Seeds of kochia (susceptible to atrazine, chlorsulfuron, and glyphosate), were collected from a field in Manhattan, Kansas, and germinated in plastic trays filled with Miracle-Gro potting mix (Scotts Miracle-Gro Company, Marysville, OH, USA) in the greenhouse (29/25 C temperature; 15/9 h light day/night, supplemented with 1200 µmol illumination using sodium vapor lamps). Individual seedlings (~4 cm tall) were transplanted to 6.5 cm × 6.5 cm × 6.5 cm pots in the greenhouse for herbicide treatments.

Nine kochia seedling (10 - 12 cm) split into replicates of three seedlings each (three seedlings for each time-point) were separately treated with field rates (1×) of three sites of action (SOA) herbicides including recommended adjuvants: atrazine, a PSII-inhibitor (Aatrex 4L^® @ 2240 g ai ha⁻¹; 1% v/v crop oil concentrate (COC)), chlorsulfuron, an ALS-inhibitor (Glean^® XP @ 18 g ai ha⁻¹; 0.25% v/v nonionic surfactant (NIS)), and glyphosate, an EPSPS inhibitor (Roundup Weathermax^® @ 840 g ae ha⁻¹; 2% w/v ammonium sulfate (AMS)). Additionally, three seedlings (controls) were treated only with adjuvants and three seedlings were left untreated (no adjuvant and no herbicide treatment). The field rate (1×) was chosen for herbicide treatments based on our preliminary studies that indicated susceptibility of kochia population to above three SOA (data not shown).

Young leaf tissue from three treated plants (biological replicates) was separately collected at 8, 24 h, and 7 d after treatment, and flash frozen in liquid nitrogen (−196 C). Additionally, tissue was also collected at 48 h to study KspsbA gene expression after atrazine treatment. Tissue from control plants was separately collected at 7 d after adjuvant treatments and used for relative comparisons with other treated time points. The untreated leaf tissue with no adjuvant or herbicide treatments was collected at 8 h time point. All the harvested tissue was stored at −80 C until RNA isolation.

The frozen leaf tissue was homogenized in liquid nitrogen using a pre-chilled mortar and pestle to prevent thawing. The powdered tissue was transferred to a 1.5 mL microcentrifuge tube and the total RNA was isolated using a TRIazol^® reagent (Thermo Fisher Scientific, Waltham, MA, USA) following manufacturer’s instructions with modifications. Ribonucleic acid was treated with DNase 1 enzyme (Thermo Fisher Scientific, Waltham, MA, USA) to remove genomic DNA contamination. The isolated RNA was stored at −80 C. The quantity and quality (integrity) of total RNA was determined using a spectrophotometer (NanoDrop 1000, Thermo Fisher Scientific) and agarose gel (1%) (Agarose Pro- ducts, MS, USA) electrophoresis.

Complementary DNA (cDNA) was synthesized from 1 μg of the total RNA using RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific). The cDNA was then diluted with molecular grade water in a 1:5 ratio and used in a quantitative PCR (qPCR) reaction to study the expression of herbicide target-site and photosynthetic genes in kochia.

Quantitative Real-Time PCR primers (Table 1) were designed using OligoAnalyzer 3.1 [https://www.idtdna.com/calc/analyzer] and by the alignment of the nucleotide sequences from species representing several plant families available in GenBank (National Center for Biotechnology Information, Bethesda, MD). Nucleotide sequence alignment was done using MultAlin software [13] . Conserved K. scoparia psbA (KspsbA) primers were designed by aligning the nucleotide sequences of K. scoparia (accession no. AY251266.1) and Poa annua (accession no. AF131887.1). K. scoparia ALS (KsALS) primers were designed based on the sequence (accession no. EU517499.1) available at GenBank. Primers for K. scoparia Rubisco (large subunit) (KsRubisco) were designed by aligning nucleotide sequences from Bassia dasyphylla (Fisch. & C. A. Mey.) Kuntze (accession no. AY270150.1), Atriplex covillei (Standl.) J.F. Macbr. (accession no. HM587609.1), Kochia americana (S. Watson) A.J. Scott (accession no. AY270103.1), Salsola genistoides Poir. (Accession no. AY270128.1), and Suaeda linifolia Pall. (accession no. HM630106.1) respectively. Previously reported primers were used for amplifying the EPSPS, CAB and PPDK genes from K. scoparia [14] [15] . In this study, we tested previously reported reference genes such as SAND (SAND family protein), UBC (Ubiquitin-Conjugating Enzyme), ARF2 (Auxin Response Factor 2) from leafy spurge (Euphorbia esula L.) [16] and CAC (Clathrin Adaptor Complex) from mustard (Brassica juncea L.) [17] based on comparative delta CT method (data not shown). However, we found that most of these top performing reference genes in other plant systems were not stably expressed under the different herbicide stress conditions in kochia, except β- tubulin.

Quantitative PCR reaction was performed using a StepOnePlus^TM real-time detection system (Thermo Fisher Scientific) to determine the expression of herbicide target-site and photosynthetic genes from K. scoparia. The qPCR reaction mix (14 µL) consisted of 8 µL of PowerUp SYBR Green master mix (Applied Biosystems, Waltham, MA, USA), 2 µL each of forward and reverse primers (5 µM), and 2 µL of cDNA. The primers for herbicide target-site and photosynthetic genes were used in the qPCR reactions (Table 1). The qPCR conditions were 95 C for 15 min and 40 cycles of 95 C for 30 s and an annealing at 59/60 C for 1 min (Table 1). A melt curve profile was included following the thermal cycling protocol to determine the specificity of the qPCR products. Single curves were obtained for the herbicide target-site and tubulin primers.

A non-destructive method for determining leaf chlorophyll index was adopted using a chlorophyll meter SPAD-502 plus (Konica Minolta Optics Inc., Japan) [18] . The SPAD meter readings are proportional to the amount of chlorophyll

present in the leaf samples [19] . Leaf chlorophyll index in kochia was measured at 6 h, 1, 2, 3, 4, 5, 6, and 7 d after treatment with different herbicide SOA.

A chlorophyll fluorometer OS30p₊ (Opti-Sciences Inc., Hudson, NH, USA) was used to measure the efficiency of photosynthetic light reaction (PSII) as indicated by chlorophyll fluorescence, which is highly sensitive to stress [20] . Chlorophyll fluorescence in kochia was measured at 6 h, 1, 2, 3, 4, 5, 6, and 7 d after treatment with different herbicide SOA.

Fold-change in gene expression (as) was calculated by the comparative Ct method [21] , relative to the adjuvant treated control samples, where △Ct = [Ct target gene − Ct reference gene]. β-tubulin was used as an endogenous reference gene (housekeeping gene) for normalizing the gene expression data [22] . A calibrator sample (untreated-no adjuvant) was also included to represent 1x expression of gene of interest. Gene expression data was analyzed using one-way ANOVA in SigmaPlot (version 12.3). Means were separated using Fisher’s LSD (p < 0.05) and standard deviations (SD) were calculated based upon three biological replicates (n = 3). Chlorophyll index and fluorescence were measured on four plants (four biological replicates, n = 4) for each herbicide treatment at each time point and the data was analyzed using one-way ANOVA in SigmaPlot (version 12.3). Means of herbicide treatments at each time point were compared with control (untreated means) using Holm-Sidak’s test (p < 0.05).

There was a >200-fold increase (p ≤ 0.001) in KspsbA gene expression at 24 h after treatment (HAT) with atrazine (Figure 1(a)). Levels of KspsbA transcript returned to the levels found in untreated plants 48 h after atrazine treatment. In contrast, transcript levels of KsCAB, KsPPDK, and KsRubisco were downregulated (p ≤ 0.020) at 8 and 24 h after atrazine treatment (Figure 1(b)). KsRubisco levels reached a peak (7-fold relative to control, p ≤ 0.001) at 7 d after atrazine treatment (Figure 1(b)).

There was a significant upregulation of the KsALS gene (~9-fold relative to control, p ≤ 0.001) immediately after chlorsulfuron treatment, followed by a 7-fold increase at 24 HAT (Figure 2). KsALS transcript levels returned to near control levels at 7 d after treatment with chlorsulfuron. The expression of KsCAB remained unaffected at 8 and 24 HAT (p ≤ 0.008), whereas KsPPDK was upregulated at 24 HAT (p ≤ 0.011) (Figure 2). KsRubisco was upregulated 2.6-fold relative to untreated control at 24 HAT with chlorsulfuron (p ≤ 0.001) (Figure 2). At 7 d after chlorsulfuron treatment, the transcript levels of KsCAB, KsPPDK,

and KsRubisco were found to be lower compared to their controls.

KsEPSPS gene was upregulated (3.3-fold, p ≤ 0.001) at 8 HAT with glyphosate, maintaining similar transcript levels (3.4-fold) at 24 HAT as well (Figure 3). There was a 10-fold increase in KsEPSPS levels 7 d after glyphosate treatment (Figure 3). Compared to other herbicide treatments, KsRubisco showed the maximum expression levels after glyphosate treatment, reaching a peak at 8 h (7-fold relative to control, p ≤ 0.001) (Figure 3). A high KsRubisco level (~5- fold) was maintained even 7 d after glyphosate treatment. In contrast, KsCAB and KsPPDK genes were relatively downregulated (~0.5-fold, p ≤ 0.001, p ≤ 0.01) 7 d after glyphosate treatment (Figure 3).

In general, upon treatment with either atrazine, chlorsulfuron, or glyphosate, there was no significant change (except at few time points) in chlorophyll index in kochia compared to untreated plants (control) during the seven-day period after herbicide treatments (Figure 4(a)). Chlorophyll fluorescence, on the other hand varied, especially after atrazine treatment (Figure 4(b)). There was a significant decline in chlorophyll fluorescence at 6 h (indicated by 0 on the x-axis), and the decreasing trend continued for the next 7 d after atrazine treatment (p ≤ 0.001) (Figure 4(b)). However, after glyphosate and chlorsulfuron treatments, the chlorophyll fluorescence remained unchanged compared to untreated kochia (Figure 4(b)).

There was a strong upregulation (~200 fold) of the KspsbA gene in kochia at 24 HAT with atrazine (Figure 1(a)). The KspsbA expression returned to the same level as in control by 48 h and continued to stay at that level 7 d after atrazine

treatment. D1 protein, encoded by the psbA gene is the target-site of PSII-inhi- bitors [31] [32] . A high Dl protein degradation and turnover rates have been reported after photoinhibition [33] . These high Dl protein turnover rates are essential for repair of the photoinhibitory damage caused by PSII-inhibitors. The high D1 protein turnover is also accompanied by corresponding increase in psbA gene transcription [34] . In this study, the high KspsbA gene expression in kochia after atrazine application can be explained by the rapid turnover of the D1 protein. The photoinhibitory damage caused by atrazine also leads to production of reactive oxygen species (ROS) such as singlet oxygen (¹O₂) [35] . The singlet oxygen and the other free radicals produced after the blockage of electron transfer by triazines result in an irreversible damage to photosynthetic machinery contributing to plant death [36] .

ALS-inhibitor herbicides suppress the biosynthesis of branched-chain amino acids (e.g. valine, leucine, and isoleucine) through the inhibition of acetolactate synthase. Transcript levels of the KsALS were approximately 9 and 7-fold higher in K. scoparia at 8 and 24 HAT (Figure 2). Several studies have determined the expression of protein after the application of ALS-inhibitors, but there is insufficient literature regarding the expression of ALS gene. A three- to four-fold variation in ALS gene expression has been reported in several tobacco organs such as leaves, seedlings, and flowers [37] . An increase in the free amino acid pool after the application of ALS-inhibitors has been reported [38] [39] [40] . This increased protein pool was attributed to an increase in protein turnover rates in response to treatment with ALS-inhibitors [41] . The higher expression levels of KsALS gene observed in this study, may lead to increased protein turnover and free amino acid pools after application of ALS-inhibitors. The increased protein turnover is due to increased protein degradation and reduced protein synthesis [42] .

Gaines et al. [43] found no difference in EPSPS transcript levels in Palmer amaranth (Amaranthus palmeri S. Wats.) 8 h after glyphosate application. Similarly, there was no induction of EPSPS transcripts in horseweed (Conyza canadensis L. Cronq.) at 21 d after treatment with glyphosate [44] . In contrast, Gao et al. [45] recently reported upregulation (two to three-fold) of EPSPS gene in annual wild soybean (Glycine soja) at 6 h after glyphosate treatment. In the current study, we observed a similar three-fold increase in KsEPSPS expression at 8 and 24 h followed by a ten-fold increase at 7 d after glyphosate treatment (Figure 3). Although both inhibit different pathways, glyphosate induces physiological responses similar to ALS-inhibitors in plants, resulting in an increased free amino acid content (as discussed above) and an impaired carbon metabolism leading to plant death [39] [42] [46] . Similar to ALS-inhibitors, glyphosate application also increases EPSPS transcript levels (Figure 3), possibly resulting in higher protein turnover and degradation.

The expression of Rubisco is dependent on various factors such as light [47] [48] , cell and tissue type [49] [50] , efficiency of photosynthetic machinery [51] , phytohormone and nutrient levels [52] [53] , and plant developmental stage [54] . Rubisco consists of two subunits (large and small), encoded by two separate genes (chloroplastic large subunit, LSU and nuclear small subunit, SSU) [55] . In this investigation the expression of KsRubisco (LSU) was determined in response to herbicides with different SOA.

KsRubisco levels varied with different herbicide treatments. KsRubisco was upregulated (6-fold) at 8 h and its transcript levels remained high (4-fold) in kochia even at 7 d after treatment with glyphosate (Figure 3). Glyphosate is known to induce physiological changes leading to increased photosynthetic rates and growth stimulations [56] . An increase in shikimate is the most commonly observed biochemical change in response to glyphosate application. Both shikimate and photosynthetic pathways are known to be closely related, as majority of carbon that is synthesized during photosynthesis feeds the shikimic acid pathway [57] . Increased photosynthetic rate observed in barley after glyphosate treatment was attributed to increased carbon fixation resulting from higher Rubisco expression and RuBP (ribulose-biphosphate) turnover rates [56] [58] . In the current study, the strong upregulation of KsRubisco at 8 HAT with glyphosate (Figure 3) supports previous reports. However, both chlorophyll index and fluorescence (F_v/F_m) remained unchanged (Figure 4(a) and Figure 4(b)) during the 7 d period after glyphosate treatment. There was a 2.6-fold increase in KsRubisco expression in kochia after chlorsulfuron treatment (Figure 2). The upregulation of Rubisco after chlorsulfuron treatment is likely to have caused as a result of the effect of ALS-inhibitors on carbon metabolism. Amino acid synthesis inhibitors such as glyphosate and chlorsulfuron, are known to induce such physiological responses in plants [42] .

In the current study, KsRubisco was downregulated (at 8 and 24 HAT with atrazine) when compared to the control, however maximum expression of KsRubisco (~7-fold relative to control) was seen 7 d after atrazine treatment (Figure 1). Previous reports suggest that the specific activity of Rubisco was significantly reduced in maize (Zea mays L.) during the first two days after atrazine treatment followed by a recovery phase after one week [59] . The high expression of KsRubisco observed in this study most likely resulted due to the initiation of recovery phase 7 d after atrazine treatment.

Similar to Rubisco, photosynthetic efficiency in plants is highly dependent on PPDK activity [60] . A reduction in the activity of PPDK would cause less production of PEP for PEPC action, leading to reduced CO₂ fixation by Rubisco in the bundle sheath cells [59] . Nemat Alla et al. [59] observed a reduction in maize PPDK level within the first 2 d after treatment with ALS-inhibitors such as rimsulfuron and imazethapyr, followed by an increase in PPDK levels for the next 10 d. In our study, KsPPDK transcript levels were increased 1.8-fold at 24 h and then decreased (0.3-fold) 7 d after treatment with chlorsulfuron relative to the control (Figure 2).

In the current study, KsPPDK was found to be downregulated relative to the control at 8, 24 h, and 7 d after atrazine treatment (Figure 1(b)). Similarly, Nemat Alla et al. [59] reported a decreased PPDK activity in maize in the first two days after atrazine application and the decreasing trend continued for the next 10 d. PSII-inhibitors such as atrazine block the electron transport chain and the ATP production in the mesophyll cells, causing severe imbalance in the C4 photosynthetic pathway.

The transcript levels of KsPPDK were relatively unaffected by glyphosate treatment (Figure 3). The regulation of carbon flux is dependent on the availability of precursors E4P (D-erythrose 4-phosphate) and PEP for the enzyme 3-Deoxy-D-arabino-heptulosonate7-phosphate (DAHP) synthase, which catalyzes the first committed step of the shikimate pathway [61] . The conversion of pyruvate to PEP by plastidic PPDK is one of the several pathways that maintain a constant supply of PEP for the shikimate pathway to function [62] [63] . The requirement for PEP should be minimum when shikimate pathway is inhibited by glyphosate, and as a result PPDK activity may not be crucial.

KsCAB showed downregulation relative to control at 8 h after atrazine treatment (Figure 1(b)). Similarly, expression of CAB gene in soybean was reported to be downregulated at 1, 2, 4, and 8 HAT with atrazine [64] . Expression of CAB is influenced by several abiotic stress factors such as salinity, metal ions, and ultraviolet radiation [65] [66] , and is known to be active during the production ROS in chloroplasts [67] [68] .

KsCAB gene expression remained unchanged at 8 and 24 h, and downregulated (0.3-fold relative to control) at 7 d after treatment with chlorsulfuron (Figure 2). The cause for the downregulation of KsCAB at 7 d after chlorsulfuron treatment is unknown. Similar downregulation of KsCAB was observed at 24 h and maintained until 7 d after glyphosate treatment (Figure 3). Jiang et al. [27] reported a downregulation of CAB genes in soybean at 24 and 72 HAT with glyphosate. The observed downregulation of KsCAB after glyphosate application may be due to inhibition of glutamine synthetase 2 (GS2), a key enzyme for the synthesis of precursor glutamine for chlorophyll biosynthesis via the C5 pathway [27] [69] [70] .

Leaf chlorophyll index is an important parameter for evaluating general plant health [19] . In the current study, we observed no change (compared to untreated control) in the leaf chlorophyll index of kochia during the one week after treatment with three SOA herbicides (Figure 4(a)). Previous studies in common lambsquarters (Chenopodium album L.) and in some other weed species have shown no change in the leaf chlorophyll index in the first week after herbicide treatment [71] [72] .

Chlorophyll fluorescence (F_v/F_m) indicates the photosynthetic capability and the quantum efficiency of the PSII, which is highly sensitive to stress [20] . In this study, significant decrease in the chlorophyll fluorescence in kochia was observed after atrazine treatment (Figure 4(b)). A decrease in F_v/F_m values suggests an increase in dissipation of energy as heat and inhibition of photosynthetic machinery [73] . Atrazine, being a PSII inhibitor, has a more direct effect on the efficiency of PSII compared to other SOA herbicides. The blockage of the electron transport chain by the atrazine treatment would eventually affect the functioning of the PSII with corresponding decrease in F_v/F_m values.

In summary, this study clearly demonstrates differential expression of target-site and photosynthetic genes in response to herbicide treatments in kochia. Herbicide target-site genes (KsALS, KspsbA, and KsEPSPS) were readily upregulated after herbicide treatment. It is important to realize that herbicide application, directly or indirectly, not only influences the targeted pathway, but also other critical pathways such as Calvin and C4 cycles leading to tissue or plant death. In future, a more elaborate gene expression study (high-throughput) using RNA sequencing tools is essential in order to elucidate the expression levels of not only photosynthetic genes after herbicide application, but also genes related to plant defense and stress response.