Lymphoma Genomic Features
AITL samples harbor recurrent somatic mutations in genes encoding epigenetic regulators such as tet methyl cytosine dioxygenase 2 (TET2), DNA methyl transferase 3 alpha and isocitrate dehydrogenase 2 [6]). All of these mutations occur frequently in numerous hematologic malignancies, whereas TET2 mutations are detected more frequently in PTCL exhibiting a TFH phenotype compared to those without a TFH phenotype [7]. The p.Gly17Val mutations in the ras homologue family member A (RHOA) (designated as the G17V RHOA mutations) are specifically found in up to 50–70% of AITL and its related lymphomas with a Tfh phenotype [6, 8, 9].
Tet2 Gene Trap Mice
Homozygous Tet2 gene trap(Tet2gt/gt) mice harbor a gene‐trap vector in the Tet2 second intron [10]. Muto et al. reported that 70% of Tet2gt/gt mice developed T‐cell lymphomas with Tfh‐like gene expression patterns around 67 weeks old [10]. DNA methylation analysis revealed that lymphoma cells of Tet2gt/gt mice exhibited increased methylation at transcriptional start sites, gene bodies and CpG islands, and decreased hydroxymethylation at transcriptional start site regions [10]. Hypermethylation of the first intronic silencer region of the Bcl6 gene reportedly has been known to inhibit CCCTC‐binding factor (CTCF) binding to this locus and promotes Bcl6 transcription [11]. Density of methylation was increased in lymphoma cells of Tet2gt/gt mice compared with control CD4+ cells [10]. Upregulation of Bcl6, encoding a master transcriptional regulator in Tfh development finally results in outgrowth of Tfh‐like cells in Tet2gt/gt mice [10]. These results suggest overall that decreased Tet2 function contributes to AITL initiation. Notably, hypermethylation of the corresponding region in BCL6 locus is also found in human PTCL samples with TET2 mutations [12].
G17V RHOA Mouse Model
To determine effects of G17V RHOA expression, multiple independent G17V RHOA model mice have been established using either retroviral transduction [13], knock‐in [14], or transgene [15, 16] approaches. These G17V RHOA model mice did not develop AITL‐like lymphomas, although increase of Tfh cell populations [14, 15] and autoimmunity [15] were observed in some lines of mice. Therefore, the appearance of oncogenic phenotypes may require additional gene mutations.
Because the G17V RHOA mutations are almost always accompanied by loss‐of‐function TET2 mutations in human AITL samples [6], mice expressing the G17V RHOA mutant in Tet2‐null background were established by using distinct approaches [13–16]. Briefly, Zang et al. transduced Tet2‐null T cells with retrovirus harboring G17V RHOA, and they performed adoptive transfer of these cells into T‐cell receptor (TCR) α‐deficient mice [13]. Cortes et al. crossed G17V RHOAconditional knock‐in (cKI), Tet2conditional knockout (cKO), and CD4CreERT2 mice, in which tamoxifen induces expression of G17V RHOA mutant and deletion of Tet2 gene in T cells, and then transplanted bone marrow cells of these mice into irradiated C57BL/6 mice, followed by immunization with sheep red blood cells [14]. Ng et al. crossed G17V RHOA transgenic mice with Tet2 cKO x Vav‐Cre mice, lacking Tet2 gene in hematopoietic cells, and OT‐II TCR mice, followed by immunization with NP‐40‐Ovalbumin [15]. Nguyen et al. crossed G17V RHOA transgenic mice with Tet2cKO x Mx1‐Cre mice, lacking Tet2 gene in hematopoietic cells with injection of polyinosinic:polycytidylic acid (pI:pC) [16]. Model mice established by Zang et al. revealed skewed T‐cell differentiation such as increase of TFH cells and decrease of T regulatory cells, and abnormal expansion of CD4+ T cells [13]. Mice established by Cortes et al., Ng et al., and Nguyen et al. developed AITL‐like lymphomas after 25, 38, and 48 weeks, respectively [14–16]. Zang et al. also reported that TET2 loss and G17V RHOA expression synergistically inactivates FoxO1: Tet2 loss leads to increase of methylation in FoxO1 promoter that suppress its transcription in CD4+ T cells, whereas the G17V RHOA mutant elevates phosphorylation of FoxO1 and promotes its translocation from the nucleus to the cytosol, where it undergoes proteasomal degradation [13]. This agrees with findings of mTORc1‐associated gene expression by Ng et al. in G17V RHOA‐expressing tumor cells, as FoxO1 suppresses mTORc1 signaling [15]. Likewise, Cortes et al. reported that G17V RHOA activates PI3K‐mTORc1 signaling in CD4+ T cells dependent on ICOS‐ICOSL signaling [14]. Accordingly, tumor cell proliferation was inhibited by everolimus, a mTOR inhibitor [15] and duvelisib, a PI3K inhibitor [14], supporting the idea that ICOS‐PI3K‐mTORc1 signaling drives Tfh cell expansion in mice. Fujisawa et al. reported that hyperactivation of TCR signaling is an essential downstream event of the G17V RHOA mutant in vitro: the G17V RHOA mutant binds to and activates VAV1, a key component of TCR signaling [17]. Nguyen et al. reported that phosphorylation of VAV1 and activation of TCR signaling were also observed in murine tumor cells [16]. Dasatinib, a multikinase inhibitor effectively prolonged the survival of mice through inhibiting the TCR signaling pathway [16].
PDX Models of Angioimmunoblastic T‐cell Lymphoma
A PDX model was established by inoculating primary AITL tumor cells and microenvironmental reactive cells into NOD/Shi‐scid, IL2Rgammanull (NOG) mice [18]. The immunohistological features of tumors in PDX mice recapitulated those of AITL patients. Additionally, human immunoglobulin G/A/M was detected in the sera of PDX mice, indicating that patient‐derived B and plasma cells were activated by AITL tumor cells in mice. Their analysis suggests that the function of TFH cells in AITL cells was reconstituted in the PDX mice.
Anaplastic Large T‐cell Lymphoma
Microscopically, anaplastic large‐cell lymphoma (ALCL) is commonly marked by large cells with abundant cytoplasm and eccentric, lobulated nuclei [1]. CD30 is highly expressed on the surface of ALCL cells [1]. ALCL is classified in two distinct diseases by the expression of ALK: ALCL, ALK positive and ALCL, ALK negative. The translocations involving ALK gene with various partner genes are essential mechanisms in ALK‐positive ALC. The most frequent translocation, t(2;5)(p23; q35), fuses a portion of the nucleophosmin (NPM)1 gene on chromosome 5q35 with a portion of ALK on chromosome 2p23 [1]. The NPM1‐ALK fusion gives rise to a chimeric protein consisting of the NPM1 N‐terminus with the ALK catalytic domain [19].
Viral and Chimeric Models
Chimeric models have been created by transplanting bone marrow cells transduced with a retroviral vector carrying NPM1‐ALK cDNA into lethally‐irradiated mice. The first model was reported by Kuefer et al., who infected 5‐fluorouracil‐treated murine bone marrow cells with NPM1‐ALK cDNA using retrovirus and then injected them into lethally irradiated BALB/cByJ mice [20]. Transplanted mice developed B‐lineage large cell lymphomas at four to six months in mesenteric lymph nodes, with metastases clearly associated with aberrant ALK activation [20]. Miething et al. later asked whether B‐cell phenotypes seen in the Kuefer at al. study were attributable to low infection efficiencies and hence low NPM1‐ALK expression. To address this possibility, they compared infection conditions employing low versus high multiplicity of infection and confirmed that lower multiplicity of infection promoted plasmacytomas around 12–16 weeks in mice, while higher multiplicity of infection caused aggressive histiocytic malignancies around 3–4 weeks [21]. Miething et al. also developed a murine model of ALCL by employing a Lox‐STOP‐Lox‐NPM1‐ALK encoding vector [22]. They then infected bone marrow cells from two sets of mice: one expressing Cre controlled by the lysozyme M‐promotor (lysM‐mice), which is active in the myeloid compartment and the other expressing Cre from the granzyme B‐promotor (GrzmB‐mice), which is mainly active in T cells. Mice transplanted with bone marrow cells from lysM‐mice developed histiocytic malignancies around four
