Overexpression of I Kappa B Alpha Without Inhibition of NF-κB Activity and Mutations in the I Kappa B Alpha Gene in Reed-Sternberg Cells

THE TRANSCRIPTION FACTOR NF kappa B (NF-κB) is a mediator of inducible gene expression in response to inflammatory stimuli.1,2 The NF-κB family comprises 5 members (p50, p52, p65, c-rel, and RelB), which form homo- and heterodimers.3 NF-κB is associated with inhibitors of the I kappa B family (IκBα, IκBβ, IκBγ, and IκBε), which are characterized by their ability to retain the transcription factor in an inactive complex in the cytoplasm. In response to external stimuli, IκBα is phosphorylated at serine residues 32 and 36 by the IκB kinase complex and subsequently degraded by the ubiquitin-proteasome pathway.2,4-9 As a consequence of IκBα degradation, NF-κB translocates into the nucleus, where it activates transcription.10 Activated NF-κB induces the transcription of its own inhibitor.11The IκBα protein is subsequently resynthesized and accumulates in the nucleus, where it dissociates NF-κB from DNA binding and contributes to its export into the cytoplasm.12 13 

Hodgkin’s disease (HD) is a malignant lymphoma characterized by the presence of mononucleated Hodgkin (H) and multinucleated Reed-Sternberg (RS) cells in a background of reactive cells comprising lymphocytes, eosinophils, plasma cells, histiocytic cells, and fibroblasts.14 Recently, we identified constitutively activated NF-κB (p50/p65) as a unique and common characteristic of HRS cells.15 Blocking of constitutive NF-κB activity by overexpression of a dominant-negative form of IκBα renders HRS cells susceptible towards apoptotic stimuli and suppressed proliferation and tumor growth of HRS cells after xenotransplantation into severe combined immunodeficiency (SCID) mice.16 These data show that constitutive NF-κB is required for apoptosis resistance and proliferation of HRS cells.

The molecular basis for constitutive nuclear NF-κB in HRS cells still needs to be investigated. The aim of this work was to identify possible molecular defects in the NF-κB regulatory system of HRS cells. In hematopoietic cells, NF-κB is mainly regulated by IκBα.17 In primary HRS cells, we found high levels of IκBα mRNA, indicating a persistently strong NF-κB–dependent transcriptional activity. To elucidate possible molecular defects in this system, we have analyzed IκBα in 7 different HD-derived cell lines. In addition, the genomic sequence of IκBα in HRS cells from 10 HD patients was analyzed using a technique involving micromanipulation of single HRS cells and analysis by polymerase chain reaction (PCR).

The analysis of 7 HD-derived cell lines showed mutations in the IκBα gene in 2 cell lines (L428 and KM-H2), which lead to C-terminally truncated proteins. These inhibitor variants may be responsible for constitutive NF-κB activity in these cell lines. Moreover, we were able to identify a mutation in the IκBα gene in HRS cells from 1 patient with HD, which encodes for a comparably defective IκBα form. Our data provide first indications that constitutive nuclear NF-κB activity in HRS cells might be a consequence of mutations in the inhibitor genes.

All cases of classic Hodgkin’s disease were selected from the files of the Institute of Pathology, Free University Berlin, and classified according to the REAL classification. The HD-derived cell lines L428, L540, L591, L1236, HDLM-2, KM-H2, and HD-MyZ were maintained in RPMI 1640 (Seromed Biochrom, Berlin, Germany), 10% heat-inactivated fetal calf serum (FCS), 2 mmol/L glutamine (GIBCO, Karlsruhe, Germany), and penicillin/streptomycin (Seromed-Biochrom, Berlin, Germany); HeLa cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) (GIBCO, Karlsruhe, Germany) supplemented with 10% FCS, 1 mmol/L sodium pyruvate, and penicillin/streptomycin.

For protein extraction 5 × 10⁶ cells were incubated in a lysis buffer containing proteinase inhibitors (complete, Mini, Boehringer-Mannheim, Germany). After 10 minutes of incubation at 4°C, the lysate was centrifuged for 5 minutes at 14,000 rpm in a microcentrifuge. Thirty micrograms protein extract were separated in 12% polyacrylamide gel containing sodium dodecyl sulfate (SDS) and blotted onto a nitrocellulose membrane (Schleicher & Schüll, Dassel, Germany) by electroblotting. Western blots were analyzed with chemiluminescence following the manufacturer’s recommendations (ECL system, Amersham, Braunschweig, Germany). Antibodies directed against the N-terminus (C-15) and the C-terminus (C-21) of IκBα were obtained from SantaCruz Biotechnology Inc (Heidelberg, Germany). The RNA was prepared using the RNeasy Kit (Quiagen, Hilden, Germany) according to the manufacturer’s recommendations.

The cDNA was synthesized under the following conditions: 1 μg total RNA was incubated with 2 μL 10 × incubation buffer, 1 mmol/L each deoxyribonucleoside-triphosphate (dNTP), 5 mmol/L MgCl₂, 50 U of RNAse inhibitor, and 500 ng Oligo-(dT)₁₅ primer. After a denaturation step (65°C for 15 minutes), 20 U of avian myeloblastosis virus (AMV) RT (Boehringer-Mannheim, Germany) were added. The reaction was incubated for 1 hour at 42°C.

To amplify full-length IκBα cDNAs, 4 different pairs of primers were designed (Table 1). The conditions for the PCR consisted of an initial denaturation step of 90 seconds at 95°C, 30 cycles of 40 seconds at 95°C, 40 seconds at 60°C, and 60 seconds at 72°C. Buffer conditions were as follows: 1.25 μmol/L MgCl₂, 200 μmol/L each dNTP, 10 pmol/L of each primer, and 1 U of Taq polymerase (InViTek, Berlin, Germany).

Twenty microliters of each PCR was analyzed on ethidium bromide–stained agarose gels (1%) and the amplificates were isolated from the gel by the glass milk technique for DNA sequencing.

Single HRS cells were isolated from CD30 immunostained frozen sections as previously described.18 In brief, single cells were extracted by hydraulic micromanipulators from the surrounding tissue and transferred into PCR tubes with a minimal volume of buffer (0.05 to 0.1 μL) covering the tissue sections. For control, an aliquot of at least 1 μL was drawn from the buffer covering the tissue sections during the cell isolation procedure and subjected to single-copy PCR.

To amplify the C-terminal portion of the IκBα gene, we designed different sets of PCR primers capable of generating 3 overlapping amplificates in a 2-step nested primer PCR (Table 2). The first PCR contained all 6 primers in 1 assay, whereas for reamplification, the 3 primer pairs were separately applied. The conditions for the first round of PCR consisted of a denaturation step of 2 minutes at 95°C, 5 cycles of 40 seconds at 95°C, 60 seconds at 58°C, and 120 seconds at 72°C, followed by 38 cycles of 40 seconds at 95°C, 60 seconds at 58°C, and 60 seconds at 72°C. The final extension lasted 10 minutes at 72°C. For reamplification, an aliquot (1.5%) of the first PCR was subjected to each of the 3 PCRs and amplified under the following conditions: an initial step of 2 minutes at 95°C and 40 cycles of 20 seconds at 95°C, 40 seconds at 58°C, and 60 seconds at 72°C followed by the final step of 10 minutes at 72°C.

The same buffer conditions were used for both the first and the second amplification round: 1.5 mmol/L MgCl₂, 200 μmol/L each dNTP, 8 pmol/L of each primer and 2 U of AmpliTaq (Perkin-Elmer, Weiterstadt, Germany). A total of 6 μL of each PCR was analyzed on ethidium bromide-stained agarose gels (1%), and the amplificates were isolated from the gel by the glass milk technique for DNA sequencing.

PCR products obtained from the HD-derived cell lines or from individual HRS cells were sequenced by the chain termination technique using fluorescence-labeled ddNTPs (BigDye; PE Applied Biosystems, Weiterstadt, Germany). The sequencing reactions were analyzed on an automated DNA sequencer (377A; Applied Biosystems) and the resulting sequences were compared with the sequence of the IκBα gene.19 20 

Radioactive in situ hybridization for the detection of IκBα mRNA was performed on paraffin sections. For this purpose, we generated a hybridization probe by cloning a portion of the IκBα cDNA (from base 540 to base 1379; M69043)19 into pKS. Radioactive-labeled run-off transcripts were prepared after linearization of the plasmid and applied to the pretreated tissue sections as previously described. In brief, dewaxed and rehydrated paraffin sections were exposed to 0.2 N HCl, 0.6 mg/mL pronase, followed by postfixation with 4% paraformaldehyde. After acetylation with 0.1 mol/L triethanolamine pH 8.0/0.25% (vol/vol) acetic anhydride and dehydration in graded ethanols, the slides were separately hybridized to 2 to 4 × 10⁵ cpm of the labeled sense and antisense probes and left overnight at 50°C. Washing and autoradiography were performed as previously described.21 

To examine IκBα mRNA expression in primary HRS cells, we performed an in situ hybridization with an IκBα-specific cDNA probe spanning nucleotide from base 540 to base 137919 in 20 HD cases. Most cases harbored abundant amounts of mRNA in the HRS cells (Fig 1, Table 3). In contrast to the HRS cells, only some reactive lymphocytes expressed low to moderate amounts of IκBα-specific transcripts (Table 3). A labeling of reactive lymphocytes with varying signal intensity was also observed in 3 cases of infectious mononucleosis. Less than 1% of the tumor cells of 5 cases of B-cell chronic lymphocytic leukemia (B-CLL) and less than 10% of the neoplastic cells of 5 cases of T-cell non-Hodgkin lymphoma (T-NHL) showed low to moderate amounts of IκBα mRNA.

We examined IκBα proteins in 7 different HD-derived cell lines (L428, L540, L591, L1236, HDLM-2, KM-H2, and HD-MyZ). Western blot analysis was performed using antibodies directed against the N- and the C-terminal epitope of IκBα.

The analysis of L428 cells with the antibody against the N-terminus showed the expression of a faster migrating IκBα form of about 30 kD (Fig 2A). When using the antibody directed against the C-terminus of IκBα, no protein was detectable in these cells (Fig 2B). Therefore, L428 cells appear to contain a C-terminally truncated form of the IκBα protein. In KM-H2 cells, no IκBα protein could be detected with both antibodies. The 5 remaining HD-derived cell lines (L540, L591, L1236, HD-LM2, and HD-MyZ) showed expression of full-length IκBα proteins of 38 kD (Fig 2).

Next, we analyzed the IκBα mRNA and the IκBα gene in HD-derived cell lines. Transcripts for IκBα were amplified by RT-PCR and sequenced. In L428 and KM-H2 cells, we detected mutated IκBα transcripts. L428 cells showed a point mutation at nucleotide position (pos) 893 of the human cDNA19 generating a preterminal stop codon (Fig 3A). When we examined KM-H2 cells, we detected 2 deletions, 1 between pos 509 and pos 613 and another between pos 618 and pos 640 (Fig 3B). These deletions result in a frame shift and are followed by a preterminal stop at pos 715. Genomic sequencing showed that the observed alterations are due to mutations in the IκBα gene. The genomic PCR detected only mutant and no wild-type IκBα sequences. These data show that the occurrence of defects of IκBα proteins in HRS cells is a consequence of mutations in the IκBα gene. In accordance with the Western blot data in both cell lines, the mutant forms of IκBα were expressed exclusively and no transcripts of wild-type IκBα were detectable.

To investigate whether IκBα mutations could also be detected in primary HRS cells, we analyzed single HRS cells from lymph node biopsy samples from patients with HD. A total of 420 individual HRS cells were isolated from frozen sections of 10 patients, and the part of the IκBα gene comprising the mutations in the cell lines (2283 nt to 4391 nt)20 was divided into 3 parts and amplified by PCR. Between 5 and 11 PCR products were obtained from each amplified region, sequenced, and compared with the IκBα gene sequence.

In 1 case, we detected a point mutation in exon V (pos 3398; TGT → TGA) causing a stop codon, with the consequence of a preterminal breakage of the protein synthesis at amino acid 214 (Table 4; Fig 4). From 72 HRS cells of this case, 22 PCR products were obtained, 11 of which comprised the affected 3′-region of the IκBα gene. In 4 of 11 cells, we detected the stop codon exclusively, whereas an additional 4 cells contained both the mutant and the wild-type IκBα gene. This alteration did not occur in the remaining 3 HRS cells. This finding indicated that only 1 allele harbored the stop codon, whereas the other allele was wild-type. The IκBα gene of reactive bystander lymphocytes isolated from the same biopsy tissue showed no alteration, indicating that this stop codon was restricted to the HRS cells.

Base substitutions in the introns were detectable in all cases, as well as in reactive lymphocytes (Table 4). However, because these modifications do not affect the coding sequence, they most likely represent irrelevant interindividual variations. In addition, the same silent base substitution was found in exon II of 4 cases (Table 4), indicative of an interindividual polymorphism.

The aim of this work was to identify molecular defects in the NF-κB regulatory system leading to constitutive NF-κB activation in HRS cells. In IκBα^−/− mice, constitutive NF-κB activation could be observed in lymphoid cells.17Hence, this inhibitor appears to play a major role in regulating NF-κB activity in the lymphoid system.

In cultured HRS cells, mRNAs for IκBα are strongly overexpressed.15 To test if this is also valid for primary HRS cells, we examined 20 cases of classic HD by a highly sensitive radioactive in situ hybridization with an IκBα-specific cDNA probe. In all cases, we detected overexpression of IκBα mRNA in the HRS cells, whereas in normal lymphoid tissues and in cases of B-CLL and T-cell NHL, no or very little amounts of IκBα mRNA were found. Therefore, high levels of IκBα are a highly characteristic feature of cultured and primary HRS cells.

Because NF-κB activates transcription of its own inhibitor, this finding reflects the high transcriptional activity of NF-κB in this lymphoid malignancy.11 It has been demonstrated that enforced expression of IκBα leads to accumulation of the protein in the nucleus.12 Nuclear IκBα is involved in the export of NF-κB into the cytoplasm, thus contributing to the termination of NF-κB activity.13 The paradoxical finding of high NF-κB activity despite the strong expression of its inhibitor indicates that the NF-κB/IκB system is severely deregulated in HRS cells. A mechanism leading to this loss of control could be a defect of the IκBα molecule. To investigate this possibility, we analyzed IκBα transcripts in 7 HD-derived cell lines and found a disruption of the coding sequence by a stop codon in the cell line L428, and by 2 deletions followed by a stop codon in the cell line KM-H2. These mutations found in the RNA were confirmed by genomic sequencing, disclosing the existence of only mutated IκBα genes in the absence of wild-type alleles. In accordance with these sequence data, we detected expression of exclusively C-terminally truncated proteins in both cell lines consistent with previous reports.22 23 

To analyze whether comparable IκBα mutations also occur in primary HRS cells, we isolated single HRS cells from CD30 immunostained tissue sections of 10 HD patients and analyzed the IκBα gene by single cell PCR. The IκBα gene of 1 case (case 1; Table 4) contained a mutation generating a preterminal stop of the translational machinery. The expected mRNA codes for a C-terminally truncated protein of about 214 amino acids comparable to that found in L428 cells. This stop codon was not found in nonmalignant lymphoid bystander cells of the same case, indicating that this mutation was specific for the HRS cells. No disruptive alterations were detectable in the IκBα genes of the HRS cells in the remaining 9 HD cases.

In contrast to cell lines with only mutant genomic sequences (L428 and KM-H2), the mutated HRS cells of case 1 contained both mutant and wild-type sequences, suggesting a monoallelic mutation of IκBα and presumably leading to coexpression of mutant and wild-type IκBα proteins. This raises the question as to the functional significance of this monoallelic mutation. It is well conceivable that the C-terminally truncated protein can block the wild-type protein. Aside from its function of retaining NF-κB in the cytoplasm, IκBα can localize in the nucleus and dissociate NF-κB-DNA complexes.13 A prerequisite for effective inhibition of NF-κB-DNA binding are the ankyrin repeats and an intact C-terminus, in which IκBα is constitutively phosphorylated at serine and threonine residues by casein kinase II.24-27 A role of the C-terminal end for the inhibition of NF-κB-DNA binding has recently been suggested by an x-ray structure analysis of the NF-κB/IκBα complex.28,29 Furthermore, IκBα contains a nuclear export signal (NES) at amino acids 264-281, which confers active shuttling of the NF-κB/IκBα complex to the cytosol.13 20 Therefore, the functionally defective truncated IκBα may protect DNA-bound NF-κB from dissociation by wild-type IκBα and interfere with the nuclear export pathway. As a consequence of this process NF-κB-DNA binding activity would be maintained in HRS cells.

Although the observed mutation is found only in 1 of 10 cases, our data provide the first indication that permanently activated NF-κB in primary HRS cells might be a consequence of gene mutations of one of its inhibitors. Studies on constitutive NF-κB activity have shown that a hypophosphorylated form IκBβ shields NF-κB from IκBα-mediated inhibition and plays an important role in permanent NF-κB activation. It is therefore tempting to speculate that defects of additional members of the IκB family contribute to the functional blockage of IκBα in the remaining cases. Further expression and molecular studies are required to investigate this possibility.

During the review process, a report that was submitted after ours has appeared dealing with the analysis of IκBα mutations in enriched RS cells. The investigators described IκBα mutations in 2 of 8 HD cases, which cause a preterminal truncation of the IκBα protein.30