Patterns of somatic mutations in V_H genes reveal pathways of clonal transformation from MGUS to multiple myeloma

Monoclonal gammopathy of undetermined significance (MGUS) is defined by key criteria: a plasma cell infiltration of less than 10% of bone marrow mononuclear cells, serum monoclonal M-protein of equal to or less than 30 g/L, absence of lytic lesions, related anemia, hypercalcemia, and renal insufficiency.1 MGUS varies in disease outcome, in some cases transforming to multiple myeloma (MM). Little is known concerning the underlying pathways of this progression, or indeed if the same clone is always involved. In part, these questions are hampered by the paucity of suitable tumor material for longitudinal analysis. More recently, evolution from MGUS to MM has been shown to exhibit a chromosome 13q abnormality.2 3 

Analysis of immunoglobulin variable (V) region genes now provides a means of identifying salient features of B-cell tumor biology. Tumor V genes retain an imprint of the clonal history of the cell of origin.4 They can reveal the impact of somatic mutation, most likely to have occurred in the germinal center (GC).5This is initiated in vitro by ligation of surface immunoglobulin (sIg) in normal and neoplastic B cells.6-8 Tumors residing in the GC display intraclonal variation in V gene sequence patterns, a product of ongoing mutational activity at this site.4 9 

Extensive data have characterized a postfollicular origin for MM, with more than 100 V_H gene sequences displaying homogeneity of clonal tumor-derived sequences.4 In contrast, 3 of 7 MGUS revealed significant intraclonal variation in V_H sequence in tumor clones, suggesting at least in a subset of MGUS a maintenance of the mutational mechanism.10 We now report 2 cases, evaluated during progression from MGUS to myeloma by V_H gene analysis. Our findings reveal that the same tumor clone evolves to malignant status, and, importantly, analysis of somatic mutations suggests different pathways of disease progression.

Case 1 presented with advanced lung cancer with pulmonary involvement and a lesion in TH10. Immunofixation identified a trace IgAλ band. Bone-marrow (BM) biopsy (sample 1/1) revealed normal composition, with plasma cells (PCs) lower than 5%. Magnetic resonance imaging revealed infiltration of most vertebra 4 months later. Serum M-component was demonstrated (total IgA, 64 g/L; M level, not determined [ND]), with 80% PCs in BM (sample 1/2). A retrospective classification of TH10 could not rule out an isolated plasmacytoma at first biopsy; however, this had excluded any clinical marrow involvement. Further, postmortem examination revealed disseminated myeloma and confirmed concomitant squamous-cell cancer with multiple pulmonary lesions, suggesting a possible metastatic TH10 spread.

Case 2 initially had an M-gradient in her serum electrophoresis (total IgG, 15.7 g/L; M level, ND; PCs, 5% of nucleated BM cells; sample 2/1) that progressed over 67 months (IgG, 26.6 g/L; BM PCs, 30%; sample 2/2).

BM aspirates were taken for magnetic activated cell sorting of CD138⁺ cells.11 Total RNA isolation from CD138⁺ cells (1 × 10⁴-2.6 × 10⁶ cells) and cDNA preparation using oligo-dT or constant region (C_H) primer were as described.12 cDNA was amplified using V_H1-7 leader primers with 3′ IgG or IgA primer. Cloning, sequence analysis, and V_H gene alignments were as reported.12 For V_H genes, between 16 to 30 clones/sample were sequenced from 2 separate polymerase chain reactions (PCRs).

For each MGUS and MM sample, 246 to 258 bp of the tumor C_Hwere amplified and cloned to establish the Taqerror rate using single 5′ complementary determining region 3 (CDR3) and Cγ or Cα primers (916-3690 bp analyzed/sample).Taq error depends primarily upon the initial amount of template available. We used the same cDNA preparation and identical amounts of template input for V_H and C_Hamplification. This corrected for variations in amount of RNA isolated and efficiency of cDNA synthesis, reflected in the range ofTaq error observed in C_H(3.4 × 10⁻⁴ to 1.6 × 10⁻³bp⁻¹). Replicate analysis confirmed Taqerror.

Case 1 utilized V_3-11 and case 2 V_4-30.2in sequential samples. These showed evidence for extensive somatic mutations (nucleotide sequences have been deposited in the European Molecular Biology Laboratory database; accession numbersAJ536045-AJ536051).

We observed intraclonal variation in sequence at the MGUS stage in case 1. Mutations in codons 107T>C and 120A>G were identified in 2 respective clones obtained from separate PCRs (Figure1). Additional single nucleotide changes were also observed (not shown), with a mutational frequency (1.6 × 10⁻³ bp⁻¹) exceeding Taqerrors in C_H (3.4 × 10⁻⁴ bp⁻¹; 11 clones from 2 PCRs). This confirms heterogeneity in MGUS.10 It implicates a cell that is undergoing continual somatic mutation following tumorigenic arrest. This may be occurring in the GC environment, suggesting a less differentiated but sIg⁺ cell.6-8 This cell has acquired a proliferative potential, perhaps due to a chromosomal event 1 (Figure2). It allows progeny with heterogeneous sequences to survive and migrate to the marrow to differentiate further. This heterogeneity also exists in clonally derived Cμ transcripts, demonstrated in only one MGUS case to date,10and implicating the stage of isotype switch in event 1.

Following transformation to myeloma in this patient (sample 1/2; Figure1), the observed ongoing mutational activity appeared to be silenced, as the frequency of single nonclonal mutations (1.5 × 10⁻³ bp⁻¹) was similar toTaq error (1.6 × 10⁻³ bp⁻¹). However, we observed 2 mutational changes at codons 68T>C and 107T>C in single clones at the MM stage that were identical to mutations at the early stage of disease (Figure 1). These 2 mutations were detected in samples taken at 2 different disease stages, negating random PCR error. Most likely, they represent residual MGUS clones as transformation was rapid. In the absence of paired samples, a very low level of nonshared mutations in some MM clones would have been assigned as Taq error in V gene analysis. Reports of stability of myeloma V gene sequence indicate the rarity of identifying such persisting residual clones by current technology.4However, there have been sporadic reports that hint at this possibilty. In 1 of 4 MM cases with more than 40% tumor cells, 2 clonally-related subclones with heterogeneity were observed,13 possibly reflecting a narrowing of a wider variation that existed previously. In a further MM case, significant intraclonal diversity was noted in the IgA variant transcripts but not the tumor isotype, suggesting emergence of a dominant clone but with residual variants.14 Given that MM is clonally homogeneous even in Cμ tumor-derived sequences,15-17 this suggests that the final transformation event (event 2, Figure 2) occurs in one cell of the population, at a site where somatic mutation is silent, possibly the bone marrow.

In case 2, we observed intraclonal homogeneity at both the MGUS and MM stage, confirming previous observations in 4 of 7 MGUS cases.10 A self-cloning phenomenon may underlie this observed sequence homogeneity in MGUS, being analyzed at a stage where hypermutation has ceased and heterogeneity has been lost because of clonal outgrowth. Transformation to myeloma then occurs at the invariant stage, involving event 2 (Figure 2).

V_H gene analysis has revealed diversity in MGUS and has suggested that transformation may involve different pathways. MGUS may derive from an tumorigenic sIg⁺ cell undergoing somatic mutation, which differentiates into Ig-secreting plasma cells. This population may persist and possibly self clone to homogeneity. However, it is susceptible to further transforming events that can occur in one cell at the heterogeneous or homogeneous stage, giving rise to homogeneous MM that is independent of the sIg⁺ population. Progression to myeloma then involves factors intrinsic to this cell.

We wish to thank Dr Judith Schuster and Irene Assmann for assistance in cell sorting.