The annotation is arguably the

The annotation is arguably the most important part of analysis, as it enables one to evaluate and interpret the content of the transcriptome assembly. In this context, the non-redundant contigs were initially run in BLAST against the Nr database showing that 55% of them displayed high similarity to certain protein sequence. The Gene ontology (GO) (http://geneontology.org/) analysis classified the non-redundant contigs into a wide range of GO categories, suggesting that the contigs represent a wide MCC950 sodium sale of transcripts from V. carinata genome, further demonstrating a good assembly. To better understand functions and interactions, all annotated genes were mapped against the KEGG database for a pathway-based analysis. As a result, 22.9% of total non-redundant contigs were assigned to a KEGG pathway. This relatively low number of genes assigned to a pathway is likely the result of imperfect annotation caused by Blast2go, although genes were present in 144 different KEGG pathways. The over-representation of some categories in the GO analysis as well as genes in some KEGG pathways has also been seen in other leaf transcriptome annotations of bromeliads (Li et al., 2016; Ma et al., 2015; Palma-Silva et al., 2016). It was performed an additional COG functional analysis. Each categorized COG term represents an ancient conserved domain, but our results implied that there is only a small proportion (28%) of non-redundant contigs with annotation for COG categories. In Bromeliaceae family, the greater amount of endopeptidases they possess has stimulated the emergence of many studies (Barrett, 1999). The majority of them have focused in the biochemical characterization and potential biotechnological application of these enzymes, being the fruit bromelain, from A. comosus, the most studied cysteine protease. In this study, we decided to expand the research and analyze the expression pattern of CysPs in leaves and inflorescence from V. carinata. The RT-qPCR analysis showed significant differences for metacaspase, pyroglutamyl peptidase, legumain and papain-like CyPs gene expression in the leaf compared to the inflorescence. Furthermore, there were differential expression patterns seen when the expression was compared from the base to the apex of the leaf (Fig. 6). The leaves of Bromeliads represent a key tissue for the plant because they have special trichome-dependent capability of absorbing water and nutrients, which allow them to grow where little water is available (Benzing, 2000). Besides this, V. carinata is considered as a tank-type (epiphytic) bromeliad under the ecophysiological classification systems of Benzing (2000). The tanks act as a mechanism for impounding organic debris for decomposition (Smith, 1989) and also house a variety of organisms including detrivores and saprophytes which help to break down the leaf litter for nutrient uptake into the plant (Vanhoutte et al., 2017). In this sense, it is not surprising to obtain different expression patterns of CysPs in the base as compared to the leaf apex, given the previously described roles that they play inside the cell. They are involved in many biological events that include senescence (De Michele et al., 2009), mobilization of proteins during germination (Szewińska et al., 2016), programmed cell death developmentally regulated and induced by biotic and abiotic stress (Ahmad et al., 2012; Grudkowska and Zagdańska, 2004; He et al., 2008; Mase et al., 2012; McLellan et al., 2009). The family of papain-like peptidases is the most investigated CysPs group, with roles more often associated with fruit ripening (van der Hoorn, 2008). It was also hypothesized, and generally accepted, that CysPs can help in the process of nitrogen and nutrient acquisition – a process that happens in the bromelian tank, consequently justifying their differential expression pattern in different parts of the leaf. Many works have reported tissue specific expression of the CysPs evaluated in this study, such as metacaspase (Kwon and Hwang, 2013; Liu et al., 2016; Wang and Zhang, 2014), pyroglutamyl peptidase (Szewczuk and Kwiatkowska, 1970), legumain (Christoff et al., 2014; Santana et al., 2016) and papain-like proteases (Martinez et al., 2009; Shi and Xu, 2009).