The importance of being duplicated in the tomato sHSP gene family

KRSTICEVIC, F.J.; PRATTA, G.R.; ARCE, D.P.; CACCHIARELLI, P.; TAPIA, E.; SPETALE, F.; PONCE, S.D,

Cali

Congreso; IV Congreso Colombiano de Bioinformática y Biología Computacional y VIII Conferencia Iberoamericana de Bioinformática; 2017

Sociedad Iberoamericana de Bioinformática

In living organisms, small heat shock proteins (sHSPs) are triggered in response to stress situations. Evolutionary relationships in sHSP genes have been investigated in rice, potato and tomato genomes highlighting the importance of tandem duplication events for the expansion of sHSP gene family in plants. Additionally, transcriptomic studies have revealed complex patterns of expression for these genes during fruit ripening. This work aims to shed light on the way the expression patterns of sHSPs genes get affected by variations in their nearby promoter architecture by taking advantage of their redundant presence. HSEs indeed appeared in the regulatory region of most up-regulated sHSPs genes during tomato fruit ripening, however some striking exceptions are observed. Conversely, regulatory regions of down-regulated, not-differentially expressed or not-expressed sHSP genes are almost depleted of HSEs. Concerning the exceptions, we note Solyc09g015000, arising from tandem duplication event in chromosome 9. This uncharacterized sHSP gene is strongly up-regulated during fruit ripening (and heat shock) but is completely depleted from HSEs. Instead, a unique short CTAGA non-HSE motif, resembling a degenerated HSE sequence, is present as confirmed by similar results in orthologous promoters from S. tuberosum and S. pennelli. Preliminary results from predicted protein-protein interactions involving the sHSP protein family and heat shock response (HSR) key factors (HSP70s, HSP90 and MBF1c) suggest that Solyc09g015000 could be mediating the stress response through a regulatory network of the ring type involving four sHSP proteins, two of them of tandem origin, and one uncharacterized protein with an ATPase domain typical of the HSP100 family. Aiming at a precise characterization of HSEs in the promoter region of tandem duplicated sHSP genes in tomato, a position weight matrix (PWM) approach is considered. For this purpose, a careful data-curation process over the gene expression data was performed. This curation process includes selection of best suitable algorithms for PWM modeling of HSEs, together with stringent PWM quality controls. Resulting PWMs were used to scan the promoter region of the 33 members in the sHSP gene family of S. lycopersicum (cv Heinz 1706) with main focus on the five physical clusters of tandem duplicated sHSPs genes. Additionally, PPIs analysis for predicting protein-protein interactions in sets of co-regulated genes during tomato fruit ripening is considered. Jointly analyses of these results contribute to improve the sparse annotation of the sHSP gene family and interactors in tomato. Gene promoter architecture characterization of sHSP gene family in S. lycopersicum cv Heinz 1706 can shed light on its functional annotation. A pattern-oriented screening of transcription factor binding sites (TFBS) families against the regulatory region of sHSP genes showed the over representation of HSE, ERE, ABA and WRKY TFBS families. To avoid false positives, only over represented TFBS of minimum size 5 nt were considered. As a result, only the HSE, ERE and ABA TFBS families remained for downstream statistical analysis with QC PWMs (p-value=1E-04) in the regulatory region of the 33 members of sHSP gene family. HSEs indeed appear in the regulatory region of most sHSPs up-regulated genes during tomato fruit ripening. Conversely, regulatory regions of down-regulated, not-differentially expressed or not-expressed sHSP genes during fruit ripening were almost depleted of HSEs. This trend on HSEs, increasing their presence for up-regulated members while depleting it for the others, holds also tandem duplicated sHSP genes. Disentangling the actual effect of the HSE presence in tandem duplicated genes may help to better understand the functionality of ubiquitous sHSP genes. In this regard, HSEs are observed for almost all up-regulated tandem duplicated sHSP genes. In particular, two physical clusters of sHSP genes, one located in chromosome 6 (Solyc06g076520, Solyc06g076540, Solyc06g076560 and Solyc06g076570) and the other in chromosome 8 (Solyc08g062340 and Solyc08g062450), show the HSE presence. Also, HSEs heavily populates the unique up-regulated member (Solyc08g078700) in the remaining physical cluster of three sHSP genes located in chromosome 8. For the other two nearby members of this cluster, HSEs are just present (Solyc08g078720) or absent (Solyc08g078710), consistently with their observed expression profiles. An exception to the apparent rule of HSE presence and sHSP up-regulation is observed in the physical cluster of two sHSP genes in chromosome 9. In this case, both sHSP genes are up-regulated but just one of them (Solyc09g015020) shows the HSE presence while the other (Solyc09g015000) seems to be completely depleted from HSEs. Finally, a heterogeneous expression pattern of tandem duplicated sHSPs genes in a phylogenetic cluster (not a physical one) with an expected HSE presence was detected. HSE presence exception to sHSP upregulation during stress, including fruit ripening, rises the question about its actual importance for triggering sHSP expression. In this regard, the study of PPIs may shed light on this issue. We wonder if Solyc09g015000 may be behave like MBF1, known to be a key factor during thermoregulation in Arabidopsis despite the HSE absence in its promoter region, PPI analysis from up-regulated genes during fruit ripening and HS was conducted. Up-regulated genes include those in the HSP90 and HSP70 families known to contribute in maintaining cellular homeostasis during stress conditions in human, Arabidopsis and tomato. HSP90 constitutes approximately 1-2 % of the total protein content in eukaryotes suggesting complex interconnections with other key regulators and cochaperones in response to stress. Moreover, HSP70 and HSP100 participate with HSP90 during the proteome response to stress also contributing to maintain cellular homeostasis in physiological and stress conditions. Despite small variations in their order of appearance, HSP90, HSP70 and HSP100 proteins are the first barrier or fence in response to stress conditions in different organisms. This observation motivates a computational strategy for uncovering the functionality of uncharacterized sHSPs in tomato, including the striking Solyc09g015000. When removing characterized HSP90, HSP70 and HSP100 proteins, a ring network with the same five nodes emerged in both ripening and HS. In this scenario, Solyc09g015000 mediates between annotated Solyc03g082420, belonging to the ring, and the well characterized sHSP Solyc05g014280, outside the ring. GO MF and CC computational analysis of Solyc03g082420 and Solyc05g014280 suggests that Solyc09g015000 is a cytosolic protein probably which binds to ribosomes or RNA contributing to the first barrier or fence in response to stress. In silico promoter analysis of the sparsely characterized sHSP gene family in tomato was performed. Quality controlled PWMs were estimated from curated transcriptome and microarray datasets. PWMs revealed unusual promoter architectures in certain uncharacterized tandem duplicated sHSPs largely reported to be associated with fruit ripening. Aiming to disentangle the functionality of these genes, we advance into the study of PPIs during different type of stresses. To uncover the actual functionality of target sHSP genes, PPI networks were analyzed in terms of stress pathways. Hence, well-known upstream, first-barrier, HSPs involved in stress responses, including HSP 100s?, HSP90s? and HSP70s?, were removed from PPI networks. As a result, background PPI networks involving just sHSPs and complimentary uncharacterized proteins were obtained. Despite the stress type, a striking ring arrangement involving four sHSPs and one HSP100-like protein emerged. Among the four sHSPs, two of them came from tandem duplication events. These newly discovered subnetworks, mostly built from sHSPs, may serve as backup stress response networks once upstream HSPs (90-70) have been defeated.