silva, A. A., Concenço, G., Silva, A. F. & Vargas, L. (2008). Glyphosate translocation in hairy fleabane (Conyza bonariensis) biotypes. Planta Daninha, 26(3), 637-643.

Green, J. M., Hazel, C. B., Forney, D. R., & Pugh, L. M. (2008). New multiple-herbicide crop resistance and formulation technology to augment the utility of glyphosate. Pest Management Science, 64(4), 332-339.

Green, M. B., LeBaron, H. M., & Moberg, W. K. (Eds.). (1990). Managing resistance to agrochemicals: from fundamental research to practical strategies. American Chemical Society.

Gressel, J. (2002). Molecular biology of weed control (Vol. 1). CRC Press.

Gressel, J., Ammon, H.U., Folgelfors, H., Gasquez, J., Kay, Q.O.N & Kees, H. (1982). Discovery and distribution of herbicide-resistant weeds outside North America. In: H.M. LeBaron & J. Gressel (eds.). Herbicide resistance in plants. John Wiley & Sons, Inc, Canada.

Haas, B. J., Papanicolaou, A., Yassour, M., Grabherr, M., Blood, P. D., Bowden, J. & MacManes, M. D. (2013). De novo transcript sequence reconstruction from RNA-seq using the Trinity platform for reference generation and analysis. Nature protocols, 8(8), 1494-1512.

Heap, I. 2017. International Survey of Herbicide Resistant Weeds. (n.d.). Retrieved August 30, 2017, from https://www.weedscience.org/

Holt, J. S., Stemler, A. J., & Radosevich, S. R. (1981). Differential light responses of photosynthesis by triazine-resistant and triazine-susceptible Senecio vulgaris biotypes. Plant physiology, 67(4), 744-748.

Jasieniuk, M., Brûlé-Babel, A. L., & Morrison, I. N. (1996). The evolution and genetics of herbicide resistance in weeds. Weed Science, 176-193.

Karlsson, L. M., & Milberg, P. (2007). Comparing after?ripening response and germination requirements of Conyza canadensis and C. bonariensis (Asteraceae) through logistic functions. Weed Research, 47(5), 433-441.

Kaspary, T. E., Lamego, F. P., Langaro, A. C., Ruchel, Q., & Agostinetto, D. (2016). Investigation of the mechanism of resistance to glyphosate herbicide in hairy fleabane. Planta Daninha, 34(3), 555-564.

Koger, C. H., & Reddy, K. N. (2005). Role of absorption and translocation in the mechanism of glyphosate resistance in horseweed (Conyza canadensis).Weed Science, 53(1), 84-89.

LeBaron, H. M. (1991). Distribution and seriousness of herbicide-resistant weed infestations worldwide. Herbicide resistance in weeds and crops, 27-43.

Lee, L. J., & Ngim, J. (2000). A first report of glyphosate?resistant goosegrass (Eleusine indica (L) Gaertn) in Malaysia. Pest Management Science, 56(4), 336-339.

Lorraine-Colwill, D. F., Powles, S. B., Hawkes, T. R., Hollinshead, P., Warner, S. A. J., & Preston, C. (2002). Investigations into the mechanism of glyphosate resistance in Lolium rigidum. Pesticide Biochemistry and Physiology, 74(2), 62-72.

Malone, J. M., Boutsalis, P., Baker, J., & Preston, C. (2014). Distribution of herbicide?resistant acetyl?coenzyme A carboxylase alleles in Lolium rigidum across grain cropping areas of South Australia. Weed Research, 54(1), 78-86

Menendez, J., & De Prado, R. (1996). Diclofop-methyl Cross-Resistance in a Chlorotoluron-Resistant Biotype of Alopecurus myosuroides. Pesticide Biochemistry and Physiology, 56(2), 123-133.

Morant, M., Bak, S., Møller, B. L., & Werck-Reichhart, D. (2003). Plant cytochromes P450: tools for pharmacology, plant protection and phytoremediation. Current opinion in biotechnology, 14(2), 151-162.

Neve, P., Diggle, A. J., Smith, F. P., & Powles, S. B. (2003). Simulating evolution of glyphosate resistance in Lolium rigidum II: past, present and future glyphosate use in Australian cropping. Weed Research, 43(6), 418-427.

Nol, N., Tsikou, D., Eid, M., Livieratos, I. C., & Giannopolitis, C. N. (2012). Shikimate leaf disc assay for early detection of glyphosate resistance in Conyza canadensis and relative transcript levels of EPSPS and ABC transporter genes. Weed Research, 52(3), 233-241.

Okada, M., & Jasieniuk, M. (2014). Inheritance of glyphosate resistance in hairy fleabane (Conyza bonariensis) from California. Weed science, 62(2), 258-266.

Parkhomchuk D, Borodina T, Amstislavskiy V, Banaru M, Hallen L, Krobitsch S, et al. Transcriptome analysis by strand-specific sequencing of complementary DNA. Nucleic Acids Res. 2009;37:e123.

Peng, Y., Abercrombie, L. L., Yuan, J. S., Riggins, C. W., Sammons, R. D., Tranel, P. J., & Stewart, C. N. (2010). Characterization of the horseweed (Conyza canadensis) transcriptome using GS-FLX 454 pyrosequencing and its application for expression analysis of candidate non-target herbicide resistance genes. Pest management science, 66(10), 1053-1062. 

Perez, A., Alister, C., & Kogan, M. (2004). Absorption, translocation and allocation of glyphosate in resistant and susceptible Chilean biotypes of Lolium multiflorum. Weed Biology and Management, 4(1), 56-58.

Pieterse, P. J. (2010). Herbicide resistance in weeds–a threat to effective chemical weed control in South Africa. South African Journal of Plant and Soil, 27(1), 66-73.

Powles, S. B., & Preston, C. (2006). Evolved glyphosate resistance in plants: biochemical and genetic basis of resistance 1. Weed Technology, 20(2), 282-289.

Powles, S. B., Lorraine-Colwill, D. F., Dellow, J. J., & Preston, C. (1998). Evolved resistance to glyphosate in rigid ryegrass (Lolium rigidum) in Australia. Weed science, 604-607.Preston, C. (2002). Common mechanisms endowing herbicide resistance in weeds. In Proceedings of the 13th Australian Weeds Conference, Plant Protection Society of South Australia, Perth, Australia (pp. 666-674).

Prieur-Richard, A. H., Lavorel, S., Grigulis, K., & Dos Santos, A. (2000). Plant community diversity and invasibility by exotics: invasion of Mediterranean old fields by Conyza bonariensis and Conyza canadensis. Ecology Letters, 3(5), 412-422.

Ramakers, C., Ruijter, J. M., Deprez, R. H. L., & Moorman, A. F. (2003). Assumption-free analysis of quantitative real-time polymerase chain reaction (PCR) data. Neuroscience letters, 339(1), 62-66.

Shaner, D., & Bridges, D. (2003). Inhibitors of aromatic amino acid biosynthesis (glyphosate). Herbicide action course. West Lafayette: Purdue University, 514-529.

Shields, E. J., Dauer, J. T., VanGessel, M. J., & Neumann, G. (2006). Horseweed (Conyza canadensis) seed collected in the planetary boundary layer. Weed Science, 54(6), 1063-1067.

Shrestha, A. (2008). Biology and management of horseweed and hairy fleabane in California. UCANR Publications.

Shrestha, A., Steinhauer, K. M., Moretti, M. L., Hanson, B. D., Jasieniuk, M., Hembree, K. J., & Wright, S. D. (2014). Distribution of glyphosate-resistant and glyphosate-susceptible hairy fleabane (Conyza bonariensis) in central California and their phenological development. Journal of pest science, 87(1), 201-209.

Simarmata, M., Kaufmann, J. E., & Penner, D. (2003). Potential basis of glyphosate resistance in California rigid ryegrass (Lolium rigidum). Weed Science, 51(5), 678-682.

Tani, E., Chachalis, D., & Travlos, I. S. (2015). A glyphosate resistance mechanism in Conyza canadensis involves synchronization of EPSPS and ABC-transporter genes. Plant Molecular Biology Reporter, 33(6), 1721-1730.

Wakelin, A. M. and Preston, C. (2005). Target-site glyphosate resistance in rigid ryegrass (Lolium rigidum Gaudin). Weed Science Society of America. Abstr. 417. Lawrence, KS: Weed Science Society of America.

Wakelin, A. M., Lorraine-Colwill, D. F., & Preston, C. (2004). Glyphosate resistance in four different populations of Lolium rigidum is associated with reduced translocation of glyphosate to meristematic zones. Weed Research, 44(6), 453-459.

Weaver, S. E. (2001). The biology of Canadian weeds. 115. Conyza canadensis. Canadian Journal of Plant Science, 81(4), 867-875.

Wu, H., Walker, S., Robinson, G., & Coombes, N. (2010). Control of flaxleaf fleabane (Conyza bonariensis) in wheat and sorghum. Weed Technology, 24(2), 102-107.

Young, B. G. (2006). Changes in Herbicide Use Patterns and Production Practices Resulting from Glyphosate-Resistant Crops1. Weed Technology, 20 (2), 301-307.

Yuan, J. S., Tranel, P. J., & Stewart, C. N. (2007). Non-target-site herbicide resistance: a family business. Trends in plant science, 12(1), 6-13.