|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
CANCER STEM CELLS |
aSurgery Branch, National Cancer Institute, Bethesda, Maryland, USA;
bDepartment of Pathology,
cMolecular Cytogenetics Core Facility,
dDepartment of Medicine,
eDepartment of Surgery, Memorial Sloan-Kettering Cancer Center, New York, New York, USA
Key Words. Cancer stem cells • Bone marrow • Solid organ cancer • Bone marrow transplantation
Correspondence: Itzhak Avital, M.D., Surgery Branch, National Cancer Institute, Building 10-Hatfield CRC, Room 3-3940, 10 Center Drive, Bethesda, Maryland 20892-1201, USA. Telephone: 301-496-4164; Fax: 301-402-1738; e-mail: avitali{at}mail.nih.gov
Received on May 24, 2007;
accepted for publication on July 31, 2007.
First published online in STEM CELLS EXPRESS August 9, 2007.
| ABSTRACT |
|---|
|
|
|---|
Disclosure of potential conflicts of interest is found at the end of this article.
| INTRODUCTION |
|---|
|
|
|---|
We demonstrate that donor-derived bone marrow cells contribute to secondary solid organ cancers developing after ABMT. However, the level of incorporation was low, which raises questions in regard to the clinical relevance.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Of the approximately 12,000 patients recorded in the prospectively maintained MSKCC bone marrow transplant database, we identified 9 patients as having developed solid organ malignancies after myeloablation, TBI, and ABMT as treatment for hematologic malignancies. Of these, four male patients, who had undergone ABMT from female donors and then developed solid organ cancers for which archival pathological material was available, were identified and are the basis of this report.
Patient characteristics are summarized in Table 1. At the time of BMT, all donors were free of cancer and younger than 18 years old. Three donors were six-human leukocyte antigen (HLA)-matched siblings (sisters), and the fourth was an HLA-matched unrelated donor. As patient 1 was unrelated, follow-up data are not available. The donor for patient 2 was lost to follow-up. The donor for patient 3 died traumatically and was apparently cancer-free at the time of death. The donor for patient 4 lives outside of the U.S. and could not be contacted for follow-up.
|
6–6.5) solution for
5–10 minutes. Plasmid clones pSV2X5 for (X-alphoid) and pY84 for (Y-alphoid) were labeled with Spectrum Green and Spectrum Orange (Vysis, Downers Grove, IL, http://www.vysis.com), respectively, by nick translation [8]. The probe mixture was prewarmed in a 70°C water bath for 5 minutes, applied to the target area, covered, and sealed with rubber cement. The slides were denatured for 5 minutes at 75°C and hybridized overnight at 37°C in HYBrite (Vysis). The slides were then washed with 2x SSC at 45°C and stained with 4,6-diamidino-2-phenylindole. We targeted autosomal probes to the 12
and 4
chromosomes' centromeric region, prepared from plasmid clones p
12H8 and GXBA11/34, respectively, and labeled with Spectrum Green and Spectrum Orange, respectively. Combined immunolabeling and FISH were performed with modifications as previously described [9]. Immunostaining was performed at 4°C overnight using fluorescein isothiocyanate (FITC)-conjugated mouse anti-human CD45 (DakoCytomation, Glostrup, Denmark, http://www.dakocytomation.com) (4 µg/ml) and FITC-conjugated mouse anti-human cytokeratin AE1:AE3 (Abcam, Cambridge, U.K., http://www.abcam.com). The slides were washed with 1x phosphate-buffered saline/0.05% Tween 20 (Fisher) and then fixed with 1% formaldehyde (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com) for 10 minutes.
Five male patients who had received same-sex (male) allo-BMT were used as positive controls for the ability to identify Y chromosomes and as negative controls for the absence of two X chromosomes. The Y-chromosome detection rate was in excess of 90%. Normal human uterine tissue was used as the positive control for the presence of two X chromosomes and as the negative control for the absence of Y chromosomes.
Tissue Analysis
Analyses of the combined FISH/immunolabeling and FISH alone were done using Zeiss AxioImaging 2 microscope (Carl Zeiss, Jena, Germany, http://www.zeiss.com) with a Zeiss AxioCam MRm charge-coupled device camera. Image acquisition was done with AxioVision version 4.0. To avoid the visual effects of nuclear stacking, further analysis was done using confocal microscopy.
The percentage of female (XX, no Y) malignant cells was determined by counting FISH-detected intranuclear X and Y spots in all tumors, eliminating CD45-positive cells in 20 high-power fields from three different experiments. In two cases (an adenocarcinoma of the lung and a squamous cell carcinoma of the larynx), counting was performed only when cells were concurrently positive for AE1:AE3 (an epithelial marker consistent with adenocarcinoma or squamous cell carcinoma).
| RESULTS |
|---|
|
|
|---|
|
|
|
|
It is hypothesized that adult bone marrow-derived stem cells participate in normal tissue regeneration leading to terminally differentiated donor-derived epithelial cells being found in patients after ABMT [11, 12]. Malignant degeneration of these differentiated cells could occur and account for our findings. We examined the normal tissue adjacent to two tumors (lung adenocarcinoma and laryngeal squamous cell carcinoma) for the presence of XX-chromosome-positive cells. We found that only 0.07% of the cells in the normal tissue could have been donor-derived.
Several areas in these four tumors demonstrated potential evidence for clonality of the malignant cells. Figure 5 demonstrates that one area of a malignancy contains predominantly XX-positive and Y-negative cells, whereas an adjacent area demonstrates predominantly XY-positive cells.
|
| DISCUSSION |
|---|
|
|
|---|
The common view of epithelial carcinogenesis is that accumulated sequential mutations in differentiated epithelial cells lead to malignancy, which can include a dedifferentiated phenotype (Knudson's hypothesis) [22]. It is difficult to accept this common view when considering, for example, the fact that the colonic mucosa is replaced every 7 days. One might pose the following question: How is it, then, that the first Knudson's hit is carried on? An alternate hypothesis was proposed by Sell and Pierce [1–3], who suggested that cancers arise during tissue regeneration as a result of maturation arrest of stem cells, resulting in aberrant differentiation but sustained and abnormal proliferation. The existence of cancer cells with stem cell-like behavior has been demonstrated in human liquid malignancies where it has been noted that <1% of the cells taken from patients with leukemia is capable of transferring the cancer to an immunodeficient animal [23–25]. Similarly, it was found that a fraction of breast cancer cells have stem cell-like characteristics in that they could give rise to several populations of cells differing phenotypically [23].
However, demonstrating the existence of malignant cells that behave like stem cells (the so-called "cancer stem cells") is not the same as proposing that stem cells participate in carcinogenesis. Recently, Houghton et al. reported that gastric cancers arising in H. felis-infected mice after TBI and BMT were derived from transplanted stem cells and not from the differentiated cells of the recipients' gastric mucosa [4]. There have been two prior reports, each describing a patient who developed a renal cell carcinoma after bone marrow transplantation and in which the authors felt that donor stem cell-tumor cell hybridization may have occurred [5, 6]. However, in these studies, CD45 staining was not done, and the epithelial nature of the donor cells was not confirmed, leaving open the possibility that the visualized cells in the tumors were donor-derived lymphocytes infiltrating into a recipient-derived malignancy.
Our results extend these findings by providing phenotypic evidence that bone marrow-derived stem cells contribute to human epithelial and mesenchymal neoplasia. Cells were counted as donor-derived malignant cells only if they were XX-chromosome positive, Y-chromosome negative, CD45 negative, and AE1:AE3 positive (lung adenocarcinoma and laryngeal squamous cell carcinoma). Although donor derived tumor infiltrating leukocytes were found, histologically malignant cells that were CD45 negative were also seen, indicating that they were not of a hematopoietic lineage.
Hyperploidies or hypoploidies have been reported in various cancers, and our results could be interpreted as a spurious finding based on both chromosomal gain and loss [10] (i.e., simultaneous loss of a Y chromosome and duplication of a X chromosome). Therefore, we performed FISH with probes to chromosomes 4 and 12 to determine the frequency at which simultaneous autosomal gain or loss had occurred. Ninety-nine percent of the female cancer cells in these four male patients expressed a diploid number of autosomes. However, this does not exclude aneuploidy of the sex chromosomes, and it provides only circumstantial support for our findings. Moreover, specimens of cancers from patients after male-to-male ABMT did not demonstrate a significant number of XX-positive cells (0.03%); there was no loss of the Y chromosome detected in the same cells. In addition, our Y-chromosome detection rate was relatively high in the control subjects (male-to-male BMT). This suggests that alterations in chromosomal ploidy occur but are unlikely to account for our observations. We concluded that it is unlikely that loss of the Y chromosome and duplication of the X chromosome as an isolated simultaneous chromosomal event occurred in all four male patients receiving stem cells from women, and in none of the five male patients receiving stem cells from men.
Only very limited follow-up was available for the donors, and therefore we cannot determine with complete certainty whether these donors harbored undetectable cancers at the time of bone marrow donation. However, the donors did not demonstrate signs of malignancy at the time of transplant and were free of cancer during the available follow-up period. Three of the cancers (Kaposi sarcoma, laryngeal squamous cell carcinoma, and lung adenocarcinoma) are uncommon in the donors' age group (all being less than 18 years old). The fourth tumor, a glioblastoma multiforme, does not appear to have been diagnosed in the donor during the 3 years she was alive after donating stem cells to her brother.
Transfer and persistence of stem cells between mothers and their unborn offspring ("maternal-fetal microchimerism") have been described [26]. It could be hypothesized that the female malignant cells could represent differentiated somatic cells arising from maternal stem cells, which later transformed into cancer. If so, it would be common to find female cells in male cancers, but we did not in the control cases. A related possibility is that transplanted bone marrow-derived stem cells differentiated into somatic cells, which then underwent malignant transformation. Although we did not have tissue available to allow us to determine the degree of integration of female cells into the distant normal tissue, we found that only 0.07% of the cells in the adjacent normal tissue near the tumors appeared to be XX positive and Y negative.
Houghton et al. [4] reported on a murine model of induced gastric cancer after TBI and reconstitution of the bone marrow with allogeneic stem cells. They found that nearly 100% of the gastric cancer cells were derived from donor stem cells. In contrast, the percentage of cancer cells in our patients derived from donor stem cells ranged from 2.5%–6%. One possible explanation for the difference between the human and the animal model may lie in the intensity of the TBI and the efficacy of stem cell ablation, both in the bone marrow and the periphery. In rodents, TBI can be performed with doses of radiation more commensurate with complete ablation of both the bone marrow and peripheral stem cells' compartment, whereas in humans such doses are not tolerated and may have not completely eradicated the host's peripheral stem cells. In fact, the patient with the lowest rate of donor cell incorporation never received TBI (Kaposi sarcoma). Another possible explanation might be the short interval between the BMT and the development of secondary cancers (3.5 years). Since initiation of cancer is thought to occur 10–20 years prior to its clinical manifestation, it is possible that by the time the cancer was apparent, the recipient's own bone marrow already contributed significantly, thus resulting in a lesser contribution by the donor bone marrow. Therefore, it is possible that, in our human study, we had at least two bone marrow-derived populations of cancer cells, one derived from the donors and the other from the recipients.
Moreover, we do not know the exact identity of the cell that originated from solid organ cancers. Houghton et al. postulated that this cell might be a mesenchymal stem cell rather than a hematopoietic stem cell, and as such it is not clear whether routine methods of marrow ablation eliminate completely these cells from the recipient's marrow. In addition, it is not clear whether these cells are contained within the typical BMT even when whole BMT is performed. Therefore, there is a certain degree of uncertainty that the relevant bone marrow-derived stem cells are completely ablated with routine regimens, which potentially underestimates the donor cells' contribution to carcinogenesis of solid organ cancers
More recently, seminal work reported by Cogle et al. demonstrated, for the first time in humans, that bone marrow-derived cells contribute to epithelial neoplasia (lung cancer and colon adenomas but not skin cancer) [27]. In their paper, they employed a more rigorous methodology (mouse models of intestinal and lung neoplasia) to evaluate the mechanisms involved in the contribution of bone marrow to epithelial neoplasia. They proposed that rather than having a direct role, bone marrow contributes to cancer as a developmental mimicry. The present work further substantiates their findings that bone marrow contributes to epithelial cancers. However, we did not study the mechanisms underlining this phenomenon and, therefore, we cannot comment on potential mechanisms in our study. Of note, all of our patients had chronic inflammatory conditions predating the development of post-ABMT solid organ cancers. Patient 1, who developed Kaposi sarcoma, had lichenoid dermatitis, eccrine hydradenitis, and chronic liver and cutaneous graft-versus-host disease (GVHD). Patient 2, who developed lung adenocarcinoma, had GVHD and obstructive chronic inflammatory changes in the lung. Patient 3, who developed laryngeal squamous carcinoma, had bronchiolitis obliterans (bronchiolitis obliterans organizing pneumonia), extensive chronic GVHD, and a long history of chronic sinusitis. Patient 4, who developed glioblastoma multiforme, had acute lymphoblastic leukemia relapse in the brain, for which he received 1,800 cGy, and GVHD. Although we strongly agree that inflammation plays a cardinal role in bone marrow contribution to neoplasia, we cannot confirm it based on our study. It is possible that the cancer itself is an adequate inflammatory stimulus that triggers bone marrow cell recruitment. The rates of bone marrow incorporation into epithelial neoplasia and secondary cancers were similar between the present work and the report by Cogle et al., which supports the developmental mimicry hypothesis. In addition to epithelial cancers, we report on one patient with nonepithelial solid organ cancer (Kaposi sarcoma). The present work confirms the report by Cogle et al., and both represent a strong support to the hypothesis that bone marrow contributes to solid organ cancers. However, the question regarding the clinical relevance of this phenomenon remains open.
The finding that only a fraction of the malignant cells was derived from the donors suggests that the relatively simple model of a cancer being derived from a single transformed cell or a single stem cell is not likely to be accurate. Instead, hypothetical models for carcinogenesis include the possibility of fusion between stem cells and transformed cells [14]. This model allows for persistent asymmetric division of malignant cells, as is seen with stem cells [25], as well as the ongoing recruitment and incorporation of stem cells into a cancer as it progresses. Although not accepted by all and yet to be definitively demonstrated, fusion can explain how stem cells can acquire such a diverse array of phenotypes. Although there is only scarce experimental evidence, one might also consider that fused cells might undergo a reduction division with subsequent asymmetrical divisions resulting in few "original" donor-derived cells amid many daughter cells, which look like the recipient cells. Unless the donor cells are genetically tagged, there is no means to identify these late fusion products as donor-derived.
Further evidence is required to demonstrate that bone marrow contributes to carcinogenesis, at least in cases of chronic inflammatory conditions (ulcerative colitis, chronic gastritis, etc.). If proven true, it will open the doors for novel strategies for cancer early detection and treatment.
| DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST |
|---|
|
|
|---|
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
D. L. Worthley, A. Ruszkiewicz, R. Davies, S. Moore, I. Nivison-Smith, L. Bik To, P. Browett, R. Western, S. Durrant, J. So, et al. Human Gastrointestinal Neoplasia-Associated Myofibroblasts Can Develop from Bone Marrow-Derived Cells Following Allogeneic Stem Cell Transplantation Stem Cells, June 1, 2009; 27(6): 1463 - 1468. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. C. Borczuk, R. L. Toonkel, and C. A. Powell Genomics of Lung Cancer Proceedings of the ATS, April 15, 2009; 6(2): 152 - 158. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Janin, H. Murata, C. Leboeuf, J.-M. Cayuela, E. Gluckman, L. Legres, A. Desveaux, M. Varna, P. Ratajczak, J. Soulier, et al. Donor-derived oral squamous cell carcinoma after allogeneic bone marrow transplantation Blood, February 19, 2009; 113(8): 1834 - 1840. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Ando, R. Abe, M. Sasaki, J. Murata, D. Inokuma, and H. Shimizu Bone Marrow-Derived Cells Are Not the Origin of the Cancer Stem Cells in Ultraviolet-Induced Skin Cancer Am. J. Pathol., February 1, 2009; 174(2): 595 - 601. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Quante and T. C. Wang Inflammation and Stem Cells in Gastrointestinal Carcinogenesis Physiology, December 1, 2008; 23(6): 350 - 359. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. J. Weiss, J. K. Kolls, L. A. Ortiz, A. Panoskaltsis-Mortari, and D. J. Prockop Stem Cells and Cell Therapies in Lung Biology and Lung Diseases Proceedings of the ATS, July 15, 2008; 5(5): 637 - 667. [Full Text] [PDF] |
||||
![]() |
S. Takaishi, T. Okumura, and T. C. Wang Gastric Cancer Stem Cells J. Clin. Oncol., June 10, 2008; 26(17): 2876 - 2882. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| STEM CELLS | THE ONCOLOGIST | CME | ALPHAMED PRESS JOURNALS |