Main

Hodgkin's lymphoma is one of the most frequent lymphomas in western countries. There are two re-cognized forms: the classical Hodgkin's lymphoma form and the lymphocyte-predominant form. Morphologically, they are characterized by a few large tumour cells within a background of activated lym-phocytes and mixed inflammatory cells. It has been demonstrated, by molecular studies, that Hodgkin's lymphoma is, in most cases, a B-cell lymphoma that originates from germinal centre1 neoplastic cells that escape from FAS-mediated apoptosis by an unknown mechanism,2 probably involving c-FLIP overexpression.3 Neoplastic cells from classical and lymphocyte-predominant Hodgkin's lymphoma usually present potentially functional clonal immunoglobulin (Ig) gene rearrangements though carrying a high load of somatic mutations.1, 2

What differentiates classical Hodgkin's lymphoma from lymphocyte-predominant Hodgkin's lymphoma is a low frequency of expression of B-cell-associated molecules (such as CD20, CD79a and J chain), and an absence of Ig light and Ig heavy messenger RNA (mRNA)4 in Hodgkin/Reed–Sternberg (HRS) cells. To explain this fact, several studies were performed. The first of them suggested crippling somatic mutations that would impede the Ig gene regulation,5, 6 but further studies demonstrated that, in most cases, there were not such mutations and that the lack of Ig expression in classical Hodgkin's lymphoma was due to a transcription machinery disruption caused by a deficiency in transcription factors, as was confirmed by transfection in cellular lines.4, 7, 8 Mutations in the Ig-octamer-binding region2 have finally been ruled out as the cause of the lack of B-specific gene expression.8

Transcription factors OCT.1 and OCT.2, and their coactivator BOB.1, are necessary for the octamer-dependent transcription of Ig and other important lymphoid-specific genes of B cells.9, 10, 11 Also, the activity of the cotransduced Ig promoter can be restored by transfection with BOB.1 and OCT.2 and/or OCT.1.7

OCT.1 is an ubiquitous factor present in a wide range of lymphoid and not lymphoid cell types. Studies on the expression of OCT.1 have demonstrated a high-level expression in pre-B cell lines.12, 13 On the other hand, OCT.2 expression is restricted to B cells and neuronal cells,14 and plays an important role in Ig promoter transactivation, as has been demonstrated by transfection experiments.15 They both participate in the control of important B-specific genes involved in proliferation and differentiation,16, 17, 18, 19, 20 such as Ig genes, CD20, CD79a and J chain. Transfection studies have demonstrated that OCT.1 can compensate the absence of OCT.2.21

Octamer functions are dependent on the presence of the B-cell-restricted coactivator BOB.1, which is a transcription factor necessary for B-cell development, inducing the formation of germinal centres.16, 22, 23

With the aim of evaluating the expression of transcription factors at the protein level in HRS cells in the largest series of cases reported until now, we performed an immunohistochemical study for OCT.1, OCT.2 and BOB.1 in 325 cases of classical Hodgkin's lymphoma, correlating the results with the expression of the B-cell markers CD20, CD79a, B-cell-specific activator protein (BSAP) and MUM.1, the presence of Epstein–Barr virus and the histo-logical subtype. For this purpose, we used the recently developed tissue microarray technology that allows the simultaneous analysis of a large number of cases and whose feasibility in Hodgkin's lymphoma analysis has been demonstrated.24, 25

Materials and methods

Classical Hodgkin's Lymphoma Samples and Cell Lines

The collaborating members of the Spanish Hodgkin's Lymphoma Study Group collected 355 cases of classical Hodgkin's lymphoma during the period between 1994 and 1998, of which 325 cases were suitable for this study (see below). Using standard tissue sections, according to the World Health Organization (WHO) classification,26 there were 184 nodular sclerosis, 112 mixed cellularity, 18 lymphocyte-rich, nine lymphocyte-depletion and two unclassificable classical Hodgkin's lymphomas. All patients were treated with standard polychemo-therapeutic regimens with or without adjuvant radiotherapy.

Paraffin blocks were selected only on the basis of the availability of suitable formalin-fixed, paraffin-embedded tissue (at least 1 mm thick).

Paraffin-embedded blocks from reactive lymphoid tissue and different B- and T-cell lymphoma samples were obtained from the Tissue Archives of Centro Nacional de Investigaciones Oncológicas (CNIO) Tumour Bank.

We also obtained five Hodgkin's lymphoma-derived cell lines (L428, HDLM2, L540, KMH2 and HD-MY-Z) from the German Collection of Microorganisms and Cell Cultures (DSMZ, Braunschweigh, Germany).

Tissue Microarrays Design

Tissue Arrayer device (Beecher Instrument, Silver Spring, MD, USA) was used to construct the tissue microarrays in CNIO.24 Briefly, the richest areas of HRS cells were marked in the paraffin blocks. In each case, two selected 1-mm-diameter cores from two different areas, were included along with 43 different controls. Thus, five different tissue microarray blocks were constructed, each containing 187 cores excepting block number 5 that contained 176 cores.

External controls were provided by cell lines, a representation of reactive lymphoid tissue (10 samples, including tonsil, spleen and lymphadenitis) and 28 samples of the most frequent non-Hodgkin's lymphomas of which two were cases of B-cell chronic lymphocytic leukaemia, two of follicular lymphoma, two of mantle cell lymphoma, two of diffuse large B-cell lymphoma, two of Burkitt's lymphoma, two of peripheral T-cell lymphoma and other two of anaplastic large-cell lymphoma. We used the previously referred cell lines, grown following standard protocols, fixed in formalin and included in paraffin as described.27 Tumour cells had to be identified in each of the two cylinders of a case to be included in the study.

Immunohistochemistry

Immunohistochemistry was performed as previously described.24 Tissue microarray blocks were sectioned with a thickness of 3 μm and dried for 16 h at 56°C before being dewaxed in xylene and rehydrated through a graded ethanol series and later washed with phosphate-buffered saline. Antigen retrieval was achieved by heat treatment in a pressure cooker for 2 min in 10 mM citrate buffer (pH 6.5). Before staining the sections, endogenous peroxidase was blocked. Immunohistochemical staining was performed on the sections using seven different antibodies: CD20 (L26, B cells; Dako), CD79a (JCB117, Dako), MUM.1 (ICSAT/M-17, Santa Cruz Biotechnology), BSAP (polyclonal, Santa Cruz Biotechnology), OCT.1 (12F11, Santa Cruz Biotechnology), OCT.2 (polyclonal, Santa Cruz Biotechno-logy) and BOB.1 (polyclonal, Santa Cruz Biotechnology). All of them were carried out in CNIO except BSAP that was performed in Ramón y Cajal Hospital. MUM.1, BSAP, OCT.1, OCT.2 and BOB.1 antibodies were used at 1:200, 1:250, 1:100, 1:500 and 1:7000 dilutions, respectively. After incubation, immunodetection was performed with the LSAB Visualization System (Dako, Glostrup, Denmark) using diaminobenzidine chromogen as substrate. Sections were counterstained with haematoxylin. Internal control was provided in each case by the reactivity of the accompanying lymphocytes, plasma cells and macrophages.

The pattern of staining was recorded as + (weak and focal positivity), ++ (intense positivity in most of cells, as strong as in normal and neoplastic germinal centre control cells) and − (completely absent), considering the expression in the nucleus of the neoplastic cells. All of the weak positive cases were associated to a focal pattern of staining.

The results were evaluated by two investigators (MGC and PMA) using uniform criteria. Identical results in each of the two cores were considered necessary to be suitable for the study.

In Situ Hybridization of Epstein–Barr Virus-Encoded RNAs (EBERs)

EBERs in situ hybridization was performed in CNIO.24 Tissue microarray sections were pretreated with proteinase K for 10 min at 37°C, and then incubated with EBER PNA Probe/FITC (Dako) for 1.5 h at 55°C in a humid chamber and afterward washed with Stringent Wash Solution for 25 min at 55°C in a water bath with shaking. Detection was performed with antifluorescein isothiocyanate (anti-FITC) antibody (clone DAK-FITC4, from Dako) diluted at 1:100. After incubation, immunodetection was performed with biotinylated antimouse immuno-globulins, followed by peroxidase-labelled streptavidin (LSAB Visualization System, Dako) with diaminobenzidine chromogen as substrate. Sections were counterstained with haematoxylin.

Statistical Study

The Pearson χ2 test and Fisher's exact test were used, where appropriate, to establish any association between the different markers included in the study. Differences were considered significant for values of P<0.05.

Results

The results of immunohistological studies on the tissue arrays, in the overall classical Hodgkin's lymphoma cases, for CD20, CD79a, OCT.1, OCT.2, BOB.1, BSAP, MUM.1 and EBERs are shown in Table 1. The positivity range of the markers studied, considering the different histological subtypes except for the two nonclassificable classical Hodgkin's lymphoma cases are displayed in Tables 2, 3, 4 and 5.

Table 1 Total (N*=325)
Full size table
Table 2 Nodular sclerosis (N*=184)
Full size table
Table 3 Mixed cellularity (N*=112)
Full size table
Table 4 Lymphocyte rich (N*=18)
Full size table
Table 5 Lymphocyte depletion (N*=9)
Full size table

CD20 and CD79a were found in a small percentage of cases (18 and 18%, respectively). None of the lymphocyte depletion cases expressed CD20. Lymphocyte-rich classical Hodgkin's lymphoma was the subtype that showed more CD79a-positive cases (43%), this association being significative.

Interestingly, BSAP was strongly expressed in all the studied cases (Figure 1), with independence of the histological subtype. MUM.1 positivity was frequently detected (98% of cases), with a strong nuclear staining (Figure 2).

Figure 1
figure 1

BSAP expression. Intense staining in nucleus of all HRS cells ( × 400).

Full size image
Figure 2
figure 2

MUM.1 expression. Intense staining in nucleus of all HRS cells ( × 400).

Full size image

OCT.1 and OCT.2 were expressed in a notorious percentage of cases. Thus, when considering all the positive cases, with independence of the grade of staining, more cases expressed OCT.2 (Figures 3 and 4) than OCT.1 (62 vs 59%). Variable staining for BOB.1 was seen in the 37% of cases (Figures 5 and 6). The histological subtype that showed a higher percentage of OCT.1, OCT.2 and BOB.1 positivity, was lymphocyte-rich, and the subtype that most frequently was negative for some of these factors, was nodular sclerosis. There was not any statistical association between the histological subtype and OCT.1, OCT.2 and BOB.1 expression.

Figure 3
figure 3

OCT.2 expression. All positive cases displayed a weak labelling as compared to normal B cells ( × 400).

Full size image
Figure 4
figure 4

OCT.2 expression. Strong labelling in a non-Hodgkin's lymphoma control case ( × 400).

Full size image
Figure 5
figure 5

BOB.1 expression. A case of weak labelling in the nucleus of several, but not all, HRS cells ( × 400).

Full size image
Figure 6
figure 6

Strong BOB.1 expression in contrast to Figure 5.

Full size image

Taking into account only the strongly positive cases (++), the percentage of transcription factors positivity was much lower: OCT.1: 11%, BOB.1: 16%, OCT.2: 0% cases. There were no cases strongly positive for OCT.2 and BOB.1 simultaneously, and only 2% of cases were strongly positive for OCT.1 and BOB.1.

We found a direct association between OCT.2 and BOB.1 expression, regardless of the grade of positivity, OCT.2 expression was also statistically associated with all the B markers studied, excepting MUM.1.

Finally, in situ hybridization for EBERs was positive in 39% of cases, being mixed cellularity the subtype that more frequently showed the presence of Epstein–Barr virus (68%) (Table 6). There was not any statistical association between Epstein–Barr virus presence and transcription factor positivity.

Table 6 In situ hybridization for EBERs
Full size table

Discussion

It is already known that HRS cells are often lack in expression of several B lineage markers, such as CD20, CD79a, OCT.2 and BOB.1.4, 7, 8, 28, 29, 30 In order to assess what previous studies have suggested about the lack of expression of transcription factors in classical Hodgkin's lymphoma, we performed an immunohistochemical analysis of OCT.1, OCT.2 and BOB.1 in 325 cases, using the tissue microarray technique and correlating the results to the histo-logical subtype, B-cell marker expression, and Epstein–Barr virus presence. We decided to study the transcription factors on the neoplastic cells at the protein level because it might be more representative of their function than at the molecular level28, 31 and because there are few immuno-histochemical studies including a large series of cases.

The reviewed studies about the expression of transcription factors are based, mostly, on cellular lines and do not specify clearly the intensity of the positivity.4, 7, 8 In our study, when only the strongly positive cases were considered, independently from the histological subtype, no cases strongly expressing OCT.2 were found. The frequency of strong OCT.1 and BOB.1 positivity was poor and even lower if their simultaneous expression was analysed, as has been described elsewhere. Conversely, when we analysed our cases, with independence of the intensity of the staining, the percentage of expression was much higher than that described before: 62% for OCT.2, 59% for OCT.1 and 37% for BOB.1. Similarly, OCT.1, OCT.2, and BOB.1 expression without immunoglobulin production have been described recently by Pileri et al32 in a set of primary mediastinal B-cell lymphomas.

We did not find any significant association between OCT.1, OCT.2 and BOB.1 expression and the histological subtype, although the higher incidence of positivity of these markers in the lymphocyte-rich cases is notorious. This finding corroborates what previous studies have suggested about the fact that lymphocyte-rich classical Hodgkin's lymphoma might be more closely related to lymphocyte-predominant Hodgkin's lymphoma30 than to the other classical subtypes or that it could be in between classical and lymphocyte-predominant Hodgkin's lymphoma.33 From the 325 analysed cases, only a low percentage of them, less than expected, expressed CD20 and CD79a. This finding could be related to the use of tissue microarrays instead of whole sections, since the expression of these two markers is of varied intensities and present on a minority of the neoplastic cells. The significant association (P<0.05) between the weak signal of OCT.2 and the expression of the B markers suggests the existence of a common B lineage downregulating factor.

BSAP is a transcription factor encoded by PAX.5 gene, involved in the establishment and the maintenance of B-cell identity.29, 34 It is expressed in all B-cell stages of development from the pro-B-cell stage to the stage previous to plasma cell, where it is downregulated.35 Previous studies35, 36, 37 found a frequent BSAP expression in the HRS cells of classical Hodgkin's lymphoma. The percentage of BSAP positivity, in our cases, was of nearly 100%, which provides an evidence for the B-cell origin of this disease. Considering that PAX.5 regulates CD20, CD79a and OCT.2 expression, the low positivity rate of these markers found in our study is surprising. In order to explain a functional defect in BSAP protein of HRS cells, some studies have been attempted in search for mutations in PAX.5 gene,29 but the absence of such mutations points to a defect of other essential factors in the control of B lineage of classical Hodgkin's lymphoma. Most of our cases expressed MUM.1, a transcription factor that belongs to the interferon regulatory factor family (IRF).38 It has been demonstrated that MUM.1 is expressed in the final step of intragerminal centre B-cell differentiation and in subsequent steps of B-cell maturation.39, 40, 41 The high frequency of expression of MUM.1 in the HRS cells of classical Hodgkin's lymphoma in our cases, and in reported series,38 suggests a similar phenotype between plasma cells and HRS cells, although the presence of BSAP expression in the latter is inconsistent with a final stage of differentiation that is seen in plasma cells.

Epstein–Barr virus presence was demonstrated by in situ hybridization in a considerable number of our cases (39%), although we did not find any significative association with the expression of transcription factors. An association between the presence of Epstein–Barr virus and a decrease in the levels of the transcription factors that regulate B-cell lineage in experimental studies has been observed,42, 43 but more studies are necessary to determine the role of Epstein–Barr virus in Ig gene expression.

In conclusion, the lack of strong labelling for transcription factors OCT.1, OCT.2 and BOB.1 in neoplastic cells of classical Hodgkin's lymphoma could explain, in part, the absence of Ig expression and B phenotype of these types of lymphomas in spite of the presence of potentially functional rearrangements of Ig genes. Nevertheless, the BSAP and MUM.1 simultaneous presence in most of our cases, and the weak expression of OCT.1, OCT.2 and BOB.1 in a notorious percentage of HRS cells, may support the hypothesis that other mechanisms could be involved in Ig gene transcription and B lineage control of classical Hodgkin's lymphoma.