Amy L. Paguirigan

Single-cell genotyping demonstrates complex clonal diversity in acute myeloid leukemia

The targeted genotyping of single acute myeloid leukemia cells is technically feasible, identifies the zygosities of concurrent mutations, and suggests that sequencing of bulk populations may underestimate clonal complexity. Cancer evolution: No simple answers Traditionally, the evolution of cancer has been explained in simple terms: A cell acquires mutations and becomes malignant and then gives rise to progeny that become more malignant as they acquire additional mutations. A key assumption of this model is that the entire cancer is derived from descendants of a single original cell. However, a new study by Paguirigan et al. challenges this paradigm by providing evidence of convergent evolution in acute myeloid leukemia. The authors analyzed individual cells from multiple patients with leukemia and demonstrated that the mutation patterns seen in each patient could not have arisen from a single ancestral cell, suggesting a need for more sophisticated models of cancer evolution to inform the development of new treatment strategies. Clonal evolution in cancer—the selection for and emergence of increasingly malignant clones during progression and therapy, resulting in cancer metastasis and relapse—has been highlighted as an important phenomenon in the biology of leukemia and other cancers. Tracking mutant alleles to determine clonality from diagnosis to relapse or from primary site to metastases in a sensitive and quantitative manner is most often performed using next-generation sequencing. Such methods determine clonal frequencies by extrapolation of allele frequencies in sequencing data of DNA from the metagenome of bulk tumor samples using a set of assumptions. The computational framework that is usually used assumes specific patterns in the order of acquisition of unique mutational events and heterozygosity of mutations in single cells. However, these assumptions are not accurate for all mutant loci in acute myeloid leukemia (AML) samples. To assess whether current models of clonal diversity within individual AML samples are appropriate for common mutations, we developed protocols to directly genotype AML single cells. Single-cell analysis demonstrates that mutations of FLT3 and NPM1 occur in both homozygous and heterozygous states, distributed among at least nine distinct clonal populations in all samples analyzed. There appears to be convergent evolution and differential evolutionary trajectories for cells containing mutations at different loci. This work suggests an underlying tumor heterogeneity beyond what is currently understood in AML, which may be important in the development of therapeutic approaches to eliminate leukemic cell burden and control clonal evolution-induced relapse.

Self-digitization microfluidic chip for absolute quantification of mRNA in single cells

Quantification of mRNA in single cells provides direct insight into how intercellular heterogeneity plays a role in disease progression and outcomes. Quantitative polymerase chain reaction (qPCR), the current gold standard for evaluating gene expression, is insufficient for providing absolute measurement of single-cell mRNA transcript abundance. Challenges include difficulties in handling small sample volumes and the high variability in measurements. Microfluidic digital PCR provides far better sensitivity for minute quantities of genetic material, but the typical format of this assay does not allow for counting of the absolute number of mRNA transcripts samples taken from single cells. Furthermore, a large fraction of the sample is often lost during sample handling in microfluidic digital PCR. Here, we report the absolute quantification of single-cell mRNA transcripts by digital, one-step reverse transcription PCR in a simple microfluidic array device called the self-digitization (SD) chip. By performing the reverse transcription step in digitized volumes, we find that the assay exhibits a linear signal across a wide range of total RNA concentrations and agrees well with standard curve qPCR. The SD chip is found to digitize a high percentage (86.7%) of the sample for single-cell experiments. Moreover, quantification of transferrin receptor mRNA in single cells agrees well with single-molecule fluorescence in situ hybridization experiments. The SD platform for absolute quantification of single-cell mRNA can be optimized for other genes and may be useful as an independent control method for the validation of mRNA quantification techniques.

Microfluidics for single-cell genetic analysis

The ability to correlate single-cell genetic information to cellular phenotypes will provide the kind of detailed insight into human physiology and disease pathways that is not possible to infer from bulk cell analysis. Microfluidic technologies are attractive for single-cell manipulation due to precise handling and low risk of contamination. Additionally, microfluidic single-cell techniques can allow for high-throughput and detailed genetic analyses that increase accuracy and decrease reagent cost compared to bulk techniques. Incorporating these microfluidic platforms into research and clinical laboratory workflows can fill an unmet need in biology, delivering the highly accurate, highly informative data necessary to develop new therapies and monitor patient outcomes. In this perspective, we describe the current and potential future uses of microfluidics at all stages of single-cell genetic analysis, including cell enrichment and capture, single-cell compartmentalization and manipulation, and detection and analyses.

Heterogeneous resistance to quizartinib in acute myeloid leukemia revealed by single-cell analysis

Genomic studies have revealed significant branching heterogeneity in cancer. Studies of resistance to tyrosine kinase inhibitor therapy have not fully reflected this heterogeneity because resistance in individual patients has been ascribed to largely mutually exclusive on-target or off-target mechanisms in which tumors either retain dependency on the target oncogene or subvert it through a parallel pathway. Using targeted sequencing from single cells and colonies from patient samples, we demonstrate tremendous clonal diversity in the majority of acute myeloid leukemia (AML) patients with activating FLT3 internal tandem duplication mutations at the time of acquired resistance to the FLT3 inhibitor quizartinib. These findings establish that clinical resistance to quizartinib is highly complex and reflects the underlying clonal heterogeneity of AML.

Paper or plastic? BCR-ABL1 quantitation and mutation detection from dried blood spots

To the editor: Tyrosine kinase inhibitors (TKIs) have dramatically changed the natural history of chronic myeloid leukemia (CML), with 10-year overall survival surpassing 80% for patients treated in chronic phase. Unfortunately, most of the world’s CML patients reside in low-resource areas, where

Self-digitization chip for single-cell genotyping of cancer-related mutations

Cancer is a heterogeneous disease, and patient-level genetic assessments can guide therapy choice and impact prognosis. However, little is known about the impact of genetic variability within a tumor, intratumoral heterogeneity (ITH), on disease progression or outcome. Current approaches using bulk tumor specimens can suggest the presence of ITH, but only single-cell genetic methods have the resolution to describe the underlying clonal structures themselves. Current techniques tend to be labor and resource intensive and challenging to characterize with respect to sources of biological and technical variability. We have developed a platform using a microfluidic self-digitization chip to partition cells in stationary volumes for cell imaging and allele-specific PCR. Genotyping data from only confirmed single-cell volumes is obtained and subject to a variety of relevant quality control assessments such as allele dropout, false positive, and false negative rates. We demonstrate single-cell genotyping of the NPM1 type A mutation, an important prognostic indicator in acute myeloid leukemia, on single cells of the cell line OCI-AML3, describing a more complex zygosity distribution than would be predicted via bulk analysis.

Abstract 2420: A population sampling approach for validating deconvolution of high throughput sequencing data to describe clonality in heterogeneous tumor samples

Proceedings: AACR 107th Annual Meeting 2016; April 16-20, 2016; New Orleans, LA This abstract describes an in silico model of a heterogeneous population sampling methodology that elucidates how to best determine which genetic variants in a population are likely to occur in the same or different clones along with an experimental system to validate the approach. While the role of clonal evolution during leukemia development and therapy has been a focus for a number of avenues of research, our abilities to deduce the clonal composition of individual samples have been limited by the use of bulk samples. Leukemia populations of unknown heterogeneity are typically described by extracting nucleic acids from the entire sample, including cells of unrelated lineages, any residual normal cells, and those cells constituting the leukemic population. Bioinformatic approaches are employed to reconstruct the identities of the clones involved using clustering of feature frequencies in sequence data, but by definition this type of approach relies on a basic set of assumptions regarding the underlying biology which may or may not be accurate at all loci for all samples. We have generated an in silico model of how to use random population sampling with either targeted high throughput sequencing (54 genes) or targeted pyrosequencing (4 genes) to better determine which assumptions are not accurate for a given patient. In addition, we have an in vitro leukemia system on which we can test the in silico model. Flow cytometric validation of three AML cell lines along with the genetic characterization of each cell type and verification of comparable growth rates in a normalized medium in vitro has been performed. Subsequent mixing of these cell lines, random sampling of cells (10-2,000) isolated via flow cytometry and DNA sequencing can be used to test our ability to deconvolve the mixture based on a bulk measurement with the knowledge of the true population composition obtained via flow cytometry. Using bulk sequencing data, we can identify potential clonal structures likely to exist in a given sample via variant allele frequency densities and k-means clustering. Given the possible clonal models, we can select variants that appear to be in the same clone. The in silico sampling model then aids in determining the ideal replicate number and sample size to best discriminate if proposed variants are truly in the same clones or not. Using our experimental system, we then verified if the predicted variants can be deconvolved using specific sampling methods. With this sampling approach in hand and a framework for testing the assumptions, we can perform a prospective analysis of genetic variants within a heterogeneous leukemia sample via high throughput sequencing, then selectively validate selected mutations for concurrence in individual cells via targeted pyrosequencing. These results suggest that we can better describe leukemia heterogeneity without the requirement for extensive single cell analyses. Citation Format: Amy L. Paguirigan, Weston D. Christensen, Zaneta J. Holman, Jordan L. Smith, Emily A. Stevens, Brent Wood, Jerald P. Radich. A population sampling approach for validating deconvolution of high throughput sequencing data to describe clonality in heterogeneous tumor samples. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 2420.

Abstract NG02: Polyclonal and heterogeneous resistance to targeted therapy in leukemia

Genomic studies in solid tumors have revealed significant branching intratumoral clonal genetic heterogeneity. Such complexity is not surprising in solid tumors, where sequencing studies have revealed thousands of mutations per tumor genome. However, in leukemia, the genetic landscape is considerably less complex. Chronic myeloid leukemia (CML) is the human malignancy most definitively linked to a single genetic lesion, the BCR-ABL gene fusion. Genome wide sequencing of acute myeloid leukemia (AML) has revealed that AML is the most genetically straightforward of all extensively sequenced adult cancers to date, with an average of 13 coding mutations and 3 or less clones identified per tumor. In CML, tyrosine kinase inhibitors (TKIs) of BCR-ABL have resulted in high rates of remission. However, despite excellent initial response rates with TKI monotherapy, patients still relapse, including virtually all patients with Philadelphia-positive acute lymphoblastic leukemia and blast crisis CML. Studies of clinical resistance highlight BCR-ABL as the sole genetic driver in CML as secondary kinase domain (KD) mutations that prevent drug binding are the predominant mechanism of relapse on BCR-ABL TKIs. In AML, a more diverse panel of disease-defining genetic mutations has been uncovered. However, in individual patients, a single oncogene can still drive disease. This is the case in FLT3 mutant AML, in which the investigational FLT3 TKI quizartinib achieved an initial response rate of ∼50% in relapsed/refractory AML patients with activating FLT3 internal tandem duplication (ITD) mutations, though most patients eventually relapsed. Confirming the importance of FLT3 in disease maintenance, we showed that 8 of 8 patients who relapsed on quizartinib did so due to acquired drug-resistant FLT3 KD mutations. Studies in CML have revealed that sequential TKI therapy is associated with additional complexity where multiple mutations can coexist separately in an individual patient (“polyclonality”) or in tandem on a single allele (“compound mutations”). In AML, we observed polyclonal FLT3-ITD KD mutations in 2 of 8 patients examined in our initial study of quizartinib resistance. In light of the polyclonal KD mutations observed in CML and AML at the time of TKI relapse, we undertook next generation sequencing studies to determine the true genetic complexity in CML and AML patients at the time of relapse on targeted therapy. We used Pacific Biosciences RS Single Molecule Real Time (SMRT) third generation sequencing technology to sequence the entire ABL KD or the entire FLT3 juxtamembrane and KD on a single strand of DNA. Using this method, we assessed a total of 103 samples from 79 CML patients on ABL TKI therapy and 36 paired pre-treatment and relapse samples from 18 FLT3-ITD+ AML patients who responded to investigational FLT3 TKI therapy. In CML, using SMRT sequencing, we detected all mutations previously detected by direct sequencing. Of samples in which multiple mutations were detectable by direct sequencing, 85% had compound mutant alleles detectable in a variety of combinations. Compound mutant alleles were comprised of both dominant and minor mutations, some which were not detectable by direct sequencing. In the most complex case, 12 individual mutant alleles comprised of 7 different mutations were identified in a single sample. For 12 CML patients, we interrogated longitudinal samples (2-4 time points per patient) and observed complex clonal relationships with highly dynamic shifts in mutant allele populations over time. We detected compound mutations arising from ancestral single mutant clones as well as parallel evolution of de novo polyclonal and compound mutations largely in keeping with what would be expected to cause resistance to the second generation TKI therapy received by that patient. We used a phospho-flow cytometric technique to assesses the phosphorylation status of the BCR-ABL substrate CRKL in as a method to test the ex vivo biochemical responsiveness of individual mutant cell populations to TKI therapy and assess functional cellular heterogeneity in a given patient at a given timepoint. Using this technique, we observed co-existing cell populations with differential ex vivo response to TKI in 2 cases with detectable polyclonal mutations. In a third case, we identified co-existence of an MLL-AF9 containing cell population that retained the ability to modulate p-CRKL in response to BCR-ABL TKIs along with a BCR-ABL containing only population that showed biochemical resistance to all TKIs, suggesting the co-existence of BCR-ABL independent and dependent resistance in a single patient. In AML, using SMRT sequencing, we identified acquired quizartinib resistant KD mutations on the FLT3-ITD (ITD+) allele of 9 of 9 patients who relapsed after response to quizartinib and 4 of 9 patients who relapsed after response to the investigational FLT3 inhibitor, PLX3397. In 4 cases of quizartinib resistance and 3 cases of PLX3397 resistance, polyclonal mutations were observed, including 7 different KD mutations in one patient with PLX3397 resistance. In 7 quizartinib-resistant cases and 3 PLX3397-resistant cases, mutations occurred at the activation loop residue D835. When we examined non-ITD containing (ITD-) alleles, we surprisingly uncovered concurrent drug-resistant FLT3 KD mutations on ITD- alleles in 7 patients who developed quizartinib resistance and 4 patients with PLX3397 resistance. One additional PLX3397-resistant patient developed a D835Y mutation only in ITD- alleles at the time of resistance, suggesting selection for a non-ITD containing clone. All of the individual substitutions found on ITD- alleles were the same substitutions identified on ITD+ alleles for each individual patient. Given that the same individual mutations found on ITD- alleles were also found on ITD+ alleles, we sought to determine whether these mutations were found in the same cell or were indicative of polyclonal blast populations in each patient. To answer this question, we performed single cell sorting of viably frozen blasts from 3 quizartinib-resistant patients with D835 mutations identified at the time of relapse and genotyped single cells for the presence or absence of ITD and D835 mutations. This analysis revealed striking genetic heterogeneity. In 2/3 cases, polyclonal D835 mutations were found in both ITD+ and ITD- cells. In all cases, FLT3-ITD and D835 mutations were found in both heterozygous and homozygous combinations. Most surprisingly, in all 3 patients, approximately 30-40% of FLT3-ITD+ cells had no identified quizartinib resistance-causing FLT3 KD mutation to account for resistance, suggesting the presence of non-FLT3 dependent resistance in all patients. To determine that ITD+ cells lacking FLT3 KD mutations observed in patients relapsed on quizartinib are indeed consistent with leukemic blasts functionally resistant to quizartinib and do not instead represent a population of differentiated or non-proliferating cells, we utilized relapse blasts from another patient who initially achieved clearance of bone marrow blasts on quizartinib and developed a D835Y mutation at relapse. We performed a colony assay in the presence of 20nM quizartinib. As expected, this dose of quizartinib was unable to suppress the colony-forming ability of blasts from this relapsed patient when compared to DMSO treatment. Genotyping of individual colonies grown from this relapse sample in the presence of 20nM quizartinib again showed remarkable genetic heterogeneity, including ITD+ and ITD- colonies with D835Y mutations in homozygous and heterozygous combinations as well as ITD+ colonies without D835Y mutations, again suggesting the presence of blasts with non-FLT3 dependent resistance. Additionally, 4 colonies with no FLT3 mutations at all were identified in this sample, suggesting the presence of a quizartinib-resistant non-FLT3 mutant blast population. To see if we could identify specific mechanisms of off-target resistance, we performed targeted exome sequencing 33-AML relevant genes from relapse and pre-treatment DNA from all four patients and detected no new mutations in any genes other than FLT3 acquired at the time of disease relapse. Clonal genetic heterogeneity is not surprising in solid tumors, where multiple driver mutations frequently occur, but in CML and FLT3-ITD+ AML, where disease has been shown to be exquisitely dependent on oncogenic driver mutations, our studies suggest a surprising amount of clonal diversity. Our findings show that clinical TKI resistance in these diseases is amazingly intricate on the single allele level and frequently consists of both polyclonal and compound mutations that give rise to an complicated pool of TKI-resistant alleles that can change dynamically over time. In addition, we demonstrate that cell populations with off-target resistance can co-exist with other TKI-resistant populations, underscoring the emerging complexity of clinical TKI resistance. Such complexity argues strongly that monotherapy strategies in advanced CML and AML may be ultimately doomed to fail due to heterogeneous cell intrinsic resistance mechanisms. Ultimately, combination strategies that can address both on and off target resistance will be required to effect durable therapeutic responses. Citation Format: Catherine C. Smith, Amy Paguirigan, Chen-Shan Chin, Michael Brown, Wendy Parker, Mark J. Levis, Alexander E. Perl, Kevin Travers, Corynn Kasap, Jerald P. Radich, Susan Branford, Neil P. Shah. Polyclonal and heterogeneous resistance to targeted therapy in leukemia. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr NG02. doi:10.1158/1538-7445.AM2015-NG02

Abstract 3686: Single cell assays of heterogeneity and clonality in leukemia

A range of hypotheses have emerged as to why imatinib (a targeted inhibitor of bcr-abl) seems unable to eliminate the most primitive chronic myeloid leukemia (CML) cells, such as bcr-abl amplification and clonal evolution. The biology of the CML imatinib-resistant cell is of critical importance to understanding why despite high percentages of initial clinical responses to imatinib (>80% of patients in chronic phase at some point obtaining complete cytogenetic remission), resistance, relapse and/or progression occurs in 20-30% of cases and most patients in remission still have residual disease detected by sensitive PCR monitoring of the CML specific bcr-abl transcripts. In acute myeloid leukemia (AML) as well, patients are typically treated with a chemotherapeutic regimen that leads to clinical responses in 80-95% of patients. However, over half of these patients will relapse and eventually die of disease suggesting that a subpopulation of leukemic cells are capable of both surviving treatment and causing regrowth of the tumor. In both diseases, it has not been definitively established what the identity of the cells responsible for resistance and relapse is, nor what combination of mutations, gene expression changes, and alternative splicing events are required for these cells’ resistant phenotype. Assessing changes in gene expression in the bulk samples provides a valuable overall picture of the phenotype of the predominant cell type is or if there are large changes in expression in less frequent cells, but will completely mask changes exhibited by rare cells. One limiting factor in our ability to study more levels of heterogeneity in leukemia subpopulations is the lack of appropriately powerful and sensitive techniques to assay more complex facets of cellular behavior in such small populations. We have developed several multiplexed, high sensitivity analysis techniques and applied them to study multiple facets of cell behavior and identity in rare cell samples, down to the single cell level. By analyzing expression and genotype in smaller samples, direct from cells, we can analyze heterogeneity in expression and alternative splicing, combined with both cell surface markers and the presence of multiple mutations. These innovative techniques have allowed us to study clonality in AML samples in detail via a robust assay that can identify multiple mutations occurring concurrently in single cells. We have also optimized a protocol to simultaneously assay gene expression of 24 in 100-500 cell samples from flow sorted leukemia cell populations. Finally multiplexed assays for alternative splice variant detection in single cells has enabled us to study whether alternative splice variant expression of several genes occurs concurrently in the same cells, or in different cells of the population. With these techniques in hand we have the unique opportunity to begin studying the biology of these diseases from a previously unrealistic angle. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3686. doi:1538-7445.AM2012-3686