Aberrant somatic hypermutation in multiple subtypes of AIDS-associated non-Hodgkin lymphoma

Acquired immunodeficiency syndrome–related non-Hodgkin lymphomas (AIDS-NHLs) represent a significant source of morbidity and mortality among HIV-infected individuals.^1,2  According to body location and histologic criteria, the pathologic spectrum of AIDS-NHL includes systemic AIDS-NHL, primary central nervous system lymphoma (AIDS-PCNSL), and primary effusion lymphoma (AIDS-PEL).^1-3  Systemic AIDS-NHLs are histologically classified into AIDS-related Burkitt lymphoma (AIDS-BL) and AIDS-related diffuse large B-cell lymphoma (AIDS-DLBCL).³  All AIDS-PCNSLs are histologically represented by DLBCL.³  Despite their clinicopathologic heterogeneity, most AIDS-NHLs derive from B cells that have experienced the germinal center (GC) reaction⁴  and, therefore, have been exposed to the mechanism of somatic hypermutation (SHM) that normally targets the immunoglobulin variable region (IgV), BCL-6, and FAS genes.^5-10  In particular, AIDS-BL and a fraction of AIDS-DLBCLs and AIDS-PCNSLs express markers typical of GC B cells, whereas AIDS-PEL and the remaining fraction of AIDS-DLBCLs and AIDS-PCNSLs are composed of post-GC B cells maturing toward the plasma cell pathway of differentiation.⁴ 

The molecular pathogenesis of AIDS-NHL is associated with heterogeneous genetic lesions, including mainly chromosomal translocations that deregulate the expression of various proto-oncogenes.²  In fact, all AIDS-BLs harbor a reciprocal chromosomal translocation between band 8q24, the site of the c-MYC proto-oncogene, and one of the immunoglobulin gene loci, causing transcriptional deregulation of the proto-oncogene.²  A significant fraction of AIDS-DLBCLs associates with chromosomal translocations of BCL-6, a proto-oncogene initially cloned from 3q27 breakpoints in DLBCL of the immunocompetent host and found to encode a zinc-finger transcriptional repressor.²  Chromosomal rearrangements lead to substitution of the BCL-6 promoter with heterologous regulatory elements that derive from different chromosomal sites, including, although not restricted to, the immunoglobulin loci.²  In addition, a significant fraction of AIDS-BLs displays deletions and mutations of the p53 tumor suppressor gene.² 

Recently, a second mechanism of genetic lesion, termed aberrant somatic hypermutation, has been identified in DLBCL of the immunocompetent host.¹¹  Normally, the SHM process introduces mainly single nucleotide substitutions into the IgV genes of GC B cells, to produce antibodies with high affinity for the antigen.^5,6  SHM requires transcription, targets sequences within 2 kb downstream from the transcription initiation site, and is not restricted to the IgV sequences.^7-14  In fact, at least 2 nonimmunoglobulin genes, BCL-6 and Fas/CD95, accumulate mutations in normal GC B cells, suggesting that this mechanism may target more genes than originally suspected.^7-10  Although the molecular basis of SHM remains largely unknown, studies from the past 2 years have identified the AID gene, encoding for the activation-induced cytidine deaminase, as an absolute requirement for SHM in humans and mice.^15-17  In more than half of the patients with DLBCL, the SHM process appears to misfire and aberrantly targets multiple loci, including well-known proto-oncogenes already implicated in the pathogenesis of lymphoid malignancies (PIM-1, PAX-5, RhoH/TTF, and c-MYC).¹¹  In these genes, the mutations are distributed in the 5′ region, including regulatory as well as coding sequences, occur independently of chromosomal translocations to the immunoglobulin genes, and display similar features to those observed in the IgV genes, although at lower rates.¹¹  In contrast to IgV and BCL-6, however, mutations of PIM-1, PAX-5, RhoH/TTF, and c-MYC do not occur at a significant level in normal GC B cells or in other B-cell malignancies, suggesting a malfunction of SHM associated to DLBCL.¹¹ 

The present study was aimed at investigating the role of aberrant somatic hypermutation in AIDS-NHL. We have performed a comprehensive analysis exploring the mutation profile of PIM-1, PAX-5, RhoH/TTF, and c-MYC, along with IgV and BCL-6, throughout the clinicopathologic spectrum of AIDS-NHL. We report that aberrant hypermutation is a frequent event in AIDS-NHL, can be detected in lymphomas displaying a GC or post-GC profile, and is ongoing at least in some cases. Overall, these findings indicate that aberrant hypermutation may represent a major contributor to lymphomaganesis in AIDS-NHL.

This study was based on 39 samples of AIDS-NHL, including 29 systemic AIDS-NHL (all primary samples), 4 AIDS-PCNSL (all primary samples), and 6 AIDS-PEL cell lines. According to the Revised European-American classification of lymphoid neoplasms (REAL) and the World Health Organization classification of hematopoietic tumors,^3,18  systemic AIDS-NHLs were classified into AIDS-DLBCL (n = 18) and AIDS-BL (n = 11). AIDS-PCNSLs were histologically classified as DLBCL. Primary samples of AIDS-NHL were derived from lymph nodes, bone marrow, peripheral blood, body cavity effusions, or other involved organs and were obtained in the course of routine diagnostic procedures. In all instances, the specimen was collected at diagnosis before specific therapy. According to morphologic and immunophenotypic analysis, the fraction of malignant cells in the pathologic specimen was 95% or more in the case of AIDS-BL and AIDS-PEL, and 70% or more in the case of AIDS-DLBCL. All AIDS-NHL primary samples displayed a major monoclonal B-cell population, as determined by immunophenotypic or immunogenotypic studies. The PEL cell lines included in this study were HBL-6, BC-2, CRO-AP/3, CRO-AP/5, CRO-AP/6, and BCBL-1.^19-23  BCBL-1 was obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health (NIH, Bethesda, MD). Gross rearrangements of the c-MYC gene, tested by conventional strategies,²⁴  were detected in all AIDS-BL cases and in 1 of 18 AIDS-DLBCL cases (case 1905).

Genomic DNA from primary samples and cell lines was isolated by cell lysis followed by digestion with proteinase K and purified by “salting out” extraction and ethanol precipitation.²⁵  In selected cases of AIDS-NHL, paired germ line DNA was also available and was included in the analysis.

The sequences of the oligonucleotides used for mutational analysis of PIM-1, PAX-5, RhoH/TTF, and c-MYC were as follows: PIM1-F, 5′-CCA CAG CCA CAG CCA CAG CC-3′, and PIM1-R, 5′-CGC AGT TGT GGC AGT GCC GC-3′ (for PIM-1); PAX5-F, 5′-CCC AGA GAC TCG GAG AAG CA-3′, and PAX5-R, 5′-AAG AGC TGA AAT GTC GCC GCC G-3′ (for PAX-5); TTF-F, 5′-CCT TAA AAG TAT TTC TTT GGT GTC-3′, and TTF-R, 5′-AAC TCT TCA AGC CTG TGC TG-3′ (for RhoH/TTF); MYC1-F, 5′-CAC CGG CCC TTT ATA ATG CG-3′, and MYC-1R, 5′-CGA TTC CAG GAG AAT CGG AC-3′ (for c-MYC exon 1 fragment); and MYC2-F, 5′-CTT TGT GTG CCC CGC TCC AG-3′, and MYC-2R, 5′-GCG CTC AGA TCC TGC AGG TA-3′ (for c-MYC exon 2 fragment). Where needed, additional primers internal to each amplicon were used, and their sequence is available on request. The oligonucleotides used for mutational analysis of BCL-6 have been previously reported.²⁶  For analysis of the rearranged IgV_H genes, a set of 6 V_H gene family–specific primers that hybridize to sequences in the V_H leader region and an antisense-degenerated J_H primer (5′-CTY ACC TGA RGA GAC RGT GAC C-3′) were used.^27,28 

Mutational analysis of the 4 proto-oncogenes was performed on selected regions previously shown to contain more than 90% of the mutations found in DLBCL of immunocompetent hosts.¹¹  Mutational analysis of PIM-1 (GenBank accession no. AF386792) was performed by DNA sequencing of a unique polymerase chain reaction (PCR) product of 901 bp encompassing exon 1 to the 3′ region of exon 4 (nucleotides +1062 to +1963). Mutational analysis of PAX-5 (GenBank accession no. AF386791) was performed by direct sequencing of a unique PCR product of 859 bp encompassing the 3′ region of exon 1B and part of intron 1 (nucleotides +630 to +1489). Mutational analysis of RhoH/TTF (GenBank accession no. AF386789) was performed by direct sequencing of a unique PCR product of 844 bp encompassing the 5′ region of intron 1 (nucleotides +265 to +1109). Mutational analysis of c-MYC (GenBank accession no. X00364) was performed by direct sequencing of 2 PCR products of 1300 bp and 582 bp encompassing, respectively, the region downstream to the major P1/P2 promoter (c-MYC exon 1; nucleotides +2289 to +3589) and the region downstream to the minor P3 promoter (c-MYC exon 2, nucleotides +4486 to +5068). PCR products were purified and directly sequenced using a commercially available kit (Thermosequenase; Amersham Life Sciences, Oxford, United Kingdom). [α-³³P]-labeled terminator dideoxynucleotides (Amersham Life Sciences) were included in the sequencing mixture. For each DNA fragment analyzed, sequencing of both strands was performed on independent PCR reactions. Sequences obtained were compared with the corresponding germ line sequences, and the first nucleotide of the reference sequence was arbitrarily defined as position +1. Nucleotide changes due to previously reported polymorphisms or occurring more than once in separate cases were considered as polymorphic variants, unless their somatic origin could be proven.¹¹ 

Mutations of the BCL-6 5′ noncoding region were assessed by PCR amplification and direct sequencing of a 739-bp genomic region located in the gene intron 1 and previously shown to harbor more than 95% of the mutations detected in B-cell lymphoma (GenBank AY189709; note that the numbering has been corrected [+1 bp] compared with the one reported in our previous publications).^8,26,29  PCR conditions have been reported in detail previously.²⁶ 

The rearranged IgV_H genes were amplified using a set of 6 V_H gene family–specific primers that hybridize to sequences in the V_H leader region and a J_H degenerated primer in separate reactions for each V_H primer.²⁶  Samples for which no clonal IgV_H gene rearrangement was obtained with leader primers were amplified with a set of 6 V_H gene family–specific primers that hybridize to sequences in framework region (FR)I.²⁸  PCR products were purified and subjected to DNA direct sequencing. The rearranged IgV_H sequences were then compared with the most homologous germ line sequences using the V-BASE sequence directory (MRC Centre for Protein Engineering, Cambridge, United Kingdom) and MacVector 6.0.1 software (Oxford Molecular Group PLC, Oxford, United Kingdom).

The presence of intraclonal heterogeneity, indicative of ongoing hypermutation, was assessed in 7 AIDS-NHL cases (4 AIDS-BL, 2 AIDS-DLBCL, and 1 AIDS-PCNSL) previously shown to harbor mutations in PIM-1, PAX-5, and/or RhoH/TTF by DNA direct sequencing. Genomic DNA was subjected to PCR amplification using the Pyrococcus furiosus (Pfu) Turbo polymerase (Stratagene, La Jolla, CA), as described previously.¹¹  Sizes of PCR products were 715 bp for PIM-1, 932 bp for PAX-5, 875 bp for RhoH/TTF, and approximately 230 bp for the rearranged IgV_H genes.¹¹  After 30 cycles (35 for IgV_H), PCR products were gel purified, incubated with 0.2 mM dATP (deoxyadenosine triphosphate) and Taq polymerase (GIBCO BRL, Karlsruhe, Germany) for 15 minutes at 72°C, and cloned into the pGEM-T vector (Promega, Madison, WI) as recommended by the manufacturer. For each gene, 25 to 65 clones derived from independent PCR reactions were sequenced and analyzed. As a control for Pfu-introduced background error, a fibroblast cell line (IMR91) and purified tonsillar naive B cells were analyzed using the same approach.¹¹ 

Infection by Epstein-Barr virus (EBV) was investigated by PCR with primers SL-1 and SL-3 representative of the EBV nuclear antigen 1 (EBNA-1) gene, as previously reported.²⁹ 

Histogenetic markers of AIDS-NHL included mutations of IgV_H and BCL-6 genes (“Sequencing analysis of IgV_H genes”) as well as protein expression of the BCL-6 and CD138 antigens that distinguish the GC phenotype from the post-GC phenotype of AIDS-NHL.^4,30  On the basis of previously reported protocols, CD138 and BCL-6 expression were revealed using the B-B4 (Serotec, Oxford, England) and the PG-B6 (Dakopatts A/S, Glostrup, Denmark) monoclonal antibodies, respectively.⁴ 

All 39 samples of AIDS-NHL were subjected to sequence analysis of PIM-1, PAX-5, RhoH/TTF, and c-MYC. Mutations of c-MYC were not considered when detected in cases harboring gross rearrangements of the gene with the immunoglobulin loci (all AIDS-BL cases and AIDS-DLBCL 1905), because these mutations may be related to the juxtaposed immunoglobulin sequences, as described for translocated c-MYC alleles in BL.^31-33 

Results are summarized in Table 1 and are graphically represented in Figure 1. Overall, mutations in at least one of the 4 genes were detected in 19 of 39 (48.7%) AIDS-NHL cases, including 10 of 18 (55.5%) AIDS-DLBCL, 4 of 11 (36.3%) AIDS-BL, 4 of 6 (66.6%) AIDS-PEL, and 1 of 4 (25%) AIDS-PCNSL. Mutations in more than one gene were found in 9 of 39 (23.1%) AIDS-NHL cases, namely 6 of 18 (33.3%) AIDS-DLBCL, 2 of 11 (18.1%) AIDS-BL, and 1 of 4 (25.0%) AIDS-PCNSL. Each of the 4 genes analyzed was altered in a significant fraction of AIDS-NHL, because PIM-1 was mutated in 5 of 39 (12.8%), PAX-5 in 8 of 39 (20.5%), RhoH/TTF in 9 of 39 (23.1%), and c-MYC in 7 of 27 (25.9%) cases. Analysis of the corresponding germ line DNA, performed in selected cases, formally confirmed the somatic nature of the mutations, as previously reported in DLBCL of the immunocompetent host. The observed changes clearly belong to the neoplastic clone and not to nonlymphomatous surrounding B cells or non-B cells according to the following lines of evidence. (1) Previous studies performed on purified naive and GC B cells obtained from reactive tonsils unequivocally demonstrated that normal B cells and fibroblasts do not undergo SHM in these loci at a detectable level.¹¹  (2) Because the analysis was performed by PCR amplification and direct sequencing, only changes present in a significant fraction (≥ 10%) of the template DNA, ie, only clonal mutations, would be detected in the electropherograms. Therefore, the mutations seen can only reflect the status of the tumor clone, as also documented by the amplification of clonal V_H gene rearrangements from the same samples. (3) The fraction of malignant cells in the biopsies was 95% or more in AIDS-BL and AIDS-PEL, and 70% or more in AIDS-DLBCL.

The presence of aberrant SHM was independent of EBV infection, because mutations of one or more of the tested proto-oncogenes were detected in 10 of 23 (43.5%) EBV-positive AIDS-NHL and in 9 of 16 (56.2%) EBV-negative cases.

Detailed characterization of phenotypic markers of histogenesis (expression of BCL-6 and CD138 antigens) was available in 23 cases of AIDS-NHL. All AIDS-BL cases displayed the BCL-6⁺/CD138^– phenotype typical of GC B cells, and all AIDS-PEL cases displayed the BCL-6^–/CD138⁺ post-GC phenotype. Among systemic AIDS-DLBCL and AIDS-PCNSL, 9 cases expressed the BCL-6^–/CD138⁺ phenotype, and 1 case expressed the BCL-6⁺/CD138^– phenotype. Comparison of histogenetic profiles with the occurrence of mutations in PIM-1, PAX-5, RhoH/TTF, and c-MYC showed the presence of aberrant hypermutation in 8 of 15 (53.3%) BCL-6^–/CD138⁺ cases and in 3 of 8 (37.5%) BCL-6⁺/CD138^– cases (not shown).

The detailed characterization of PIM-1, PAX-5, RhoH/TTF, and c-MYC mutations in AIDS-NHL is reported in Table 2, and their general features are summarized in Table 3.

A total of 68 sequence variants were detected in 19 AIDS-NHL cases. The average frequency of mutations, calculated taking into account only mutated cases, ranged from 0.05 × 10^–2/bp in the case of c-MYC exon 1 to 0.43 × 10^–2/bp in the case of c-MYC exon 2 (Table 3), with no significant differences among the clinicopathologic categories of AIDS-NHL. The overwhelming majority of the mutations included single base-pair substitutions (n = 62), whereas deletions/insertions of a short DNA stretch were observed in 5 cases (Table 3). Of the 62 single base-pair substitutions observed, 36 were transitions and 26 were transversions, with a transition-to-transversion ratio of 1.38 (expected 0.5; P < .0001; chi-square test) (Table 3). Analysis of the nucleotide exchange pattern, when considering together all the nucleotide substitutions observed in PIM-1, PAX-5, RhoH/TTF, and c-MYC, showed a G+C/A+T ratio of 1.82 (Table 3), similar to that reported in human out-of-frame IgV rearrangements and in 2 in vitro mutating BL cell lines.^30,34,35  Differences in the individual transition/transversion and G+C/A+T ratios (eg, c-MYC exon 2 and RhoH/TTF, respectively) may be related to the limited number of mutational events counted. Analogously, although the low number of mutations prevents the assessment of statistical significance, the frequency of mutations targeting G residues within an RGYW motif was in all genes higher than expected (PIM-1, 14% versus 3.4%; RhoH/TTF, 9% versus 4.5%; PAX-5, 14% versus 4.2%; c-MYC exons 1 and 2, 21% versus 3.4%).

In PIM-1 and c-MYC, a number of the mutations were located in coding exons and led to amino acid substitutions, with potential functional consequences. In particular, 4 PIM-1 mutations in 3 AIDS-NHL cases represented missense mutations affecting the gene exon 4 (Figure 2A). Because this exon also encodes for the serine-threonine kinase domain, the mutations predict a change in the structure and, possibly, the function of the Pim-1 protein (Figure 2A).^36,37  In c-MYC, 18 missense mutations were found in 4 cases within the sequences of exon 2, which encode the transactivation domain and represent a mutational hot spot in translocated c-MYC alleles in BL and mouse plasmacytoma (Figure 2B).^33,38-40  Notably, 2 of these changes (Pro57Ser in AIDS-DLBCL case 2378 and Pro60Ser in AIDS-PCNSL case 2256) cluster around the evolutionary conserved and functionally critical phosphorylation sites Thr58 and Ser62 (Figure 2B).^39,41  Several studies have shown that mutations affecting or flanking these residues abolish phosphorylation of the c-Myc transactivation domain, resulting in defective regulation by p107, protein stabilization, and increased transforming activity.^41-43  Thus, alterations introduced by aberrant hypermutation are likely to have a profound effect on the function of this proto-oncogene.

A functional immunoglobulin V_HDJ_H rearrangement was obtained from 22 of 39 (56.4%) AIDS-NHL cases, whereas in 3 additional cases a stop-codon was detected within the originally productive V_H rearrangement, leading to a crippled immunogloublin sequence (Table 2). In one case, only an out-of-frame IgV_HDJ_H rearrangement was amplified. No PCR product could be obtained in the remaining 13 cases. The rearranged IgV_HDJ_H genes were found to be somatically mutated in all but one case, at frequencies ranging from 2.2% to 22.6% of the nucleotides examined (Table 2).

Mutations of BCL-6 were detected in 24 of 39 (61.5%) AIDS-NHL cases, including 11 of 18 (61.1%) AIDS-DLCL, 5 of 11 (45.4%) AIDS-BL, 3 of 4 (75.0%) AIDS-PCNSL, and 5 of 6 (83.3%) AIDS-PEL. Mutations are detailed in Table 2.

Overall, the average mutation frequency of PIM-1, PAX-5, RhoH/TTF, c-MYC exon 1, and c-MYC exon 2 was 50- to 100-fold lower than that observed in the IgV_H genes and 2- to 5-fold lower than that observed in BCL-6 (Table 3), as already observed in DLBCL of the immunocompetent hosts. Comparison of the mutation frequencies within each case revealed that the occurrence of PIM-1, PAX-5, RhoH/TTF, and c-MYC mutations did not correlate with an increased frequency of mutations in the IgV_H genes, supporting the hypothesis that aberrant hypermutation may be due to a qualitative rather than to a quantitative defect of the somatic hypermutation mechanism.¹¹ 

To investigate whether the aberrant mutational activity is ongoing in AIDS-NHL, we selected a subset of cases (n = 7) shown by direct sequencing to carry mutations in PIM-1, PAX-5, and/or RhoH/TTF. These cases were analyzed for the presence of intraclonal heterogeneity in the 4 proto-oncogenes as well as in the IgV_H genes by sequencing cloned PCR products generated using the proofreading Pfu polymerase. As a control for background error rate, the fibroblast IMR91 cell line and naive B cells purified from human tonsils were used.¹¹  Statistical significance was assessed by the Student t test.

The results showed that, in all tumors analyzed, 1 or 2 predominant alleles recapitulate the mutations observed by direct sequencing of the PCR product, confirming their presence in the tumor clone and revealing their frequent biallelic distribution (not shown). In addition, a significant number of intraclonal variants were found in the PIM-1 and RhoH/TTF sequences from 2 of the 7 cases examined. In the PIM-1 gene of AIDS-DLCL 2378, 27 of 65 clones (41.5%) carried one or more additional mutations (corresponding to 34 events in 46 475 bp sequenced) (Table 4 and Figure 3). Accordingly, this same case also displayed ongoing mutations in the rearranged IgV_H gene (Table 4), consistent with a derivation of the mutations from a common mechanism. A second case, AIDS-DLBCL 1220, exhibited a total of 5 point mutations, each unique to a single clone, which were distributed in 5 of 50 sequences analyzed and, therefore, did not differ significantly from the expected polymerase error frequency (0.004 to 0.008 × 10^–2/bp). This result prevents us from conclusively assessing whether there was low-rate ongoing mutation activity in this case. No evidence of intraclonal diversification was found in the PIM-1, PAX-5, and RhoH/TTF genes of the remaining cases (Table 4). These data indicate that the aberrant hypermutation activity may be ongoing in at least some cases of AIDS-NHL.

The aim of this study was to investigate the presence of mutations targeting the PIM-1, PAX-5, RhoH/TTF, and c-MYC genes throughout the clinicopathologic spectrum of AIDS-NHL. We report that mutations of these 4 proto-oncogenes represent a frequent event in AIDS-related lymphomas, display molecular features typical of the IgV-associated somatic hypermutation process, associate with different AIDS-NHL histotypes, and may be ongoing in some cases.

On the basis of several parameters, the molecular profile of aberrant hypermutation in AIDS-NHL is similar to that detected in DLBCL of the immunocompetent host and closely resembles the one reported for IgV and BCL-6 mutations.^7-14  First, mutations in PIM-1, PAX-5, RhoH/TTF, and c-MYC are predominantly represented by single nucleotide substitutions, with rare deletions and insertions. Second, a preference for transitions over transversions, a bias for specific targeting motifs (RGYW/WRCY), and an elevated ratio of G+C over A+T substitutions was observed in all of the 4 genes. As already reported in DLBCL from non-AIDS patients, the mutation frequency in PIM-1, PAX-5, RhoH/TTF, and c-MYC was lower than that observed in IgV and BCL-6 sequences in the same cases. Finally, mutations appear to be independent of translocations with the immunoglobulin genes, because, as formally documented for c-MYC, they occurred in cases devoid of gross alterations of the respective loci. Because mutations at the 4 loci analyzed have not been found in normal GC B cells, these data provide further support to the hypothesis that these mutations result from an abnormal activity of the SHM mechanism normally active in GC B cells.¹¹ 

The identification of an aberrant hypermutation activity throughout the clinicopathologic spectrum of AIDS-NHL expands the types of aggressive lymphoma associated with this molecular abnormality. In fact, aberrant hypermutation was found preferentially associated to DLBCL (> 50% of the cases analyzed) in the immunocompetent host, with only few FL and BL cases harboring isolated mutations in the RhoH/TTF and PAX-5 genes.¹¹  Conversely, in the context of AIDS-NHL aberrant hypermutation appears to affect at similar frequencies AIDS-DLBCL (both systemic and primarily localized to the central nervous system), AIDS-BL, and AIDS-PEL. This difference may be related to the relatively small number of BL cases examined to date in immunocompetent as well as in immunodeficient hosts. Alternatively, it may reflect true differences in the pathogenesis of sporadic versus endemic BL, as also suggested by other molecular differences between the 2 disease variants.^1,2  The detection of an aberrant SHM activity in AIDS-NHL reinforces the concept that this molecular mechanism is a frequent and selective feature of aggressive lymphomas, because mutations of PIM-1, PAX-5, RhoH/TTF, and c-MYC are rare or absent in indolent lymphomas.¹¹ 

The genes targeted by aberrant hypermutation in AIDS-NHL encode proteins with different functions, including both signal transducers (PIM-1 and RhoH/TTF) and transcription factors (PAX-5 and c-MYC).^33,36,44-46  As previously noted, these 4 genes represent known proto-oncogenes that have been already implicated in lymphoma-associated chromosomal translocations. On the basis of the type and distribution of the mutations, it can be predicted that they may have relevant functional consequences, with 2 distinct modalities. First, because they cluster around the gene 5′ regulatory regions, they may deregulate gene transcription, as formally documented in the case of c-MYC.³²  Second, in the case of c-MYC and PIM-1, in which the process also affects coding sequences introducing amino acid substitutions, the mutations may alter the biochemical and/or structural properties of the respective protein. In particular, in vitro and in vivo studies have previously shown that mutations in the c-MYC transactivation domain can deregulate its function by interfering with its phosphorylation, protein stability, or the repression of its transactivation activity by the Rb-related protein p107.^41-43  These findings suggest that at least some of the translocation-independent mutations in AIDS-DLBCL may have been selected for their ability to alter c-MYC function, as is the case for translocated c-MYC alleles in BL.³³ 

Finally, our results shed some light on the timing of aberrant hypermutation in AIDS-lymphomagenesis. Although in most AIDS-NHL cases the IgV somatic hypermutation process seems to be no longer active, as indicated by the absence of ongoing mutations in approximately 70% of the samples (G.G., unpublished observation, 2001), all of the 4 cases analyzed in this study showed intraclonal heterogeneity in the IgV genes; this finding may be related to the small number of cases investigated or may reflect a selection bias based on the presence of aberrant SHM. In addition, isolated AIDS-NHL cases were found which display intraclonal diversity in both IgV and non-IgV genes (Table 4 and Figure 3). This observation indicates that the tumor clone may continue to be exposed to aberrant SHM and to accumulate additional, intraclonal variants. This prolonged mutational activity, and the likelihood that it may affect many more genes than those identified so far, suggests that aberrant SHM may represent a major contributor to genetic damage during AIDS-associated lymphomagenesis.

We thank Daniela Vivenza, Michaela Cerri, and Annarita Conconi, from the Hematology Unit, Division of Internal Medicine, Department of Medical Sciences, Amedeo Avogadro University of Eastern Piedmont, for help in various parts of the study.