The concept of clonal diversity is becoming well accepted as a hallmark of cancer. As a tumor grows and progresses, the genetic landscape of the cell population can change. These changes are largely due to random errors occurring during each cell division or through mutational events stimulated by various exposures. When one of these random events occurs in the right locus it will result in a survival advantage for all of the subsequent offspring of that initial cell. Understanding the underpinnings of clonal diversity may prove to be an essential part of the treatment plan for patients, helping to guide drug selection and to determine the percentage of clones that may be responsive/resistant to specific treatments. Moreover, many of the therapies used to treat myeloma will likely induce mutations through their mechanisms of action or through unexpected secondary effects. Understanding the effects of individual therapies and specific combinations on the underlying mutation rates that drive the diversity of a tumor population will help to identify regimens that increase the underlying mutation rate and put the patient at an increased risk of developing an aggressive clone. These changes can be identified by next generation sequencing of the bulk tumor population compared to single cell clones that have been selected from that population. In order to identify the diversity of mutations found in the bulk tumor population, we propose that single cell cloning the parent population, and then sequencing and comparing across several individual clones will give a better idea of the random variety of mutations present in individual cells that originate from the same parent population.

To identify the diversity present in a random population of myeloma cells we selected the human myeloma cell line KMS-18 as a model system. We sorted single cells from the KMS-18 parent population by FACS with the selection criteria based solely on the viable, single cells. These individually sorted cells expanded over a period of weeks until the population was large enough to be collected for analysis (target approximately 5E6 cells). Four of these single cell clones were selected (SCC_04, SCC_10, SCC_16, SCC_18) for analysis. We prepared whole genome libraries and captured a 3.2Mb region using the Agilent SureSelect Kinome capture kit. The final capture libraries were sequenced on the Illumina MiSeq platform to an average target region depth of 200X.

Results were filtered to identify the number of mutations present exclusively in one subclone compared to another. Such events either existed in the original single cell or occurred early in the expansion of the single cell clone. To limit the analysis to events present in the original single cell or very early in the doubling process we identified the variants that were found at a frequency of >20%. Many of these events were present in multiple single cell clones that could define the clonal relationship of each original cell, however, 10% of these variants were unique to a single subclone. On average we observed 1.6 mutations per Mb of the target region. If this same mutation rate holds true across the entire genome, we would expect to see over 5000 unique mutations between any two random cells taken from a bulk tumor sample.

Studies are currently ongoing to examine clonal diversity between generations of subclones. Further studies are also underway to look at changes in clonal diversity between different myeloma subtypes, with the hypothesis that more aggressive subtypes like t(4;14) and MAF may lead to a more diverse clonal population. If a more diverse clonal population correlates with more aggressive tumor subtype, then this returns full circle to the question of appropriate therapies, and if certain therapies may indeed increase diversity in the tumor population and result in a more aggressive relapse of the disease.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution