EGR1

12.5.1 EGRs

Early growth response genes (EGR1-4) are the prototypical examples of IEGs, as their expression is increased by acute and ascending doses of cocaine (Piechota et al., 2010). Induction of EGR1 expression by acute cocaine is mediated by D1Rs (Drago, Gerfen, Westphal, & Steiner, 1996), while chronic cocaine treatment reduces EGR1 expression in CPu, NAc shell, and prelimbic cortex (Ennulat, Babb, & Cohen, 1994), and recovers again during withdrawal (Hammer & Cooke, 1996). These changes could serve as a biochemical basis to induce drug craving. EGR3 is part of a feedforward loop that enhances the rewarding properties of cocaine, as it triggers the upregulation of CREB and FOSB, and also induces the expression of CAMK2A, NR4A2, and SIRT1 in D1 MSNs in the NAc. In contrast, the association of EGR3 mRNA with ribosomes is reduced in D2 neurons (Chandra et al., 2015). EGR3 also interacts with the histone methylation enzyme G9a (EHMT2) and the DNA methylation enzyme DNMT3A (Chandra et al., 2015).

4 Human Variants in Genes for Experimental Myopia

Reduced or increased expression of many genes has been implicated in experimental myopia in the chicken, rat, and monkey.^86,87 Of them, alteration of ZENK expression (human EGR1 OMIM 128990) consistently has been shown to be involved in ocular growth and refraction. Upregulation of ZENK in retinal glucagon amacrine cells is assumed to create a STOP signal to inhibit axial eye growth, whereas downregulation of ZENK is associated with axial eye growth in animals.^88–90 Egr1 knockout mice had longer eyes and a relative myopic shift in refraction.⁹¹ However, there is no evidence showing association of human myopia with ERG1 variants. Similarly, whether most genes implicated in experimental myopia in animal studies have a role in human myopia are still uncertain.

73.2.5 Hormones

2.6.1 Epigenetic Programming of the GR by Early Life Experiences

Among the very first candidates implicated in stress-related epigenetic regulation was the GR due to its critical role in restraining HPA-axis activity. Human NR3C1 comprises multiple first exons among which four exons are located in the distal promoter region (A_1–3 and I) and ten (D, J, E, B, F, G, C_1–3, and H) in the proximal composite promoter region [53]. The distal and proximal promoter regions locate 30 and 5 kb upstream of exon 2 that encodes the start codon. A CGI consisting of a high frequency of CpG residues straddles the proximal promoter and most alternative promoters are expressed in a tissue-specific and partly species-specific manner. The proximal NR3C1 promoter is highly conserved between humans, rats, and mice and thus allows for comparing epigenetic marking in response to different environmental exposures across species [53].

As early as 2004, a seminal study in rats evidenced epigenetic programming of Nr3c1 in response to differences in the quality of postnatal maternal care [54]. Pups reared by high-care taking mothers—defined by frequent pup licking and grooming—showed an enhanced turnover of serotonin in hippocampal tissues, in which GR expression was site-specifically increased. Once bound to G-protein coupled receptors, serotonin triggers intracellular cAMP production that in turn induces expression of the transcription factor EGR1 (early growth response 1, also called NGFI-A). EGR1 recognizes a DNA element at Nr3c1 exon 1.7 (rat homologue to 1F in human) and facilitates the recruitment of general coactivators harboring histone acetyltransferase activity. This event promotes chromatin opening and Nr3c1 transcription in response to high quality maternal care. Importantly, the same chain of events also initiates lasting demethylation at the exonic EGR1 DNA-binding site and thus leaves a molecular memory trace of the early life experience. This maternal-care dependent hypomethylation directs long-term increases in hippocampal GR expression and contributes to an improved negative feedback regulation and associated behavior [54]. However, epigenetic programming of Nr3c1 can be reversed by timely pharmacological or behavioral interventions and thus illustrates the dynamic nature of experience-dependent methylation marks [55].

These findings from rats were subsequently translated to humans by investigating the NR3C1 methylation status of adults with a history of early child abuse who died by suicide [56]. Hypermethylation of exon 1F was detected in the hippocampus of suicide completers with a history of child maltreatment and correlated with lower GR expression. In contrast, suicide completers without a history of childhood maltreatment showed neither a change in hippocampal NR3C1 exon 1F methylation or in gene expression. Functionally, hypermethylation impaired EGR1 binding and subsequent transactivation of NR3C1. Together, these findings indicate that ELA can confer site-specific epigenetic programming of NR3C1and disrupt negative feedback regulation of the HPA axis in rodents and humans. Although these initial studies suggested that early life stress (ELS) targets preferentially exon 1F, this view has to be revised in light of later work that revealed epigenetic programming of multiple proximal promoter regions. In this context, expression of the noncoding exons 1B, 1C, and 1H, reported significant decrease in the hippocampus of suicide completers with a history of childhood abuse compared to nonabused suicides and controls [57]. This study also showed that NR3C1 1C methylation correlated inversely with GR expression in agreement with previous results on 1F in rats. What is the mechanism that coordinates lasting changes in multiple GR transcripts in response to ELS?

Part of the answer has been gained from a recent study in mice that used postnatal MS as a well-established model of ELS [58]. Separated pups develop sustained HPA-axis hyperactivity and behavioral impairments in conjunction with hypomethylation of hypothalamic Avp and pituitary Pomc (see subsequent section) [59,60]. Notably, GR expression was unaltered in mice hippocampus and pituitary, which represent major sites of negative feedback regulation. In contrast, multiple exon 1 transcripts were lastingly upregulated in the adult PVN of maternally separated mice and caused a net increase in total GR transcripts. This event translated into higher GR protein expression and an enhanced regulation of downstream glucocorticoid target genes. Interestingly, enhanced hypothalamic Nr3c1 expression correlated with hypermethylation of the proximal CGI shore region. Functional analyses showed that the shore region encodes an insulator function thought to shield the proximal promoter from upstream regulatory influences. The shore region also contained a methylation-sensitive DNA-binding site for the transcriptional regulator YY1, which enforced insulator function upon binding. As a result, ELS dynamically regulated YY1 binding at the CGI shore region, associated insulator function, and transcription of multiple proximal GR transcripts (Fig. 2.3).

Figure 2.3. Model of Nr3c1 Transcript Regulation by ELS in the PVN

The proximal Nr3c1 promoter contains a DNA-binding site for the transcriptional repressor YY1 at the CGI-shore region. CGIs spanning the proximal Nr3c1 promoter are depicted above the different nontranslated exons boxed in grey and numbered according to Bockmühl et al. [53]. In control mice (Ctrl), YY1 binding occurs in the absence of DNA methylation (open lollipop) and confers transcriptional repression of multiple GR-transcripts within the proximal promoter region. In contrast, increased methylation (black lollipop) at the YY1 DNA-binding site in response to ELS prevents YY1 binding and relieves transcriptional repression. As a result, multiple GR transcripts are upregulated and translated into GR protein.

Interestingly, ELS-induced Nr3c1 upregulation was detected solely in parvocellular Crh-, but not in Avp-expressing neurons. Moreover, Crh expression was unaffected under resting conditions. In contrast, exposure to chronic mild stress led to a robust up-regulation of Crh in control mice but not in mice with a history of ELS (Fig. 2.4). Taken together, this work illustrates that the effects of cell type-specific epigenetic programming of Nr3c1 can stay masked unless challenged by reexposure to stress. If so, ELS may actually protect against later stress. Prompted by these studies several groups have analyzed peripheral blood mononuclear cells (PBMC) from different populations of individuals who were exposed to varying forms of ELA. Although clinical phenotypes of the respondents, measures of ELA, and methylation analysis varied among these investigations, most of them agree on altered NR3C1 methylation patterns in peripheral tissues raising the prospect of a potential biomarker for individuals exposed to ELA [61]. Overall, findings on epigenetic programming of NR3C1 suggest that the effects depend on the quality of the initial stressor [maternal care vs. maternal separation (MS)], and are cell type-specific (hippocampal vs. PVN neurons), and context-dependent (naive vs. reexposed conditions).

Figure 2.4. Cell Type-Specific, Epigenetic Programming of Hypothalamic Nr3c1

Mouse neurons in the PVN can express either Crh and Avp or solely Crh. Exposure to ELS leads to epigenetic programming and sustained upregulation of Avp when compared to controls (Co) (see also paragraph on Avp). Additionally, ELS induces epigenetic programming and upregulation of GR in Crh-expressing neurons. In adult mice, application of chronic stress results in upregulation of Crh in control mice, but not in mice with a history of ELS. As a result, ELS partly protects against chronic stress in adulthood as evidenced by a faster return to baseline glucocorticoid levels and less behavioral impairments.

3.1 Cancer Metabolism

Cancer cells reconstruct their metabolism to accommodate the drastic phenotypical change during malignant transformation. The common aim of an altered metabolism in cancer is to acquire necessary nutrients from frequently nutrient-deficient tumor environment. The altered metabolism can either sustain or promote growth, survival, dissemination, and long term maintenance of the cancer cells. One of the common features of the cancer metabolism is the increased uptake of glucose, and metabolization of the glucose through aerobic glycolysis in which glucose is fermentated to lactate. This feature is termed as the Warburg effect. This general effect is brought out by aberrant expression of metabolic genes. A study showed that lncRNA H19 was upregulated by EGR1 in liver cancer, thereby leading to the induction of pyruvate kinase M2 in liver cancer cells. Pyruvate kinase M2 is essential for the Warburg effect (Li et al., 2015a). In bladder cancer, increased lncRNA UCA1 level activated mTOR signaling to activate hexokinase 2 which functioned as a mediator of glycolysis (Li et al., 2014b). LncRNA lincRNA-p21 could modulate hypoxia-induced Warburg effect. LincRNA-p21 dissociated Hypoxia inducible factor (HIF)-α protein from VHL protein, and attenuated ubiquitination of HIF-α. The accumulation of HIF-α protein then promoted glycolysis under hypoxia (Yang et al., 2014a). Another lncRNA that regulated Warburg effect is named Colorectal neoplasia differentially expressed (CRNDE) that is activated in early CRC. CRNDE was downregulated during the treatment of insulin and insulin-like growth factors. It was shown that CRNDE regulated gene targets that participated in glucose and lipid metabolism, which could promote the change of metabolism of CRC cells to aerobic glycolysis (Ellis et al., 2014). In ovarian cancer, lncRNA ceruloplasmin (NRCP) was upregulated in tumor tissue which promoted tumor progression. It was reported that NRCP also regulated cancer metabolism via increasing cancer cell glycolysis (Rupaimoole et al., 2015).

Wilms’ Tumor 1

The transcription factor Wilms Tumor 1 (WT1) plays a pivotal role in early development of the primordial gonad and the bipotential internal genital duct systems prior to sex determination; one WT1 isoform appears to be required for specific development of the testis from the bipotential gonad. In renal tissue, WT1 acts as a tumor suppressor.^40,41 WT1 is a member of the early growth response (EGR) family of transcription factors (proteins expressed early in the cell cycle, at G0 to G1 transition) and acts as a transcriptional activator or repressor, depending on the cellular or chromosomal context.^42-45 There appears to be a general requirement for WT1 in the formation of organs derived from the intermediate mesoderm, particularly the differentiation of glomerular epithelial cells and gonadal primordium.⁴⁶ WT1 is implicated in gonadal and genital development by analysis of human mutations and transgenic mice with WT1 deletions or mutations.^40,47

The human WT1 gene is a complex locus at 11p13 that in fact consists of two genes, WT1 and WIT1, expressed from opposite DNA strands.^48,49 The function of the WIT1 transcript is unknown, but a role as an antisense regulator of WT1 has been postulated. The highly conserved WT1 gene spans 50 kb and contains 10 exons that can be alternatively spliced to yield four distinct mRNA species of approximately 3 to 3.5 kb each. The primary WT1 protein is a 429–amino-acid (∼50 kD) transcription factor with 4 contiguous Cys2-His2 zinc finger domains (encoded by exons 7 to 10) and an amino terminus rich in proline and glutamine, typical of certain transcription factors (Fig. 118-2, A). Zinc fingers 2, 3, and 4 of WT1 have more than 60% amino acid identity with the zinc fingers of the EGR1 transcription factor. There are separate domains for transcriptional repression (amino acids 85 to 124) and activation (amino acids 181 to 250). These regions are distinct from the DNA-binding domain, and their activities are probably mediated by protein-protein interaction. The four mRNA species encode four major proteins, WT1 A to D, differing mainly on the basis of the presence or absence of 17 amino acids in the central region of the protein (in placental mammals) and a lysine/threonine/serine triad (abbreviated by the single-letter code for amino acids as KTS) between the third and fourth zinc fingers (see Fig. 118-2, A). Perhaps as many as 32 different isoforms of WT1 result from additional variations in translational start site, either 5′ or 3′ of the main translation initiator.^40,41,47 Two of the four major WT1 isoforms contain the KTS sequence (designated +KTS), and two do not (designated –KTS). The KTS amino acid triad alters the spacing between the third and fourth zinc fingers, thereby changing the DNA-binding specificity, likely preventing its binding to the typical EGR1-like DNA binding sequence. The –KTS and +KTS isoforms also have differential expression patterns within cell nuclei and appear to have distinct but somewhat overlapping roles. The fact that all transcripts are expressed at similar levels suggests that each encoded protein makes a significant contribution to WT1 function, and interactions between the proteins, each of which may have distinct targets and functions, may be important in the control of cellular proliferation and differentiation exerted by WT1. The smaller –KTS isoform has greater transcriptional activation potential than the +KTS isoform.

Various forms of WT1 regulate SRY, DAX1, SF1, and AMH expression (Fig. 118-2, B).^50-53 WT1 –KTS isoforms associate and synergize with SF1 to promote AMH expression; WT1 –KTS can also activate the DAX1 promoter. However, DAX1 antagonizes the synergy between SF1 and WT1, most likely through a direct protein-protein interaction with SF1, suggesting that WT1 and DAX1 functionally oppose each other in testis development by modulating SF1-mediated transactivation. The fact that WT1 can upregulate expression of DAX1, which in turn antagonizes WT1/SF1-mediated stimulation of AMH expression, suggests that the relative dosages of WT1 –KTS and DAX1 and the timing of their expression during embryogenesis are vital in the delicate balance of transcription factor activity required for gonadal development. The –KTS isoforms of WT1 also upregulate SRY gene expression through the EGR1-like sequence in the core promoter.

The expression of WT1 in human fetal development occurs during the period between days 28 to 70 of gestation.⁵⁴ During this time, WT1 is expressed mainly in mesodermally derived tissues—kidneys, gonads, and mesothelium—but is also expressed in spinal cord and brain—tissues of ectodermal origin. In midtrimester human embryos, there is strong expression in kidneys and gonads. WT1 expression is limited to Sertoli cells in adult testes. This expression pattern is also observed in mouse, corresponding to embryonic days 10.5 to 15, and upregulation occurs prior to the expression of nuclear transcription factors NR5A1 and NR0B1.^54-56

Human mutations highlight the developmental importance of WT1. Denys-Drash syndrome is associated with heterozygous germ line mutations in WT1 (mainly in the zinc finger encoding regions) in more than 90% of cases. The disorder is genetically dominant; no patients have been described with mutations in both alleles of the gene. Complete deletion of WT1 produces milder genital variations (cryptorchidism and/or hypospadias in 46,XY individuals) than does a mutation that encodes expression of an abnormal WT1 protein (46,XY sex reversal with streak gonads), suggesting a dominant-negative mechanism of action of mutant WT1 proteins, perhaps due to abnormal DNA binding.⁵⁷ Alterations in gonadal development have also been reported in 46,XX individuals with WT1 mutations, underscoring the role of WT1 in both male and female gonadal development. Detailed description of these mutations is beyond the scope of this chapter but can be found in a comprehensive review.⁵⁸

It seems likely that the various forms of WT1 may function differentially in the different genetic contexts of testicular versus ovarian development. Consistent with this concept, WT1 expression is differentially regulated during development, depending on the sexual differentiation of the gonad. Mice homozygous for knockout of the entire WT1 gene had failure of renal, gonadal, and adrenal development.^46,59,60 Failure of development was observed at embryonic day 11 of gestation secondary to apoptosis of cells in the metanephric blastema resulting in failure of the growth of the ureteric bud metanephric kidney. Homozygous knockouts of WT1 are nonviable, probably due to abnormal development of the mesothelium, heart, and lungs.

The developmental phenotype of mice with WT1 gene knockouts varies depending on the specific isoform of the transcription factor targeted⁵³ and the magnitude of loss of expression (i.e., heterozygous versus homozygous deletion). A single wild-type allele is adequate for urogenital normal development. In contrast, mice completely lacking the WT1 –KTS isoform had tiny streak gonads in both males and females, associated with reduced Dax1 expression. The mice also had abnormal development of the internal genital ducts and severely impaired renal development. Male mice homozygous for deletion of the +KTS isoforms (which retain normal levels of the –KTS isoforms) showed complete XY sex reversal, their embryonic gonads having the morphologic appearance of ovaries, associated with a dramatic reduction of gonadal Sry and Sox9 expression and female-type Dax1 expression. These data demonstrate distinct functions for the WT1 +/– KTS isoforms and place the Wt1 +KTS variants as likely regulators of Sry in the sex determination pathway.

6.1 Cellular Reprogramming

During carcinogenesis certain apoptotic and differentiation pathways are epigenetically inactivated in order to assure tumor cell survival, being many of these pathways also required for the antineoplastic effect of most chemotherapeutic agents. By reversing epigenetic changes, epigenetic therapy can result in cell reprogramming that, in turn, may trigger cell differentiation, cell death, and/or senescence among other possible phenotypes. Cellular reprogramming may also restate pathways that lead to increased chemotherapy sensitivity.

For example, it has been reported that sequential treatment of drug-resistant breast cancer cells first with decitabine followed by doxorubicin reverses this resistance in more than 80% of treated cells [24]. Decitabine pretreatment induces a depletion in the DNMT1 protein levels that relieves transcriptional repression of p21 (which is responsible for cell cycle regulation), thereby increasing its expression [24,25]. This DNMT1 reduction plays also an indirect role in p21 induction via reexpression of methylation inactivated transcription factors such as EGR1, SMAD3, and HES6, which interact with p21 promoter and increase its expression in tumor cells [25–27]. Induction of p21 causes cells to undergo G2/M arrest that might result in accumulation of Topoisomerase II. Since stabilization of Topoisomerase II is also an effect of doxorubicin, this pathway reactivation results in increasing doxorubicin tumor cytotoxic effects [28,29].

It is worth nothing that reactivation of tumor cell-specific pathways may be more therapeutically crucial than reaching global epigenetic changes (i.e., increase in histone acetylation or DNA hypomethylation). In DLBCL, this is the case for the transforming growth factor β (TGFβ) pathway, specifically for its intracellular transducer SMAD1. TGFβ belongs to the superfamily of growth factors, involves in the control of cell growth, proliferation, differentiation, apoptosis, and homeostasis of normal B cells. These functions are mediated through proteins from the SMAD family that transduce the extracellular signals from the TGFβ ligands to the nucleus where they activate target genes [30]. We recently found that SMAD1 is epigenetically silenced by DNA hypermethylation in malignant B cells, allowing these cells to sustain the oncogenic phenotype. Moreover, we also demonstrated that DNA methylation-mediated silencing of SMAD1 results in chemoresistance in DLBCL. Treatment of DLBCL patients with the DNMTi azacitdine hypomethylates and reactivates SMAD1 expression. This makes the TGFβ pathway to become functional again, similarly to what is seen in B cells. Lymphoma cells exhibiting this gained functionality are now responsive to growth inhibitory signals through the TGFβ pathway, and more importantly, to chemotherapeutic drugs [23] (Figure 28.2). Therefore, cellular epigenetic reprogramming restores a “normal” tumor suppressor pathway in lymphoma cells that causes higher vulnerability to chemotherapy agents. Although speculative, patients with genetic lesions affecting TGFβ and/or SMAD1 could less likely benefit from this approach and will not be selected for such therapy.

Figure 28.2. A depiction of the model proposed for the mechanism-based epigenetic chemosensitization of aggressive lymphomas upon low-dose DNMT inhibitor treatment. SWING: Senescence with incomplete growth arrest.

Activation of AMPK by PGF2α

Early studies established that PGF_2α binds to and activates its cognate G_q protein-coupled receptor, PTGFR. This leads to the rapid activation of phospholipase C, which leads to increases in both cytoplasmic Ca^2 + and activation of protein kinase C. These initial events contribute to the activation of additional protein kinase cascades like the mitogen-activated protein kinases (ERK1/2, p38, and JNK) [88], which contribute to the induction of early responses genes like FOS, JUN, EGR1, and ATF3 [11]. While these early response genes have been implicated in the luteolytic response to PGF2α, it is not clear how or whether they impact steroidogenic events in luteal cells. The developmental-specific expression of PKC and CAMKK2 isoforms, proteins involved in Ca^2 + homeostasis, and AMPK have been implicated in the cellular mechanisms of acquisition of luteolytic capacity by bovine corpus luteum [83,84]. Based on these observations, it seems reasonable to predict that PGF2α could activate at Ca^2 +/CAMKK2 pathways leading to the activation and phosphorylation of AMPK on Thr172. Bowdridge et al. [78] recently reported that PGF2α rapidly (2 min) and transiently stimulated the phosphorylation of AMPK on the Ser485 site in dispersed bovine luteal cells. The response was prevented by treatment with STO-609, a CAMKK2 inhibitor. Treatment with STO-609 also prevented the inhibitory effect of PGF2α on progesterone synthesis in overnight incubations of dispersed luteal cells. In recent studies using bovine large luteal cells, we have observed that PGF2α rapidly stimulates AMPK on the stimulatory Thr172 residue and the inhibitory Ser485 residue (Hou, Zhang, Talbott, and Davis, unpublished). The observation that PGF2α can target multiple sites on AMPK is consistent with findings that PGF2α can activate multiple pathways in luteal cells, pathways linked to calcium signaling and pathways linked to PKC/MAPK/MTOR signaling [89]. While additional studies are needed to determine exactly how PGF2α regulates AMPK in luteal cells, it seems clear that activation of AMPK with pharmacologic tools disrupts progesterone synthesis. Studies are needed to determine whether AMPK is activated in vivo during natural and PGF2α-induced luteolysis. The reduction in luteal blood flow coupled with increase in hypoxia and inflammatory mediators during luteal regression may elevate AMPK activity resulting in decreased HSL activity and reduced mobilization of cholesterol from luteal lipid droplets.

Reactivation and Replication

Although EBV in cancer cells is found mostly in latency, a small number of lytically-infected cells promote carcinogenesis through the release of growth factors and oncogenic cytokines. Reactivation of the virus from latency, which occurs spontaneously at low levels, can be enhanced by chemical or biological inducers, such as phorbol esters, calcium ionophore, inhibitors of histone deacetylases or DNA methylases, or by cross-linking immunoglobulin. Expression of two immediate early virus transactivators, Zta, the product of the BZLF1 open reading frame, and Rta, the BRLF1 gene product, initiates the productive replication cycle. This then proceeds, as does all herpesvirus replication, with synthesis of early proteins, which are primarily replication enzymes including a DNA polymerase, replication of the DNA by a rolling circle mechanism and synthesis of late structural proteins.

EBV reactivation is also modulated by host factors. Cells change the ability of Zta and Rta to trigger EBV reactivation through posttranslational mechanisms. Phosphorylation is the most common modification by which modulating the transcriptional potential of transcription factors regardless of whether they are encoded by the host cell or the virus (Kenney and Mertz, 2014). Accumulating data indicate PKC, MAPK, and PI3-K signaling pathways are involved in BCR induction of the EBV lytic cycle. The ataxia telangiectasia mutated (ATM) activation that occurs in response to DNA damage or oxidative stress which are initial effects of carcinogenic factors, has been shown to induce EBV reactivation through a p53-dependent mechanism (Goswami et al., 2012). P53 may support EBV reactivation through direct interactions with transcription factor Sp1 protein bound to Zta promoter (Zp), as well as increase in EGR1 (Kenney and Mertz, 2014). The cellular transcription factor B-lymphocyte-induced maturation protein 1 (BLIMP1, also known as PRDI-BF1 and PRDM1), a key regulator in both epithelial and B-cell differentiation, induces reactivation of the virus out of latency into lytic replication in a variety of cancerous epithelial cell types and some B cell types. BLIMP1 cooperatively with KLF4, both are differentiation-dependent cellular transcription factors, induce the expression of LMP1. The LMP1 expression is in vitro required for efficient lytic reactivation in epithelial cells by enhancing expression of the Zta and Rta proteins (Reusch et al., 2015; Nawandar et al., 2017).

Capsids are assembled in the nucleus and packaged with unit length DNA cleaved from concatamers by an ATP-driven terminase-packaging complex. The mature capsids leave the nucleus by budding through the inner nuclear membrane into the perinuclear space, which is contiguous with the endoplasmic reticulum, acquiring a primary envelope in the process. Enveloped capsids then fuse out of the perinuclear space into the cytoplasm, losing the primary envelope. Tegument proteins are added to the capsid and interaction of tegument proteins with the cytoplasmic tails of glycoproteins clustered in the membrane of a late golgi compartment membrane or late endosomal membrane is thought to position the tegumented virus beneath membranes rich in virus glycoproteins into which it buds. As a result, budding virus now reenters the secretory pathway and is released from the cell by exocytosis. The cell eventually dies. Some of these details have been worked out for EBV but much of the information on later stages of assembly and egress is based on analogy with other herpesviruses (Baines, 2011), particularly the α-herpesviruses, which replicated more synchronously and are easier to study (Johnson and Baines, 2011).

Some of the proteins expressed in this lytic cycle have functions that are not essential to replication itself, but can influence the immediate host environment. These include the BCRF1 protein, a virus interleukin 10 homolog, the BARF1 protein, a soluble receptor for human colony stimulating factor 1, the BNRF2a protein, which blocks the transporter of antigenic peptides required for antigen presentation by MHC class I and the BILF1 protein which binds to MHC class I molecules and alters their trafficking, impeding export of new complexes and enhancing endocytosis and degradation of existing ones. Zta targets the MHC class II antigen presentation pathway by downregulating the class II transactivating protein CIITA, the BGLF5 protein has a general host shutoff function which limits new MHC protein expression and some of the miRNAs can influence cell proteins involved in antiviral responses (Long et al., 2011).

Ectopic versus Eutopic Endometrium

In ovarian endometriomas, compared to paired eutopic endometrium during the proliferative phase of the cycle, genes regulating structural proteins and immune-related genes have been observed to be dysregulated.¹⁴⁸ The significance of the upregulation of genes associated with immune functions and structural proteins may reflect immune dysfunction contributing to the pathogenesis of the disorder and for anchoring lipid droplets for steroid-hormone biosynthesis by endometriosis tissue, respectively.¹⁴⁸ A study using subtractive cDNA libraries and samples from subjects with minimal/mild peritoneal disease (revised American Fertility Society [rAFS] scores stage I/II) and moderate to severe peritoneal disease (stage III/IV) during either the proliferative or secretory phase revealed the following¹⁴⁹: Several genes and gene families were identified, including extracellular matrix/cell adhesion proteins, ribosomal proteins, transcriptional regulators (including JUN and EGR1), RNA processing, signaling intermediates, cell-cycle regulators (CDK2), GDP/GTP binding proteins, metabolism, and other functions. Some of these genes are known to participate in estradiol action and have antiapoptotic actions, suggesting that they may play a role in the pathophysiology of the disorder and may be diagnostic biomarkers or candidate therapeutic targets.

Epithelial cells isolated by laser capture microscopy (LCM) from eutopic endometrium and ectopic endometrium (endometrioma and peritoneal lesions) and subsequent analysis using cDNA microarrays with 9600 genes revealed that samples cluster by site (ovarian endometrioma versus nonovarian disease), underscoring that ovarian and peritoneal endometriosis have different complements of gene expression and may represent different disorders.¹⁵⁰ Both types of endometriosis had lower expression of genes involved in cell adhesion, WNT signaling, and induction of apoptosis; both types had higher expression of genes involved in acute-phase response, cell proliferation, cell-cycle regulation, and regulation of transport. Differences were observed in expression of genes associated with glycoprotein function, response to oxidative stress, and G protein–coupled receptor signaling. This may represent a response to the proinflammatory environment in which the lesions exist. In fact, IL-8 was upregulated, and IL-15 and PDGF-RA were downregulated, consistent with other studies.^151,152 Members of the MAPK pathway and oxidative stress pathways were upregulated in endometriosis lesions compared to normal endometrium. The study by Wu and co舉workers¹⁵⁰ underscores the differences in gene expression in ovarian and peritoneal disease and some important pathways that have been implicated in the pathogenesis of endometriosis.³⁴

Matsuzaki and associates¹⁵² characterized the transcriptome of deep endometriosis (i.e., rectovaginal disease) compared to ovarian and peritoneal disease using cDNA microarrays and LCM. PDGF-RA, PKCb1, and JAK1 were upregulated, and sprouty 2 and MAP kinase kinase 7 were downregulated in rectovaginal endometriosis stromal cells, supporting a role for the RAS/RAF/MAPK signaling pathway in the pathogenesis of the disorder. In endometriosis epithelial cells, COUP-TF2 and PGE2/EP3 were downregulated, and since these are negative regulators of aromatase, their downregulation may contribute to the known estradiol synthesis that occurs in some endometriotic lesions (see earlier). Tyrosine kinase receptor B (TRkB) in endometriosis epithelium and the serotonin transporter (5HTT) and the mu opioid receptor (MOR) in endometriosis stromal cells were upregulated genes that may be candidates in the pathophysiology of pain in endometriosis.¹⁵²

Using a cDNA microarray approach, Taylor and colleagues^153,154 found differential responsiveness to IL-1β, an inflammatory cytokine, in endometrial stromal cells from eutopic endometrium and endometriomas. IL-1 β downregulated Tob-1 in endometriotic stromal cells but had minimal effect on normal stromal cells, suggesting that this cytokine promotes growth of endometriotic lesions through inhibition of Tob-1 and an association of IL-1 β with altered cell-cycle gene expression in cells derived from endometriotic implants.^153,154

Overall, these data highly suggest that ectopic and eutopic endometrium differ in their gene expression in subsets of endometriosis (e.g., peritoneal, deep lesions, and endometriomas) and demonstrate a dysregulation in ectopic endometrium of genes that are members of the Ras, MAP kinase, and PI3 kinase signaling pathways. This is consistent with the mouse model of overexpression of K-ras and conditional PTEN deletion in ovarian surface epithelium that resulted in endometriosis (and endometrioid ovarian carcinoma).³⁴