| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Biochemistry, Molecular Biology, and Cell Biology, Northwestern University, Evanston, Illinois 60208
ABSTRACT
An expressed-sequence tag database search has identified three rat cDNA clones in the prolactin/growth hormone family, including a homologue of mouse proliferin-related protein (PRP). The encoded proteins of the two novel clones, designated prolactin-like proteins L (PLP-L) and M (PLP-M), are predicted to be synthesized as precursors of 229 and 227 amino acids, modified by N-linked glycosylation, and secreted as mature glycoproteins of 199 and 200 residues, respectively. Murine homologues to PLP-L and PLP-M were also identified. The open reading frame of rat PRP encodes a precursor protein of 245 amino acids and predicts a secreted 215-amino acid glycoprotein with 81% identity to mouse PRP. All three rat mRNAs are expressed in the placenta, and expression is not detected in other tissues. PLP-L mRNA expression is observed from Days 11–20, with highest levels at Day 13; highest levels of PLP-M are observed from Day 11 until parturition, with peak levels also on Day 13; and highest levels of PRP are also observed from Day 11 until term, with maximal expression on Day 17. All three genes are most highly expressed in invasive trophoblast cells lining the central placental vessel. The identification of molecular markers for endovascular trophoblasts serves to highlight the invasive nature of rodent placentation and may prove useful for future studies of placental function.
cytokines, placenta, trophoblast
INTRODUCTION
The mouse and rat provide the principal laboratory models for reproduction and development, and many processes in rodents translate readily to the human. Important differences, though, have been noted between rodent and human placentation. In particular, during the formation of the chorioallantoic placenta, human trophoblasts aggressively degrade the uterine epithelium and underlying structures [1], whereas rodent decidualization anticipates trophoblast invasion by removing barriers such as the uterine epithelium and by remodeling the decidual vasculature through angiogenesis [2]. Qualitatively, rodent trophoblasts are commonly considered less overtly invasive [3], although a subpopulation of rodent trophoblasts does migrate extensively into the uterus, establishing a structure called the placental vessel [4, 5].
A common feature of the rodent and human placenta is that invasive trophoblasts displace maternal endothelial cells and come into direct contact with maternal blood. These cells express proteases, adhesion molecules, and growth factors associated with their invasive phenotype and their establishment in an endovascular environment [6, 7]. Aberrant regulation of trophoblast invasiveness may be clinically manifested as preeclampsia, intrauterine growth retardation, or placenta accreta [8–10]. As a critical checkpoint for further development, placental defects in securing maternal vascular support are frequently evidenced as embryonic morbidity and mortality in targeted gene-disruption experiments [7]. Thus, the characterization of trophoblast differentiation and invasion in the rodent provides a framework for interpreting mutant murine phenotypes and for studying elements of human disease.
Among the most useful markers for differentiating classes of rodent trophoblasts are hormones in the prolactin (PRL)/growth hormone family [11–16], including at least 18 distinct proteins expressed in either the rat or mouse [17–31]. These hormones display a spectrum of cell-specific and temporally specific patterns of expression and are therefore useful in elucidating the factors that promote trophoblast differentiation and that regulate broad programs of trophoblast-specific gene expression. Some of these hormones, including placental lactogen (PL) I and proliferin (PLF), are expressed early in gestation, specifically in trophoblast giant cells [32, 33]. Placental lactogen II is also synthesized only by giant cells but in the latter half of pregnancy [18]. Proliferin-related protein (PRP) is expressed in both giant cells and spongiotrophoblasts [34], whereas rat decidual/trophoblast prolactin-related protein (d/tPRP) and mouse prolactin-like protein (PLP) B switch sites of expression from the maternal decidua to the extraembryonic trophoblasts during gestation [35, 36]. Prolactin-like protein J expression is detected exclusively in the decidua from shortly after implantation until midgestation [37].
The receptors and biological functions of many of these placental PRL-related hormones remain to be elucidated. Among the members with demonstrated bioactivity are PL-I and PL-II, which bind to the PRL receptor and elicit PRL-like actions, including the induction of progesterone production by the corpus luteum [38]. In the mouse, the angiogenic protein PLF and the antiangiogenic hormone PRP are important modulators of blood vessel development during pregnancy [39]. PLF accounts for the majority of soluble angiogenic activity secreted by midgestation placental cultures, and a reduction in placental production of PLF is associated with decreased decidual neovascularization [40]. Conversely, PRP is a potent antiangiogenic protein secreted from mid to late gestation [39] that may limit endothelial invasiveness by creating a barrier zone of predominantly antiangiogenic activity at the border between maternal and fetal tissue [11]. Prolactin-like protein A has recently been shown to bind to natural killer cells and to inhibit their cytolytic activity [41]. The murine proteins PLP-E and PLP-F, which share 54% identity at the amino acid level, have been found to bind to splenic and bone marrow megakaryocytes, and PLP-E has been shown to stimulate megakaryocyte differentiation [42].
Seven members of the PRL family (PLP-C, PLP-Cv, PLP-C
, PLP-Crp, PLP-D, PLP-H, and d/tPRP) with undefined receptors and functions can be grouped into the PLP-C subfamily on the basis of a high degree of amino acid sequence identity and a six-exon, five-intron gene structure [12]. Expression of the majority of the PLP-C family thus far appears to be limited to the rat. The exceptions include the rat protein d/tPRP, which has an identified murine homologue, dPRP [36], and PLP-C, which is unique to the mouse [27].
Prolactin-like protein J and PLP-B are the remaining characterized members of the PRL family with unknown functions, but during the preparation of this manuscript, the nucleotide sequences of several additional members of the rat PRL family were published [43]. These newest members, termed PLP-I, PLP-K, and PLP-L, are all predicted to be secreted glycoproteins. The expression of one of these hormones, PLP-L, is characterized in this report, along with the identification and characterization of two previously unreported, related hormones. These three hormones prove to be excellent markers of invasive trophoblasts in the rat. Future analysis is expected to reveal additional endocrine activities essential for the physiological changes that occur during pregnancy and the molecular components that contribute to trophoblast invasiveness.
MATERIALS AND METHODS
Animals and Animal Care
Pregnant Sprague-Dawley rats (Charles River Laboratories, Wilmington, MA) at defined gestational stages were maintained on days with 14L:10D, with lights-on at 0600 h. Food and water were freely available. All procedures were approved by Northwestern University's animal care and use committee.
Database Screening and DNA Sequence Analysis
The nonmouse, nonhuman expressed sequence tag (EST) database from the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/blast/blast.cgi?Jform=0) was searched for sequences similar to mouse PLF by using BLAST (Basic Local Alignment Search Tool). When EST cDNA sequences were found that were similar to that of PLF, they were screened for identity to previously characterized family members found in the nonredundant database consisting of GenBank+ EMBL+ DDBJ+ PDB sequences (but no EST, STS, GSS, or HTGS sequences) by using BLAST. EST cDNAs of novel rat PRL family mRNAs were obtained from Genome Systems (St. Louis, MO). Plasmid DNAs were purified from bacterial lysates by alkaline lysis. With regard to PLP-L and PLP-M, the Marathon cDNA amplification kit and Advantage cDNA PCR kit (Clontech, Palo Alto, CA) were utilized according to the manufacturer's instructions for 5'-directed, anchored PCR to obtain cDNAs with complete open reading frames. The nucleotide sequences of PLP-M, PLP-L, and rat PRP were determined by using primers corresponding to both vector-specific and internal oligonucleotides, except in the case of the anchored PCR products, which were sequenced directly as an uncloned pool using 5'-directed primers. Sequencing was performed on a ABI 310 PRISM genetic analyzer using Big Dye fluorescent terminators (Perkin-Elmer, Forest City, CA). The major open reading frame, which begins with the initial ATG in each clone, and the predicted translation product were identified by using GeneWorks (Intelligenetics, Mountain View, CA). Putative signal sequences and their cleavage sites were identified by alignment to the sequences of previously characterized hormones in the PRL family; signal sequence cleavage in this protein family is typically after a serine residue in the consensus sequence (-6)WENVASXP(+2).
RNA Filter and In Situ Hybridization
Total RNA was prepared from rat tissues with Tri-Reagent (Sigma Chemical Co., St. Louis, MO) according to the manufacturer's instructions. At least two separate placental RNA preparations were analyzed for each gestational time point. For filter hybridizations, 20 µg of total RNA per lane were separated on 1% formaldehyde-agarose gels, transferred to nylon membranes (Schleicher and Schuell, Keene, NH), and exposed to ultraviolet light to cross-link the RNA to the membranes. Digoxigenin-labeled riboprobes were synthesized by using either T7 or T3 RNA polymerase in a reaction of 20 µl containing 0.5 µg linearized template DNA, 1x transcription buffer (Roche Molecular Biochemicals, Indianapolis, IN), 40 U ribonuclease inhibitor (Promega, Madison, WI), 400 µM nucleoside triphosphates (adenosine triphosphate, cytidine triphosphate, and guanosine triphosphate), 100 µM uridine triphosphate (UTP), and 40 µM digoxigenin-11-UTP at 37°C for 60 min. Template DNA was removed by adding 1 µl RQ1 deoxyribonuclease (Promega) and incubating at 37°C for 10 min. Probes were isolated by ethanol precipitation. Hybridizations were carried out for 12–16 h at 68°C in ULTRAhyb prehybridization and hybridization buffer (Ambion, Austin, TX). After hybridization, membranes were washed twice in 2x SSC (1x SSC is 150 mM NaCl and 15 mM Na citrate); twice in 0.5x SSC, 0.1% SDS at 65°C; and twice in 0.1x SSC, 0.1% SDS at 65°C. Membranes were incubated in blocking buffer (100 mM Tris-HCl pH 7.5, 150 mM NaCl, and 5% blocking reagent from Roche) at room temperature for 30–60 min. Alkaline phosphatase-conjugated antibody against digoxigenin (Roche Molecular Biochemicals) was diluted 1:20 000 in blocking buffer and added to the membranes for 30 min at room temperature. Membranes were washed twice for 15 min in washing buffer (100 mM Tris-HCl pH 7.5, 150 mM NaCl) and equilibrated for 5 min in detection buffer (20 mM Tris-HCl, pH 9.5; 100 mM NaCl; and 50 mM MgCl2). Finally, membranes were placed between acetate sheets for the application of the chemiluminescent substrate CDP-Star (Roche) diluted 1:500 in detection buffer.
Tissues for in situ hybridization analysis were rapidly frozen on dry ice and stored at -80°C before preparing 20-µm sections on a cryostat. Tissue sections were air dried before fixation in 4% paraformaldehyde in PBS for 5 min. After incubating in 2x SSC for 5 min, the sections were dehydrated by a series of 50%, 70%, 95%, and 100% ethanol incubations, each for 3 min. In situ hybridization was performed in a humidified chamber at 47°C for 12–16 h in a solution containing 50% formamide; 0.3 M NaCl; 10 mM Tris-HCl, pH 8.0; 1 mM EDTA; 1x Denhardt solution (200 µg/ml each of Ficoll 400, polyvinylpyrrolidone, and BSA), 10% dextran sulfate, 10 mM dithiothreitol, 0.5 mg/ml yeast transfer RNA, 0.5 mg/ml polyA RNA, and 10 ng/ml denatured riboprobe. After hybridization, slides were washed in 2x SSC for 10 min, followed by 20 µg/ml RNase A digestion at 37°C for 1 h. Slides were then washed in 0.5x SSC for 30 min, followed by a 1–2 h wash in 0.1x SSC at 65°C. After incubating in blocking buffer (2x SSC, 0.05% Triton X-100, and 0.1% BSA) for 1 h, alkaline phosphatase-conjugated antibody against digoxigenin was diluted 1:500 in blocking buffer and added to the tissue sections for 1 h at 37°C. Slides were washed for 10 min each in washing buffer and in detection buffer, followed by incubation at 37°C with the alkaline phosphatase chromogenic substrate solution containing 250 µg/ml nitro blue tetrazolium and 225 µg/ml 5-bromo-4-chloro-3-indolyl-phosphate (Roche) until adequate staining was observed. At least three placentas were analyzed with each probe for each time point.
RESULTS
Identification of PLP-L, PLP-M, and Rat PRP
The amino acid sequence of PLF was used to screen an EST database containing nonhuman, nonmouse cDNAs translated in all six reading frames to identify closely related sequences. Three novel rat cDNA clones (RSPBE14, ui-r-c0-il-b-11-0-ui, and ui-r-a1-dz-d-09-0-ui) were identified. Anchored PCR was used to extend the available EST cDNAs to generate complete open reading frames. Beginning with the initial ATG sequence, ui-r-c0-il-b-11-0-ui was found to encode a 227-amino acid protein; RSPBE14, a 229-residue protein; and ui-r-a1-dz-d-09-0-ui, a 245-amino acid protein. During the preparation of this manuscript, another group published the nucleotide sequences of several rat PRL family members [43]. Clone ui-r-c0-il-b-11-0-ui is very similar to what was termed PLP-L, and therefore, we have also designated this clone as PLP-L, but significant differences between the two sequences exist and will be discussed below. Continuing with the current convention of naming new PRL family members with sequential letters, we have named RSPBE14 PLP-M. The third clone, ui-r-a1-dz-d-09-0-ui, shares 81% predicted amino acid identity with mouse PRP and so is likely the rat homologue.
Using the rat PLP-L sequence as a probe, a mouse homologue (clone c0002a07) was identified in an EST database. 5' coding sequence was missing from the mouse clone, so PCR using a primer derived from the 5' untranslated region of rat PLP-L was used to obtain a complete open reading frame. The mouse and rat sequences share 81% identity at the amino acid level (Fig. 1). The mouse PLP-L open reading frame codes for a 230-amino acid protein that is predicted to be secreted as a 200-residue protein. A predicted consensus site for N-linked glycosylation (Asn-X-Ser/Thr) exists in the mature rat protein at residue 20, whereas the mature mouse protein is predicted to be modified at residues 20, 31, and 82. Most of the consensus cysteine residues found in the majority of cytokine superfamily members are present: in the case of both rat and mouse PLP-L, cysteines are found at positions 3, 11, 58, and 177. Evolutionarily, PLP-L appears to be distinct from any of the PRL subfamilies because its closest relative is PRL, to which it is 31% identical.
|
A mouse homologue for PLP-M was also found in an EST database. Clone 1434324 was obtained and found to be lacking a 5' coding sequence. PCR using 5' untranslated information from the rat sequence resulted in a complete open reading frame. The mouse and rat protein forms of PLP-M are 81% identical (Fig. 2). Analysis of the sequences suggests that the first 29 residues of both the rat and mouse proteins serve as a signal sequence. N-linked glycosylation is predicted at a consensus site in the mature mouse protein at residue 107; no N-linked glycosylation consensus sites are predicted to be present in the rat protein. PLP-M has the classical pattern of cysteine residues for PRL-related hormones, with cysteines at positions 3, 11, 58, 174, 191, and 199 in both the mature rat and mouse proteins. Comparisons to PRL and related hormones confirm that PLP-M is novel; the secreted form of PLP-M is most closely related to PLF, with which it shares 39% amino acid identity.
|
Mouse PRP is predicted to be synthesized as a preprotein of 244 amino acids and secreted as a 214-amino acid protein that is N-linked glycosylated at residues 4, 19, and 61 [44]. The open reading frame of rat PRP encodes a 245-amino acid protein, of which the first 30 amino acids are predicted to serve as a signal sequence. The mature mouse and rat forms of PRP are quite similar, sharing 81% amino acid identity, including cysteines in the same locations (Fig. 3). As with the mouse protein, mature rat PRP is predicted to be glycosylated at asparagines 4 and 19; in contrast, rat PRP lacks the consensus glycosylation site found at residue 61 in the mouse protein.
|
The complete cDNA sequences for PLP-L, PLP-M, and rat PRP have been submitted to GenBank. The corresponding accession numbers are AF226607 for rat PLP-L, AF226611 for mouse PLP-L, AF226608 for rat PLP-M, AF226610 for mouse PLP-M, and AF226609 for rat PRP.
Temporal Pattern of Expression of Rat PLP-L, PLP-M, and PRP
In placental tissue, PLP-L was found to be expressed first on Day 11 and at peak levels on Day 13, with expression diminished to a very low level by Day 20 and undetectable at term (Fig. 4A). Prolactin-like protein M has a similar expression pattern: mRNA was first observed on Day 11 with peak levels on Day 13, but in contrast to PLP-L, expression persisted throughout the remainder of pregnancy (Fig. 4B). Rat PRP expression was detected between Day 11 and term; however, in contrast to PLP-M, expression levels gradually increased to Day 17, remained high on Day 20, and abruptly diminished thereafter, with little hybridization observed at parturition (Fig. 4C). In the rat, expression of PLP-L, PLP-M, and PRP appears to be restricted to placental tissue on the basis of a lack of hybridization to RNA isolated from a variety of tissues, including heart, lung, liver, ovary, testis, spleen, and pituitary (data not shown).
|
Cell Type-Specific Expression of PLP-L, PLP-M, and Rat PRP
To identify the cells that synthesize PLP-L, PLP-M, and PRP in the rat placenta, in situ hybridization analysis was performed on rat tissues from Day 13 and Day 20 of gestation. Strikingly, PLP-L was found to be highly expressed on Day 13 in cells lining the central placental vessel (Fig. 5A). Additional hybridization is observed in the giant cell layer and in cells underlying the placental vessel within the spongiotrophoblast layer (Fig. 5A). On Day 20, expression within the central placental vessel remains the predominant region of PLP-L synthesis (Fig. 6A). The weak hybridization signal at Day 20 is consistent with the decline in PLP-L mRNA levels in total placental extracts (Fig. 4A).
|
|
PLP-M expression is also highly expressed by invasive trophoblasts lining the central placental vessel on Day 13 (Fig. 5C). Other sites of PLP-M synthesis were also observed, including giant trophoblast cells, spongiotrophoblasts, and labyrinthine trophoblasts (Fig. 5C). Endovascular expression of PLP-M persists at Day 20, as does expression by cells in the giant cell and labyrinth layers; of note is the diminished expression by cells in the spongiotrophoblast layer (Fig. 6C).
In the mouse, PRP is synthesized in both the giant cell and spongiotrophoblast layers [34]. In contrast, rat PRP was found to be expressed on Day 13 by the outer rim of giant cells, whereas expression by spongiotrophoblasts was undetectable (Fig. 5E). The expression by giant cells was at a very low level relative to the amount of synthesis observed in the endovascular trophoblasts lining the placental vessel. Note that the tortuous path of the central placental vessel results in both longitudinal and transverse sectioning of the vessel, resulting in a discontinuous hybridization pattern. At Day 20, little endovascular or giant cell hybridization remains (Fig. 6E); the bulk of the hybridization noted in total placental extracts likely corresponds to faint but broad labyrinthine expression.
DISCUSSION
During rodent placentation, modification of maternal uterine tissue occurs to afford the developing conceptus a means for nutrient and waste exchange. After blastocyst implantation on Day 4 of gestation, the antimesometrial decidua is degraded by placental trophoblasts, allowing for the diffusion of gases and nutrients from the decidua. As the metabolic needs of the conceptus increase, the mesometrial pole of the decidua undergoes extensive anticipatory vascular modification over Days 7–9 [2]. At this time, uterine barriers to ectoplacental cone invasion are removed by uterine lymphocytes of the natural killer lineage [45], after which trophoblasts invade and line the burgeoning maternal blood sinus forming the placental vessel.
Invasive trophoblasts can be distinguished from other trophoblast lineages in their expression of proteases and cell adhesion proteins associated with achieving an epithelial to endothelial conversion. Degradation of maternal stroma is thought to be enhanced by trophoblastic secretion of several factors including matrix metalloproteinase 9, urokinase-type plasminogen activator, and cathepsins B and L [7]. In humans, the endothelial cell surface markers platelet-endothelial cell adhesion molecule 1, vascular cell adhesion molecule 1, and integrins vß3 and 1ß1 are all up-regulated by invasive trophoblasts [46]. Targeted gene disruption experiments show that none of these factors is absolutely required for proper invasion and placentation in mice [7, 47], suggesting the presence of functional redundancy among factors or functional divergence between species.
Shortly after the establishment of the hemochorial rat placenta, the mRNAs encoding PLP-L, PLP-M, and PRP are highly expressed by the endovascular trophoblasts of the placental vessel. Maximal expression of PLP-L and PLP-M is observed on Day 13, which coincides with the earliest histological observation of maximal maternal penetration by endovascular trophoblasts [5]. Features predicted to be shared by PLP-L, PLP-M, rat PRP, and other proteins of the cytokine superfamily include the presence of a cleaveable signal sequence, asparagine-linked carbohydrate modifications, and conserved cysteine residues. From an evolutionary perspective, PLP-L appears to form a unique branch of the PRL family tree, whereas PLP-M shares 39% identity with PLF, a degree of similarity shared between PL-I and PL-II. The nucleotide sequence of PLP-L reported recently [43] differs from the sequence we obtained by three nonsequential nucleotides. This disparity results in a 16-amino acid divergence between the two sequences, including the addition of an extra amino acid in the earlier report. Although both sequences originated from the ui-r-c0-il-b-11-0-ui EST clone, only the sequence reported here is similar to the predicted mouse translation product in the region where the two rat PLP-L sequences diverge; 13 of 15 amino acids are conserved between the mouse and rat sequences in this report, whereas no identities exist in this region between the mouse homologue and the earlier reported rat PLP-L.
In addition to being markers of invasive trophoblasts, the expression patterns of PLP-L, PLP-M, and PRP serve to highlight differences in gene expression among trophoblasts. At Day 13 and Day 20 of rat gestation, PLP-M is broadly expressed by many trophoblast cell populations, with the exception of the spongiotrophoblasts to which little hybridization is observed at Day 20. PLP-L expression is predominantly confined to cells surrounding the placental vessel and giant cells. Early PRP expression is limited to invasive trophoblasts, with faint hybridization to cells lining the maternal-fetal interface. By Day 20, though, labyrinthine expression was seen to predominate. The differential expression of PLP-L and PRP compared with that of PLP-M by the endovascular trophoblasts of late gestation may prove useful in identifying important genetic regulatory elements. In sum, it appears that in addition to secreting factors to modify the extracellular matrix and switching integrins to adhere to each new locale, invasive rodent trophoblasts express several cytokines in the PRL family.
Endovascular trophoblasts are well situated to deposit factors into the circulation to affect maternal physiology. It is possible that other previously characterized PRL family members that are expressed during midgestation are also synthesized by endovascular trophoblasts. Because of the central location of the placental vessel, though, tissue sections taken from the periphery of the discoid placenta may fail to contain endovascular cells, and even central sections may miss this structure because of its winding topology. By identifying PLP-L, PLP-M, and PRP as markers of these endovascular trophoblasts, it becomes possible to define the factors that regulate gene expression in these cells. These regulatory components, from extracellular signaling molecules to transcription factors, in turn represent candidates for directing the fate of this class of trophoblasts and their invasive properties.
ACKNOWLEDGMENTS
We wish to thank Diane M. Mayer and Janelle Roby for expert technical assistance and Jiandie Lin for insightful camaraderie.
FOOTNOTES
1 This work was supported by NIH grants HD29962 and HD24518, by the NIH P30 Research Center on Fertility and Infertility at Northwestern University (HD28048), and by the Robert H. Lurie Comprehensive Cancer Center (P30 CA60553). 
2 Correspondence: Daniel I.H. Linzer, Department of Biochemistry, Molecular Biology, and Cell Biology, Northwestern University, 2153 Sheridan Road, Evanston, IL 60208. FAX: 847 467 1757; dlinzer{at}nwu.edu
Accepted: March 23, 2000.
Received: February 2, 2000.
REFERENCES
v integrins. Cell 1998; 95:507–519.[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  A. Ozturk, A. Fresnoza, A. Savoie, H. W. Duckworth, and M. L. Duckworth Defining Regulatory Regions in the Rat Prolactin Gene Family Locus Using a Large P1 Genomic Clone Endocrinology, November 1, 2003; 144(11): 4742 - 4754. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. O. Wiemers, L.-j. Shao, R. Ain, G. Dai, and M. J. Soares The Mouse Prolactin Gene Family Locus Endocrinology, January 1, 2003; 144(1): 313 - 325. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
