Análisis Descriptivo

As described previously, diffuse large B-cell lymphoma arises from the clonal expansion of B cells in the GC (Klein and Dalla-Favera, 2008). During their

development in GC, naïve B-cells are the subject of complex processes of somatic hypermutation and class switch recombination in order to develop antibodies with higher affinity and capable of distinct functions (Muramatsu et al., 2000). These processes are associated with single and double DNA strands breaks and can succeed successfully only with switched off DNA damage responses, pro-apoptotic mechanisms and inhibition of activation and differentiation. Additionally, during the GC reaction B- cells undergo very rapid proliferation with <12 hours doubling time and are at constant risk of mutations (Victora and Nussenzweig, 2012). If the class-switch recombinations affect not only constant immunoglobulin genes, they can cause chromosomal translocations. Furthermore if the somatic hypermutations are not limited to variable regions of immunoglobulin genes, they can cause aberrant hypermutations in other genes including oncogenes and tumour suppressor genes and contribute to lymphomagenesis (Pasqualucci and Dalla-Favera, 2015).

Clonally rearranged Ig heavy and light chains genes are detectable. They show somatic hypermutations in the variable regions (Swerdlow et al., 2008).

Approximately 50% of DLBCL cases show aberrant somatic hypermutations targeting multiple genetic loci, including PIMI, c-MYC (v-myc avian myelocytomatosis oncogene homolog), RHOH/TTF (ARHH) and PAX5. These mutations can contribute to the oncogenesis of this lymphoma (Pasqualucci et al., 2001).

The translocation involving the BCL6 gene is the most common in DLBCL NOS, accounting for up to 30% of cases (Offit et al., 1994) (Ohno and Fukuhara, 1997). Translocation of the BCL2 gene occurs in 20-30% of cases (Weiss et al., 1987) (Lipford et al., 1987) and a c-MYC re-arrangement was observed in approximately 10% of cases in an unselected series (Yunis et al., 1989). The c-MYC break partner is an Ig gene in 60% and a non Ig gene in 40% of cases (Hummel et al., 2006). Approximately 20% of cases with a c-MYC translocation have a concurrent BCL2 and or a BCL6 translocation (Hummel et al., 2006). These cases are characterised by very high proliferation rates (>90% Ki67), and it is suggested that they can be better categorized as “B-cell lymphomas, unclassifiable with features intermediate between DLBCL and BL” (Swerdlow et al., 2008).

Gene expression profiling (GEP) identified two distinctive subgroups in DLBCL: germinal centre B-cell-like (GCB) and activated B-cell like (ABC) (Alizadeh et al., 2000) (Rosenwald et al., 2002). The GCB subgroup has the gene expression profile of GC B-cells and the ABC subgroup has the profile of activated peripheral B-

cells. Initially a third group was defined (termed type 3) but it has turned out to be a collection of undefined cases and not a distinct group (Hummel et al., 2006) (Wright et al., 2003). The two defined groups are characterised by different chromosomal aberrations. The ABC cases have frequent gains of 3q, 18q21-q22 and losses of 6q21- q22 and by contrast GCB cases show frequent gains at 12q12 (Bea et al., 2005) (Tagawa et al., 2005) and BCL2 rearrangements (Huang et al., 2002).

The recent expansion of sequencing technologies, allowing studies on the whole genome / exome, has delivered a new comprehensive and unbiased view of the genetics basis of DLBCL (Pasqualucci and Dalla-Favera, 2015). The above described known abnormalities have been confirmed and put in the context of oncogenic pathways and new alternations have been described. Importantly, it has been revealed that the coding genome has approximately 50 – 100 lesions/case and great variability as compared with the genomes of other B-cells malignancies like CLL or ALL (Schneider et al., 2011).

Using next generation sequencing methods the genetic abnormalities in DLBCL, NOS can be classified in the context of the cell of origin theory in: i) those commonly affecting both GCB- and ABC-DLBCL e.g. genes involved in epigenetic mechanisms like acetylation/deacetylation and methylation/demethylation of histones and DNA,

BCL6, changes affecting genes involved in immune surveillance mechanism or TP53;

ii) the genetic lesions characteristic to GCB-DLBCL e.g. MYC and BCL2, PTEN and mir-17-92 micro RNA cluster and iii) genetic lesions to ABC-DLBC e.g. activation of

NF-κβ or block in terminal B cell differentiation (Pasqualucci and Dalla-Favera, 2015).

i) Genetic abnormalities affecting both GCB- and ABC-DLBCL. Up to 30% of DLBCL, NOS cases have mutations in acetyltransferase CREBBP and rarely in acetyltransferase EP300 (Goodman and Smolik, 2000). Both enzymes modify lysine residues by acetylation on both histone and non-histone nuclear proteins (Goodman and Smolik, 2000). This can result in both activation and deactivation of the modified protein. It has been postulated that acetylotransferases contribute to lymphomagenesis via impaired acetylation of BCL6 and p53, which leads to constitutive activation of the oncoprotein and inhibition of tumour suppressor (Bereshchenko et al., 2002). The changes affecting acetyltransferases are important as they can be targeted by a new group of drugs called inhibitors of deacetylases.

Another group of affected proteins involved in epigenetic mechanisms are methyltransferases (Pasqualucci and Dalla-Favera, 2015). These enzymes transfer methyl groups on histones and other nuclear proteins. The most commonly affected

methyltransferase is MLL2. Mutations to the MLL2 gene can potentially lead to broad effects on chromatin regulations, which may contribute to lymphomagenesis. Importantly, the MLL2 mutations have been also found in FL cases (Morin et al., 2011).

Alternations deregulating BCL6 are the main mechanism of transformation in DLBCL, NOS. The chromosomal translocations involving the BCL6 locus are present in approximately 35% of DLBCL cases, more frequently involving ABC-DLBCL (Iqbal et al., 2007). The BCL6 locus is also a target for point mutations observed in >70 cases (Shen et al., 1998). The mutations in the first noncoding exon are found exclusively in DLBCL and they deregulate BCL6 expression by disruption of the auto-regulatory loop by which BCL6 inhibits and regulates its own transcription (Wang et al., 2002) and by inhibition of binding of IRF4 and transcriptional repression (Saito et al., 2007). The activity of BCL6 can be also increased by mutation in the MEF2B transcription factor, a very potent BCL6 transcriptional activator (Ying et al., 2013). These mutations promote the function of MEF2B. In 5% of DLBCL cases the loss of function mutations of

FBX011 inhibit proteosomal degradation of BCL6 (Duan et al., 2012).

The loss of the immune surveillance mechanism is affecting a significant number of DLBCL cases (Pasqualucci and Dalla-Favera, 2015). The beta-2- microglobuline (β2M) is a non-variable unit of the HLA class I complex and is not expressed on the tumour cells in approximately 60% of DLBCL cases. In 29% of DLBCL cases the β2M is lost because of mutation affecting directly its gene and in the remaining 30% due to another mechanism involving its expression (Challa-Malladi et al., 2011). The other gene involved in immune surveillance of DLBCL is CD58, a member of the immunoglobulin superfamily and a ligand of the CD2 receptor on T-cells participating in their adhesion and activation (Pasqualucci and Dalla-Favera, 2015).

Among other lesions affecting DLBCL the mutations of the Tumour Protein 53 (TP53) gene are important and common (found in approximately 20% of DLBCL cases) (Monti et al., 2012). Tumour Protein 53 is a protein which is crucial in preventing cancer formation in multicellular organisms and TP53 is classified as a tumour suppressor gene. The role of TP53 in prevention of oncogenesis is associated with its function in i) activation of DNA repair, ii) arresting the cells at the G1/S point due to DNA damage, iii) initiation of apoptosis and iv) response to short telomeres. Mutations of FOXO1 transcription factor are also of importance and are present in all subtypes of DLBCL (Trinh et al., 2013).

ii) Genetic abnormalities affecting GCB-DLBCL. The genetic lesions characteristic to GCB-DLBCL were very poorly characterized until recently. The majority of research focused on chromosomal translocations of MYC, BCL2, PTEN and mir-17-92 micro RNA cluster (Trinh et al., 2013) (Saito et al., 2009) (Lenz et al., 2008).

As described above, in 35% of DLBCL the translocation (14;18) leads to ectopic expression of BCL2 (a key anti-apoptotic molecule) by juxtaposing the BCL2 in direct neighborhood of the Ig locus and disturbing negative suppression by BCL6 (Saito et al., 2009). Additionally, 40% of DLBCL cases without (14;18) translocation co-express BCL2 and BCL6 as a result of i) deregulation of Miz1 (a molecule connecting BCL6 to

BCL2), ii) aberant somatic hypermutation to BCL2 promotor side and iii) mutations to BCL2 coding sequence (Saito et al., 2009).

Approximately 15% of DLBCL cases show a translocation t(8;14) involving the

c-MYC transcription factor gene and the Ig gene locus (Kramer et al., 1998).

Amplification of the MIGH1 region encoding the microRNA 17-92 cluster is present in around 12% of GCB-DLBCL (Lenz et al., 2008). The microRNAs from this cluster enhance the lymphomagenetic function of c-MYC (He et al., 2005). The cluster also promoted the oncogenesis by inhibition of tumour suppressor gene PTEN and the pro- apoptotic protein BIM (He et al., 2005). The deletion of PTEN leads to the activation of AKT. The AKT can be also activated by phosphatidylinositol 3 kinase (PI3K) leading to inhibition of apoptosis, promotion of cell growth, cell motility and angiogenesis (Abubaker et al., 2007).

The mutations involving the histone methyltransferase EZH2 gene have been reported in 22% of GCB DLBCL patients (Morin et al., 2010). These mutations usually cause the increased activity of the enzyme resulting in increased GCB hyperplasia and induction of DLBCL in cooperation with BCL2 (Béguelin et al., 2013). Importantly the

EZH2 inhibitors just entered the clinical trials (Roschewski et al., 2014).

Approximately 20% of GCB-DLBCL show structural damaging mutations in various components of G-proteins e.g. GNA12, SIPR2, ARHGEF1 and P2RY8 resulting in increased GC B cells survival and dissemination (Muppidi et al., 2014).

iii) Genetic abnormalities affecting ABC-DLBCL. Genetic lesions to ABC- DLBCL are better characterized and they include two main groups: activation of NF-κβ and block in terminal B cell differentiation (Pasqualucci and Dalla-Favera, 2015). Additionally, the recurrent lesions involve the BCL2 locus as describe above and the

Genetic lesions contributing to constitutive activation of NF-κβ are usually caused by mutations activating continuously the BCR (B-cell receptor) signaling pathway via Ig superfamily members CD79B and CD79A or CARD11, a BCR signalosom complex (Davis et al., 2010). BCR signaling is responsible for canonical activation of NF-κβ. Mutations of myeloid differentiation primary response gene 88

(MYD88) gene, a universal activating protein responsible for NF-κβ activating in the TLR (toll like receptor) pathway, can lead to constitutive activation of NF-κβ and JAK/STAT transcriptional responses (Ngo et al., 2011). Finally the activation of NF-κβ can follow the mutation, inactivating negative regulators of NF-κβ like mutation to

TNFAIP3 gene encoding A20 a dual function ubiquitin modification enzyme involved

in the NF-κβ responses triggered by BCR and TLR stimulations (Ngo et al., 2011). The bi-allelic mutations of TNFAIP3 inhibit the function of A20 leading to inappropriately prolonged activation of NF-κβ.

The second group of genetic aberrations characteristic to ABC-DLBCL are genetic lesions preventing terminal differentiation of B-cells into plasma cells. The B- cell requires for this final step PRMD1, a sequence-specific transcriptional repressor (Shapiro-Shelef et al., 2003). The mutations to PRMD1 prevent this differentiation step. Approximately 25% of ABC-DLBCL have lost of PRDM1 gene owning to truncating mutations, missense mutations and / or genomic deletions (Pasqualucci et al., 2006).