Abstract
Abstract 3666
Activation-induced cytidine deaminase (AID) is essential for somatic hypermutation (SHM) and class switch recombination (CSR) in B cells. Deamination of cytosine residues by AID leads to double-strand breaks and point mutations in immunoglobulin (Ig) genes by subsequent mechanisms. Physiological AID expression is tightly regulated and restricted to germinal center (GC) B cells. AID acts not exclusively on Ig loci but effects aberrant SHM (ASHM) throughout the genome (Liu et al., 2008). High-fidelity mutation repair mechanisms protect most non-Ig AID target genes against mutation accumulation. Some genes including BCL6, the most highly mutated non-Ig gene, also undergo error-prone repair and consequently accumulate aberrant somatic mutations (Liu et al., 2009). In contrast to normal B cells, AID is constitutively expressed in follicular lymphoma (FL) and diffuse large B cell lymphoma (DLBCL) (Greeve et al., 2003; Lossos et al., 2004). These lymphoma entities are derived from GC B cells and continue to accumulate somatic mutations in the immunoglobulin heavy variable region (IGHV) through ongoing SHM (Kosmas et al., 1998). In FL, the 5` noncoding region of BCL6 is altered by aberrant somatic mutations. Histological transformation of FL to DLBCL is associated with ASHM of this gene segment (Pasqualucci et al., 2001; Lossos et al., 2000). To further elucidate the role of AID in FL pathogenesis, we performed precise quantitative correlations between AID expression, frequency of ongoing SHM of IGHV and ASHM of BCL6 in 16 isotype-switched (IgG-expressing) and 23 non-switched (IgM-expressing) FL.
AID expression was measured by quantitative real-time PCR. SHM of IGHV and ASHM of the 5` noncoding region of BCL6 were determined by PCR-based cloning and sequencing, followed by alignment with their consensus and germ-line sequences, respectively. Detected mutations were further categorized into consensus mutations, defined by their presence in the majority of clones, and sporadic mutations, defined by their presence in the minority of clones. Consensus and sporadic mutations are taken to represent previous and current AID activity, respectively.
All FL had readily detectable AID transcripts, albeit at highly variable levels. Overall, AID expression levels did not differ between isotype-switched and non-switched FL (p=0.99). Sporadic IGHV mutations were slightly more prevalent in switched FL (p=0.026). Consensus IGHV mutations (p=0.99) and both sporadic and consensus mutations in BCL6 did not differ between the FL subgroups (p=0.91 and p=0.77, respectively). In non-switched FL, AID expression levels were positively correlated with sporadic (r=0.44, p=0.041) and consensus IGHV mutations (r=0.42, p=0.049) and sporadic BCL6 mutations (r=0.51, p=0.016). Consensus mutations in BCL6 were independent of AID levels (r=0.23, p=0.30). In switched FL, AID expression was inversely correlated with consensus IGHV mutations (r= -0.71, p=0.004). There was no correlation between AID expression and sporadic IGHV mutations (r= -0.055, p=0.85), nor with sporadic or consensus BCL6 mutations (r=0.056, p=0.84 and r= -0.058, p=0.82, respectively).
Only in non-switched FL, AID expression levels determine its enzymatic activity with respect to expected mutations in its physiological and unphysiological target genes IGHV and BCL6 in a manner that appears to be quantitatively independent from AID cofactors or repair mechanisms. In contrast, the rates of ongoing IGHV mutations and of ASHM in isotype-switched FL are not directly dependent on AID expression levels. The unexpected inverse correlation to consensus IGHV mutations points to efficient high-fidelity mutation repair or alternative mechanisms that suppress AID activity in isotype-switched FL. Therefore, our data demonstrate hitherto unrecognized biological differences between isotype-switched and non-switched FL. These results predict that the combination of IgM isotype with high AID expression define an FL subgroup that is genetically less stable over time than their IgG-expressing or AID low-expressing counterparts. As a consequence, we postulate that continuous AID activity and the resulting ASHM of BCL6 and perhaps other proto-oncogenes cause a more aggressive clinical behaviour and an inferior prognosis in this FL subgroup, including a predisposition to histological transformation.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.