Читать книгу The Peripheral T-Cell Lymphomas - Группа авторов - Страница 49

Neoplasias have a distinct pattern of disrupted pathways, which are the result not only of genetic alterations, but also of heritable patterns of disrupted gene expression. Epigenetics refers to these clonal changes in patterns of gene expression that are mediated by mechanisms that do not alter the primary DNA sequence. Epigenetic changes in tumors mostly result in inappropriate gene silencing and effect a diverse variety of cellular functions, often affecting the tumor itself and its interaction with the microenvironment and immune response [1].

Epigenetic changes generally are the result of two processes: (i) DNA methylation (DNA global hypomethylation and promoter‐localized hypermethylation) and (ii) histone modifications. The proteins responsible for the alterations characteristic of the cancer epigenome are the enzymes that catalyze DNA methylation, the proteins that bind methylated DNA at promoters and contribute to transcriptional silencing, and the chromatin modifier enzymes that catalyze histone acetylation, deacetylation, methylation, and demethylation.

Adding another layer of complexity, not only can gene transcription be altered within the abnormal cell by abnormal changes in these processes, cancers can harbor specific mutations of key epigenetic controlling genes. For example, the methylation of cytosine residues to 5‐methylcytosine (5mC) is mediated by DNA methyltransferases (DNMTs), and gain‐of‐function mutations of DNMT3A have been identified frequently in hematologic malignancies. A good example of this might be Sézary syndrome [2–5]. Other such mutations discovered within TCLs include IDH2, TET2, MLL2, KMT2A, KDM6A, CREBBP, and EP300 genes [3, 6–10]. Thus, TCLs appear to be a group of diseases with significant epigenetic vulnerabilities that can be exploited in rational therapeutic approaches.

The best understood, and currently the most clinically relevant, of these derangements involves histone modification mediated by HDAC proteins. HDACs are classified by their homology to yeast HDACs. To date, 18 HDACs have been described, of which 11 are zinc‐dependent metalloproteinases belonging to classes I, II, and IV. Specific HDAC enzymes in these classes constitute the focus of most HDAC research, which, usually in conjunction with other co‐repressors, deacetylate the terminal amino‐moieties of lysine on the histone protein [11]. The acetylation status of histone depends on the balance between deacetylase activity and histone acetyltransferase (HAT) activity in the cell. Deacetylation results in a relatively closed chromatin conformation that leads to repressed transcription [11]. Thus, HDAC inhibitors (HDACi) are generally considered to be transcriptional activators [12]. However, gene expression profiling has demonstrated that as many genes may be repressed as de‐repressed after exposure to an HDACi. This is likely to be a consequence of the direct and indirect effects of these drugs on other transcriptional regulators and cell signaling pathways, and/or the dynamic and complex interrelations between chromatin remodeling and regulated gene transcription [13, 14].

HDACi are currently classified according to their chemical structure, and each agent varies in its ability to inhibit individual HDACs (Table 3.1). HDACi share a common pharmacophore containing a cap, connecting unit, linker, and a zinc‐binding group that chelates the cation in the catalytic domain of the target HDAC [15]. Examples of the pan‐deacetylase inhibitors include vorinostat (suberoylanilide hydroxamic acid, SAHA), panobinostat (LBH589), chidamide, and trichostatin A, which inhibit class I, II, and IV HDACs, while valproate, entinostat (MS‐275) and romidepsin (depsipeptide, FK228) are considered relatively class I‐specific, while tubacin is considered an HDAC6‐specific inhibitor (Table 3.1). These specificities are relative, as, in vivo, the concentration of these drugs achieved is typically orders of magnitude above the concentration showing to inhibit the enzyme in cell‐free systems. These supratherapeutic concentrations are likely to increase the spectrum of individual HDAC enzymes inhibited by any given HDAC inhibitor. Recent efforts in drug development have also focused on discretely targeting specific HDACs. HDAC6 for example, is predominantly [16, 17] but not exclusively [13] localized to the cytoplasm. Inhibition of HDAC6 is associated with very specific effects, including those on cell motility, the proteasome and aggresome pathways, which are considered by some investigators to be responsible for much of the cytotoxicity of the HDACi in general. This is one example of how different HDACi may vary in their HDAC selectivity, as the pan‐HDACi include HDAC6 among their targets, while the “class 1‐selective” HDACi (such as romidepsin) do not. Such differences may provide a rationale for the development of novel, highly HDAC‐specific agents.

Selective inhibition of HDAC3 has also emerged as a novel mechanism for therapy in B‐cell lymphomas that express the CREBBP mutation. CREBBP codes for the CREB binding protein which catalyzes histone acetylation. CREBBP mutations are highly recurrent in B‐cell lymphomas and either inactivate (i.e. loss of function) its histone acetyltransferase domain or truncate the protein. CREBBP plays a critical role in supporting p53 dependent tumor suppressor functions and is inhibited by viral basic leucine zipper transcription factor HBZ found in many lymphomas [18]. CREBBP mutations are direct targets of the BCL6/HDAC3 onco‐repressor complex and HDAC3 selective inhibitors reverse CREBBP mutant aberrant epigenetic programming. By restoring the function of these genes, HDAC3 inhibitors restore the ability of tumor infiltrating lymphocytes to kill diffuse large B‐cell lymphoma (DLBCL) cells in an MHC class I‐ and II‐dependent manner [19].

For now, it is more convenient to group the HDACi in commercial development into the pan‐HDACi and those that are relatively more class 1‐specific. It is important to recognize that while some HDACi may be more class specific, these drugs, as noted above, typically achieve high micromolar concentrations in the plasma, and likely inhibit a broader array of HDAC in the clinical setting.

Table 3.1 Classes of histone deacetylase inhibitor (HDACi).

HDACi class	HDAC specificity
HDAC Cellular Distribution	Nuclear	Nuclear, Cytoplasmic	Cytoplasmic	Nuclear
HDAC Class	I	IIa	IIb	IV
HDAC	1	2	3	8	4	5	7	9	6	10	11
Short chain fatty acids	Butyrate
	Valproate
Hydroxamic acid derivative	Tubacin
	Belinostat (PXD101)
	Panobinostat (LBH589)
	Trichostatin A
	Vorinostat (suberoylanilide hydroxamic acid, SAHA)
Benzamide	Mocetinostat (MGCD0103)
	Etinostat (MS‐275)
Cyclic tetrapeptide	Romidepsin (depsipeptide)

HDACi induce a plethora of molecular and extracellular effects that alone or in combination, result in potent anti‐cancer activities [1]. The drug targets, HDACs, are diverse and exhibit different substrate specificities and biological functions. Despite the rapid expansion of literature in this area, we remain uncertain of the mechanistic hierarchy of the HDACi. For example, while it is clear that HDACi induce apoptosis associated with altered transcription of proteins involved in the intrinsic and extrinsic pathways, other mechanisms are in play, such as those relating to the aggresome/proteasome system. Through hyperacetylation of histone and non‐histone targets, HDACi can induce quite diverse effects on cellular function. These include: (i) altering immune responses through effects on the host and/or target cells; (ii) inducing permanent (i.e. senescence) or temporary (quiescence) cell‐cycle arrest, usually at the G1/S transition; (iii) inhibiting angiogenesis, and (iv) inducing apoptosis and autophagy [20–24]. HDACi not only induce injury to the cell, they also modulate its ability to respond to stressful stimuli. Moreover, it is important to appreciate that the anti‐tumor effect of these drugs is due to targeting not only the tumor cell itself, but also the tumor microenvironment and the immune milieu.

The question is why do TCLs and other hematologic malignancies show a selection pressure for specific patterns of epigenetic modification? It is well recognized that epigenetic modulation is responsible for sustaining the adaptive transcriptional memory for both central and tissue resident memory T cells [25]. Specifically, this provides the means for different T helper subsets to transcribe inducible genes more rapidly [26].

A key clue that the TCLs represent an “epigenetic disease” came from the empiric finding that HDACi exhibit consistent and reproducible activity (e.g., a response rate of approximately 25%) across all TCL subtypes. While only one in four patients can expect to respond to HDACi, the duration of response seen across these drugs is impressive, typically more than one year. Preclinical studies have confirmed that this vulnerability to HDACi can be exploited in combination. Many studies have now established that the HDACi exhibit potent synergy with a host of other drugs known as “active” in TCL, including HMA, pralatrexate, proteasome inhibitors, and aurora A kinase inhibitors, and could serve as a cornerstone for future platform development [27–33].

Cutaneous TCL (CTCL) was the first malignancy for which HDACi were approved. CTCL is a disease where a panoply of epigenetic abnormalities might explain why HDAC inhibition is associated with clinical benefit in this disease [5]. While mutations in DNMT3A, IDH2, TET2, MLL2, KMT2A, KDM6A, CREBBP, and EP300 genes have been well recognized for years, it remains unclear whether any of these genetic factors portend differential response to HDAC inhibitors [3, 6–10].

DNMT3A functions as a DNA methyltransferase catalyzing cytosine methylation of CpG islands in promoters leading to transcriptional silencing. While mutations in DNMT3A have been identified in about 11–33% of patients with PTCL due to missense or nonsense mutations [34, 35], the mutations frequently coexist with mutated TET2, which may ultimately lead to transcriptional repression [36].

The TET2 gene encodes an alpha‐ketoglutarate dependent dioxygenase, which converts 5mC to 5‐hydroxymethylcytosine (5hmC), 5‐formylcytosine (5fC), and 5‐carboxylcytosine (5caC) [37–39]. Oxidation of 5mC is part of a demethylation pathway that influences transcriptional activation, where hypermethylation leads to silencing of gene expression, while hypomethylation leads to gene expression. The TET family of proteins are also known to be important in T‐cell differentiation, where loss of function mutations lead to TCL with follicular helper cell‐like features [40–42]. Interestingly, TET2 mutations are seen in up to 70% of PTCL patients. Specifically, it is estimated that TET mutation are found in 42–83% of patients with AITL, 28–48.5% of patients with PTCL not otherwise specified (PTCL‐NOS), and 10% of patients with adult T‐cell leukemia/lymphoma (ATLL) [34–36, 40–43]. Patients with PTCL‐NOS who carry the TET2 mutation often present with a follicular helper type of PTCL, a newly described entity that exhibits many similarities with AITL. Moreover, there can be interactions between the various epigenetic modulators. For example, wild‐type DNMTA3 binding occurs when there is less CpG density compared to TET1 binding, which occurs at a relatively higher CpG density. Both affect the polycomb repressive complex 2 (PCR2)‐mediated methylation of lysine 27 of histone H3 (H3K27), which leads to enrichment of trimethylation at H3K27 and ultimately influences gene expression [44].

Mutations in IDH2, especially at the R172 residue, have also been identified. Mutations in IDH2 alter the catalytic reactions of the Krebs cycle. Wild‐type IDH converts isocitrate to α‐ketoglutarate, a key co‐factor in the oxidative demethylase reactions which remove methyl groups from DNA. Mutant IDH2 converts isocitrate to 2‐hydroxyglutarate, which is an oncogenic metabolite that cannot function as an obligatory cofactor of TET catalytic functions [34, 45–51]. Mutations in IDH2 and TET2 reduce 5hmC levels due to global hypermethylation of promoters and 5′‐cytosine‐phosphate‐guanine‐3′ (CpG) islands (i.e. leading to transcriptional repression and gene silencing), likely contributing to peripheral T‐cell lymphomagenesis [49, 50]. While there are only modest data for the role of HMA in TCL, emerging data suggest that they do have marked single‐agent activity in AITL and appear to synergize with HDACi in both preclinical and clinical PTCL experiences [52].

The catalytic subunit of PCR2 is the enhancer of zeste homolog 2 (EZH2) and has been found to be mutated and overexpressed in several subtypes of PTCL. Mutations in EZH2 have been shown to produce transcriptional silencing, leading to lymphomagenesis; inhibition of EZH2 of PTCL lines has induced marked cell growth arrest via marked upregulation of genes involved in cell cycle [10, 51, 53] In primary cutaneous ALCL (C‐ALCL), EZH2 mutations repress antitumor immunity via suppression of the C‐X‐C motif chemokine ligand 10/receptor 3 axis, which is important in T‐cell migration in inflammatory conditions, including PTCL [53].

Another example of an epigenetic lesion that may contribute to PTCL pathogenesis is KMT2D (also known as MLL2) [3, 51]. KMT2D encodes histone H3K4 methyltransferase. Mutations in this gene have been seen in 42% of PTCL in general, with approximately 25 and 36% of patients with the AITL and PTCL‐NOS subtypes carrying the mutation [54]. Figure 3.1 shows common epigenetic mutations, targets, and the drugs that can modulate this biology (Figure 3.2).

In addition to mutations in genes directly effecting epigenetic processes, TCLs commonly have mutations in pathways that can directly or indirectly affect this biology. For example, a gene found to be commonly mutated in PTCL is RHOA. While RHOA is not thought to be involved in maintaining the PTCL epigenome, it does belong to the Rho family of small GTPases, a group of Ras‐like proteins involved in intracellular signaling. Gain of function mutations in RHOA are seen in 50–70% of AITL compared with 15% in ATLL (with its own unique distribution of mutations) and are associated with cell proliferation and invasiveness [8, 35, 43]. In patients with AITL, 70% of cases are associated with a specific mutation in RHOA G17V, which appears to be correlate with paraneoplastic autoimmunity and lymphopenia [8, 55]. Preclinical models have shown that mice with a TET2 mutation as the initiating event and RHOA G17V mutation as the second hit develop TCLs with histologic and immunophenotypic features of AITL [35, 55–57]. Murine lymphoma models have shown that RHOA G17V leads to a reduced threshold for T‐cell receptor activation and can augment the PI3K‐AKT‐mTOR signaling signature [56]. In these murine models, treatment with duvelisib, the PI3K inhibitor, can produce significant reductions in tumor burden, while treatment with everolimus improved survival. This mouse model confirms that these mutations are pathogenic and provide rationale for interrogation of drugs targeting the epigenome as a therapeutic target in PTCL.

Figure 3.1 Common epigenetic targets in peripheral T‐cell lymphoma. TET = ten eleven translocation protein, 2HG = 2‐hydroxyglutarate, IDH2 = isocitrate dehydrogenase 2, 5hmC = 5‐hydroxymethylcytosine, EED = embryonic ectoderm development, SUZ12 = suppressor of zeste 12, EZH2 = enhancer of zeste 2, PCR2 = polycomb repressive complex 2, K27me3 = trimethylation at lysine 27 of histone 3 (aka H3K27me3), HDAC = histone deacetylase.

In addition to the importance of epigenetic mutations in PTCL lymphomagenesis, and the therapeutic rationale, recent evidence suggests that surveillance of these mutations may be useful in the monitoring of TCL response to therapy. Cell‐free DNA (cfDNA) is increasingly being validated for the purpose of detection of minimal residual disease, as well as for the genetic profiling of a known malignancy. Mutations in genes with established roles in epigenetic regulation such as TET2, DNMT3A, and IDH2 have been found to be 83% concordant between cfDNA and tissue biopsy samples in patients with AITL, and may emerge as a potential marker for detecting minimal residual disease [58].