Masarykova univerzita přírodovědecká fakulta diplomová práce brno 2017 Veronika Krmeská masarykova univerzita

2. del11q

3. trisomy 12

4. normal (no cytogenetic aberration)

5. del13q14 (the best prognosis, the longest overall survival)


Figure 2. Overall survival in each of Dohner categories (Dohner et al., 2000).

1.1.3. Microenvironment

It is not just intracellular aberrations that are responsible for disease progression or resistance to the therapy. According to in vitro observations it was hypothesized that tissue microenvironment has an important role in CLL pathogenesis. In in vitro culture, CLL cells quickly undergo apoptosis, but this is changed when they get in contact with stromal cells or when soluble factors are added (Burger et al., 2009).

CLL cells come in contact with stromal cells through their chemokine receptor CXCR4. Stromal cells produce CXCL12 (SDF-1) ligand and attract B-cells by chemotaxis. Cells then migrate beneath stromal cell layer in process called pseudoemperipolesis. CXCR4 is then internalized and downregulated (Burger et al., 1999).

When gene expression profiles of CLL cells from different compartments (peripheral blood, bone marrow, lymph nodes) were compared, Herishanu et al. (2010) found that CLL cells from lymph nodes displayed upregulated BCR signalling and NF-κB activation.

It is also hypothesized that microenvironment co-evolves with the CLL-cells and because of that it represents a potential target for therapy (Shain et al., 2015).

1.2. Long non-coding RNAs

Amount of DNA in human genome that encodes proteins is approximately 1,5 %, but transcribed part of genome is 75 %. This transcribed part that doesn´t encode any protein contains non-coding RNAs (ncRNAs).

As mentioned above, miRNAs already proved their importance in cell processes. Besides the miRNAs and other short non-conding RNAs belong to to the group of ncRNAs also long non-coding RNAs.

LncRNAs are defined by length more than 200 nt and by the lack of any coding potential. Most of them are transcribed by Polymerase II and capped and polyadenylated like mRNAs, although the main difference is that lncRNAs possess higher-order structure, which is in tight relationship with their function.

They can be subdivided into many groups according to different features they carry. To filter important ones, we will look at their position in genome in relation to the nearest coding gene and their fucntion.

Depending on genomic position, lncRNAs can be either sense, antisense, intergenic, intragenic or bidirectional (figure 3). Here, bidirectional means that lncRNA is located on the opposite strand of the protein coding gene whose transcription is initiated less than 1000 base pairs (bp) away.

Figure 3. Different types of lncRNA (orange) according to the nearest coding gene (blue); (downloaded 15.10.2016 from http://www.wayenbio.com/jinqidongtai-202461-16698-item-45824.html).
1.2.1. Function of lncRNAs

LncRNAs affect many cellular processes from cell cycle through transcription and translation to imprinting and many more. Because of that, to be more accurate, we need to differentiate between transcriptional and posttranscriptional regulation (figure 4; Yang et al., 2014).

1. Transcriptional function

1. 
   Activator - or "a signal" lncRNAs in this group can either act themselves as transcription factors (TFs) or they can activate other TFs and induce transcription.
   
2. 
   Repressor - or "a decoy". Their function is exactly the opposite as in the previous group. They bind to a TF and disrupt TF-DNA interaction, so the transcription stops. E.g. GAS5 (growth-arrest-specific 5) binds to glucocorticoid receptor (GR) and prevents expression of GR-induced genes (Kino et al., 2010).
   
3. 
   Guide - lncRNAs recruit chromatin-modifying enzymes to target genes either in cis (on neighboring genes) or trans (distant genes) manner. E.g. XIST (X-inactive specific transcript) recruits PRC2 (polycomb repressive complex 2) and represses expression of genes on future inactive X cromosome (Engreitz et al., 2013).
   
4. 
   Scaffold - lncRNA brings together different molecules (RNAs, proteins) and helps to form ribonucleoprotein complexes. E.g. lncRNA HOTAIR acts as a bridge between PRC2 and LSD1/CoREST demethylase complex (Tsai et al., 2010).
   

2. Postranscriptional function

5. 
   RNA editing regulator - lncRNA in antisense orientation can bind its pre-mRNA and recruit enzymes that can affect RNA structure and coding potential. E.g. several lncRNAs and ADAR protein (Yang et al., 2013).
   
6. 
   RNA splicing regulator - lncRNA interacts with splicing factors. E.g. MALAT1 (metastasis associated in lung adenocarcinoma transcript 1) and serine/arginine (SR) splicing factors (Tripathi et al., 2010).
   
7. 
   Small RNA harbor - lncRNA can contain miRNA or other small ncRNA hairpin in their structure. E.g. MALAT1 contains small tRNA-like (transfer RNA-like) molecule called mascRNA (MALAT1-associated small cytoplasmic RNA) (Wilusz et al., 2008).
   
8. 
   MiRNA sequester - lncRNA can bind miRNAs and prevent their interaction with mRNA. E.g. lncRNA PTENP1 sequesters miR-20 which is supposed to target PTEN (Poliseno et al., 2010).
   
9. 
   MiRNA blocker - in this case lncRNA competes with miRNA for the same binding site on mRNA.
   
10. 
    RNA degradation regulator - lncRNA binds directly to mRNA and can mediate its degradation. E.g. 1/2-sbsRNAs and Staufen protein (Gong et al., 2011).
    
11. 
    Translational efficiency regulator - when bounded together, lncRNA can modulate (positively or negatively) efficiency of translation of mRNA.
    

Figure 4. Functions of lncRNAs (Yang et al., 2014).
1.2.2. CLL and selection of the profiled lncRNAs

In this section, we are going to look at several lncRNAs that were chosen to be explored more in our CLL cohort (see Results). These lncRNAs were chosen according to their relationship either to protein p53 or members of the Bcl-2 protein family.

1.2.2.1. XIST

XIST is probably one from the oldest discovered lncRNAs. This 17 kb transcript was first time described in 1991 (Brown et al., 1992).

It has a key role in X chromosome inactivation in female mammals (Penny et al., 1996). It is transcribed from the future inactive chromosome (Xi) from XIC locus (X-inactivating centre). After transcription, XIST binds PRC2 complex and together they spread inactivation of one of the X chromosomes by K27H3 trimethylation.

XIST was chosen for our further research based on the study, where authors deleted XIST gene in blood compartment of female mice (Yildirim et al., 2013). As a result, they observed Xi reactivation but, more importanly, hyperproliferation of all hematopoietic lineages and splenomegaly.

We hypothesized that XIST could explain slight inequality in CLL gender distribution (males:females = 2:1).
1.2.2.2. GAS5

LncRNA GAS5 located on chromosome 1q25.1 plays role in sensitizing cells to apoptosis upon nutrient starvation through GR (Kino et al., 2010).

GR is located in the cytoplasm and normally, when it binds to its ligand, travels to nucleus and binds to glucocorticoid response elements (GRE) through DNA-binding domain (DBD). There, it attracts coactivators and induces transcription of glucocorticoid-responsive genes (Chrousos et al., 2005).

GAS5 accumulates in cytoplasm upon cellular stress such as limited nutrients and growth factors. There it serves as a decoy and binds to DBD of GR, therefore GR cannot activate its target genes, which includes inhibitors of caspases like cIAP2 (figure 5).

Figure 5. GAS5 function (Li et al., 2016).

The choice of this lncRNA was based on the study from Isin et al. (2014) where the aim was the investigation of circulating lncRNAs in B-cell neoplasms. They selected 5 lncRNAs (also TUG1, MALAT1 and lincRNA-p21 - see below) and found out significant decrease of GAS5 in patients with multiple myeloma (MM).

1.2.2.3. TUG1

Taurine upregulated gene 1 (TUG1) located at chromosome 22q12.2 was originally discovered during genomic screen in developing mouse retinal cells treated with taurine, where it is necassary for formation of photoreceptors (Young et al., 2005).

TUG1 was found downregulated in human non-small cell lung cancer, where it acts as a tumor suppressor, since after inhibiting TUG1, cells started actively proliferating. It binds PRC2 and is able to modulate homeobox B7 expression, but more importantly, TUG1 expression is dependent on p53 (Zhang et al., 2014).

Finally, selection of TUG1 was affected by Isin et al. (2014) study, where they found significant correlation between plasma lncRNA levels and MM and CLL clinical stage.

1.2.2.4. H19

H19 is imprinted long non-coding transcript on locus 11p15.5. It is expressed from maternal alelle and regulated through imprinting control regions (ICR). When CTCF protein (CCCTC-binding factor) binds to ICR, it activates expression of H19, but when ICR is methylated and CTCF cannot bind to it, expression of Igf2 (insulin-like growth factor 2) is induced (figure 6).

Figure 6. H19/Igf locus (Wallace et al., 2007).
H19 is expressed prevalently during the embryogenesis and after the birth its expression is shut off in most of the tissues, although it was found to be reactivated during tissue regeneration and tumorigenesis (Lottin et al., 2002).

Speaking about tumorigenesis, it is not completely clear if H19 plays a role of oncogene or tumor suppressor, since it was found to be upregulated in solid cancer tissues such as in liver, colon or breast carcinomas (Fellig et al., 2005). But in hematopoietic disorders is seems that H19 act as a tumor suppresor. Loss of imprinting was observed in acute myeloid leukemia at this locus accompanied by bialllelic expresion of Igf2. Also H19 levels were decreased in chronic myeloid leukemia (Bock et al., 2003) and polycythemia vera (Nunez et al., 2000).

Lastly, it is important to note that H19 is negatively regulated by p53. Matouk et al. (2010) found that H19 is induced under the hypoxic stress conditions in the cells with mutated p53, but not in the wild type cells.
1.2.2.5. MALAT1

MALAT1, also called NEAT2 (Nuclear-Enriched Abundant Transcript 2), is multifunctional lncRNA located at 11q13.q.

First time described in 2003 by Ji et al., it got its name according to association with metastasis in lung cancer. Since then, MALAT1 was reported to be upregulated in many other types of cancer.

Besides the lncRNA itself, MALAT1 transcript gives rise to the 61 nt short tRNA-like mascRNA. At very low levels unprocessed 7 kb MALAT1 is expressed and located in nucleus. Yet primarily, MALAT1 is cleaved at 3´ end by RNase P resulting in 6,7 kb MALAT1 and pre-mascRNA. RNase Z then cleaves pre-mascRNA and three nucleotids CCA are added at the end (figure 7). Processed MALAT1 is transported to nuclear speckles and mascRNA into cytoplasm (Wilusz et al., 2008).

Figure 7. Processing of the 3´end of MALAT1 gives rise to mascRNA (Wilusz et al., 2008).
Nuclear speckles serve as a storage and assembly site for the pre-mRNA splicing machinery. MALAT1 was found to be enriched in those structures and to modulate levels of active SR splicing factors, thus, affecting alternative splicing of genes (Tripathi et al., 2010). This is possibly tissue-specific, since MALAT1 was observed to regulate alternative splicing in HeLa cells but not in lung cancer cells (Gutschner et al., 2013).

As mentioned above, MALAT1 is a multifunctional lncRNA and affects a lot of different signaling pathways or gene expression according to the type of the cell. In lung adencarcinoma MALAT1 enhances cell motility through upregulation of migration associated genes, e. g. CTHRC1 (collagen triple helix repeat containing 1) and CCT4 (chaperonin-containing tailless complex polypetide, subunit 4) but has no effect on cell proliferation (Tano et al., 2010).

Another studies suggested that MALAT1 is involved in cell cycle progression. When MALAT1 was depleted in human fibroblasts, an increase in G1-phase cells was observed together with a decrease in S-phase cells. But HeLa cells upon MALAT1 depletion showed G2/M arrest. At the same time, p53 was activated, which lead to the hypothesis that MALAT1 serves as a p53 repressor (Tripathi et al., 2013).

Hu et al. (2015) studied G2/M phase arrest more closely on esophageal cancer cells and found out that MALAT1 affects activity of ATM/CHK2 pathway. This pathway is active upon DNA damage and is able to stop cell cycle progression. When MALAT1 was inhibited, higher levels of phosphorylated ATM and CHK2 were shown on Western blot, while no change in total levels was detected. Taking this from the opposite direction, upregulated MALAT1 leads to dephosphorylation, thus inactivation of ATM/CHK2 pathway, which contributes to proliferation of cells and tumor growth.

Similarly, MALAT1 affects PI3K/Akt pathway in osteosarcoma cells and MAPK/ERK pathway in gallbladder cancer. Knockdown of MALAT1 in osteosarcoma cells by siRNAs downregulates levels of phosphorylated Akt but not its total levels (Dong et al., 2015) and in gallbladder cancer cells, inhibition of MALAT1 lead to reduction in phosphorylated MEK1/2, Erk1/2 and MAPK but, once again, no change in total levels was observed (Wu et al., 2014).

MALAT1 was even found involved in NF-κB pathway during epithelial-mesenchymal transition but not in burn injury, as Li et al. (2015) described when working on modulatory network of NF-κB. Akt, MAPK/ERK and NF-κB pathways are also activated during BCR signalization in B-cells, as mentioned above.

Few members of the Bcl-2 protein family are influenced by MALAT1 expression too. Guo et al. (2010) inhibited MALAT1 transcription in human cervical cancer cells and detected an increased expression of caspases 3, 8 and Bax protein but also a decrease in expression of Bcl-2 and Bcl-XL. Taken together, MALAT1 depletion had pro-apoptotic effects. In another study, MALAT1 was found to act as a competing endogenous RNA (ceRNA) by sponging miR-363-3p, which is an inhibitor of Mcl-1, thus resulting in higher expression in anti-apoptotic Mcl-1 (Patel et al., 2014).

Lastly, it is important to note, that MALAT1 expression was investigated during hematopoiesis. It was revealed that MALAT1 helps to keep proliferation potential of early-stage hematopoietic cells and that it downregulates during differentiation (Ma et al., 2015).

1.2.2.6. LincRNA-p21

LincRNA-p21 (large intergenic ncRNA, p21 according to the neighboring gene) at locus 6p21.2 is induced by p53 and acts as a transcriptional repressor of p53 target genes.

LincRNA-p21 binds to the heterogeneous nuclear ribonucleoprotein K (hnRNP-K) and together they repress genes that are supposed to be downregulated by p53 (figure 8). It was also found that lincRNA-p21 regulates apoptosis. When siRNA for either p53 or lincRNA-p21 were used, increase in viablity of doxorubicine treated cells was observed. Taken together, lincRNA-p21 acts in a pro-apoptotic manner in the p53 pathway (Huarte et al., 2010).


Figure 8. LincRNA-p21 function (Baldassare et al., 2012).
Just recently, study of lincRNA-p21 in CLL was realeased by Blume et al. (2015). They conducted DNA damage experiments on CLL cells and found out, similarly, that lincRNA-p21 is dependent on p53 and that this lncRNA decreases cell viability. Also, interestingly, they observed that cells with del11q showed impaired induction of lincRNA-p21.
1.2.2.7. INXS

INXS or ABALON (apoptotic BCL2L1-antisense long non-coding RNA) is a long ncRNA located at 20q11.21. More precisely, it is located on the opposite strand of BCL-X (BCL2L1) gene.

BCL-X is a member of Bcl-2 protein family. It has two isoforms with opposite functions, BCL-XL is a longer form with anti-apoptotic function and BCL-XS is shorter with pro-apoptotic function.

INXS is decreased in many cancer cell lines (kidney, liver, beast, prostate) and according to DeOcesano-Pereira et al. (2014) it causes a shift in splicing in favor of BCL-XS form. INXS binds Sam68 splicing-modulator complex and together they target BCL-X pre-mRNA. Moreover, they found that INXS expression is upregulated by apoptosis-inducing agents such as UV-C light, serum reduction and sulforaphane and it is required for activity of caspases 3 and 9.

Although, just recently, authors published erratum for their study, since they found out some non-specifity in binding of their BCL-XL primers. Nevertheless, the main objective stays untouched and that is INXS acting pro-apoptotic by increasing amount of BCL-XS protein form.

2. AIMS

• 
  the investigation of selected lncRNAs expression in CLL primary samples and selected cell lines derived from B-cells
  

• 
  the association of lncRNAs expression with CLL prognostic factors and expression of Bcl-2 protein family members
  

6. SOUHRN
Cílem této diplomové práce "Úloha dlouhých nekódujících RNA v biologii chronické lymfocytární leukémie" bylo zkoumání vybraných lncRNA (H19, XIST, GAS5, TUG1, MALAT1, LincRNA-p21 a INXS) a jejich role v CLL. Exprese byla analyzována jak na pacientských vzorcích, tak u B-buněčných linií.

Exprese lncRNA byla změřena v kohortě CLL pacientů a analyzována v souvislosti s klinickými a biologickými parametry (IgVH mutační status, del17p, del11q, trizomie 12, del13q14, pohlaví). Exprese XIST a GAS5 byla asociována s pohlavím, TUG1 s trizomií 12 a MALAT1 s del11. Exprese H19 nebyla detekována vůbec. Dále byla porovnána exprese mezi kontrolními pacientskými vzorky a vzorky po indukci poškození DNA in vitro (fludarabin), u kterých byla zjištěna upregulace třech lncRNA (GAS5, TUG1, MALAT1).

Několik vzorků CLL pacientů bylo na základě předešlých studií vybráno pro prozkoumání vztahu mezi lncRNA MALAT1 a INXS se členy rodiny Bcl-2 proteinů (Mcl-1, Bcl-XL, Bcl-2). Žádná asociace nebyla mezi těmito proteiny a lncRNA nalezena.

Také byla analyzována regulace lncRNA za různých podmínek souvisejících s aktivací maligních B-lymfocytů (aktivita STAT3, fibronektin, přítomnost mikroprostředí). Bylo zjištěno, že lncRNA MALAT1 je zvýšeně exprimována po kokultivaci se stromální buněčnou linií kostní dřeně HS5.

Naše výsledky neodhalily prognostický význam u žádné ze zkoumaných lncRNA, nicméně naznačují odlišnou roli pro MALAT1 u B-buněk cirkulujících v periferní krvi a buněk migrujících do sekundárních lymfatických orgánů.


7. SUMMARY
Aim of this thesis titled "The role of long-noncoding RNAs in the biology of chronic lymphocytic leukemia" was the investigation of selected lncRNAs (H19, XIST, GAS5, TUG1, MALAT1, LincRNA-p21 and INXS) and their role in CLL. This was performed in cohort of CLL patient samples and B-cell cell lines.

LncRNAs expression levels were measured in patients cohorts and analyzed in relation to clinical and biological parameters (IgVH mutation status, del17p, del11q, trisomy 12, del13q14, gender). XIST and GAS5 expression was associated with gender, TUG1 with trisomy 12 and MALAT1 with del11q. Expression of H19 was not detected in CLL. Then, control patient´s samples and samples after DNA-damage induction in vitro (fludarabine) were compared. Upregulation of three lncRNAs (GAS5, TUG1, MALAT1) in fludarabine treated cells was detected.

Several patient samples were selected according to previous studies to explore the relationship between lncRNAs MALAT1 and INXS and members of Bcl-2 protein family (Mcl-1, Bcl-XL, Bcl-2). No association with any of the mentioned proteins and lncRNAs was found.

LncRNAs regulation under different conditions related to the malignant B-cell activation (STAT3 activity, fibronectin, presence of microenvironment) was examined as well. It was found that lncRNA MALAT1 is upregulated after cocultivation with bone marrow stromal cell line HS5.

Our results did not reveal prognostic importance for any of the selected lncRNAs, although they suggest different role of MALAT1 in B-cells circulating in peripheral blood and B-cells migrating to the secondary lymphatic organs.

8. LITERATURE
Baldassarre, A. & Masotti, A. 2012. Long non-coding RNAs and p53 regulation. Int J Mol Sci, 13(12):16708-16717.
Binet, J.L., Auquier, A., Dighiero, G., Chastang, C., Piguet, H., Goasguen, J., Vaugier, G., Potron, G., Colona, P., Oberling, F., Thomas, M., Tchernia, G., Jacquillat, C., Boivin, P., Lesty, C., Duault, M.T., Monconduit, M., Belabbes, S. & Gremy, F. 1981. A new prognostic classification of chronic lymphocytic leukemia derived from a multivariate survival analysis. Cancer, 48(1):198-206.
Blume, C.J., Hotz-Wagenblatt, A., Hüllein, J., Sellner, L., Jethwa, A., Stolz, T., Slabicki, M., Lee, K., Sharathchandra, A., Benner, A., Dietrich, S., Oakes, C.C., Dreger, P., te Raa, D., Kater, A.P., Jauch, A., Merkel, O., Oren, M., Hielscher, T. & Zenz, T. 2015. p53-dependent non-coding RNA networks in chronic lymphocytic leukemia. Leukemia, 29(10):2015-2023.
Bock, O., Schlué, J. & Kreipe, H. 2003. Reduced expression of H19 in bone marrow cells from chronic myeloproliferative disorders. Leukemia, 17(4):815-816.
Brown, C.J., Hendrich, B.D., Rupert, J.L., Lafrenière, R.G., Xing, Y., Lawrence, J. & Willard, H.F. 1992. The human XIST gene: analysis of a 17 kb inactive X-specific RNA that contains conserved repeats and is highly localized within the nucleus. Cell, 71(3):527-542.
Burger, J.A., Burger, M. & Kipps, T.J. 1999. Chronic lymphocytic leukemia B cells express functional CXCR4 chemokine receptors that mediate spontaneous migration beneath bone marrow stromal cells. Blood, 94(11):3658-3667.
Burger, J.A., Quiroga, M.P., Hartmann, E., Bürkle, A., Wierda, W.G., Keating, M.J. & Rosenwald, A. 2009. High-level expression of the T-cell chemokines CCL3 and CCL4 by chronic lymphocytic leukemia B cells in nurselike cell cocultures and after BCR stimulation. Blood, 113(13):3050-3058.