The complexity of signal transduction in higher plants is a result of spatial and temporal signal separation, cross talk between signaling networks, and extensive genetic redundancy. As a result, classical forward genetic screens have proven ineffective at elucidating the signaling cascades responsible for plant responses to many different external stimuli. While the biochemical mechanisms for metal detoxification are well understood, the transcriptional regulation is still largely unknown. Recently, point mutants in a gene called sulfur limitation1were identified that play an important role in the regulation of the sulfate assimilation pathway under sulfur limiting conditions . SLIM1 is a central transcriptional regulator that is reported to activate the SULTR1;2 high affinity sulfate transporter in Arabidopsis under -S conditions. The transcript levels of SULTR1;2 are nearly abolished under –S conditions in the slim1-1 and slim1-2 mutants. Metabolite accumulation was also affected in the SLIM1 mutants, where slim1-1 and slim1-2 show a significant decrease in GSH levels compared to wild type when grown on low sulfur media. These findings indicate the importance of SLIM1 in sulfur metabolism. The transcriptional regulation of heavy metal-induced gene expression remains largely unknown in plants. While SLIM1 is a central regulator of the sulfate assimilation pathway, it is unclear if SLIM1 plays a role in activation of sulfate assimilation during metal detoxification. The complex regulation of sulfate assimilation and the unknown role of SLIM1 in metal detoxification raises the question whether SLIM1 acts together with other yet unknown transcription factors to be part of a large regulatory network . However, the large number of transcription factor families and the potential partially unequal redundancy of certain family members represent a challenge. Recently, a small library became available which addresses genetic redundancy in Arabidopsis using artificial micro-RNAs.

Targeting multiple homologous transcription factors allowed for screening of large groups of genes,metal greenhouse benches while reducing or eliminating functional overlap. Unfortunately, the initial screen was unable to narrow down potential transcription factors involved in metal detoxification. A larger library of amiRNA lines is now available that may allow us to gain a deeper understanding of specific transcriptional activators and repressors involved in metal detoxification. In this dissertation, we present work aimed at uncovering the genes and mechanisms underlying the cadmium and arsenic transcriptional responses using a variety of approaches. To identifying transcription factors that mediate Cd/As-induced gene expression in plants, we performed genome wide yeast one-hybrid analyses on the promoter fragments of four Cd/As-induced genes, specifically the SULTR1;2, APR2, γ-ECS, and GS promoters. Our goal was to identify the unknown master transcriptional regulators and repressors that control the promoters of multiple Cd/As-induced genes. We cloned the 2.1 kb promoter region previously characterized for SULTR1;2 as well as a 2.1 kb promoter fragment from APR2. The promoter fragments were shorter for g-ECS and GS due to the presence of upstream genes. The g-ECS promoter was approximately 1.4 kb in length, while the GS promoter was 800 bp. We used these promoter fragments to screen approximately 2,100 Arabidopsis transcription factors for DNA-protein interactions in a yeast one-hybrid assay. The GS, γ- ECS, and APR2 promoters were screened using a previously published transformation based protocol. The SULTR1;2 promoter was screened using a mating-based protocol. Because we had little experience with the new protocol, we screened the SULTR1;2 promoter 3 times to determine the level of reproducibility of this method. Figure 1 shows a heat map of the β-gal values determined from each of the 6 screens we performed. In general, the mating-based protocol identified more putative DNA-protein interactions; however, there was a low degree of reproducibility between the 3 replicate experiments. For the 3 transformation-based screens, a smaller number of putative interactions were identified. Interestingly, a number of transcription factors were identified in more than one screen. For simplicity, we averaged the β-gal values between the 3 SULTR1;2 screens and compared with the β-gal values obtained from the remaining screens. We compared the results in two different ways. First, we considered interactions having z-scores ≥ 2.0, which is traditionally used in yeast one-hybrid analyses. Secondly, we performed outlier detection analysis using an R script. Figure 2 shows Venn diagrams indicating the number of interactions identified and the overlap between the screens using both the zscore method and the outlier detection method .

In general, the outlier detection algorithm identified more putative interactions and more overlapping interactions between the different promoters than the z-score method. However, in both cases the total number of putative interactions was unexpectedly high. To look more closely at the interactions identified by these screens, we decided to focus on the three interactions from each screen with the highest total β-gal values. Table 1 shows the 3 interactions with the highest β-gal values from each of the 4 yeast one hybrid screens along with the number of putative interactions with a β-gal value of 2.0 or greater. We obtained T-DNA insertion lines in these genes and performed root elongation assays on the individual mutant lines to determine their sensitivity to cadmium. Only one mutant, Salk_094437C, was identified as having a cadmium-dependent phenotype. We obtained an additional T-DNA disruption mutant in the At2G46270 gene. Our initial screen of these T-DNA mutants revealed that both independent alleles of ANAC102 as showing Cd tolerance . No differences in growth were observed in the absence of Cd. The genome-wide yeast one-hybrid screen identified numerous putative DNA protein interactions. The three SULTR1;2 promoter screens showed a low level of reproducibility . However, there was significant overlap between several of the clusters. Interestingly, the remaining three screens showed far fewer interactions than the SULTR1;2 promoter screens. Our hypothesis that a limited number of transcriptional regulators are responsible for the co-regulation of these four cadmiuminducible genes was partially supported by the level of overlap identified between the four promoter screens . Unfortunately, the large number of putative overlapping interactions makes it difficult to determine which individual transcription factor or group of transcription factors may play a significant role in mediating early transcriptional responses to cadmium and arsenic. To further narrow down the most promising candidates, we also ranked the hits from the yeast one-hybrid screens and focused on the three hits with the highest Z-scores . Initial root-elongation screens using mutants in these 12 genes identified a single mutant in the ANAC102 gene having a long-root phenotype on cadmium . Further characterization showed that two different mutant alleles, anac102-1 and anac102-2 both had similar root elongation phenotypes on cadmium. Unfortunately, both of these mutants also exhibited long-roots when grown without cadmium, which suggests the phenotype is not cadmium or arsenic specific. These results suggest that additional screening of higher order mutants may be required to identify the genes that specifically mediate cadmium and arsenic responses in plants. Seeds were sterilized and plated on plates containing one-quarter strength Murashige and Skoog standard medium , 1 mM MES, 1% phytoagar ,rolling greenhouse benches and the pH adjusted to 5.6 with 1.0 M KOH.

Sterilized nylon mesh with a 200 µm pore size was placed on the surface of the media prior to sowing the seeds. The seeds were then acclimated with cold treatment at 4 o C for 48 hours, and grown under growth room conditions for five days . Seedlings were then transferred to quarter strength MS, 1 mM MES, 1% agar plates containing 20 µM CdCl2. For hydroponic cultures, plants were grown on ¼ MS plates for 6 weeks under growth chamber conditions as described above. Seedlings were then transferred from ¼ MS plates to polystyrene rafts floating on hydroponics solution. The composition of the hydroponics solution is as follows: KNO3 0.5mM, Ca2 0.25mM, MgSO4 1mM, KH2PO4 1mM, Fe-Na-EDTA 100μM, H3BO3 50μM, MnCl2 19μM, CuSO4 1μM, ZnCl2 10μM, MoO4Na2 0.02μM, pH 5.8. The nutrient solutions were replaced every seven days. For root growth analysis Arabidopsis seedlings were grown horizontally on one quarter-strength Murashige and Skoog basal medium , 1 mM MES, 1% agar for a period of four days. Seedlings and nylon mesh were then transferred to MS plates supplemented with 100 µM CdCl2 for an additional four days. Seedlings used for shoot growth analyses were grown horizontally under 16/8-h day/night period at 23°C. To measure metal accumulation, plants were grown under hydroponic conditions in liquid media until bolting stage . Shoots were rinsed 3 times in deionized water. For ICP-OES analyses of root tissue, roots were rinsed in deionized water, washed in 100 mL of 100 mM CaCl2 on an orbital shaker for 5 minutes at approximately 135 rpm , and then washed in 100 mL of deionized water on the orbital shaker for an additional 3 minutes. Root tissues were separated below the hydroponic sponge, approximately 3 cm below cotyledons. Both shoot and root tissues were dried at 60oC overnight. The dry weight was recorded, and then the tissues were digested in 70% trace metal grade nitric acid overnight. Samples were then boiled for 30 minutes to ensure complete digestion and diluted to a final concentration of 5% nitric acid with deionized water. To measure the metal accumulation of plate-grown plants, Arabidopsis seedlings were grown in minimal media for 4 days then exposed to 100 µM metal by transferring seedlings with nylon mesh to minimal media with metal. Plants were allowed to continue growing in arsenic-containing medium for 6, 12, 24 and 48 h. The plants were then washed briefly with distilled water twice. The plant seedlings were dried in a drying oven at 60°C overnight, the dry weight was measured, and the dry material was digested by boiling in concentrated nitric acid . These digests were then diluted with deionized water and the metal content of the digested samples was analyzed with a Varian Vista Inductively Coupled Plasma Atomic Emission Spectrometer . Total concentrations of metals were normalized based on the dry weight of the plant samples. Plants used for Inductively Coupled Plasma-Optical Emission Spectroscopy and Fluorescence HPLC analysis were grown under hydroponic conditions. Plants were grown under hydroponic conditions at 24o C under a 16/8-h day/night period and hydroponic media were replaced every 3-4 days. After reaching the bolting stage the hydroponic media was replaced with the same media to which 20 mM CdCl2 was added for a period of four days for cadmium accumulation analyses, whereas the hydroponic media containing 20 µM CdCl2 was added for a period of two days for induction of PCs. Shoot and root tissues were rinsed three times in deionized water for fluorescence HPLC coupled to mass spectrometry analyses. Thiols, including PCs were analyzed using the method described by Jobe et al. . To measure metal accumulation, plants were grown under hydroponic conditions in hydroponic media 2, 125 mM Fe-EDTA, 3.5 mM H3BO3, 1.1 mM MnCl2, 100 mM ZnSO4, 12.5 mM NaMoO4, 500 mM NaCl, 900 mM CoCl2 until the bolting stage. Cadmium treated plants grown under hydroponic conditions were first washed and then separated into root and shoot tissues. Shoot tissues were separated immediately below the cotyledons and above the hydroponic sponge. Root tissues were separated below the hydroponic sponge, approximately 3 cm below the cotyledons. Shoots were rinsed three times in deionized water. For ICP-OES analyses of root tissue, roots were rinsed in deionized water, washed in 100 mL of 100 mM CaCl2 on an orbital shaker for 5 minutes at approximately 135 rpm , and then washed in 100 mL of deionized water on the orbital shaker for an additional three minutes. Both shoot and root tissues were dried at 60oC overnight. The dry weight was recorded, and then the tissues were digested in 70% trace metal grade nitric acid overnight. Samples were then boiled for 30 minutes to ensure complete digestion and diluted to a final concentration of 5% nitric acid with deionized water. Pharmaceuticals have been increasingly prescribed for the past 30 y, and prescription rates have almost tripled in just the past 14 y . In 2013, animals grown for human consumption were treated with 9.1 × 106 kg of antibiotics; of those, 6.6 × 106 kg were used for the purpose of increasing production .