Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Somatic mutations in leukocytes infiltrating primary breast cancers

Somatic mutations in leukocytes infiltrating primary breast cancers www.nature.com/npjbcancer All rights reserved 2374-4677/15 ARTICLE OPEN Somatic mutations in leukocytes infiltrating primary breast cancers 1,8 2,8 3 1 2 4 1 Maria Kleppe , Elizabeth Comen , Hannah Y Wen , Lennart Bastian , Brian Blum , Franck T Rapaport , Matthew Keller , 5 4 6 6 5 1,3 3 1,3 Zvika Granot , Nicolas Socci , Agnès Viale , Daoqi You , Robert Benezra , Britta Weigelt , Edi Brogi , Michael F Berger , 1,3 1,7 2 Jorge S Reis-Filho , Ross L Levine and Larry Norton BACKGROUND: Malignant transformation requires the interaction of cancer cells with their microenvironment, including infiltrating leukocytes. However, somatic mutational studies have focused on alterations in cancer cells, assuming that the microenvironment is genetically normal. Because we hypothesized that this might not be a valid assumption, we performed exome sequencing and targeted sequencing to investigate for the presence of pathogenic mutations in tumor-associated leukocytes in breast cancers. METHODS: We used targeted sequencing and exome sequencing to evaluate the presence of mutations in sorted tumor- infiltrating CD45-positive cells from primary untreated breast cancers. We used high-depth sequencing to determine the presence/ absence of the mutations we identified in breast cancer-infiltrating leukocytes in purified tumor cells and in circulating blood cells. RESULTS: Capture-based sequencing of 15 paired tumor-infiltrating leukocytes and matched germline DNA identified variants in known cancer genes in all 15 primary breast cancer patients in our cohort. We validated the presence of mutations identified by targeted sequencing in infiltrating leukocytes through orthogonal exome sequencing. Ten patients harbored alterations previously reported as somatically acquired variants, including in known leukemia genes (DNTM3A, TET2, and BCOR). One of the mutations observed in the tumor-infiltrating leukocytes was also detected in the circulating leukocytes of the same patients at a lower allele frequency than observed in the tumor-infiltrating cells. CONCLUSIONS: Here we show that somatic mutations, including mutations in known cancer genes, are present in the leukocytes infiltrating a subset of primary breast cancers. This observation allows for the possibility that the cancer cells interact with mutant infiltrating leukocytes, which has many potential clinical implications. npj Breast Cancer (2015) 1, 15005; doi:10.1038/npjbcancer.2015.5; published online 10 June 2015 8,9 INTRODUCTION mediate primary tumor growth and metastasis. Recent evidence suggests that tumor-associated stromal cells and infiltrating In the past decade, targeted, whole-exome, and whole-genome leukocytes function differently than circulating or bone marrow sequencing studies have delineated a spectrum of somatic 3,10,11 1,2 resident hematopoietic cells. In particular, several studies mutations in human malignancies. These include large-scale have indicated that the content of lymphoid and myeloid cells sequencing studies in breast cancer, which have identified 12–15 infiltrating breast cancers correlates with clinical outcome. recurrent mutations in genes and pathways that contribute to We recently demonstrated that some older individuals have malignant transformation and to therapeutic response. The clinically inapparent, clonal hematopoiesis characterized by general approach to limit high-throughput genomic studies to recurrent, somatic mutations in TET2 (ref. 16). Of note, Tet2 loss cancer cells themselves is, however, in contrast to burgeoning in the hematopoietic compartment leads to increased self-renewal evidence that cancer cells interact with their microenvironment, 17–19 and myeloid bias of hematopoietic cells. These data, and including stromal cell constituents, infiltrating white blood cells, recent genomic studies of circulating hematopoietic stem cells and circulating inflammatory cytokines originating from local and 3,4 from normal volunteers and from The Cancer Genome Atlas distant sites. The integrity of the genome in these non-cancer patients suggest that morphologically normal hematopoietic cells cellular elements is therefore germane. Indeed, previous studies acquire mutations over time, most commonly in known leukemia have shown that stromal cells found in breast cancers are 20,21 disease alleles. The observation that normal individuals may characterized by site- and cell-type-specific epigenetic alterations, harbor oncogenic mutations in hematopoietic cells and the and reports suggesting the presence of somatic mutations in the 5–7 interaction between epithelial cancer cells and infiltrating leuko- tumor microenvironment are also on record. This suggests that cells other than the cancer cells themselves can acquire properties cytes raises the possibility of clonal selection in infiltrating leuko- that contribute to tumorigenesis. In addition to tissue-specific cytes. Hence, we sought to define whether tumor-infiltrating stromal cells, circulating and tumor-infiltrating leukocytes can leukocytes in breast cancer would harbor somatic mutations, and 1 2 Human Oncology and Pathogenesis Program, Memorial Sloan Kettering Cancer Center, New York, NY, USA; Breast Cancer Service, Department of Medicine, Memorial Sloan 3 4 Kettering Cancer Center, New York, NY, USA; Department of Pathology, Memorial Sloan Kettering Cancer Center, New York, NY, USA; The Bioinformatics Core, Memorial Sloan 5 6 Kettering Cancer Center, New York, NY, USA; Cancer Biology and Genetics Program, Memorial Sloan Kettering Cancer Center, New York, NY, USA; The Genomics Core, Memorial Sloan Kettering Cancer Center, New York, NY, USA and Leukemia Service, Department of Medicine, Memorial Sloan Kettering Cancer Center, New York, NY, USA. Correspondence: L Norton (nortonl@mskcc.org) These authors contributed equally to this work. Received 14 March 2015; revised 24 March 2015; accepted 26 March 2015 © 2015 Breast Cancer Research Foundation/Macmillan Publishers Limited Mutations in breast cancer-infiltrating leukocytes M Kleppe et al whether these would be enriched in the tumor as compared with Laser-capture microdissection of tumor cells peripheral blood. Ten consecutive 8-μm-thick nuclear fast red-stained representative sections of the tumors were subjected to laser-assisted microdissection on a PALM Robot MicroBeam laser microdissection system (Zeiss, Thornwood, NY, USA), as previously described. First, non-neoplastic cells, PATIENTS AND METHODS including inflammatory cells, stroma, and normal breast, were ablated. We Patient materials subsequently microdissected only histologically unequivocal neoplastic Breast cancer samples were collected from consecutive patients with cells from each sample under a microscope. Tissue was microdissected primary triple-negative breast cancer who underwent surgery at Memorial directly into extraction buffer, and DNA was extracted using the DNeasy Sloan Kettering Cancer Center between 2012 and 2013 (Table 1). Patients Blood and Tissue Kit (Qiagen, Valencia, CA, USA) and quantified with the treated with neoadjuvant chemotherapy were excluded from the study. Qubit Fluorometer (Invitrogen, Life Technologies, Norwalk, CT, USA). Non-triple-negative breast cancers showing prominent lymphocytic infiltrate in core biopsies were also included. All specimens were sectioned and DNA extraction and whole-genome amplification processed for routine pathological examination. Hematoxylin and eosin- DNA was extracted using the QiaAmp DNA kit (Qiagen) following the stained slides were reviewed by breast pathologists to establish the manufacture’s instructions. Buccal swabs were processed using the diagnoses. Estrogen receptor, progesterone receptor and human epidermal QiaAmp DNA Mini kit (Qiagen) following the manufacture’s instructions. growth factor receptor 2 (HER2) status was evaluated by immunohisto- The quality of DNA samples was analyzed with the Agilent Bioanalyzer chemistry according to the American Society of Clinical Oncology/College of 2100 (Agilent Technologies, Santa Clara, NY, USA). Samples with insuffi- 22,23 American Pathologists guidelines. HER2 fluorescence in situ hybridiza- cient amount of DNA (o500 ng) were whole-genome amplified using the tion was performed in one case with equivocal results by immunohis- REPLI-g Mini kit (Qiagen) prior to further use in downstream applications. tochemistry. Evaluation of tumor-infiltrating leukocytes was performed as QPCR was performed to assess the quality of WGA DNA. described by Loi et al. Tumor-infiltrating leukocytes were scored as following: extensive, ⩾ 50% infiltration of either stromal or intratumoral Exome sequencing and targeted capture sequencing lymphocytes; moderate = 5–10%; minimal, ⩽ 5%. Buccal swab samples were DNA extracted from sorted CD45-positive tumor-infiltrating leukocytes and collected from each patient. Mononuclear cells and granulocytes were buccal swabs (Supplementary Table 1) was sheared to an average size of isolated from peripheral blood following a standard Ficoll protocol. A detailed 180 ± 80 bp for exome sequencing. For DNA library preparation, 200–250- description on clinicopathological features of each patient is listed in Table 1. bp fragments were selected and subjected to PCR amplification. The library was then hybridized to the Agilent SureSelect Human All Exon Kit (Agilent Isolation and processing of tumor-infiltrating cells Technologies) and sequencing was performed on the SOLiD 3plus or SOLiD 4 (Applied Biosystems, Grand Island, NY, USA). Targeted sequencing All patients included in this study gave informed consent. Fresh tumor of tumor-infiltrating leukocytes and matched germline DNA of each cells, stromal cells, and tumor-infiltrating leukocytes were dissociated from patient was performed as previously described. the primary tumors by scraping the cutting surface 5–10 times with a surgical scalpel blade. Cell material was collected by rinsing the blade in phosphate-buffered saline. Cells were spun down and resuspended in red PCR and 454 sequencing analysis cell lysis buffer to remove red blood cells prior to staining with an anti- Sequence reactions were performed on DNA extracted from mononuclear human CD45-PE-Cy7- or CD45-APC-Cy7-conjugated flow antibody in FACS cells, granulocytes, laser-capture-microdissected tumor cells, and buffer (phosphate-buffered saline supplemented with 2% bovine serum tumor-infiltrating leukocytes. All PCR reactions were performed using albumin). Cells were stained for 20 min in the dark at room temperature, amplicon-specific fusion primers. Fusion primers contained next to the washed once with FACS buffer, and passed through a filter. 4′,6-diamidino- template-specific sequence a directional primer at the 5′-end followed by a 2-phenylindole was added before sorting to discriminate live and dead multiplex identifier for barcode sample identification. Samples from 6–8 cells. CD45-positive cells were then purified using a FACSAriaIII Cell Sorter different patients were mixed, processed for 454 deep sequencing, and run (MSKCC Flow Core Facility, Memorial Sloan Kettering Cancer Center, New on a Genome Sequencer FLX instrument (454 Life Sciences, Branford, CT, York, NY, USA). USA). Data were mapped with BWA MEM (BWA is freely available at http:// Table 1. Summary of clinicopathological features ID Age (years) Type Level of lymphocytic CD45 (%) Size (cm) HG NG Mitosis OG LVI LN ER PR HER2 FISH infiltration 1 40 IDC NOS Moderate 36.6 2 3 3 3 3 No No 0 0 0 — 2 72 IDC NOS Moderate 11.89 1.5 3 3 3 3 No No 0 0 0 — 3 37 IDC NOS Extensive 12.5 4.5 3 3 3 3 No No o1% o1% 1+ to 2+ 1.3 4 35 IDC NOS Moderate 0.4 5 2 3 3 3 Yes Yes 95% 90% 3+ — 5 64 ILC (C/P) Minimal 5.0 1 3 3 1 2 No No 99% 10% 1+ — 6 62 Apocrine Moderate 0.6 3.3 2 3 2 2 Yes Yes 0 0 0 — 7 83 IDC NOS Moderate 1.4 3.1 3 3 2 3 Yes No 0 0 0 — 8 35 IDC NOS Moderate 19.1 2.3 3 3 3 3 No No 0 0 0 — 9 39 IDC NOS Extensive 40.95 3 3 3 3 3 No No 0 0 1+ — 10 62 IDC NOS Moderate 7.6 1.8 3 3 3 3 Yes No 0 0 1+ — 11 53 IDC NOS Minimal 0.7 1.9 3 3 2 3 No NA 0 0 1+ — 12 88 Mucinous Moderate 0.3 6.6 2 1 1 1 No Yes 95% 60% 0 — 13 56 IDC NOS Moderate 1.4 2.5 3 3 3 3 Yes No 5% 5% 3+ — 14 65 IDC NOS Minimal 1 2.1 3 3 3 3 Yes NA 0 5% 1+ — 15 72 IDC NOS Moderate 3 1.3 3 3 3 3 No No 0 0 0 — Abbreviations: ER, estrogen receptor; FISH, fluorescence in situ hybridization; HER2, human epidermal growth factor receptor 2; HG, histological grade; IDC, invasive ductal carcinoma; ILC, invasive lobular carcinoma; LN, lymph node involvement; LVI, lymphovascular invasion; NA, not sampled; NG, nuclear grade; NOS, not otherwise specified; PR, progesterone receptor; TIL, tumor-infiltrating lymphocyte; OG, overall grade. Scoring criteria for the level of lymphocytic infiltration are defined in PATIENTS AND METHODS. Patient with concurrent astrocytoma (WHO III). Ipsilateral breast cancer recurrence. npj Breast Cancer (2015) 15005 © 2015 Breast Cancer Research Foundation/Macmillan Publishers Limited Mutations in breast cancer-infiltrating leukocytes M Kleppe et al bio-bwa.sourceforge.net; ver 0.7.4) to the full human genome. Multiple were considered true variants independent of the alternate allele mapping reads (mapping quality scores = = 0) were removed and then the frequency. Following these criteria, we identified a total of 261 BAM files were processed for base recalibration using the GATK toolkit somatic mutations (median of 5) in 14 of the 15 patients (93%). (https://www.broadinstitute.org/gatk/; ver 3.1). Mutations were called Importantly, we identified a total of 47 somatic mutations in 10 of using GATK HaplotypeCaller, which found only two events. In addition, the 15 patients affecting genes commonly mutated in hemato- the read pileups were counted at each of the known mutation sites for logical malignancies with an allelic fraction ranging from 2.9 to each sample to compute the actual depth of both the reference and 51.8% (Table 2; Supplementary Table 3). variant allele, and to compute the non-reference allele frequency for each site. Sequencing analysis of laser-capture-microdissected tumor cells We next investigated whether the mutations observed in tumor- Variant detection infiltrating leukocytes were specific to the hematopoietic cells, or Paired-end reads were aligned to the human hg19 genome with BWA whether these alleles were observed in purified breast cancer 0.6.2-r126 (ref. 26). Local realignment at indel regions and baseQ racalibration was done using the GATK suite version 2.8-1 and following cells. We performed PCR and high-depth (13,327 ± 6,897, Supple- recommendations of its authors. Variants in the targeted tumor–normal mentary Table 4) 454 sequencing on laser-capture-dissected sample pairs were called with MuTect version 1.1.4 (ref. 26). We only breast cancer cells to look for the presence of specific mutations considered variants that passed the standard Mutect filters and were not that we detected in infiltrating white blood cells (Table 2). present in the non-somatic databases (non-clinical variants from dbSNP, Importantly, for the microdissection, a system where the laser cuts NHLBI exome sequencing project, and our own internal collection of around the areas selected for microdissection, without damaging normal tissues). We focused on variants that we could validate either the cells of interest, was employed. We identified two TP53 because they were published and characterized as somatic (in COSMIC) or mutations (patient 2: TP53 p.R248L and patient 14: TP53 p.R283P) because they were present in more than 10% of the reads of the that were present in purified breast cancer cells, suggesting these corresponding exome sequencing data from the same sample. mutations originated from the epithelial, malignant clone (Supple- mentary Table 4). By contrast, all other tested somatic mutations detected in tumor-infiltrating leukocytes, including in known RESULTS leukemia genes (DNMT3A, TET2, and BCOR, Table 2) were not Targeted sequencing analysis of infiltrating white blood cells identified in breast cancer cells consistent with their origin in the We obtained fresh samples of 15 untreated primary breast cancers leukocyte component (Supplementary Table 4). The two TET2 (Table 1) and performed fluorescent-activated cell sorting to mutations were likely pathogenic as we identified a nonsense allele separate CD45-positive leukocytes from CD45-negative epithelial (TET2 p.Q1702*), which we validated by PCR and sequencing, and a cells (Figure 1a). Patients with neoadjuvant chemotherapy were mutation in a highly conserved residue in TET2 commonly mutated not studied to exclude the effects of chemotherapy on mutational in myeloid malignancies (TET2 p.E1874K). Mutations in the burden. Of the 15 patients, 10 had triple-negative breast cancer, transcriptional co-repressor BCOR, which is targeted by somatic 2 had estrogen receptor-positive disease, HER2-positive disease, mutations in myeloid leukemia, was identified in one patient. 2 had estrogen receptor-positive disease, HER2-negative disease and 1 had estrogen receptor-negative disease, progesterone Sequencing analysis of circulating leukocytes receptor-positive disease, and HER2-negative disease (Table 1). Most of the mutations identified in tumor-infiltrating leukocytes To obtain coverage for genes with known roles in malignant and not in breast cancer cells displayed mutant allelic fractions transformation, we performed capture-based sequencing of 15 ranging from 5 to 20%. This observation suggests that these paired tumor-infiltrating leukocyte and matched germline (buccal mutations were present in enriched subclones and were not rare swab) DNA samples (Table 1). We used two capture-based alleles occurring in a minority of hematopoietic stem cells as platforms containing all exons of 600 and 341 genes, respectively, previously reported in normal donors. To determine whether the and including genes implicated in hematopoietic malignancies population of tumor-infiltrating leukocytes would be enriched for and in epithelial malignancies (Supplementary Table 2 and subclones harboring somatic mutations, we used high-depth Patients and methods). Targeted capture of tumor and germline (mononuclear cells: 45,640 ± 7,486, granulocytes: 44,891 ± 7,632, DNA was performed at a mean depth of 384.98 ± 115.27 and Supplementary Table 5) 454 sequencing to look for the presence 261.50 ± 160.07 (Supplementary Table 1). Variants in the targeted of the mutations in tumor-infiltrating leukocytes in the circulating tumor–normal sample pairs were called with MuTect version 1.1.4 leukocytes from these patients. We were able to prospectively (ref. 26). Variants identified by the targeted sequencing platforms obtain peripheral blood samples in compliance with the federal that passed the recommended MuTect filters and were not found Health Insurance Portability and Accountability Act of 1996 in any of the somatic databases (non-clinical variants from dbSNP, (HIPAA) and Institutional Review Board-approved manner from NHLBI exome sequencing project, and our own internal collection seven patients in which we had identified somatic mutations in of normal tissue) were annotated as high-confidence variants. This their tumor-infiltrating leukocytes (Table 2). One mutation (patient approach identified candidate variants in known cancer genes, 2: DNMT3A p.Y533C, variant allele frequency: 0.73) was detectable including in BCOR, NOTCH2, TET2, NF1, EZH2, and JAK1 (Figure 1b). in circulating leukocytes (both mononuclear cells and granulo- Of importance, mutations in these genes have previously been cytes). Of note, the mutation in DNMT3A observed in tumor- implicated in the pathogenesis of hematologic malignancies. infiltrating leukocytes and in the peripheral blood was present at These data suggest mutations in known cancer genes are present 25-fold-higher mutant allele fraction in the tumor-infiltrating in the white blood cells infiltrating a subset of breast cancers. leukocytes compared with circulating leukocytes (Supplementary Table 5). The remaining 12 mutations were not detectable by Confirmation of identified variants using exome sequencing sequencing in circulating leukocytes, likely due to the limits of our We next performed exome sequencing of tumor-infiltrating sequencing platform. We cannot exclude that these other leukocytes (mean depth of 118.12 ± 41.29, Supplementary Table mutations were present in circulating cells at low allele burden, 1). We integrated candidate variants identified by the targeted or alternatively or additionally, in stem/progenitor cells in the sequencing panels with our exome sequencing data. We bone marrow from these patients. Taken together, these data considered candidate variants true if more than 10% of the reads demonstrate that somatic mutations are highly enriched in tumor- presented the alternate allele. Candidate variants identified by infiltrating leukocytes compared with the overall hematopoietic capture-based sequencing and previously reported in COSMIC compartment. © 2015 Breast Cancer Research Foundation/Macmillan Publishers Limited npj Breast Cancer (2015) 15005 Mutations in breast cancer-infiltrating leukocytes M Kleppe et al Figure 1. Sequencing analysis of 15 primary breast cancers identified somatically acquired mutations in tumor-infiltrating leukocytes. (a) Gating scheme for fluorescent-activated cell sorting of CD45-positive hematopoietic cells. DAPI was included as live-dead stain. Cell doublets were excluded prior to gating on APC-Cy7 (not shown). DNA extracted from the CD45-positive fraction was analyzed using three independent sequencing platforms. (b) Representative integrated genomics viewer image showing the presence of acquired mutations. Reads that do not match the reference nucleotide are colored. Gray bar chart on top displays the read depth. Reference nucleotide and protein sequence are depicted for each mutation. Variant allele frequency (VAF) and the number of altered and total reads are shown (alt|total, VAF). DAPI, 4′,6-diamidino-2-phenylindole; GFP, green fluorescent protein. DISCUSSION Table 2. Somatic mutations in known cancer genes Despite an increasing appreciation of the role of tumor-infiltrating Sample Gene Mutation Frequency inflammatory cells in solid cancers, studies to date have largely focused on cell non-autonomous interactions between tumor cells 1 EP300 p.G1777C 0.06 and stromal cells, including macrophages, neutrophils, and lympho- 2 DNMT3A p.Y533C 0.185 3 EZH2 p.A483S 0.46 cytes. In this study, we used high-throughput, next-generation TP53 p.M169I 0.029 sequencing data to demonstrate that leukocytes with somatic 4 BCOR p.P1156L 0.49 mutations in known cancer genes infiltrate many primary breast EPHA7 p.G592S 0.14 cancers. We identified and validated somatic mutations, often WT1 p.T278I 0.11 TET2 p.Q1702* 0.06 affecting known leukemia (DNTM3A, TET2,and BCOR) in tumor- EGFR p.A871E 0.042 infiltrating leukocytes but not in the cancer cells of 7 of the 15 5 ALK p.R1209Q 0.21 patients. In one case, the mutation found to be restricted to the ETV6 p.P25S 0.038 tumor-infiltrating leukocytes was also detected in the circulating 6 NOTCH2 p.P1101T 0.18 NF1 p.Q2434H 0.099 leukocytes of the same patients, but at a significantly lower SMARCA4 p.D694E 0.087 frequency. These observations provide direct evidence that at least 12 TET2 p.E1874K 0.17 some cases of primary breast cancer are infiltrated by leukocytes Mutations listed in this table were identified by targeted sequencing with with somatic mutations in genes known to be associated with an allele frequency of ⩾ 10%. Mutations occurring at a lower frequency hematologic malignancies. Notably, a subset of these mutations are were included if previously reported in COSMIC. in genes that regulate the epigenetic and transcriptional state of npj Breast Cancer (2015) 15005 © 2015 Breast Cancer Research Foundation/Macmillan Publishers Limited Mutations in breast cancer-infiltrating leukocytes M Kleppe et al hematopoietic cells. Furthermore, it suggests that the leukocytes possibility that some mutations detected in tumor biopsies might harboring such somatic mutations can be enriched within originate in hematopoietic or other stromal cells and not from the the tumor. primary tumor or metastasis, in particular studies seeking to Recent studies have highlighted the prognostic role of tumor- identify minor subclones harboring specific mutations or to infiltrating white blood cells. Studies conducted by Loi et al. and characterize intratumor genetic heterogeneity. Given that most Adams et al. recently showed that the quantification of tumor- samples subjected to whole-exome and whole-genome analysis infiltrating lymphocytes is a prognostic marker for patients with comprises a mixture of cells and stromal cells (e.g., leukocytes, triple-negative breast cancer treated with conventional che- fibroblasts, adjacent normal breast tissue) and that somatic motherapy. To date, studies investigating the prognostic role of mutations may also be present in leukocytes, if subclonal infiltrating lymphocytes have not included sequencing data on mutations are identified, defining whether these mutations are the population of infiltrating cells. In this respect, our work present in tumor cells or leukocytes would be required. In suggests it is critical to further refine our mutational and addition, subsequent studies will need to delineate which functional understanding of infiltrating hematopoietic cells. hematopoietic subsets acquire somatic mutations, and whether Increasingly, immunotherapies across multiple types of tumors the functional effects of these disease alleles varies based on the have focused on harnessing innate immune surveillance against specific hematopoietic context. Future studies with larger patient tumor initiation and progression. The observation that a subset cohorts will be needed to delineate which tumor-infiltrating of cancers harbor mutated infiltrating leukocytes may have hematopoietic cells are most commonly characterized by somatic implications on prognosis and on the response to cytotoxic, mutations, and to determine whether mutations in specific targeted, and immune-mediated therapies. It will be important to hematopoietic subsets can impact tumor growth and therapeutic pursue studies to enumerate functional interactions between response. In conclusion, the identification of somatic mutations in breast cancer cells and leukocytes with somatic mutations, breast cancer tumor-infiltrating leukocytes challenges the current through cell–cell contact and/or paracrine secretion of cytokines paradigm that although aberrant, the tumor microenvironment and chemokines. In addition, it will be important to better would not be targeted by clonal somatic alterations. The clinical delineate the relationship between tumor formation and clonal implications regarding carcinogenesis, clinical course, and evolution in infiltrating hematopoietic cells in the setting of response to treatment of these findings warrants further study. immunotherapy. It is not yet known whether the acquisition of These studies will likely reveal novel opportunities for cancer mutations within leukocytes impacts tumor initiation and/or prevention, detection, prognostication, and the development of progression. If subsequent functional studies demonstrate that therapies, which may be directed not only to cancer cells, but also clonal leukocytes contribute to cancer progression, this would lead to efforts to target, in parallel, epithelial tumor cells and to cells from the microenvironment harboring somatic genetic infiltrating leukocytes in order to increase therapeutic efficacy. alterations. In the previously published studies by Adams et al. and Loi et al. of tumor-infiltrating lymphocytes and breast cancer ACKNOWLEDGMENTS prognosis, the majority of patients received anthracyclines. We are grateful to the MSK Center for Molecular Oncology for their assistance with Secondary hematologic malignancies as a result of chemotherapy, sequencing studies. including alkylating agents and anthracyclines, remain a signifi- cant risk associated with systemic therapy for epithelial 29,30 tumors. Our work raises the possibility that some patients CONTRIBUTIONS may be at increased risk for secondary leukemias based on the MKl, EC, RLL, and LN conceived the study. MKl, EC, ZG, RB, AV, MFB, BW, JSR-F, RLL, presence of oncogenic mutations in infiltrating white cells, which and LN designed experiments. MKl, MKe, LB, AV, DY, and ZG performed experiments. pre-exist before systemic therapy, and which are selected for by MKl, LB, FTR, NS, and MFB analyzed whole-exome and targeted sequencing data. BW cytotoxic chemotherapy. This will need to be examined in and JSR-F performed laser-capture microdissection experiments. HYW and EB prospective studies, and it will be important to determine whether screened cases and performed clinicopathological analysis. BB and EC recruited the presence of mutant leukocyte clones at the time of diagnosis patients and collected samples. MKl, EC, RLL, and LN wrote the manuscript. should impact therapeutic decisions in different malignant contexts. Our data resonate with the observations that by the age 70 COMPETING INTERESTS years, at least 5% of people have known leukemia mutations in a The authors declare no conflict of interest. subset of circulating hematopoietic cells. This includes muta- tions in the same genes, such as in DNMT3A, TET2, and BCOR, which we identified in tumor-infiltrating leukocytes in our study. FUNDING However, we identified a much higher proportion of patients with This work was supported by the Breast Cancer Research Foundation, by the Cure mutations in their tumor-associated leukocytes (93%, 14 out of Breast Cancer Foundation, and by NCI 1R01CA151949-01 to RLL. MKl is a fellow of the 15), including in patients at a younger age. These data suggest Leukemia and Lymphoma Society and was previously supported by an EMBO these mutations are not solely a function of age-related mutations, Postdoctoral Fellowship. but rather represent earlier somatic events in patients with breast cancer or tumor-specific enrichment of mutant hematopoietic REFERENCES cells. It will be important to identify whether patients harboring 1 Cancer Genome Atlas Network. Comprehensive molecular portraits of human leukemogenic mutations in infiltrating leukocytes are at increased breast tumours. Nature 2012; 490:61–70. risk for future leukemias, and whether systemic therapies 2 Ellis MJ, Ding L, Shen D, Luo J, Suman VJ, Wallis JW et al. Whole-genome analysis including cytotoxic chemotherapy increase the selective advan- informs breast cancer response to aromatase inhibition. Nature 2012; 486: tage of pre-existing mutant hematopoietic clones. Moreover, it will 353–360. be important to determine how mutant breast cancer cells might 3 Acharyya S, Oskarsson T, Vanharanta S, Malladi S, Kim J, Morris PG et al. A CXCL1 influence the behavior and clonal selection of mutant hemato- paracrine network links cancer chemoresistance and metastasis. Cell 2012; 150: poietic cells, and vice versa, as their mutual presence may lead to 165–178. clonal changes in both compartments. 4 Karnoub AE, Dash AB, Vo AP, Sullivan A, Brooks MW, Bell GW et al. Mesenchymal Although our studies identified mutations in purified leuko- stem cells within tumour stroma promote breast cancer metastasis. Nature 2007; cytes, studies of en bloc tumor samples will need to consider the 449:557–563. © 2015 Breast Cancer Research Foundation/Macmillan Publishers Limited npj Breast Cancer (2015) 15005 Mutations in breast cancer-infiltrating leukocytes M Kleppe et al 5 Kurose K, Gilley K, Matsumoto S, Watson PH, Zhou XP, Eng C. Frequent somatic 20 Xie M, Lu C, Wang J, McLellan MD, Johnson KJ, Wendl MC et al. Age-related mutations in PTEN and TP53 are mutually exclusive in the stroma of breast car- mutations associated with clonal hematopoietic expansion and malignancies. Nat cinomas. Nat Genet 2002; 32:355–357. Med 2014; 20: 1472–1478. 6 Hu M, Yao J, Cai L, Bachman KE, van den Brule F, Velculescu V et al. Distinct epi- 21 Jaiswal S, Fontanillas P, Flannick J, Manning A, Grauman PV, Mar BG et al. Age- genetic changes in the stromal cells of breast cancers. Nat Genet 2005; 37:899–905. related clonal hematopoiesis associated with adverse outcomes. N Engl J Med 7 Allinen M, Beroukhim R, Cai L, Brennan C, Lahti-Domenici J, Huang H et al. 2014; 371:2488–2498. Molecular characterization of the tumor microenvironment in breast cancer. 22 Hammond ME, Hayes DF, Dowsett M, Allred DC, Hagerty KL, Badve S et al. Cancer Cell 2004; 6:17–32. American Society of Clinical Oncology/College Of American Pathologists 8 Granot Z, Henke E, Comen EA, King TA, Norton L, Benezra R. Tumor entrained guideline recommendations for immunohistochemical testing of estrogen and neutrophils inhibit seeding in the premetastatic lung. Cancer Cell 2011; 20:300–314. progesterone receptors in breast cancer. J Clin Oncol 2010; 28: 2784–2795. 9 Grivennikov SI, Greten FR, Karin M. Immunity, inflammation, and cancer. Cell 2010; 23 Wolff AC, Hammond ME, Hicks DG, Dowsett M, McShane LM, Allison KH et al. 140:883–899. Recommendations for human epidermal growth factor receptor 2 testing 10 Orimo A, Weinberg RA. Stromal fibroblasts in cancer: a novel tumor-promoting in breast cancer: American Society of Clinical Oncology/College of American cell type. Cell Cycle 2006; 5: 1597–1601. Pathologists clinical practice guideline update. J Clin Oncol 2013; 31: 11 Li HJ, Reinhardt F, Herschman HR, Weinberg RA. Cancer-stimulated mesenchymal 3997–4013. stem cells create a carcinoma stem cell niche via prostaglandin E2 signaling. 24 Westbury CB, Reis-Filho JS, Dexter T, Mahler-Araujo B, Fenwick K, Iravani M et al. Cancer Discov 2012; 2:840–855. Genome-wide transcriptomic profiling of microdissected human breast 12 Mahmoud SM, Paish EC, Powe DG, Macmillan RD, Grainge MJ, Lee AH et al. tissue reveals differential expression of KIT (c-Kit, CD117) and oestrogen Tumor-infiltrating CD8+ lymphocytes predict clinical outcome in breast cancer. J receptor-alpha (ERalpha) in response to therapeutic radiation. J Pathol 2009; 219: Clin Oncol 2011; 29: 1949–1955. 131–140. 13 Mohammed ZM, Going JJ, Edwards J, Elsberger B, McMillan DC. The relationship 25 Iyer G, Hanrahan AJ, Milowsky MI, Al-Ahmadie H, Scott SN, Janakiraman M et al. between lymphocyte subsets and clinico-pathological determinants of survival in Genome sequencing identifies a basis for everolimus sensitivity. Science 2012; patients with primary operable invasive ductal breast cancer. Br J Cancer 2013; 338:221. 109: 1676–1684. 26 Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler 14 Loi S, Sirtaine N, Piette F, Salgado R, Viale G, Van Eenoo F et al. Prognostic and transform. Bioinformatics 2009; 25: 1754–1760. predictive value of tumor-infiltrating lymphocytes in a phase III randomized 27 McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A et al. adjuvant breast cancer trial in node-positive breast cancer comparing the addi- The Genome Analysis Toolkit: a MapReduce framework for analyzing tion of docetaxel to doxorubicin with doxorubicin-based chemotherapy: BIG 02- next-generation DNA sequencing data. Genome Res 2010; 20: 1297–1303. 98. J Clin Oncol 2013; 31: 860–867. 28 Page DB, Naidoo J, McArthur HL. Emerging immunotherapy strategies in 15 Adams S, Gray RJ, Demaria S, Goldstein L, Perez EA, Shulman LN et al. Prognostic breast cancer. Immunotherapy 2014; 6:195–209. value of tumor-infiltrating lymphocytes in triple-negative breast cancers from two 29 Wolff AC, Blackford AL, Visvanathan K, Rugo HS, Moy B, Goldstein LJ et al. phase III randomized adjuvant breast cancer trials: ECOG 2197 and ECOG 1199. J Risk of marrow neoplasms after adjuvant breast cancer therapy: the national Clin Oncol 2014; 32: 2959–2966. comprehensive cancer network experience. J Clin Oncol 2015; 33: 340–348. 16 Busque L, Patel JP, Figueroa ME, Vasanthakumar A, Provost S, Hamilou Z et al. 30 Valentini CG, Fianchi L, Voso MT, Caira M, Leone G, Pagano L. Incidence of Recurrent somatic TET2 mutations in normal elderly individuals with clonal acute myeloid leukemia after breast cancer. Mediterr J Hematol Infect Dis 2011; 3: hematopoiesis. Nat Genet 2012; 44: 1179–1181. e2011069. 17 Moran-Crusio K, Reavie L, Shih A, Abdel-Wahab O, Ndiaye-Lobry D, Lobry C et al. Tet2 loss leads to increased hematopoietic stem cell self-renewal and myeloid transformation. Cancer Cell 2011; 20:11–24. This work is licensed under a Creative Commons Attribution 4.0 18 Quivoron C, Couronne L, Della Valle V, Lopez CK, Plo I, Wagner-Ballon O et al. TET2 International License. The images or other third party material in this inactivation results in pleiotropic hematopoietic abnormalities in mouse and is a article are included in the article’s Creative Commons license, unless indicated recurrent event during human lymphomagenesis. Cancer Cell 2011; 20:25–38. otherwise in the credit line; if the material is not included under the Creative Commons 19 Ko M, Bandukwala HS, An J, Lamperti ED, Thompson EC, Hastie R et al. Ten-Eleven- license, users will need to obtain permission from the license holder to reproduce the Translocation 2 (TET2) negatively regulates homeostasis and differentiation of material. To view a copy of this license, visit http://creativecommons.org/licenses/ hematopoietic stem cells in mice. Proc Natl Acad Sci USA 2011; 108: 14566–14571. by/4.0/ Supplementary Information accompanies the paper on the npj Breast Cancer website (http://www.nature.com/npjbcancer) npj Breast Cancer (2015) 15005 © 2015 Breast Cancer Research Foundation/Macmillan Publishers Limited http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png npj Breast Cancer Springer Journals

Loading next page...
 
/lp/springer-journals/somatic-mutations-in-leukocytes-infiltrating-primary-breast-cancers-9dmU76zRH5
Publisher
Springer Journals
Copyright
Copyright © 2015 by The Author(s)
Subject
Biomedicine; Biomedicine, general; Cancer Research; Oncology; Human Genetics; Cell Biology
eISSN
2374-4677
DOI
10.1038/npjbcancer.2015.5
Publisher site
See Article on Publisher Site

Abstract

www.nature.com/npjbcancer All rights reserved 2374-4677/15 ARTICLE OPEN Somatic mutations in leukocytes infiltrating primary breast cancers 1,8 2,8 3 1 2 4 1 Maria Kleppe , Elizabeth Comen , Hannah Y Wen , Lennart Bastian , Brian Blum , Franck T Rapaport , Matthew Keller , 5 4 6 6 5 1,3 3 1,3 Zvika Granot , Nicolas Socci , Agnès Viale , Daoqi You , Robert Benezra , Britta Weigelt , Edi Brogi , Michael F Berger , 1,3 1,7 2 Jorge S Reis-Filho , Ross L Levine and Larry Norton BACKGROUND: Malignant transformation requires the interaction of cancer cells with their microenvironment, including infiltrating leukocytes. However, somatic mutational studies have focused on alterations in cancer cells, assuming that the microenvironment is genetically normal. Because we hypothesized that this might not be a valid assumption, we performed exome sequencing and targeted sequencing to investigate for the presence of pathogenic mutations in tumor-associated leukocytes in breast cancers. METHODS: We used targeted sequencing and exome sequencing to evaluate the presence of mutations in sorted tumor- infiltrating CD45-positive cells from primary untreated breast cancers. We used high-depth sequencing to determine the presence/ absence of the mutations we identified in breast cancer-infiltrating leukocytes in purified tumor cells and in circulating blood cells. RESULTS: Capture-based sequencing of 15 paired tumor-infiltrating leukocytes and matched germline DNA identified variants in known cancer genes in all 15 primary breast cancer patients in our cohort. We validated the presence of mutations identified by targeted sequencing in infiltrating leukocytes through orthogonal exome sequencing. Ten patients harbored alterations previously reported as somatically acquired variants, including in known leukemia genes (DNTM3A, TET2, and BCOR). One of the mutations observed in the tumor-infiltrating leukocytes was also detected in the circulating leukocytes of the same patients at a lower allele frequency than observed in the tumor-infiltrating cells. CONCLUSIONS: Here we show that somatic mutations, including mutations in known cancer genes, are present in the leukocytes infiltrating a subset of primary breast cancers. This observation allows for the possibility that the cancer cells interact with mutant infiltrating leukocytes, which has many potential clinical implications. npj Breast Cancer (2015) 1, 15005; doi:10.1038/npjbcancer.2015.5; published online 10 June 2015 8,9 INTRODUCTION mediate primary tumor growth and metastasis. Recent evidence suggests that tumor-associated stromal cells and infiltrating In the past decade, targeted, whole-exome, and whole-genome leukocytes function differently than circulating or bone marrow sequencing studies have delineated a spectrum of somatic 3,10,11 1,2 resident hematopoietic cells. In particular, several studies mutations in human malignancies. These include large-scale have indicated that the content of lymphoid and myeloid cells sequencing studies in breast cancer, which have identified 12–15 infiltrating breast cancers correlates with clinical outcome. recurrent mutations in genes and pathways that contribute to We recently demonstrated that some older individuals have malignant transformation and to therapeutic response. The clinically inapparent, clonal hematopoiesis characterized by general approach to limit high-throughput genomic studies to recurrent, somatic mutations in TET2 (ref. 16). Of note, Tet2 loss cancer cells themselves is, however, in contrast to burgeoning in the hematopoietic compartment leads to increased self-renewal evidence that cancer cells interact with their microenvironment, 17–19 and myeloid bias of hematopoietic cells. These data, and including stromal cell constituents, infiltrating white blood cells, recent genomic studies of circulating hematopoietic stem cells and circulating inflammatory cytokines originating from local and 3,4 from normal volunteers and from The Cancer Genome Atlas distant sites. The integrity of the genome in these non-cancer patients suggest that morphologically normal hematopoietic cells cellular elements is therefore germane. Indeed, previous studies acquire mutations over time, most commonly in known leukemia have shown that stromal cells found in breast cancers are 20,21 disease alleles. The observation that normal individuals may characterized by site- and cell-type-specific epigenetic alterations, harbor oncogenic mutations in hematopoietic cells and the and reports suggesting the presence of somatic mutations in the 5–7 interaction between epithelial cancer cells and infiltrating leuko- tumor microenvironment are also on record. This suggests that cells other than the cancer cells themselves can acquire properties cytes raises the possibility of clonal selection in infiltrating leuko- that contribute to tumorigenesis. In addition to tissue-specific cytes. Hence, we sought to define whether tumor-infiltrating stromal cells, circulating and tumor-infiltrating leukocytes can leukocytes in breast cancer would harbor somatic mutations, and 1 2 Human Oncology and Pathogenesis Program, Memorial Sloan Kettering Cancer Center, New York, NY, USA; Breast Cancer Service, Department of Medicine, Memorial Sloan 3 4 Kettering Cancer Center, New York, NY, USA; Department of Pathology, Memorial Sloan Kettering Cancer Center, New York, NY, USA; The Bioinformatics Core, Memorial Sloan 5 6 Kettering Cancer Center, New York, NY, USA; Cancer Biology and Genetics Program, Memorial Sloan Kettering Cancer Center, New York, NY, USA; The Genomics Core, Memorial Sloan Kettering Cancer Center, New York, NY, USA and Leukemia Service, Department of Medicine, Memorial Sloan Kettering Cancer Center, New York, NY, USA. Correspondence: L Norton (nortonl@mskcc.org) These authors contributed equally to this work. Received 14 March 2015; revised 24 March 2015; accepted 26 March 2015 © 2015 Breast Cancer Research Foundation/Macmillan Publishers Limited Mutations in breast cancer-infiltrating leukocytes M Kleppe et al whether these would be enriched in the tumor as compared with Laser-capture microdissection of tumor cells peripheral blood. Ten consecutive 8-μm-thick nuclear fast red-stained representative sections of the tumors were subjected to laser-assisted microdissection on a PALM Robot MicroBeam laser microdissection system (Zeiss, Thornwood, NY, USA), as previously described. First, non-neoplastic cells, PATIENTS AND METHODS including inflammatory cells, stroma, and normal breast, were ablated. We Patient materials subsequently microdissected only histologically unequivocal neoplastic Breast cancer samples were collected from consecutive patients with cells from each sample under a microscope. Tissue was microdissected primary triple-negative breast cancer who underwent surgery at Memorial directly into extraction buffer, and DNA was extracted using the DNeasy Sloan Kettering Cancer Center between 2012 and 2013 (Table 1). Patients Blood and Tissue Kit (Qiagen, Valencia, CA, USA) and quantified with the treated with neoadjuvant chemotherapy were excluded from the study. Qubit Fluorometer (Invitrogen, Life Technologies, Norwalk, CT, USA). Non-triple-negative breast cancers showing prominent lymphocytic infiltrate in core biopsies were also included. All specimens were sectioned and DNA extraction and whole-genome amplification processed for routine pathological examination. Hematoxylin and eosin- DNA was extracted using the QiaAmp DNA kit (Qiagen) following the stained slides were reviewed by breast pathologists to establish the manufacture’s instructions. Buccal swabs were processed using the diagnoses. Estrogen receptor, progesterone receptor and human epidermal QiaAmp DNA Mini kit (Qiagen) following the manufacture’s instructions. growth factor receptor 2 (HER2) status was evaluated by immunohisto- The quality of DNA samples was analyzed with the Agilent Bioanalyzer chemistry according to the American Society of Clinical Oncology/College of 2100 (Agilent Technologies, Santa Clara, NY, USA). Samples with insuffi- 22,23 American Pathologists guidelines. HER2 fluorescence in situ hybridiza- cient amount of DNA (o500 ng) were whole-genome amplified using the tion was performed in one case with equivocal results by immunohis- REPLI-g Mini kit (Qiagen) prior to further use in downstream applications. tochemistry. Evaluation of tumor-infiltrating leukocytes was performed as QPCR was performed to assess the quality of WGA DNA. described by Loi et al. Tumor-infiltrating leukocytes were scored as following: extensive, ⩾ 50% infiltration of either stromal or intratumoral Exome sequencing and targeted capture sequencing lymphocytes; moderate = 5–10%; minimal, ⩽ 5%. Buccal swab samples were DNA extracted from sorted CD45-positive tumor-infiltrating leukocytes and collected from each patient. Mononuclear cells and granulocytes were buccal swabs (Supplementary Table 1) was sheared to an average size of isolated from peripheral blood following a standard Ficoll protocol. A detailed 180 ± 80 bp for exome sequencing. For DNA library preparation, 200–250- description on clinicopathological features of each patient is listed in Table 1. bp fragments were selected and subjected to PCR amplification. The library was then hybridized to the Agilent SureSelect Human All Exon Kit (Agilent Isolation and processing of tumor-infiltrating cells Technologies) and sequencing was performed on the SOLiD 3plus or SOLiD 4 (Applied Biosystems, Grand Island, NY, USA). Targeted sequencing All patients included in this study gave informed consent. Fresh tumor of tumor-infiltrating leukocytes and matched germline DNA of each cells, stromal cells, and tumor-infiltrating leukocytes were dissociated from patient was performed as previously described. the primary tumors by scraping the cutting surface 5–10 times with a surgical scalpel blade. Cell material was collected by rinsing the blade in phosphate-buffered saline. Cells were spun down and resuspended in red PCR and 454 sequencing analysis cell lysis buffer to remove red blood cells prior to staining with an anti- Sequence reactions were performed on DNA extracted from mononuclear human CD45-PE-Cy7- or CD45-APC-Cy7-conjugated flow antibody in FACS cells, granulocytes, laser-capture-microdissected tumor cells, and buffer (phosphate-buffered saline supplemented with 2% bovine serum tumor-infiltrating leukocytes. All PCR reactions were performed using albumin). Cells were stained for 20 min in the dark at room temperature, amplicon-specific fusion primers. Fusion primers contained next to the washed once with FACS buffer, and passed through a filter. 4′,6-diamidino- template-specific sequence a directional primer at the 5′-end followed by a 2-phenylindole was added before sorting to discriminate live and dead multiplex identifier for barcode sample identification. Samples from 6–8 cells. CD45-positive cells were then purified using a FACSAriaIII Cell Sorter different patients were mixed, processed for 454 deep sequencing, and run (MSKCC Flow Core Facility, Memorial Sloan Kettering Cancer Center, New on a Genome Sequencer FLX instrument (454 Life Sciences, Branford, CT, York, NY, USA). USA). Data were mapped with BWA MEM (BWA is freely available at http:// Table 1. Summary of clinicopathological features ID Age (years) Type Level of lymphocytic CD45 (%) Size (cm) HG NG Mitosis OG LVI LN ER PR HER2 FISH infiltration 1 40 IDC NOS Moderate 36.6 2 3 3 3 3 No No 0 0 0 — 2 72 IDC NOS Moderate 11.89 1.5 3 3 3 3 No No 0 0 0 — 3 37 IDC NOS Extensive 12.5 4.5 3 3 3 3 No No o1% o1% 1+ to 2+ 1.3 4 35 IDC NOS Moderate 0.4 5 2 3 3 3 Yes Yes 95% 90% 3+ — 5 64 ILC (C/P) Minimal 5.0 1 3 3 1 2 No No 99% 10% 1+ — 6 62 Apocrine Moderate 0.6 3.3 2 3 2 2 Yes Yes 0 0 0 — 7 83 IDC NOS Moderate 1.4 3.1 3 3 2 3 Yes No 0 0 0 — 8 35 IDC NOS Moderate 19.1 2.3 3 3 3 3 No No 0 0 0 — 9 39 IDC NOS Extensive 40.95 3 3 3 3 3 No No 0 0 1+ — 10 62 IDC NOS Moderate 7.6 1.8 3 3 3 3 Yes No 0 0 1+ — 11 53 IDC NOS Minimal 0.7 1.9 3 3 2 3 No NA 0 0 1+ — 12 88 Mucinous Moderate 0.3 6.6 2 1 1 1 No Yes 95% 60% 0 — 13 56 IDC NOS Moderate 1.4 2.5 3 3 3 3 Yes No 5% 5% 3+ — 14 65 IDC NOS Minimal 1 2.1 3 3 3 3 Yes NA 0 5% 1+ — 15 72 IDC NOS Moderate 3 1.3 3 3 3 3 No No 0 0 0 — Abbreviations: ER, estrogen receptor; FISH, fluorescence in situ hybridization; HER2, human epidermal growth factor receptor 2; HG, histological grade; IDC, invasive ductal carcinoma; ILC, invasive lobular carcinoma; LN, lymph node involvement; LVI, lymphovascular invasion; NA, not sampled; NG, nuclear grade; NOS, not otherwise specified; PR, progesterone receptor; TIL, tumor-infiltrating lymphocyte; OG, overall grade. Scoring criteria for the level of lymphocytic infiltration are defined in PATIENTS AND METHODS. Patient with concurrent astrocytoma (WHO III). Ipsilateral breast cancer recurrence. npj Breast Cancer (2015) 15005 © 2015 Breast Cancer Research Foundation/Macmillan Publishers Limited Mutations in breast cancer-infiltrating leukocytes M Kleppe et al bio-bwa.sourceforge.net; ver 0.7.4) to the full human genome. Multiple were considered true variants independent of the alternate allele mapping reads (mapping quality scores = = 0) were removed and then the frequency. Following these criteria, we identified a total of 261 BAM files were processed for base recalibration using the GATK toolkit somatic mutations (median of 5) in 14 of the 15 patients (93%). (https://www.broadinstitute.org/gatk/; ver 3.1). Mutations were called Importantly, we identified a total of 47 somatic mutations in 10 of using GATK HaplotypeCaller, which found only two events. In addition, the 15 patients affecting genes commonly mutated in hemato- the read pileups were counted at each of the known mutation sites for logical malignancies with an allelic fraction ranging from 2.9 to each sample to compute the actual depth of both the reference and 51.8% (Table 2; Supplementary Table 3). variant allele, and to compute the non-reference allele frequency for each site. Sequencing analysis of laser-capture-microdissected tumor cells We next investigated whether the mutations observed in tumor- Variant detection infiltrating leukocytes were specific to the hematopoietic cells, or Paired-end reads were aligned to the human hg19 genome with BWA whether these alleles were observed in purified breast cancer 0.6.2-r126 (ref. 26). Local realignment at indel regions and baseQ racalibration was done using the GATK suite version 2.8-1 and following cells. We performed PCR and high-depth (13,327 ± 6,897, Supple- recommendations of its authors. Variants in the targeted tumor–normal mentary Table 4) 454 sequencing on laser-capture-dissected sample pairs were called with MuTect version 1.1.4 (ref. 26). We only breast cancer cells to look for the presence of specific mutations considered variants that passed the standard Mutect filters and were not that we detected in infiltrating white blood cells (Table 2). present in the non-somatic databases (non-clinical variants from dbSNP, Importantly, for the microdissection, a system where the laser cuts NHLBI exome sequencing project, and our own internal collection of around the areas selected for microdissection, without damaging normal tissues). We focused on variants that we could validate either the cells of interest, was employed. We identified two TP53 because they were published and characterized as somatic (in COSMIC) or mutations (patient 2: TP53 p.R248L and patient 14: TP53 p.R283P) because they were present in more than 10% of the reads of the that were present in purified breast cancer cells, suggesting these corresponding exome sequencing data from the same sample. mutations originated from the epithelial, malignant clone (Supple- mentary Table 4). By contrast, all other tested somatic mutations detected in tumor-infiltrating leukocytes, including in known RESULTS leukemia genes (DNMT3A, TET2, and BCOR, Table 2) were not Targeted sequencing analysis of infiltrating white blood cells identified in breast cancer cells consistent with their origin in the We obtained fresh samples of 15 untreated primary breast cancers leukocyte component (Supplementary Table 4). The two TET2 (Table 1) and performed fluorescent-activated cell sorting to mutations were likely pathogenic as we identified a nonsense allele separate CD45-positive leukocytes from CD45-negative epithelial (TET2 p.Q1702*), which we validated by PCR and sequencing, and a cells (Figure 1a). Patients with neoadjuvant chemotherapy were mutation in a highly conserved residue in TET2 commonly mutated not studied to exclude the effects of chemotherapy on mutational in myeloid malignancies (TET2 p.E1874K). Mutations in the burden. Of the 15 patients, 10 had triple-negative breast cancer, transcriptional co-repressor BCOR, which is targeted by somatic 2 had estrogen receptor-positive disease, HER2-positive disease, mutations in myeloid leukemia, was identified in one patient. 2 had estrogen receptor-positive disease, HER2-negative disease and 1 had estrogen receptor-negative disease, progesterone Sequencing analysis of circulating leukocytes receptor-positive disease, and HER2-negative disease (Table 1). Most of the mutations identified in tumor-infiltrating leukocytes To obtain coverage for genes with known roles in malignant and not in breast cancer cells displayed mutant allelic fractions transformation, we performed capture-based sequencing of 15 ranging from 5 to 20%. This observation suggests that these paired tumor-infiltrating leukocyte and matched germline (buccal mutations were present in enriched subclones and were not rare swab) DNA samples (Table 1). We used two capture-based alleles occurring in a minority of hematopoietic stem cells as platforms containing all exons of 600 and 341 genes, respectively, previously reported in normal donors. To determine whether the and including genes implicated in hematopoietic malignancies population of tumor-infiltrating leukocytes would be enriched for and in epithelial malignancies (Supplementary Table 2 and subclones harboring somatic mutations, we used high-depth Patients and methods). Targeted capture of tumor and germline (mononuclear cells: 45,640 ± 7,486, granulocytes: 44,891 ± 7,632, DNA was performed at a mean depth of 384.98 ± 115.27 and Supplementary Table 5) 454 sequencing to look for the presence 261.50 ± 160.07 (Supplementary Table 1). Variants in the targeted of the mutations in tumor-infiltrating leukocytes in the circulating tumor–normal sample pairs were called with MuTect version 1.1.4 leukocytes from these patients. We were able to prospectively (ref. 26). Variants identified by the targeted sequencing platforms obtain peripheral blood samples in compliance with the federal that passed the recommended MuTect filters and were not found Health Insurance Portability and Accountability Act of 1996 in any of the somatic databases (non-clinical variants from dbSNP, (HIPAA) and Institutional Review Board-approved manner from NHLBI exome sequencing project, and our own internal collection seven patients in which we had identified somatic mutations in of normal tissue) were annotated as high-confidence variants. This their tumor-infiltrating leukocytes (Table 2). One mutation (patient approach identified candidate variants in known cancer genes, 2: DNMT3A p.Y533C, variant allele frequency: 0.73) was detectable including in BCOR, NOTCH2, TET2, NF1, EZH2, and JAK1 (Figure 1b). in circulating leukocytes (both mononuclear cells and granulo- Of importance, mutations in these genes have previously been cytes). Of note, the mutation in DNMT3A observed in tumor- implicated in the pathogenesis of hematologic malignancies. infiltrating leukocytes and in the peripheral blood was present at These data suggest mutations in known cancer genes are present 25-fold-higher mutant allele fraction in the tumor-infiltrating in the white blood cells infiltrating a subset of breast cancers. leukocytes compared with circulating leukocytes (Supplementary Table 5). The remaining 12 mutations were not detectable by Confirmation of identified variants using exome sequencing sequencing in circulating leukocytes, likely due to the limits of our We next performed exome sequencing of tumor-infiltrating sequencing platform. We cannot exclude that these other leukocytes (mean depth of 118.12 ± 41.29, Supplementary Table mutations were present in circulating cells at low allele burden, 1). We integrated candidate variants identified by the targeted or alternatively or additionally, in stem/progenitor cells in the sequencing panels with our exome sequencing data. We bone marrow from these patients. Taken together, these data considered candidate variants true if more than 10% of the reads demonstrate that somatic mutations are highly enriched in tumor- presented the alternate allele. Candidate variants identified by infiltrating leukocytes compared with the overall hematopoietic capture-based sequencing and previously reported in COSMIC compartment. © 2015 Breast Cancer Research Foundation/Macmillan Publishers Limited npj Breast Cancer (2015) 15005 Mutations in breast cancer-infiltrating leukocytes M Kleppe et al Figure 1. Sequencing analysis of 15 primary breast cancers identified somatically acquired mutations in tumor-infiltrating leukocytes. (a) Gating scheme for fluorescent-activated cell sorting of CD45-positive hematopoietic cells. DAPI was included as live-dead stain. Cell doublets were excluded prior to gating on APC-Cy7 (not shown). DNA extracted from the CD45-positive fraction was analyzed using three independent sequencing platforms. (b) Representative integrated genomics viewer image showing the presence of acquired mutations. Reads that do not match the reference nucleotide are colored. Gray bar chart on top displays the read depth. Reference nucleotide and protein sequence are depicted for each mutation. Variant allele frequency (VAF) and the number of altered and total reads are shown (alt|total, VAF). DAPI, 4′,6-diamidino-2-phenylindole; GFP, green fluorescent protein. DISCUSSION Table 2. Somatic mutations in known cancer genes Despite an increasing appreciation of the role of tumor-infiltrating Sample Gene Mutation Frequency inflammatory cells in solid cancers, studies to date have largely focused on cell non-autonomous interactions between tumor cells 1 EP300 p.G1777C 0.06 and stromal cells, including macrophages, neutrophils, and lympho- 2 DNMT3A p.Y533C 0.185 3 EZH2 p.A483S 0.46 cytes. In this study, we used high-throughput, next-generation TP53 p.M169I 0.029 sequencing data to demonstrate that leukocytes with somatic 4 BCOR p.P1156L 0.49 mutations in known cancer genes infiltrate many primary breast EPHA7 p.G592S 0.14 cancers. We identified and validated somatic mutations, often WT1 p.T278I 0.11 TET2 p.Q1702* 0.06 affecting known leukemia (DNTM3A, TET2,and BCOR) in tumor- EGFR p.A871E 0.042 infiltrating leukocytes but not in the cancer cells of 7 of the 15 5 ALK p.R1209Q 0.21 patients. In one case, the mutation found to be restricted to the ETV6 p.P25S 0.038 tumor-infiltrating leukocytes was also detected in the circulating 6 NOTCH2 p.P1101T 0.18 NF1 p.Q2434H 0.099 leukocytes of the same patients, but at a significantly lower SMARCA4 p.D694E 0.087 frequency. These observations provide direct evidence that at least 12 TET2 p.E1874K 0.17 some cases of primary breast cancer are infiltrated by leukocytes Mutations listed in this table were identified by targeted sequencing with with somatic mutations in genes known to be associated with an allele frequency of ⩾ 10%. Mutations occurring at a lower frequency hematologic malignancies. Notably, a subset of these mutations are were included if previously reported in COSMIC. in genes that regulate the epigenetic and transcriptional state of npj Breast Cancer (2015) 15005 © 2015 Breast Cancer Research Foundation/Macmillan Publishers Limited Mutations in breast cancer-infiltrating leukocytes M Kleppe et al hematopoietic cells. Furthermore, it suggests that the leukocytes possibility that some mutations detected in tumor biopsies might harboring such somatic mutations can be enriched within originate in hematopoietic or other stromal cells and not from the the tumor. primary tumor or metastasis, in particular studies seeking to Recent studies have highlighted the prognostic role of tumor- identify minor subclones harboring specific mutations or to infiltrating white blood cells. Studies conducted by Loi et al. and characterize intratumor genetic heterogeneity. Given that most Adams et al. recently showed that the quantification of tumor- samples subjected to whole-exome and whole-genome analysis infiltrating lymphocytes is a prognostic marker for patients with comprises a mixture of cells and stromal cells (e.g., leukocytes, triple-negative breast cancer treated with conventional che- fibroblasts, adjacent normal breast tissue) and that somatic motherapy. To date, studies investigating the prognostic role of mutations may also be present in leukocytes, if subclonal infiltrating lymphocytes have not included sequencing data on mutations are identified, defining whether these mutations are the population of infiltrating cells. In this respect, our work present in tumor cells or leukocytes would be required. In suggests it is critical to further refine our mutational and addition, subsequent studies will need to delineate which functional understanding of infiltrating hematopoietic cells. hematopoietic subsets acquire somatic mutations, and whether Increasingly, immunotherapies across multiple types of tumors the functional effects of these disease alleles varies based on the have focused on harnessing innate immune surveillance against specific hematopoietic context. Future studies with larger patient tumor initiation and progression. The observation that a subset cohorts will be needed to delineate which tumor-infiltrating of cancers harbor mutated infiltrating leukocytes may have hematopoietic cells are most commonly characterized by somatic implications on prognosis and on the response to cytotoxic, mutations, and to determine whether mutations in specific targeted, and immune-mediated therapies. It will be important to hematopoietic subsets can impact tumor growth and therapeutic pursue studies to enumerate functional interactions between response. In conclusion, the identification of somatic mutations in breast cancer cells and leukocytes with somatic mutations, breast cancer tumor-infiltrating leukocytes challenges the current through cell–cell contact and/or paracrine secretion of cytokines paradigm that although aberrant, the tumor microenvironment and chemokines. In addition, it will be important to better would not be targeted by clonal somatic alterations. The clinical delineate the relationship between tumor formation and clonal implications regarding carcinogenesis, clinical course, and evolution in infiltrating hematopoietic cells in the setting of response to treatment of these findings warrants further study. immunotherapy. It is not yet known whether the acquisition of These studies will likely reveal novel opportunities for cancer mutations within leukocytes impacts tumor initiation and/or prevention, detection, prognostication, and the development of progression. If subsequent functional studies demonstrate that therapies, which may be directed not only to cancer cells, but also clonal leukocytes contribute to cancer progression, this would lead to efforts to target, in parallel, epithelial tumor cells and to cells from the microenvironment harboring somatic genetic infiltrating leukocytes in order to increase therapeutic efficacy. alterations. In the previously published studies by Adams et al. and Loi et al. of tumor-infiltrating lymphocytes and breast cancer ACKNOWLEDGMENTS prognosis, the majority of patients received anthracyclines. We are grateful to the MSK Center for Molecular Oncology for their assistance with Secondary hematologic malignancies as a result of chemotherapy, sequencing studies. including alkylating agents and anthracyclines, remain a signifi- cant risk associated with systemic therapy for epithelial 29,30 tumors. Our work raises the possibility that some patients CONTRIBUTIONS may be at increased risk for secondary leukemias based on the MKl, EC, RLL, and LN conceived the study. MKl, EC, ZG, RB, AV, MFB, BW, JSR-F, RLL, presence of oncogenic mutations in infiltrating white cells, which and LN designed experiments. MKl, MKe, LB, AV, DY, and ZG performed experiments. pre-exist before systemic therapy, and which are selected for by MKl, LB, FTR, NS, and MFB analyzed whole-exome and targeted sequencing data. BW cytotoxic chemotherapy. This will need to be examined in and JSR-F performed laser-capture microdissection experiments. HYW and EB prospective studies, and it will be important to determine whether screened cases and performed clinicopathological analysis. BB and EC recruited the presence of mutant leukocyte clones at the time of diagnosis patients and collected samples. MKl, EC, RLL, and LN wrote the manuscript. should impact therapeutic decisions in different malignant contexts. Our data resonate with the observations that by the age 70 COMPETING INTERESTS years, at least 5% of people have known leukemia mutations in a The authors declare no conflict of interest. subset of circulating hematopoietic cells. This includes muta- tions in the same genes, such as in DNMT3A, TET2, and BCOR, which we identified in tumor-infiltrating leukocytes in our study. FUNDING However, we identified a much higher proportion of patients with This work was supported by the Breast Cancer Research Foundation, by the Cure mutations in their tumor-associated leukocytes (93%, 14 out of Breast Cancer Foundation, and by NCI 1R01CA151949-01 to RLL. MKl is a fellow of the 15), including in patients at a younger age. These data suggest Leukemia and Lymphoma Society and was previously supported by an EMBO these mutations are not solely a function of age-related mutations, Postdoctoral Fellowship. but rather represent earlier somatic events in patients with breast cancer or tumor-specific enrichment of mutant hematopoietic REFERENCES cells. It will be important to identify whether patients harboring 1 Cancer Genome Atlas Network. Comprehensive molecular portraits of human leukemogenic mutations in infiltrating leukocytes are at increased breast tumours. Nature 2012; 490:61–70. risk for future leukemias, and whether systemic therapies 2 Ellis MJ, Ding L, Shen D, Luo J, Suman VJ, Wallis JW et al. Whole-genome analysis including cytotoxic chemotherapy increase the selective advan- informs breast cancer response to aromatase inhibition. Nature 2012; 486: tage of pre-existing mutant hematopoietic clones. Moreover, it will 353–360. be important to determine how mutant breast cancer cells might 3 Acharyya S, Oskarsson T, Vanharanta S, Malladi S, Kim J, Morris PG et al. A CXCL1 influence the behavior and clonal selection of mutant hemato- paracrine network links cancer chemoresistance and metastasis. Cell 2012; 150: poietic cells, and vice versa, as their mutual presence may lead to 165–178. clonal changes in both compartments. 4 Karnoub AE, Dash AB, Vo AP, Sullivan A, Brooks MW, Bell GW et al. Mesenchymal Although our studies identified mutations in purified leuko- stem cells within tumour stroma promote breast cancer metastasis. Nature 2007; cytes, studies of en bloc tumor samples will need to consider the 449:557–563. © 2015 Breast Cancer Research Foundation/Macmillan Publishers Limited npj Breast Cancer (2015) 15005 Mutations in breast cancer-infiltrating leukocytes M Kleppe et al 5 Kurose K, Gilley K, Matsumoto S, Watson PH, Zhou XP, Eng C. Frequent somatic 20 Xie M, Lu C, Wang J, McLellan MD, Johnson KJ, Wendl MC et al. Age-related mutations in PTEN and TP53 are mutually exclusive in the stroma of breast car- mutations associated with clonal hematopoietic expansion and malignancies. Nat cinomas. Nat Genet 2002; 32:355–357. Med 2014; 20: 1472–1478. 6 Hu M, Yao J, Cai L, Bachman KE, van den Brule F, Velculescu V et al. Distinct epi- 21 Jaiswal S, Fontanillas P, Flannick J, Manning A, Grauman PV, Mar BG et al. Age- genetic changes in the stromal cells of breast cancers. Nat Genet 2005; 37:899–905. related clonal hematopoiesis associated with adverse outcomes. N Engl J Med 7 Allinen M, Beroukhim R, Cai L, Brennan C, Lahti-Domenici J, Huang H et al. 2014; 371:2488–2498. Molecular characterization of the tumor microenvironment in breast cancer. 22 Hammond ME, Hayes DF, Dowsett M, Allred DC, Hagerty KL, Badve S et al. Cancer Cell 2004; 6:17–32. American Society of Clinical Oncology/College Of American Pathologists 8 Granot Z, Henke E, Comen EA, King TA, Norton L, Benezra R. Tumor entrained guideline recommendations for immunohistochemical testing of estrogen and neutrophils inhibit seeding in the premetastatic lung. Cancer Cell 2011; 20:300–314. progesterone receptors in breast cancer. J Clin Oncol 2010; 28: 2784–2795. 9 Grivennikov SI, Greten FR, Karin M. Immunity, inflammation, and cancer. Cell 2010; 23 Wolff AC, Hammond ME, Hicks DG, Dowsett M, McShane LM, Allison KH et al. 140:883–899. Recommendations for human epidermal growth factor receptor 2 testing 10 Orimo A, Weinberg RA. Stromal fibroblasts in cancer: a novel tumor-promoting in breast cancer: American Society of Clinical Oncology/College of American cell type. Cell Cycle 2006; 5: 1597–1601. Pathologists clinical practice guideline update. J Clin Oncol 2013; 31: 11 Li HJ, Reinhardt F, Herschman HR, Weinberg RA. Cancer-stimulated mesenchymal 3997–4013. stem cells create a carcinoma stem cell niche via prostaglandin E2 signaling. 24 Westbury CB, Reis-Filho JS, Dexter T, Mahler-Araujo B, Fenwick K, Iravani M et al. Cancer Discov 2012; 2:840–855. Genome-wide transcriptomic profiling of microdissected human breast 12 Mahmoud SM, Paish EC, Powe DG, Macmillan RD, Grainge MJ, Lee AH et al. tissue reveals differential expression of KIT (c-Kit, CD117) and oestrogen Tumor-infiltrating CD8+ lymphocytes predict clinical outcome in breast cancer. J receptor-alpha (ERalpha) in response to therapeutic radiation. J Pathol 2009; 219: Clin Oncol 2011; 29: 1949–1955. 131–140. 13 Mohammed ZM, Going JJ, Edwards J, Elsberger B, McMillan DC. The relationship 25 Iyer G, Hanrahan AJ, Milowsky MI, Al-Ahmadie H, Scott SN, Janakiraman M et al. between lymphocyte subsets and clinico-pathological determinants of survival in Genome sequencing identifies a basis for everolimus sensitivity. Science 2012; patients with primary operable invasive ductal breast cancer. Br J Cancer 2013; 338:221. 109: 1676–1684. 26 Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler 14 Loi S, Sirtaine N, Piette F, Salgado R, Viale G, Van Eenoo F et al. Prognostic and transform. Bioinformatics 2009; 25: 1754–1760. predictive value of tumor-infiltrating lymphocytes in a phase III randomized 27 McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A et al. adjuvant breast cancer trial in node-positive breast cancer comparing the addi- The Genome Analysis Toolkit: a MapReduce framework for analyzing tion of docetaxel to doxorubicin with doxorubicin-based chemotherapy: BIG 02- next-generation DNA sequencing data. Genome Res 2010; 20: 1297–1303. 98. J Clin Oncol 2013; 31: 860–867. 28 Page DB, Naidoo J, McArthur HL. Emerging immunotherapy strategies in 15 Adams S, Gray RJ, Demaria S, Goldstein L, Perez EA, Shulman LN et al. Prognostic breast cancer. Immunotherapy 2014; 6:195–209. value of tumor-infiltrating lymphocytes in triple-negative breast cancers from two 29 Wolff AC, Blackford AL, Visvanathan K, Rugo HS, Moy B, Goldstein LJ et al. phase III randomized adjuvant breast cancer trials: ECOG 2197 and ECOG 1199. J Risk of marrow neoplasms after adjuvant breast cancer therapy: the national Clin Oncol 2014; 32: 2959–2966. comprehensive cancer network experience. J Clin Oncol 2015; 33: 340–348. 16 Busque L, Patel JP, Figueroa ME, Vasanthakumar A, Provost S, Hamilou Z et al. 30 Valentini CG, Fianchi L, Voso MT, Caira M, Leone G, Pagano L. Incidence of Recurrent somatic TET2 mutations in normal elderly individuals with clonal acute myeloid leukemia after breast cancer. Mediterr J Hematol Infect Dis 2011; 3: hematopoiesis. Nat Genet 2012; 44: 1179–1181. e2011069. 17 Moran-Crusio K, Reavie L, Shih A, Abdel-Wahab O, Ndiaye-Lobry D, Lobry C et al. Tet2 loss leads to increased hematopoietic stem cell self-renewal and myeloid transformation. Cancer Cell 2011; 20:11–24. This work is licensed under a Creative Commons Attribution 4.0 18 Quivoron C, Couronne L, Della Valle V, Lopez CK, Plo I, Wagner-Ballon O et al. TET2 International License. The images or other third party material in this inactivation results in pleiotropic hematopoietic abnormalities in mouse and is a article are included in the article’s Creative Commons license, unless indicated recurrent event during human lymphomagenesis. Cancer Cell 2011; 20:25–38. otherwise in the credit line; if the material is not included under the Creative Commons 19 Ko M, Bandukwala HS, An J, Lamperti ED, Thompson EC, Hastie R et al. Ten-Eleven- license, users will need to obtain permission from the license holder to reproduce the Translocation 2 (TET2) negatively regulates homeostasis and differentiation of material. To view a copy of this license, visit http://creativecommons.org/licenses/ hematopoietic stem cells in mice. Proc Natl Acad Sci USA 2011; 108: 14566–14571. by/4.0/ Supplementary Information accompanies the paper on the npj Breast Cancer website (http://www.nature.com/npjbcancer) npj Breast Cancer (2015) 15005 © 2015 Breast Cancer Research Foundation/Macmillan Publishers Limited

Journal

npj Breast CancerSpringer Journals

Published: Jun 10, 2015

References