Germline mutations of the perforin gene are a frequent occurrence in childhood anaplastic large cell lymphoma

Authors


Abstract

BACKGROUND.

Monoallelic and biallelic mutations of the PRF1 gene have been reported in some cases of childhood lymphoma. Anaplastic large cell lymphoma (ALCL) accounts for 10% to 15% of all childhood lymphomas. To assess the possible role of PRF1 mutations in ALCL, the authors screened a series of patients collected by the Associazione Italiana di Ematologia Oncologia Pediatrica (AIEOP).

METHODS.

The authors investigated 44 patients with ALCL by direct sequence of the PRF1 gene. To address the issue of the prevalence of the most frequently observed PRF1 mutations in the control population, the authors examined a series of 400 healthy white control subjects for the 272C>T mutation (A91V).

RESULTS.

A total of 6 different mutations were identified in 12 patients (27.3%). Eleven patients had 1 mutation whereas 1 patient was found to have 2 mutations. Of the 6 PRF1 mutations identified, 2 were novel mutations: 529C>T (resulting in R177C) and 1471G>A (resulting in D491N). The remaining 4 mutations were previously described; in particular, the 272C>T mutation (resulting in the A91V amino acid change) was found in 8 patients, whereas the 368G>A (R123H), 695G>A (R232H), and 1262T>G (F421C) mutations were all found in 1 case each. Overall, the incidence of PRF1 mutations was found to be significantly higher in patients with ALCL compared with 400 control subjects, among whom only heterozygous A91V was observed in 41 subjects (10.2%) (chi-square test, 10.9; P <.01).

CONCLUSIONS.

Patients with childhood ALCL have a higher probability of being a carrier of a PRF1 mutation compared with healthy controls, suggesting a possible predisposing role. Cancer 2007. © 2007 American Cancer Society.

Anaplastic large cell lymphoma (ALCL) was first described approximately 20 years ago by Stein et al.1 This disease covers the majority of cases diagnosed historically as malignant histiocytosis,2, 3 some cases of Hodgkin disease, and other types of non-Hodgkin lymphoma. It is characterized by the proliferation of neoplastic lymphoid cells that coexpress several activation antigens such as CD30 (Ki-1), the epithelial membrane antigen (EMA), and interleukin-2 (IL-2) receptor.2, 4 With respect to lymphocyte lineage markers, it is widely accepted that the majority of the cases are associated with a T-cell or null cell immunophenotype. According to the recent World Health Organization (WHO) classification of lymphomas,5 ALCL is now included as a subgroup of mature T-cell lymphoma.

The chromosomal translocation (2;5) (p23;q35) is associated with ALCL.6 This produces fusion of part of the nucleophosmin NPM gene on 5q35 to the ALK receptor tyrosine kinase gene on 2p23.7 This fusion gene is responsible for expression of the NPM-ALK protein detected by the ALK-1 monoclonal antibody, which therefore is used as a diagnostic tool.

ALCL accounts for 10% to 15% of all childhood lymphomas.8 Its clinical features include a predominance of systemic symptoms and an unusually high frequency of extranodal involvement, particularly of the skin. Hemophagocytosis is observed in a proportion of cases of ALCL.9–12

Perforin plays a key role in the cytotoxicity of natural killer (NK) and cytotoxic T lymphocytes (CTL).13 It is stored as an active protein in specialized secretory lysosomes, known as lytic granules, of NK and CTL.14 At the time of target cell recognition, lytic granules polarize and release their contents at the immunologic synapse.15 Secreted perforin inserts into the lipid bilayer, polymerizing to form pores in the membranes of target cells and allowing the entry of a series of granzymes that activate apoptotic pathways in the target.13 Perforin binds the target cells via a MAC domain that is able to bind phospholipid head groups in the presence of calcium. The MAC domain is hidden during biosynthesis by a carboxy-terminal propiece containing a large N-linked glycan, rendering perforin inactive.16 Removal of the propiece activates perforin by revealing the MAC domain. Processing occurs in a post-Golgi compartment by an as-yet unidentified processing enzyme. Perforin is then stored as an active protein in the lytic granules.17

Perforin gene (PRF1) mutations have been associated with an autosomal recessive immune deficiency, familial hemophagocytic lymphohistiocytosis of type 2 (FHL2, MIM 603553).18 Some of the presenting clinical features of FHL2 also may be common to ALCL, including fever, lymphadenomegaly, skin rash, and hemophagocytosis.19, 20 Therefore, some connection between the 2 disorders has been suggested.21, 22 Furthermore, the association between the PRF1 mutation and conditions other than FHL2, including lymphoma, has recently been reported. Clementi et al. reported 4 patients with Hodgkin or non-Hodgkin lymphoma who had biallelic PRF1 mutations and 4 additional patients with monoallelic PRF1 mutations.23 Mutations decreasing the function of the Fas death receptor cause autoimmune lymphoproliferative syndrome (ALPS) with autoimmune manifestations, spleen/lymph node enlargement, and expansion of CD4/CD8-negative T cells. Dianzani autoimmune lymphoproliferative disease (DALD) is a variant lacking this expansion. Among 14 ALPS patients, 28 DALD patients, and 816 controls, PRF1 N252S amino acid substitution was found in 2 ALPS patients, and an A91V amino acid substitution was reported in 6 DALD patients. N252S conferred an odds ratio (OR) of 62.7 (P = .0016) for ALPS and A91V conferred an OR of 3 (P = .016) for DALD. The copresence of A91V and variations of the osteopontin gene previously associated with DALD conferred an OR of 17 (P = .0007) for DALD.24 These data indicate that some perforin amino acid changes, either alone or in combination with other mutations of genes involved in lymphocyte survival or functional activity, may be present in patients with lymphoma or disorders related to cell apoptosis, supporting the concept that perforin also plays a key role in the mechanisms of immune surveillance that prevent tumor growth and/or development.

Based on this background, we decided to asses the frequency of PRF1 mutations in children with ALCL compared with a healthy control population. To address this issue, we studied a relatively large series of cases prospectively enrolled in the current trial of the Associazione Italiana di Ematologia Oncologia Pediatrica (AIEOP).

MATERIALS AND METHODS

Diagnosis

The diagnosis of ALCL was based on the morphologic and immunologic criteria defined in the WHO classification for non-Hodgkin lymphoma.5 The diagnosis was centrally reviewed and each case was assigned to 1 of the morphologic subtypes of ALCL.25 Immunohistochemistry was performed on routine sections and on frozen sections, when available, with a panel of monoclonal antibodies. All the cases were investigated with the ALK1 antibody directed toward the NPM/ALK protein associated with the t(2;5) translocation.26 All patients were examined for expression of the NPM-ALK transcript.

Inclusion Criteria

Patients included in the study were diagnosed as having systemic ALCL and were treated according to the current AIEOP protocol between 1997 and 2004. Approval for these studies was obtained from the Institutional Review Board.

Genetic Study

Genomic and mRNA sequences of the PRF1 gene were retrieved from the National Center for Biotechnology Information (NCBI). Genomic DNA was prepared from the peripheral blood samples obtained from the patients. To analyze the PRF1 gene, the 2 coding exons and the adjacent intronic regions were amplified and directly sequenced, in both directions, with the BigDye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems).

Perforin has 3 exons, and the coding region of the gene was amplified from exon 2 and exon 3. Exons 2 and 3 were amplified in 2 segments (Fig. 1) each and the primers used were as follows:

  • 3231F 5′CATGGCTTCCCAGAGCCCAAG3′,

  • 3596R 5′CTGGTTACATGGCGCTGGCAG3′,

  • 3449F 5′TCCCAGTGGACACACAAAG3′,

  • 3917R 5′AGCAGCCTCCAAGTTTGATTGG3′,

  • 4969F 5′CAGCTGAGGTCTCTCTCTTC3′,

  • 5645R 5′TGGTGACCGCTGAGCCATGGC3′,

  • 5581F 5′GACTGCAGCCGGCCGTGCCCA3, and

  • 6143R 5′CCTTTCCAAGCTCACTGTTC3 ′.

Figure 1.

Primer position for the amplification strategy and details of the mutations found.

Sequences obtained by the ABI PRISM 3130 Sequence Detection System (Applied Biosystems) were analyzed and compared with the reported gene structure using the dedicated software SeqScape (Applied Biosystems).

To address the issue of the prevalence of the most frequently observed PRF1 mutation 272C>T (A91V) in the control population, we examined a series of 400 healthy white control subjects. To search for this individual mutation by “minisequencing,” we amplified PFR1 exon 2. Polymerase chain reaction products were used to perform a primer single-base extension reaction (SNaPshot Kit; Applied Biosystems). The reaction is based on a specific minisequencing primer, which is exactly 1 base short of the mutation position and fluorescent ddNTP. The SNP primer sequence for the A91V mutation is the following: 5′CCCTCCAGCGCCTGCCTCTGG3′. The product obtained is separated by electrophoresis and the color of the peaks obtained makes it possible to identify the genotype (GeneScan Analysis Software; Applied Biosystems).

Statistical Analysis

The difference in the frequency of mutations between patients and controls was tested using the chi-square method. A P value <.05 was considered to be significant.

RESULTS

Study Population

We investigated a total of 44 patients in whom systemic ALCL had been histologically diagnosed. All the patients were of Italian origin. No PRF1 mutations were found in the coding exons and adjacent regions in 32 patients, whereas a total of 6 different mutations were identified in the remaining 12 patients (27.3%). Eleven patients were found to have 1 mutation, whereas 1 patient had 2 mutations (Table 1).

Table 1. Presenting Features of 12 Children With Anaplastic Large Cell Lymphoma Who Also Carried PRF1 Mutations
Case No.MutationEffectSex/Age, yInvolved sitesImmunophenotypeNPM/ALK
  • CNS indicates central nervous system.

  • *

    Not previously reported.

1272C>TA91VGirl/7Skin, mediastinumT-cellPositive
3272C>TA91VGirl/11CNST-cellPositive
4368G>AR123HBoy/14MediastinumT-cellPositive
10272C>TA91VGirl/8Lymph nodesT-cellNegative
17272C>TA91VBoy/10Mediastinum, lymph nodesT-cellPositive
24272C>TA91VGirl/11Skin, mediastinumUndeterminedPositive
26272C>TA91VBoy/7Mediastinum, lymph nodesT-cellPositive
27529C>T*R177CBoy/1CNST-cellPositive
281471G>A*D491NBoy/12MediastinumT-cellPositive
401262T>GF421CBoy/13Lymph nodes, spleenT-cellPositive
43272C>TA91VBoy/10Lymph nodes, mediastinumT-cellPositive
44272C>TA91VGirl/13Lymph nodes, bone marrowT-cellPositive
695G>AR232H

Of the 6 PRF1 mutations identified, 2 were novel mutations: 529C>T, resulting in R177C, and 1471G>A, resulting in D491N. The remaining 4 mutations were previously described; in particular, the 272C>T mutation, resulting in the A91V amino acid change, was found in 8 patients. The remaining 3 mutations (ie, 368G>A [R123H], 695G>A [R232H], and 1262T>G [F421C]) had been observed previously (unpublished data and references27–31) (Fig. 2).

Figure 2.

Electrophorograms indicating the 6 mutant nucleotides and their changes in amino acid code in the patients with a mutation.

Frequency of PRF1 Mutations in Controls

Heterozygous A91V was observed in 41 of 400 controls, accounting for a prevalence of 10.2%. Of 800 chromosomes, 41 272T alleles were observed.

The incidence of any PRF1 mutation was significantly higher in patients with ALCL compared with control subjects (chi-square test, 10.9; P <.01).

DISCUSSION

The possible role of some predisposing genetic mutations in the development of leukemia and lymphoma is currently the focus of many investigations. Escape from immune surveillance is considered a major putative mechanism.32–34 Perforin-mediated cellular cytotoxicity is a highly preserved mechanism for the killing of virus-infected and neoplastic cells. In the current study, we found that PRF1 mutations were present in 27% of children with ALCL. This percentage was much higher than in our normal control population. This finding is also in keeping with the recent report by Clementi et al. of an 8% frequency of PRF1 mutations in 816 normal controls.23 Therefore, we can conclude that PRF1 mutations can be found more frequently in patients with ALCL compared with healthy controls.

The 272C>T substitution resulting in the A91V amino acid change was found to be the most frequent single mutation in patients with ALCL. This is also in keeping with data from patients with non-Hodgkin lymphoma.23 The pathogenic role of A91V in hemophagocytic lymphohistiocytosis has been widely debated, and the current opinion shared by different groups is that it is not sufficient per se to develop FHL type 2.35 Nevertheless, several lines of evidence have documented that the A91V mutation is not neutral with regard to perforin function. Voskoboinik et al. reported that in basophil leukemia cells in a rat model, A91V perforin was expressed at reduced levels compared with wild-type perforin, resulting in a partial loss of lytic capacity.36 Trambas et al. demonstrated that A91V perforin from a patient with hemophagocytic lymphohistiocytosis (HLH) was not recognized using an antibody raised against native perforin (δG9), but was readily detected using an antibody raised against a peptide epitope (2d4), suggesting that the epitope recognized by δG9 is destroyed by A91V. A91V perforin undergoes conformational changes and impaired cleavage, which most likely explains the reduced cytotoxicity in CTL and NK cells and is also likely to contribute to the pathogenesis of HLH.37 Taken together, these data suggest that A91V cannot be considered only as a neutral polymorphism but rather as a functional polymorphism–or a mutation–that affects the function of perforin.

In conclusion, the results of the current study demonstrate that patients with childhood ALCL have a higher probability of being carrier of a PRF1 mutation compared with healthy controls. Therefore, a possible predisposing role for genetic variants affecting perforin-mediated cellular cytotoxicity machinery, also for this lymphoma subtype, can be hypothesized. Extending this study to larger series of patients with ALCL as well as to adult patients is warranted.

Acknowledgements

We are grateful to all the Italiana di Ematologia Oncologia Pediatrica (AIEOP) clinical centers participating to the AIEOP-NHL clinical trial.

Ancillary