Tissue specific expression profile of Mediator genes in Arabidopsis

Abstract

Mediator is a gigantic multiprotein complex required for transcription of almost all the protein coding genes. In this report, we have analyzed the transcript abundance of 31 Med genes in different tissues of Arabidopsis. Our analysis revealed the tissue specific differential expression profile of many Med subunit genes suggesting they might be contributing in the formation, maturation or function of that specific tissue. Moreover, we also found increase or decrease in the expression level of several Med subunits during the same duration of specific processes (for example flowering) indicating probable enrichment of a particular arrangement of selected subunits during that process. Thus, this study suggests that not only specific Med subunits have functional relevance in specific processes, but specific arrangements of Med subunits might also play significant role in some processes in Arabidopsis or other plants.

Mediator was discovered in yeast as a necessary component for transcription of a protein coding gene.^1,2 Later on, Mediator was purified and characterized to be a gigantic multiprotein complex required by RNA polymerase II for the transcription of its target genes.^3-5 It was found to play important role in facilitating the assembly of the preinitiation complex.^6-8 Genetic, biochemical and bioinformatic analyses revealed its existence in all the eukaryotes ranging from simple unicellular yeast to complex multicellular mammals and plants.^8-11 In yeast, there are 25 subunits which make Mediator, whereas in animals and plants the number is higher.^11,12 Biochemical and biophysical studies in fungi and metazoans confirmed the modular structure of this complex, wherein the subunits compose four different modules namely Head, Middle, Tail and Kinase.^5,8,13-15 Head, Middle and Tail constitute the core part of the complex, whereas the Kinase module reversibly associates with the core part. Subunits constituting the Head and Middle modules are found to interact with RNA pol II and components of transcriptional machinery, whereas subunits of Tail module establish contact with diverse transcription factors.^13,16-18 Thus, Mediator provides an interface to relay the regulatory signals from the transcriptional regulators to the transcriptional machinery. The subunits of Kinase module can interact with the subunits of the Head and Middle modules, and so affect the interaction between the complex and the RNA pol II. That is why in many cases, presence of Kinase module in the Mediator was found to be associated with the repression of the gene expression.¹⁹

Within modules, the subunits may form sub-modules of more compact structures. For instance, subunits of Middle module Med9, Med4 and Med7 form a submodule, whereas Med16, Med23 and Med24 were found to form a tight-knit triad in the Tail module.^14,20,21 Thus, presence of higher or substoichiometric concentration of some selected subunits can have significant effect on the overall composition and function of the complex. Earlier, we reported hormones and abiotic stresses affecting transcript abundance of specific Mediator subunit genes in Arabidopsis seedlings.²² In this study, in order to understand the role of specific Med subunits in different organs of a plant, we have extended the expression analysis of Med genes across different vegetative and reproductive tissues of Arabidopsis.

Arabidopsis seedlings and plants were grown at 22°C in 16 h light and 8 h dark cycle. Root and TR leaves (leaves from 3 rosette leaves stage) were harvested from 15 d old seedlings grown in MS medium, whereas CR leaves (rosette leaves from rosette growth completed plants) and stems were harvested from 30 d old plants grown on the mixture of soil and vermiculite. Flowers from plants of two different ages were taken; first open flowers (FOF) were obtained from 32 d old plants in which bolting had just happened, whereas other set of open flowers (FC) were harvested from 45 d old plants in which more than 50% flowers were open and no new flower seemed to be appearing anymore. Siliques were harvested from 45 d old plants. We used quantitative real-time RT-PCR to analyze the expression level of Med genes in these tissues. Primers used in this study were taken from earlier report.²² Readings with ACT1 (At1g13440) and GAPC2 (At3g18780) gene primers were used for normalization. For each experiment, two controls were used to cross-check the data and enhance the reliability of the data.

We analyzed the transcript level of 31 AtMed genes in different tissues of Arabidopsis. For comparative analysis we used transcript abundance of AtMed2 as reference (Fig. 1). In all the tissues selected for this study, transcript levels of half or more of the AtMed genes used in this study were found to be expressed at similar level; within range of 0.5 to 1.5 fold of expression level of AtMed2 (Table S1). This goes well with the fact that Med subunits work together as a component of the Mediator complex. However, we did notice fluctuation in the expression level of some subunits in specific tissues. In root, AtMed36 was found to be highly expressed (7.0 times of AtMed2; Table S1), and expression level of AtMed11, AtMed20, AtMed7, AtMed21, AtCycC and AtMed37 were moderately higher (Fig. 1A; Table S1). On the other hand, AtMed8, AtMed17, AtMed31, AtMed14, AtMed15, AtMed23, AtMed12, AtMed13, AtMed25, AtMed34 and AtMed35 were expressed at very low level in Arabidopsis root. In stem, transcript level of AtMed37 (7.6 fold of AtMed2 expression) was found to be much higher than any other Med genes tested in this study (Fig. 1B; Table S1). Additionally, AtMed11 from Head, AtMed7 from Middle and AtCdk8 and AtCycC from Kinase module were also found to be expressed at higher level. We found that some subunits like AtMed17, AtMed18, AtMed31, AtMed14, AtMed15, AtMed12, AtMed25, AtMed34, AtMed35 and AtMed36 are expressed at lower level in stem. In TR leaves from young plants, transcripts of AtMed7, AtCycC and AtMed37 were found to be more abundant, 3.6, 7.03 and 8.6 times of AtMed2, respectively (Fig. 1C; Table S1). However, transcript levels of AtMed17, AtMed10, AtMed31, AtMed14, AtMed15, AtMed23, AtMed12, AtMed13, AtMed25, AtMed34 and AtMed35 were much less in this tissue. In CR leaves, relative expression pattern of few AtMed genes changed. The transcript level of AtMed11, AtMed18, AtMed4, AtMed7, AtMed21, AtMed3, AtMed15, AtCdk8, AtCycC and AtMed37 were found to be moderately higher (Fig. 1D). Thus, relative expression level of AtMed18, AtMed3 and AtMed15 increased in leaves during transition from 3 rosette leaf to rosette completion (Fig. 1C and D). Similar change in expression level of AtCdk8 was observed. On the other hand, expression of AtMed37 showed the reverse trend in these tissues. In FOF flowers, many AtMed genes (such as AtMed8, AtMed17, AtMed18, AtMed19, AtMed22, AtMed9, AtMed10, AtMed31, AtMed5, AtMed14, AtMed16, AtMed23, AtMed12, AtMed13, AtMed25, AtMed34 and AtMed36) were found to be expressed at very low level (Fig. 1E). The pattern did not change much in FC flowers from older plants, though expression levels of AtMed7 and AtMed37 increased significantly (Fig. 1F). Thus, during maturation of Arabidopsis flower, expression of AtMed7 and AtMed37 increased. In siliques, expression level of AtMed37 was found to be much more than any other AtMed genes tested (6.53 times of AtMed2 expression; Fig.  1G; Table S1). Comparison of expression level of each AtMed gene with that of AtMed2 in different tissues is provided in Table S1. In all, plant specific AtMed37 was found to be most abundant Med transcript in all the Arabidopsis tissues we tested. It will be interesting to study the role of Med37 in plants. On the other hand, in comparison to other Med subunit genes, Med17, Med31, Med14, Med12, Med34 and Med35 were found to be expressed at very low level in comparison to other Med genes in all the Arabidopsis tissues we used in this study.

Figure 1. Quantitative real-time RT-PCR analysis of transcript abundance of Med genes in root (A), stem (B), TR leaf (C), CR leaf (D), FOF flower (E), FC flower (F) and silique (G) of Arabidopsis. The relative expression level of AtMed2 was used as reference and is shown in white bar. Mean values from three independent experiments are shown. Each independent experiment was performed with three technical replicates. Error bars represent standard deviations. Grey, white, black shaded and white shaded bars represent AtMed genes coding for subunits of Head, Middle, Tail and Kinase modules, respectively. Bars representing subunits whose positions are not known, are highlighted with border line. For statistical analysis, one way ANOVA test was performed to determine the significance of difference. * and ** indicate significant difference at p < 0.05 and p < 0.01, respectively.

In order to get comparative idea of differences in transcript abundance of Med genes in different organs, we used expression level in stem as the reference (Table S2). We found that the expression levels of most of the Med genes in different organs tested in this study were similar but few, which are discussed here (Fig. 2). Among the genes coding for subunits of Head module, expression of AtMed18 and AtMed22 showed some variation. Transcript level of AtMed18 was found to be significantly more in CR leaves, FC flowers and siliques (Fig. 2A). AtMed22 was found to be expressed more in FC flowers and siliques. On the other hand, some Head Med subunits of Arabidopsis were found to be expressed at very low level in some organs. In the leaves of 30 d old Arabidopsis plants, expression of AtMed6, AtMed17 and AtMed22 were found to be low. In flowers of 32 d old plants, transcript abundance of AtMed8 and AtMed17 were reduced. Among the genes coding for subunits of Middle module, AtMed9 was found to be expressed at higher level in siliques whereas AtMed21 transcript was more abundant in FOF flowers (Fig. 2B). Although expression of AtMed31 was found to be very low in all the tissues tested in this study, its level was significantly low in root and CR leaves (Fig. 2B). Among the genes coding for subunits of Tail module, AtMed3 was found to be expressed at higher level in leaves, flowers and siliques (Fig. 2C). On the other hand, transcript level of AtMed5 and AtMed23 was found to be low in these tissues. The most noticeable variation in the expression level was observed in the case of AtMed15. Its expression was quite prominent in leaves (CR) of older Arabidopsis plants (Fig. 2C), suggesting its important role in later part of plant growth and development. Moreover, its expression in the flowers of the young plants was higher, which reduced to less than half in the flowers of older plants. Similar pattern of expression was noticed for AtMed14 in flowers. There was not much variation in the expression level of Kinase module Med subunit genes in these tissues (Fig. 2D). Among the genes coding for Med subunits whose positions are not yet known, Med36 showed significantly higher expression level in Arabiodopsis roots (4.38 times of expression level in stem; Table S2). This suggests that AtMed36 might be playing very important role in root (Fig. 2E). Expression of Med34 was found to be higher in flowers of older Arabidopsis plants.

Figure 2. Quantitative real-time RT-PCR analysis of transcript abundance of Med genes coding for subunits of Head (A), Middle (B), Tail (C) and Kinase (D) modules in different tissues of Arabidopsis. Expression levels of AtMed genes coding for subunits whose positions are not known are shown in (E). The relative expression level in stem was used as reference. Mean values from three independent experiments are shown. For each experiment, three technical replicates were used. Error bars represent standard deviations. For statistical analysis, one way ANOVA test was performed to determine the significance of difference. * and ** indicate significant difference at p < 0.05 and p < 0.01, respectively.

Next, we compared the transcript levels of AtMed genes in leaves from Arabidopsis plants of different ages (Table S3). In comparison to 15 d old seedlings, leaves of 30 d old plants were found to express more of AtMed15 and AtMed18 (Fig. 3A, upper panel; Table S3). This suggests probable role of these two subunits in maturation of Arabidopsis leaves. However, a large set of Med genes (16 Med genes representing subunits of all the four modules) were found to be downregulated in the leaves during later stages of growth (Fig. 3A, middle and lower panels). Thus, it seems that when leaf is fully matured and its cells are completely differentiated, expression of Med genes is reduced. This is just like the status of Med subunits in differentiated cells of metazoans.²³ We also compared expression of Med genes in flowers from young plant (32 d old) to that in old plants (45 d) (Table S4). In the maturing flowers, expression of AtMed37, AtMed18 and AtMed22 were significantly upregulated (Fig. 3B, upper panel). It will be interesting to investigate if they have some specific roles to play. Only few Med genes (AtMed15, AtMed14, AtCycC, AtCdk8, AtMed4 and AtMed2) were found to be significantly downregulated in the Arabidopsis flower during this period (Fig. 3B, lower panel). Flowers are never an inert organ. At the later stage of maturation, flowers are primed to produce seeds, and so there are enough cells actively carrying the process of transcription in abundance.

Figure 3. Quantitative real-time RT-PCR analysis showing change in the transcript level of Med genes in the leaves (A) and flowers (B) of Arabidopsis plants of different ages. For TR, CR, FOF and FC, please see the text. Mean values from three independent experiments are shown. For each experiment, three technical replicates were used. Error bars represent standard deviations. For statistical analysis, one way ANOVA test was performed to determine the significance of difference. * and ** indicate significant difference at p < 0.05 and p < 0.01, respectively.

Considering the function of Mediator complex in basic process of RNA pol II mediated transcription, we expected the expression level of each subunit to be more or less similar. However, contrary to this, we found that not all the subunits were expressed at similar level in any given tissue, suggesting a possibility of existence of different Mediator complexes in different tissues or cell types. Also, there were noticeable variations in the expression level of many Med genes across different tissues of Arabidopsis, suggesting a possibility of tissue or cell type specific functions of these genes. Comparing the expression level of AtMed genes in the tissues like leaves or flowers from Arabidopsis plants of different ages gives an indication that some Med genes might be playing important role during development and maturation of plant organs. Increase and decrease in the level of several Med subunits during the same time frame of a process (say flower maturation) suggest probable enrichment of a particular arrangement of selected subunits during that process. It will be interesting to see if transcript profiles of AtMed genes match/differ with their protein profiles.

Glossary

Abbreviations:

Med

Mediator

AtMed

Arabidopsis thaliana Mediator

TR leaves

leaves from 3 rosette leaves stage

CR leaves

rosette leaves from rosette growth completed plants

FOF

first open flower

FC flowers

flowers of plants whose flowering is completed and 50% flowers are open

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.