Functional evidence for derivation of systemic histiocytic neoplasms from hematopoietic stem/progenitor cells

Histiocytic neoplasms are heterogeneous disorders marked by tissue infiltration and accumulation of pathologic macrophage-, dendritic cell (DC)–, or monocyte-derived cells.¹  Over the last 5 years, recurrent mutations activating MAPK signaling have been identified in adults and children affected by Langerhans cell histiocytosis (LCH)^2-7  or Erdheim-Chester disease (ECD).^7,8  While the discovery of these mutations has impacted treatment,^7,9-13  these discoveries may also be useful in identifying the cellular origins of LCH and ECD, which have been debated for decades.^14,15 

LCH was historically thought to arise from epidermal Langerhans cells^14,16  due to phenotypic similarities between LCH and normal Langherans cells. Recent data however suggest that LCH may be derived from myeloid precursor cells bearing somatic mutations in MAPK pathway members,¹⁷  including BRAF^V600E mutations in ∼50% of patients, as well as mutations in NRAS,¹⁸ KRAS,¹⁸ ARAF,^3,5,7  and MAP2K1.^4,5,7,19  Gene expression profiling of CD207⁺ DCs from LCH patients revealed that LCH cells are more akin to bone marrow (BM)–derived, immature myeloid DCs than epidermal Langerhans cells.²⁰  Moreover, BRAF^V600E mutations have been identified in CD11c⁺ myeloid DC precursors in LCH, CD14⁺ monocytes in LCH and ECD, and BM CD34⁺ hematopoietic stem/progenitor cells (HSPCs) in a proportion of LCH patients. Finally, expression of the BRAF^V600E mutation in committed DC precursors is sufficient to drive LCH-like disease in mice.¹⁷  Despite these data, however, functional evidence linking the cell of origin for histiocytosis to HSPCs in actual patient materials does not currently exist. Therefore, we attempted to understand the cell of origin of histiocytic neoplasms using genetic and functional analyses of HSPCs from histiocytosis patients. We identify that CD34⁺ cells in histiocytosis patients have functional self-renewal potential and describe the first successful patient-derived xenograft of a histiocytic neoplasm.

BM and/or peripheral blood (PB) cells were obtained from 41 patients with ECD or mixed histiocytosis (MH). A total of 37 patients had a BRAF^V600E mutation, and 1 patient had KRAS^G12S mutation within histiocytosis-infiltrated tissues (supplemental Table 1, available on the Blood Web site); 10 patients had a concurrent myeloid neoplasm (chronic myelomonocytic leukemia or myeloproliferative neoplasm) in addition to histiocytosis (a recently identified clinical overlap²¹ ).

BM and PB cell subsets were purified by fluorescence-activated cell sorting or immunomagnetic selection. Purity was consistently >90% (supplemental Methods). CD34⁺ BM cells were assayed in methylcellulose (Methocult GF+ H4435, STEMCELL Technologies) and/or grown in liquid culture conditions allowing differentiation into myeloid dendritic cells (mDCs) or plasmacytoid dendritic cells.²²  In some experiments, single CD34⁺ cells were expanded in liquid culture for 10 days.²³ 

Detection of BRAF^V600E and BRAF^WT alleles within DNA extracted from BM, PB, or colonies was performed by real-time polymerase chain reaction (PCR) and/or picodroplet digital PCR as described previously.¹⁸  Mutations in colony-forming unit-macrophage (CFU-M), colony-forming unit-granulocyte (CFU-G), and burst-forming unit-erythroid (BFU-E) were analyzed by pools of 10 colonies.

Xenografts were performed by direct intrafemoral injection²⁴  of a variable number of CD34⁺ cells from BM mononuclear cells from histiocytosis patients into sublethally irradiated (200 cGy) nonobese diabetic severe combined immunodeficiency IL2Rgamma-null mice with transgenic expression of human interleukin-3, stem cell factor, and granulocyte-macrophage colony-stimulating factor (NSGS)²⁵  mice. Human engraftment was monitored by monthly flow cytometric analysis of PB until mice became moribund. Thereafter, animals were sacrificed, and tissues were analyzed using flow cytometry, histology, immunohistochemistry, and sequencing to detect driver mutations identified in the histiocytosis/co-occurring myeloid neoplasm. Studies were approved by the ethics committees of University Hospital La Pitié-Salpêtrière and Memorial Sloan Kettering Cancer Center and by the Institutional Review Board of the French Regulatory Agency and conducted in accordance to the Declaration of Helsinki. All patients signed informed consent.

We first performed genotyping for the BRAF^V600E mutation in blood cells from ECD or MH known to harbor the BRAF^V600E mutations within infiltrated tissues. BRAF^V600E mutations were detected in 71% (10/14) of blood CD14⁺ monocytes (Figure 1A). The BRAF^V600E mutation was more frequent within CD14⁺ CD16^low classical²⁶  than CD14⁺ CD16^high nonclassical monocytes. These results confirm prior detection of MAPK pathway mutations in CD14⁺ cells of children with systemic LCH¹⁷  and in 1 patient with ECD.¹⁸  The BRAF^V600E mutation was also detected in T (CD3⁺) and B lymphocytes (CD19⁺ or CD20⁺) and/or polymorphonuclear (CD14⁻ CD15⁺ CD16⁺) cells in some patients (Figure 1A). The BRAF^V600E mutant allele frequency (BRAF-MAF) was low in all cell populations and ranged from 0.01 to 2.9 within monocytes (Figure 1B). The detection of the BRAF^V600E mutation in multiple mature blood cells suggests that the oncogenic event involved a multipotent BM progenitor.

We next investigated BM progenitor populations for BRAF^V600E in patients with ECD (n = 22) or MH (n = 8). BRAF^V600E was detected within purified CD34⁺ BM cells from 5 out of 15 patients (33%) known to have BRAF^V600E in lesional histiocytes (Figure 1C). The BRAF-MAF was always <1% and lower in the CD34⁺ CD38⁻ multipotent progenitor/long-term hematopoietic stem cell compartment than in the CD34⁺ CD38⁺ myeloid progenitor compartment (Figure 1B). Given these low BRAF-MAFs in HSPCs, we further sought to evaluate if mutant BRAF could be detected in progeny derived from BRAF mutant HSPCs. Through evaluation of colonies obtained upon seeding CD34⁺ BM cells in methylcellulose clonogenic cultures, BRAF^V600E was detected in CFU-M (3/14) and CFU-G (1/13), but not in BFU-E colonies. BRAF^V600E mutations were also detected within in vitro–derived plasmacytoid dendritic cells (2/13), but not in mDCs. We also analyzed colonies derived from single CD34⁺ BM cells of 6 patients with BRAF^V600E ECD. Only 3 out of 459 colonies (0.65%) were mutated for BRAF (supplemental Figure 1), which confirms the low BRAF-MAF detected by picodroplet digital PCR. The patient with MH who had 2 BRAF-mutated colonies (#P10) was further analyzed for TET2 and SRSF2 mutations, which were known to exist in a concurrent acute myelomonocytic leukemia (AMML) in this patient. Interestingly, the presence of colonies with BRAF/TET2 comutation, as well as with TET2 mutation alone, revealed that the mutation in TET2 occurred before the BRAF mutation and suggests a common clonal origin for both the histiocytosis and the AMML (Figure 1D). This was further reinforced by the detection of the same BRAF, TET2, and SRSF2 mutations within LCH cells in a skin biopsy from this patient (supplemental Figure 2).

Next, to test the self-renewal and differentiation capacity of HSPCs from histiocytoses patients, CD34⁺ cells were purified from 8 patients, including 6 with ECD or MH plus a concomitant myeloid neoplasm and 2 with ECD alone (supplemental Table 1). A total of 0.1 to 0.8 × 10⁶ CD34⁺ cells from each patient were transplanted into sublethally irradiated NSGS²⁴  mice. Successful engraftment of human myeloid cells was observed in blood of recipient mice xenografted from 50% of patients (4 out of 8) after a median of 90 days (range: 60-120 days; supplemental Figure 3A). Of these 4 samples, CD34⁺ cells from 1 ECD patient sample induced histiocytosis-like lesions, as shown by infiltration of human CD45⁺ (hCD45) cells coexpressing the human myeloid and monocyte lineage markers hCD33 and hCD14 in the BM, spleen, liver, and lung (Figure 2A-B; supplemental Figure 3B). Immunohistochemistry revealed that tissues were infiltrated by an hCD45⁺ hCD163⁺ hCD68⁺ hS100⁺ hCD1a⁻ population of foamy histiocytes, characteristic of ECD (Figure 2C). Furthermore, genomic analysis of hCD45⁺ cell DNA from the engrafted animal’s BM and spleen revealed the same KRAS^G12S mutation present in the patient’s ECD lesion (Figure 2D). All other mice that successfully engrafted with human cells revealed either no evidence of histiocytosis histologically and/or no evidence of the histiocytosis-associated mutation.

This study provides the first functional evidence that CD34⁺ cells from patients with histiocytosis can functionally initiate disease. Xenotransplantation successfully recapitulated the unique histologic, immunophenotypic, and genetic characteristics of human ECD. Moreover, these data confirm prior results suggesting that somatic mutations affecting MAPK signaling detectable in histiocytosis lesions are also detectable in the CD34⁺ compartment and in mature blood cells of a proportion of histiocytosis patients. Further analysis of HSPCs from histiocytosis patients for additional somatic mutations in histiocytosis lesional cells may therefore be very helpful in clarifying the cells of origin of systemic histiocytoses. Moreover, these data, together the wide clinical heterogeneity of histiocytoses, also suggest the possibility of alternate cells of origin for these conditions beyond HSPCs.

The authors thank Isabelle Plo for help in quantification of the JAK2 mutation, Laetitia Claër for contribution in cell sorting, and Catherine Le Gall, Dominique Péchaud, Yolaine Pothin and Mariama Bakari and Rim Ben Jannet for contribution to BRAF mutation detection.

B.H.D. is supported by the American Society of Hematology Senior Research Training Award for Fellows and the New York State Council on Graduate Medical Education Empire Clinical Research Investigator Program Fellowship. A.Y. is supported by the Aplastic Anemia and MDS Research Foundation and the Lauri Strauss Leukemia Foundation. E.L.D. and O.A.-W. are supported by grants from the Erdheim Chester Disease Global Alliance and the Histiocytosis Association. O.A.-W. is supported by grants from the Edward P. Evans Foundation, the Department of Defense Bone Marrow Failure Research Program (BM150092 and W81XWH-12-1-0041), National Institutes of Health, National Heart, Lung, and Blood Institute (R01 HL128239), a National Institutes of Health K08 Clinical Investigator Award (1K08CA160647-01), the Josie Robertson Investigator Program, a Damon Runyon Clinical Investigator Award, an award from the Starr Foundation (I8-A8-075), the Leukemia and Lymphoma Society, and the Pershing Square Sohn Cancer Research Alliance. J.-F.E. is supported by grants from Association pour la Recherche et l'Enseignement en Pathologie. V.T. is supported by Ligue Nationale Contre le Cancer (Equipe labelisée EL2016.LNCC/VaT). O.A.B. is supported by Ligue Nationale Contre le Cancer (Equipe labelisée).

Contribution: B.H.D., D.R.-W., C.B., A.Y., M.M., Z.H.-R., N.T., N.O., A.D., C.Y.P., G.G., V.T., O.A.B., O.A.-W., F.M.L., and J.-F.E. performed the experiments; F.C.-A., M.P., R.R., F.U., L.L.F., E.L.D., T.P., F.C., Z.A., and J.H. collected patient material and clinical data; and B.H.D., O.A.-W., J.H., and J.-F.E. wrote the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Jean-François Emile, Service de Pathologie and EA4340, Hôpital Ambroise Paré, 9 Avenue Ch de Gaulle, 92104 Boulogne, France; e-mail: jean-francois.emile@uvsq.fr; and Julien Haroche, Département de Médecine Interne, Hôpital de la Pitié-Salpêtrière, 47-83 Boulevard de l’Hôpital, 75013 Paris, France; e-mail: julien.haroche@aphp.fr.