1. Clinical Significance of DNA Variants in Chronic Myeloid Neoplasms
Rebecca F. McClure, Mark D. Ewalt, Jennifer Crow, Robyn L. Temple-Smolkin, Mrudula Pullambhatla, Rachel Sargent, Annette S. Kim 
The Journal of Molecular Diagnostics 
Volume 20, Issue 6, Pages (November 2018)
DOI: /j.jmoldx
Copyright © 2018 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

2. Figure 1
Overview of cellular processes with components reported to have coding gene variants in CMNs. External growth factors and cytokines bind and activate their respective transmembrane receptors, triggering activation of pathways (predominantly composed of kinase cascade proteins), leading to activated key intermediaries, such as extracellular signal-regulated kinase (ERK), AKT1, TP53, and mTORC1 or mTORC2. These intermediaries interact with many targets located throughout the cell until signals reach effectors (eg, transcription factors, mitochondrial proteins, and structural and transport proteins) that ultimately facilitate the major cell functions of proliferation, differentiation, survival, or apoptosis. Signaling through JAK/STAT/ERK is most commonly used by the three main cytokine receptors known to be important for normal myeloid hematopoiesis: erythropoietin receptor (EPOR), predominantly for development of erythrocytes; thrombopoietin receptor (TPOR; gene called MPL), predominantly for development of megakaryocytes; and colony-stimulating factor 3 receptor (CSF3R), predominantly for development of neutrophils. RUNX1 is a master hematopoietic transcriptional regulator, involved in both the formation of hematopoietic stem cells and differentiation, the latter with fine-tuning assistance by CEBPA (Figure 2). Core binding factor β binds RUNX1 to enhance its DNA affinity but does not directly interact with the DNA. ETV6 is a strong transcriptional repressor, and it is also implicated in interactions with corepressors that recruit histone deacetylases, such as NCOR2. TP53 is activated by diverse forms of stress (cytokine deprivation, DNA damage, and others) and is the main proapoptotic proponent in the cell's continual balance of proapoptotic and antiapoptotic factors. TP53 induces the expression of PPM1D (data not shown), a phosphatase that negatively regulates the mitogen-activated protein kinase pathway and consequently p53-mediated transcription (Figure 3) and apoptosis. It initiates cell cycle arrest and activates proapoptotic members of the BCL2 family in the mitochondria through direct and indirect mechanisms, the latter through its role as a transcription factor, whereby it increases transcription of proapoptotic BCL2 family members and directly represses transcription of antiapoptotic family members. The cohesion complex (SMC1, SMC3, RAD21, and STAG) forms a ring-like structure that encircles DNA without directly binding it, providing sister chromatid cohesion during processes that require DNA looping, such as transcription, meiosis, mitosis, recombination, and ribosomal biogenesis (Supplemental Figure S1). Nucleophosmin shuttles proteins between the nucleus and cytoplasm, playing a key role in ribosomal biogenesis, cell cycle regulation (eg, maintaining the stability and function of p53), activation of cyclin-dependent kinases, and regulation of the mitotic spindle. Old or dysfunctional proteins may be disposed of through ubiquitin tagging (Supplemental Figure S2), followed by degradation, either in the proteasome or through autophagy. The protein substrates affected by the CBL family include a wide range of tyrosine kinases acting in signal transduction pathways, including KIT, FLT3, and JAK2. SET and its binding partner, SETBP1, play poorly elucidated roles in hematopoiesis, although both are known to have oncogenic properties when functionally abnormal. Similarly, the role of PHF6 (data not shown) is poorly understood but appears to act as a tumor suppressor, regulating transcription of signaling genes and rRNA. CALR is a multicompartmental protein [extracellular matrix, outer cell surface, cytosol, endoplasmic reticulum (ER), and nucleus] that regulates a wide array of cellular processes, including, among others, normal and abnormal protein movement through the ER, calcium signaling, and transcription modulation. However, variant CALR functions in a different manner than the wild type, traveling with TPOR from the ER to the cell surface and activating the JAK/STAT pathway. For detailed information on the alterations noted in this figure in myeloid neoplasia (Supplemental Tables S1, S2, and S5). Proteins in red are discussed extensively in the text and are either on the list of core genes or are included in other prognostic scoring systems.
The Journal of Molecular Diagnostics  , DOI: ( /j.jmoldx )
Copyright © 2018 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

3. Figure 2
Epigenetic modification of DNA and histones with an emphasis on components reported to have coding gene variants in CMNs. Epigenetics broadly encompasses processes that affect gene expression without altering the underlying DNA sequence and includes methylation of both DNA and histones. DNA methylation is mediated by DNA methyl transferases (DNMTs) and occurs at cytosine residues of CpG dinucleotides located within CG-rich regions of gene promoters. DNMT1 is responsible for maintaining patterns of methylation during DNA replication, whereas DNMT3A and DNMT3B are de novo methyltransferases. Both DNMT3A and DNMT3B are expressed in hematopoietic stem cells and involved in self-renewal and differentiation, although only DNMT3A is expressed in more mature elements. Cytosine methylation results in recruitment of transcriptional repressors and gene silencing. Cytosine methylation marks can be lost during DNA replication or actively removed by the TET family proteins, including TET2. TET enzyme activity is inhibited by 2-hydroxyglutarate (2-HG), which is generated by the variant form of isocitrate dehydrogenase (IDH), either from the cytosolic gene, IDH1, or the mitochondrial homolog, IDH2. Histone methylation alters chromatin structure, directly affecting accessibility to transcriptional activators and repressors and allowing recruitment of additional epigenetic regulators, such as DNMTs. The polycomb repressor complex 2 (PRC2) consists of four core members (JARID2, EED, SUZ12, and EZH1 or EZH2) and silences chromatin by methylating histone H3, with activity further augmented by ASXL1. BCOR is another polycomb repressor complex (composed of BCOR, BCORL1, and others), which represses transcription by ubiquitylating histone H2A. Opposing the action of the polycomb repressor complexes are a variety of histone deubiquitinases and histone demethylases, including KDM6A. For detailed information on the alterations noted in this figure in myeloid neoplasia (Supplemental Table S3).
The Journal of Molecular Diagnostics  , DOI: ( /j.jmoldx )
Copyright © 2018 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

4. Figure 3
Spliceosome mechanisms with an emphasis on components reported to have coding gene variants in CMNs. For genes containing more than one exon (>90% of all genes), excision of introns through splicing of mRNA is required before translation. The primary machinery required for this process is a complex of five small nuclear RNAs (snRNAs) coupled with >150 small nuclear ribonucleic proteins (snRNPs), collectively termed the spliceosome.6 Two basic types of splice site sequences are known that recruit slightly different versions of spliceosome: The major splice sequence (used for approximately 99% of splices) recruits the major spliceosome (composed of snRNPs U1, U2, U4, U5, and U6). Intronic material to be excised is bookended by splice sites at their 5′ and 3′ ends (5′ss and 3′ss, respectively). The 5′ss is characterized by a GU sequence, whereas the 3′ss is characterized by AG. Between these sites are two additional key recognition sequences, a conserved adenosine, the branch point site (BPS) located between the two ends; and a polypyrimidine tract (PPT), located just 5′ of the 3′ss. The initial step in the splicing process is the recognition of the 5′ss by U1 snRNP. Simultaneously, there is recognition of the PPT by the serine-arginine (SR)–rich splicing factors SRSF1 or SFSR2 as well as the U2 auxiliary factor, U2AF2 (formerly U2AF65). In addition, U2AF1 (formerly U2AF35) recognizes the AG of the 3′ss along with ZRSR2, an SR factor with zinc finger activity. Together, these form the early (E) complex. The U2 auxiliary factors guide the U2 snRNP to the BPS to form complex A. Binding of the U2 snRNA to the BPS, regulated by the SF3a (subunits 1 to 3) and SF3b (subunits 1 to 6) complexes, generates a bulge by a single base noncomplementarity that exposes the 2′-hydroxyl of the conserved adenosine for the transesterification to the 5′-phosphate of the guanosine adjacent to the end of the 5′ exon. The complex U4/U5/U6 tri-snRNP is then recruited through interaction of U5 to the 3′ss, resulting in an initial B complex, and ultimately forming B* complex, the active spliceosome, with loss of U1 and U4. After the first catalytic step, the entirety is termed the C complex. The final transesterification results in fusion of the two exons and release of the intron in a lariat configuration. For detailed information on the alterations noted in this figure in myeloid neoplasia (Supplemental Table S4).
The Journal of Molecular Diagnostics  , DOI: ( /j.jmoldx )
Copyright © 2018 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

5. Figure 4
Heat map of incidence and prognostic significance of gene variants by CMN entity. Commonly mutated genes in CMNs are listed with their overall incidence as the number in each box. The color coding represents a synthesized, overall view of the clinical value of variants in each gene. To generate the heat map, key prognostic matrix points (progression, including fibrosis and transformation to acute leukemia; leukemia-free survival; and overall survival) were given a score from 1 to 3 as follows: Unclear, no convincing reports OR contradictory reports OR fewer than five patients reported (labeled as unclear prognostically); 1, two or fewer articles OR multiple contradictory reports OR <25 patients reported; 2, two or more articles AND no contradictory reports AND ≥25 patients reported; and 3, four or more articles AND no contradictory reports AND ≥100 patients reported. The articles used to generate the score could not all be referenced in this article because of space limitations. Asterisks indicate studies that reviewed IDH1 and IDH2 mutations together and their conclusions are listed herein; daggers, percentages for chronic myelomonocytic leukemia (CMML) and atypical chronic myelogenous leukemia (aCML) may reflect the inclusion of cases of chronic neutrophilic leukemia (CNL), whereas other references have suggested that CSF3R may be highly sensitive and specific for CNL; double daggers, ranges for JAK2 incidence in polycythemia vera (PV) reflect the incidence of the JAK2 p.V617F mutation only and reflect many earlier studies that contained lower frequencies of JAK2 mutations. In addition, the prognostic designation for JAK2 in PV, essential thrombocythemia (ET), and primary myelofibrosis (PMF) is related largely to high variant allele fraction JAK2 mutations, not the presence of a JAK2 mutation itself; section mark, the incidence of KIT mutations in systemic mastocytosis (SM) reflects the incidence in mast cells and not necessarily the incidence in all marrow compartments, particularly in SM with an associated hematopoietic neoplasm; paragraph symbol, SF3B1 mutations are found in 10% to 33% of cases of myelodysplastic syndrome (MDS) overall but are present in 50% to 83% of cases of MDS with ring sideroblasts (RSs); double vertical symbol, PPM1D prognosis (shaded for MDS only) and frequency are only related to the high incidence of PPM1D mutations found in clonal hematopoiesis of indeterminate potential that are associated with therapy-related myeloid neoplasms, specifically, after autologous stem cell transplantation. MDS/MPN-RS-T, MDS/myeloproliferative neoplasm with RSs and thrombocytosis.
The Journal of Molecular Diagnostics  , DOI: ( /j.jmoldx )
Copyright © 2018 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

6. Figure 5
Schematic of common changes acquired in the differentiation and progression of CMNs. Although many and varied passenger variants are evident at the time of development of CMNs, the founder mutations in CMNs are enriched for variants in genes involved with epigenetic regulation and the spliceosome. TET2 and ASXL1 are particularly prevalent in the former category. In the latter category, the four main spliceosome genes (SRSF2, SF3B1, U2AF1, and ZRSR2) are commonly recurrent in myelodysplastic syndrome (MDS), whereas only SRSF2 is particularly prevalent in all CMNs and SF3B1 in entities involving ring sideroblasts (RSs). Although this base appears to characterize MDS, in other CMNs, additional secondary variants are recurrently identified and confer the disease-defining features of those entities, such as CSF3R in chronic neutrophilic leukemia (CNL) or JAK2 and CALR in the non– chronic myelogenous leukemia (CML) myeloproliferative neoplasms (MPNs). Finally, during disease progression, many CMNs acquire additional variants in TP53 or SETBP1. Specifically, in MDS, additional variants in signal transduction pathway members (eg, NRAS, KRAS, FLT3, and KIT), myeloid transcription regulators (eg, RUNX1 or ETV6), or additional epigenetic regulators (eg, IDH1/2, DNMT3A, and EZH2) are common findings in more high-risk disease or secondary acute myeloid leukemia. In the non-CML MPNs, biallelic or loss of heterozygosity of JAK2 leads to high variant allele fractions (VAFs) in association with progressive polycythemia and fibrosis. aCML, atypical CML; ALL, acute lymphoblastic leukemia; CMML, chronic myelomonocytic leukemia; MDS-RS-T, MDS-RS-thrombocytosis; n, progression mutation(s); SM, systemic mastocytosis; x, variable number of passenger mutations; y, founder mutation(s); z, secondary mutation(s).
The Journal of Molecular Diagnostics  , DOI: ( /j.jmoldx )
Copyright © 2018 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

7. Supplemental Figure S1
Cohesin complex details. Cohesin is a ring-like complex of four subunits that encircles DNA without directly binding it, providing sister chromatid cohesion. Cohesin functions during meiosis, mitosis, and other processes that require DNA looping, including transcriptional regulation, locus rearrangement by recombination, and ribosomal biogenesis.7,8 The subunits consist of two structural maintenance of chromosome (SMC) proteins, SMC1 and SMC3, and sister chromatid cohesin proteins, RAD21 and stromal antigen STAG. During mitosis, the four subunits adhere initially to a single chromatid in G1 at centromeres and telomeres, mediated by the loading factors NIPBL and MAU2. The PDS5-WAPL complex then facilitates the unloading of the cohesin from the sister chromatids, allowing for segregation during anaphase. For detailed information on the alterations noted in this figure in myeloid neoplasia, see Supplemental Table S5.
The Journal of Molecular Diagnostics  , DOI: ( /j.jmoldx )
Copyright © 2018 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions

8. Supplemental Figure S2
Protein ubiquitination details. Ubiquitination is the common pathway for regulated proteolysis and is mediated by three enzymes, E1 to E3. E1 activates the 76–amino acid ubiquitin protein through initial adenylation, followed by thioester covalent attachment of the carboxy terminus of ubiquitin to a cysteine residue of E1. Transthioesterification occurs to shuttle ubiquitin to a cysteine on E2, the ubiquitin-conjugating enzyme, followed by transfer of the ubiquitin moiety to a lysine on the substrate. This final transfer from E2 to substrate is mediated by the E3 ligases. There are >600 E3 ligases known in humans, falling within three broad categories of the really interesting new gene (RING) E3s, homologous to the E6AP carboxy terminus E3s, and RING between RING E3s.9 Within the RING family is a small family of casitas B-lineage lymphoma (Cbl) ubiquitin E3 ligases, composed of just three members, c-Cbl, Cbl-b, and Cbl-c. The protein substrates affected by the Cbl family include a wide range of tyrosine kinases acting in signal transduction pathways, including epidermal growth factor receptor, c-KIT, FLT3, GRB2, and JAK2. For detailed information on the alterations noted in this figure in myeloid neoplasia, see Supplemental Table S1. NEDD, neural–precursor-cell-expressed developmentally down-regulated 8; PPi, pyrophosphate; Ub, ubiquitin.
The Journal of Molecular Diagnostics  , DOI: ( /j.jmoldx )
Copyright © 2018 American Society for Investigative Pathology and the Association for Molecular Pathology Terms and Conditions