Side population cells in Hodgkin lymphoma

Side population (SP) cells are defined as cells that efficiently extrude the Hoechst 33342 dye, and hence remain negative for this fluorescent marker in flow cytometric analysis of life cells [1]. Among bone marrow cells, SP cells are enriched for hematopoietic stem cells, and, therefore, Hoechst 33342 negativity was initially used to enrich these cells before other markers became available. Further interest in SP cells was raised when such cells were also found at low frequency in several types of solid tumors. Importantly, cells with features of cancer stem cells seem to be enriched among SP cells [1,2]. SP cells (as well as cancer stem cells) show increased resistance to chemotherapeutic drugs, and there is hence much interest in finding ways to target these cells by novel therapeutic approaches. The chemoresistance and Hoechst 33342 extrusion of SP cells are actually caused by the same feature of these cells: SP cells express various members of the ABC transporter family, in particular the multidrug resistance gene 1 (MDR1) and ABCG2 (breast cancer resistance protein, BCRP) [1,2]. These transporters expel the Hoechst dye as well as various chemotherapeutic agents from the cells.

In this issue of Leukemia and Lymphoma, Shafer and colleagues report for the first time on the identification of SP cells in Hodgkin lymphoma (HL) cell lines and primary HL tumor cells [3]. HL is one of the most frequent lymphomas in the Western world, and is characterized by rare, pathognomonic tumor cells, the Hodgkin and Reed–Sternberg (HRS) cells. These cells are in nearly all cases derived from B cells, although they have lost most of the B cell-typical gene expression pattern [4]. In general, HL can be treated successfully by conventional chemo- and/or radiotherapy, but about 10% of patients relapse, and these are then more difficult to treat [5].

SP cells were identified in two of four HL cell lines investigated, accounting for less than 0.5% of cells in the two positive lines [3]. When studying cell suspensions from six HL biopsies, SP cells were found in each of the lymphomas, again mostly at a frequency below 1%. The SP cells were enriched among CD30⁺CD3⁺CD19⁻CD20⁻ cells, the expected phenotype of HRS cells in flow cytometry, considering that the seeming CD3 positivity most likely derived from T cells tightly sticking to HRS cells [6]. The authors confirmed increased transcript levels for MDR1 and ABCG2 in the two SP-positive HL cell lines L428 and HDLM2, and showed that SP cells are enriched in these cell lines when they are treated with the cytotoxic drug gemcitabine. The increased chemoresistance of SP cells was further supported by showing that sorted SP cells survived gemcitabine treatment better than non-SP cells. As SP and non-SP cells showed a similar survival and proliferation in the absence of gemcitabine, a direct toxic effect of the Hoechst dye in non-SP cells, as it is known for some types of cells, could be excluded. Thus, Shafer et al. identified a small subpopulation of SP cells in HL cell lines that show increased resistance to a cytotoxic drug.

If also SP cells of the HRS clones in vivo are more chemoresistant than the bulk of the HRS cells, these cells may be responsible for relapses of HL occurring in a fraction of patients. Thus, it would be important to find therapeutic approaches that target these cells. To this end, the authors performed an additional set of experiments and showed that treatment of the HL cell lines L428 and HDLM2 with the DNA demethylating agent 5-aza-2′-deoxycitidine increased expression of the tumor-associated antigens SSX2, MAGEA4, and survivin in the HRS cells, mostly more pronounced in the SP cells than in non-SP cells. Importantly, when the 5-aza-2′-deoxycitidine-treated cells were incubated with MAGEA4-specific cytotoxic T cells, significant cell killing was induced, and the frequency of SP cells among the HRS cells diminished dramatically, indicating that SP cells can be targeted by this immunological therapeutic approach.

The identification and initial characterization of SP cells in HL is very intriguing, and will certainly stimulate further studies, because numerous questions remain to be clarified. First, does the fact that two of four HL cell lines apparently lack SP cells mean that these cells are not essential for the expansion of the HRS cell population in general? Second, as SP cells are enriched for cancer stem cells in several other tumors, it is an exciting issue to study whether the HL SP cells have a better survival and a higher proliferative potential than the non-SP cells. Are SP cells more aggressive and have a higher clonogenic potential than non-SP cells when transferred into immunodeficient mice? Third, although SP cells were enriched among CD30⁺CD19⁻ cells in HL lymph nodes, in each of the six HLs studied actually more than 50% of SP cells did not show this phenotype. What are these cells? Do the CD30-negative SP cells in HL belong to the HRS cell clone, or does the HL microenvironment induce a SP phenotype in some non-HRS cells? Are SP cells perhaps also present in non-malignant lymph nodes? An analysis for clonal immunoglobulin gene rearrangements in HRS cells and CD30⁺ and CD30⁻ SP cells would be a feasible approach to clarify the relationship between CD30⁺ and CD30⁻ SP cells and HRS cells in HL. Fourth, and related to the previous point, it will be important to bring the present findings together with a recent study by Jones and colleagues reporting on HRS precursor cells that lack CD30 expression but instead have a B cell phenotype with CD20-positivity [7]. These cells were also found in HL cell lines, and showed features of lymphoma stem cells, including a higher clonogenic potential than the other HRS cells [7]. Thus, it will be essential to clarify the relationship between the putative CD30⁻CD20⁺ HRS precursor cells reported by Jones et al. and the mostly CD30⁺CD20⁻ SP cells identified by Shafer and coworkers. However, it should be noted that it has been questioned whether the data presented by Jones et al. indeed demonstrated the existence of CD30⁻CD20⁺ cells belonging to the HRS tumor clone in HL patients [8]. Other work actually argued against the regular existence of HRS cell clone members among CD30-negative cells in HL lymph nodes [9]. Finally, if it can be confirmed that also in HL patients, SP cells belonging to the HRS cell clone exist, and that these cells show increased chemoresistance, it will be important to develop therapeutic strategies able to target these cells. By showing that induced DNA demethylation of SP cells from HL cell lines renders these cells more vulnerable for killing by cytotoxic T cells, Shafer and coworkers presented a first rationale for such a strategy.

Declaration of interest: The author reports no conflicts of interest. The author alone is responsible for the content and writing of the paper.