CG&D
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cell Growth & Differentiation

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Houldsworth, J.
Right arrow Articles by Chaganti, R. S. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Houldsworth, J.
Right arrow Articles by Chaganti, R. S. K.
Cell Growth & Differentiation Vol. 13, 257-264, June 2002
© 2002 American Association for Cancer Research

Expression Profiling of Lineage Differentiation in Pluripotential Human Embryonal Carcinoma Cells1

Jane Houldsworth, Simon C. Heath, George J. Bosl, Lorenz Studer and R. S. K. Chaganti2

Cell Biology [J. H., R. S. K. C.], and Cellular Biochemistry and Biophysics Programs [L. S.], and Departments of Epidemiology and Biostatistics [S. C. H.], Medicine [G. J. B., R. S. K. C.], and Neurosurgery [L. S.], Memorial Sloan-Kettering Cancer Center, New York, New York 10021


    Abstract
 TOP
 Abstract
 Introduction
 Results and Discussion
 Materials and Methods
 References
 
Pluripotential human embryonal carcinoma (EC) cell linesundergo differentiation programs resembling those occurring in embryonal stem cells during development. Expression profiling was performed during the terminal differentiation of the EC cell line, NTera2/Clone D1 by all-trans-retinoic acid. Time-response analysis via clustering of >12,000 human transcripts revealed distinct stages in the transition from an EC cell to neuronal progenitor cells expressing patterning markers compatible with posterior hindbrain fates followed by the appearance of immature postmitotic neurons with an evolving synaptic apparatus. Global analysis of gene expression allows monitoring cell fate and differentiation of EC cells in vitro and may provide insight into human embryonal stem cell development.


    Introduction
 TOP
 Abstract
 Introduction
 Results and Discussion
 Materials and Methods
 References
 
Adult human male germ cell tumors are unique in their display of histopathologies that resemble different stages of human development, and thus comprise a model system for the study of molecular mechanisms involved in human ES3 cell development (1) . Cell lines derived from such tumors, in particular, EC, have provided an invaluable in vitro resource in which molecular events regulating cell fate/lineage decision can be studied (2) . The EC cell lines can be maintained in an undifferentiated state in vitro, and undergo spontaneous or morphogen-induced differentiation programs. Some display the ability to differentiate along both somatic and extra-embryonic lineages placing them as equivalents of cells in the inner cell mass of the developing embryo (3 , 4) , whereas others exhibit a more restricted pluripotential differentiation program (5 , 6) . Global perspectives of the expression patterns of such cell lines during different differentiation programs may lead to the identification of genes that play important roles in lineage decision/cell fate. Yet other EC cell lines have limited or no capacity for differentiation (7) , studies of which may ultimately reveal master regulators of differentiation induction in the developing zygote (1) .

To date, the most widely characterized pluripotential EC cell line is NT2/D1 (8) , which displays the ability to differentiate along different somatic lineages dependent on the agent or morphogen. In the presence of RA, NT2/D1 irreversibly differentiates along a neuronal lineage (8 , 9) that in recent years has been exploited as a source of human postmitotic neurons for experimental studies (10) . Other morphogens such as BMP-2, BMP-4, and BMP-7 have been reported to induce NT2/D1 cells to differentiate into non-neuronal lineages of an epithelial-phenotype (11, 12, 13) , whereas yet other agents such as hexamethylene bisacetamide cause differentiation into a non-neuronal program of undefined lineage (5) . The molecular events that regulate the induction of such lineage decisions in regard to known mammalian morphogens (RA and the BMP family) are as yet unknown and perhaps mimic those occurring during human ES cell development. To identify potential candidate genes for such regulatory roles, comprehensive expression profiling of the NT2/D1 cell line was performed during RA-induced differentiation using oligonucleotide microarray hybridization. Of the >12,000 transcripts represented on the array, 953 exhibited a change (predominantly an increase) in level of expression during the differentiation program. Examination of the expression profiles during the differentiation program revealed a loss of stem cell phenotype associated with the appearance of a neuronal progenitor cell with regional characteristics of the posterior hindbrain and finally immature postmitotic neurons with evidence for an evolving synaptic apparatus. Together, this comprehensive expression profile analysis of the pluripotential NT2/D1 cell line during differentiation induction has provided an insight into genes that on functional analysis may have similar roles during normal human embryonic development.


    Results and Discussion
 TOP
 Abstract
 Introduction
 Results and Discussion
 Materials and Methods
 References
 
Expression Profiling of RA-induced Differentiation of NT2/D1 Cells.
Exposure of NT2/D1 cells to RA leads to a flattening of the cells, noticeable by day 3, which was followed by a marked decrease in the cell doubling time between days 3 and 6, and the appearance of neurons around day 12 (13) . At designated times during the differentiation program, cultures were collected, the RNA extracted and submitted for expression profiling using Affymetrix oligonucleotide arrays representing >12,000 human transcripts. Because morphological changes were evident by day 3 of treatment, five profiles (6, 12, 24, 48, and 72 h) were obtained during this period. It was reasoned that genes that have roles in lineage decision and slowing of cell growth associated with differentiation may be up- or down-regulated at earlier times rather than later. On the other hand, at later times, genes that are involved with neural identity, biochemistry, and function were expected to be either induced or down-regulated. During this period, cultures were collected at 6, 10, and 14 days of RA treatment. To identify transcripts of which the expression levels were altered as a result of treatment, the profiles of RA-treated cultures were compared with those obtained from mock-treated cultures. The entire differentiation program and expression profiling was performed in triplicate, and the primary expression data for each collection time and for each experiment are available (see supporting information). Only those transcripts that exhibited a >=2-fold change (increase or decrease) in expression in all three of the replicates for at least one collection time were considered further. A total of 953 transcripts fulfilled these criteria (see supporting information). The majority of these transcripts (800) could be assigned a gene name and/or a function; however, a significant portion of hypothetical transcripts remain with an unknown function.

Northern hybridization analysis of selected genes was performed to confirm changes in transcript level as identified by oligonucleotide microarray expression profiling. Fig. 1Citation shows the hybridization signals obtained for six transcripts shown by expression profiling to be either down-regulated during the differentiation program (CCND1 and MYCN), or up-regulated (CDKN1A, NEFL, PAX6, and ANXA1). These arbitrarily tested transcripts show a good concordance in the profile of expression as assayed by microarrays and Northern hybridization.



View larger version (46K):
[in this window]
[in a new window]
 
Fig. 1. Comparison of the expression profiles of transcripts by Northern and oligonucleotide microarray hybridization. Northern hybridization was performed with probes for the indicated transcripts on RNAs obtained in a representative replicate of the three experiments performed and used for microarray expression analysis. The hybridization signal obtained for ACTB was performed to control for loading. The log-transformed expression profile obtained for the respective experiment for each transcript using oligonucleotide microarrays is shown.

 
Temporal Expression Clusters.
To characterize patterns of transcript expression during the time course of differentiation, k-means clustering was performed on the transcripts of which the expression was altered at least once during the program. The fold change in expression of each of the 953 transcripts at each collection time for the three triplicates was log transformed, and the mean used for k-means clustering. Fig. 2Citation shows the 20 expression profiles obtained using this clustering algorithm. Within each cluster, the profile shown represents the weighted mean of the normalized vectors of the transcripts within that cluster. Whereas each transcript is represented by an independent probe set on the array and considered as such for clustering, it must be noted that 73 transcripts were found to be represented by at least two independent probe sets based on Affymetrix groups of homologous sequences. Of these, 39 were similarly clustered, whereas 34 were placed in different clusters, suggesting that the latter may not be representative probe sets of the same gene. Also, three transcripts were represented on the array in addition to an alternatively spliced transcript that in one case clustered with the primary transcript. One gene was represented by four probe sets and one alternative splice that all clustered together.



View larger version (46K):
[in this window]
[in a new window]
 
Fig. 2. Temporal expression clusters during RA-induced differentiation of NT2/D1 cells. K-means clustering was performed on 953 transcripts where transcript vectors were normalized so that clustering was performed according to pattern rather than absolute level of expression. The profiles of the 20 expression patterns are shown, with the number of transcripts assigned to each cluster in parentheses. Each data point represents the weighted mean of the normalized transcript vectors within a cluster.

 
Such a temporal cluster analysis revealed three clusters of transcripts of which the expression was either up-regulated (clusters A and B) or down-regulated (cluster K) and returned to near undifferentiated levels. Cluster A represented a group of 19 transcripts (comprising 14 known and 2 unknown genes) that were markedly induced at the first collection time (6 h after addition of RA), peaked at 48 h, and returned to near undifferentiated levels by day 6. Contained within cluster A were the transforming growth factor ß-related secreted cytokines LEFTA and LEFTB, which are known to be induced by RA and involved in the Nodal loop of left-right asymmetric expression of genes such as Pitx2 that are important during mammalian development (14 , 15) . Whereas the induction of ID1 (an inhibitory transcriptional regulator of the basic helix-loop-helix family) after RA treatment has been documented in both NT2/D1 and other cell lines, its functional significance is as yet unknown (13 , 16) . The other transcriptional regulator within cluster A, TCF7, has been suggested to act as a transcriptional repressor of ß-catenin-TCF4 target genes such as CCND1 and MYC (17) , both of which exhibit a decline in expression during the differentiation program (clusters S and K, respectively). Also included are a number of genes that are involved in signal transduction both at upstream (BLNK, ARHGAP8, and PARG1) and downstream (MAPKAP2 and MKP4) sites in signaling cascades. Changes in the levels of neurofilament H transcripts (intermediate filament specific for neurons) during the differentiation program clustered in this group. Whereas the promoter of NEFH has not as yet been shown to contain a retinoic acid response element (18) , detectable levels of the protein are evident in few cells by day 2 after RA treatment and persist as a marker for NT2/D1 neurons (9) . Other clusters of transcripts (clusters C and E) exhibited an early increase in transcript level although not to the same extent as those in cluster A. In those clusters, the transcripts remained elevated throughout the differentiation program (Fig. 2)Citation . Of all of the transcripts of which expression remained up-regulated at day 14 of treatment (clusters C-J and clusters L-O), only relatively few (clusters J and O) were induced late (after day 6).

Fewer transcripts were observed to be down-regulated than up-regulated during the differentiation program (Fig. 2)Citation . A cluster of down-regulated transcripts with a temporal profile comparable with that for cluster A was not identified. Rather, transcripts exhibiting an early down-regulation tended to remain so throughout the differentiation program (cluster P) or, over time, the levels gradually increased above those in undifferentiated cells (clusters H and M). Using this clustering algorithm, no cluster of transcripts was identified to be down-regulated only late in the differentiation program (after day 6).

Retinoid Response.
Within the transcripts of which the expression was altered during the differentiation program were a number encoded by genes known to be involved in the response of cells to RA such as the retinoid receptors and cofactors, as well as those involved in the cellular metabolism of retinoids. Fig. 3Citation shows the expression patterns of seven such transcripts during RA treatment. The observed increase in transcript levels for the retinoid nuclear receptors RARA and RARB, along with a cofactor (CRABP2) that is important for delivering RA onto the receptor and acting as a cofactor, are consistent with their purported functions (19 , 20) . The remaining four genes are involved with regulation of the metabolism (CYP26A1 and SDR1) and transport (RBP1 and CRABP1) of retinoids.



View larger version (38K):
[in this window]
[in a new window]
 
Fig. 3. Expression patterns of transcripts during RA-induced differentiation of NT2/D1 cells. The expression patterns of select transcripts are shown, with each data point representing the unnormalized mean of the log-transformed f values for the triplicate experiments.

 
Proliferation/Differentiation.
As in other differentiation programs, RA-induced differentiation of NT2/D1 cells is associated with a slowing of the cell cycle. In this study, a number of transcripts were altered during RA treatment, of which the function could be directly implicated in cell cycle control (Fig. 3)Citation . Transcripts of which the levels were increased during the differentiation program included the cell cycle inhibitor, CDKN1A/WAF1 (Figs. 1Citation and 3Citation ), for which an increase in this differentiation program has been documented previously at both the mRNA and protein levels (13) . Two members of the RB gene family were induced at the transcript level, which in the case of RB1 was not mimicked at the protein level.4 The transcript levels of two cyclin gene family members were altered during the differentiation program. CCNDG2 transcript levels were increased (Fig. 3)Citation consistent with an up-regulation reported during growth arrest (21) . On the other hand, CCND1 transcript levels declined (Figs. 1Citation and 3Citation ), which is also reflected at the protein level (22) . Control of expression of CCND1 appears to be an important step in slowing of cell cycle, and few transcription factors have been implicated in this function that were observed to be altered in this study. One of these includes TCF7 as described above (17) . An induction of the levels of TOB1 was also observed, which is a member of the so-called PC3/BTG/TOB family of genes that mediate growth arrest before differentiation (possibly through Rb) and after genotoxic damage (23 , 24) . It has also been reported recently to antagonize BMP signaling (25) . Interestingly PC3/BTG2, BTG1, and BTG3 (clusters C, C, and G, respectively) were all induced. The precise molecular mechanism whereby these genes effect growth arrest is not entirely clear at present. Alterations in the levels of transcripts that encode proteins with essential functions during cell cycle were predominantly down-regulated during differentiation, in agreement with their respective roles in DNA replication (e.g., MCM7/CDC47), nucleotide homeostasis (e.g., ATIC and UMPK), chromatin assembly (e.g., CHAF1A), and chromosome segregation (e.g., STK12). Interestingly, the levels of only three transcripts of which their function is directly involved with apoptosis were altered during differentiation. These included the proapoptotic genes BNIP3L and PEG3 (cluster G), and CFLAR/FLIP (cluster N).

Associated with the slowing of the cell cycle and progression toward terminal differentiation, microarray expression analysis revealed alterations in the expression of genes, among others, of which their products function in affecting cytoskeleton remodeling (at signal transduction and structural levels) and glycolytic/oxidative pathways of energy metabolism, respectively (see supporting information). Such changes may underlie some of the morphological and metabolic alterations reported to occur in this EC cell line.

Embryonic Development.
Induction of terminal differentiation of NT2/D1 cells by RA is coupled with a loss of EC markers. In the past, these have included the liver isoform of alkaline phosphatase, stage-specific embryonic antigens (globoseries glycolipids), and high molecular weight glycoproteins, all of which are easily detected by antibodies (2) . Whereas fewer studies have identified EC markers at the transcript level, POU5F1/OCT3/OCT4 is a well-described marker of ES cells (26) . In this study, the transcript levels of POU5F1 declined after an initial delay (Fig. 3)Citation , consistent with loss of a pluripotent phenotype during differentiation.

Much research effort has focused on identifying genes that have functional roles in regulating mammalian development. Within the study presented here, the transcript levels of a number of such genes encoding morphogens, cytokines/growth factors, glycoproteins, and transcriptional regulators that have known roles in regulating spatial patterns of downstream effector genes and, thus, pattern formation in the developing embryo, were altered. Members of the BMP family of morphogens have been documented to induce differentiation of NT2/D1 cells along lineages distinct from that induced by RA (11, 12, 13) . During the differentiation program induced by RA, the transcript levels of BMP7 were found to increase late, and, thus, may not play an immediate early role in RA-induced differentiation. In addition, an up-regulation of TOB1 was detected consistent with its recently identified role as an inhibitor of cell growth and BMP signal transduction (Fig. 3Citation ; Ref. 25 ). Thus, BMPs appear to have minor roles in RA-induced differentiation of NT2/D1 cells as discussed elsewhere (12) . Expression profiling performed during this differentiation program has now implicated roles for other cytokines/growth factors that have documented roles in mammalian development. These include LEFTA and LEFTB, TDGF1/CRIPTO, and FGF4 (Fig. 3)Citation . Only one member of the WNT family of genes that encode glycoproteins functioning in pattern formation in the developing embryo, WNT2B, exhibited an alteration in the transcript level (Fig. 3)Citation , confirming a previous report (27) . Two receptors for wnts of the frizzled family (FZD6 and FZD7) exhibited earlier increases in expression (Fig. 3)Citation than did WNT2B. Clearly the precise roles of such genes in this and other differentiation programs remain to be confirmed through function-based assays.

A variety of transcriptional activators and repressors have been implicated in mammalian development in regulating the spatial expression of genes controlling pattern formation within the developing embryo. These include the well-studied HOX clusters, of which the expression patterns in NT2/D1 cells during RA treatment have been mostly described (28) . Fig. 3Citation shows a sequential activation of expression of these genes, correlating with their physical order in the cluster from 3' to 5' (28) . The expression of these and other homeotic genes have been reported to be negatively regulated by members of the polycomb group of genes (29) of which three (BMI1, EDR2, and SCML2) were found to be elevated in this study (Fig. 3)Citation . In this model system, changes in the transcript levels of other homeobox-containing genes were identified [MEIS1, MRG1, and ATBF1 (cluster E); DLX5 and PBX1 (cluster C); MEOX1 (cluster H); PMX1 (cluster O); PROX1 (cluster G); TCF2 (cluster D); and TGIF (cluster P)]. Of note, only one of these (TGIF) exhibited a decline in transcript level during the differentiation program, possibly related to its reported role as a repressor of RXR responsive elements within promoters (30) .

Neural Differentiation.
Because NT2/D1 differentiates predominantly along a neuronal lineage on exposure to RA, markers associated with induction of this lineage have been identified as reported previously (2 , 8 , 9) . Such studies indicated that the resulting embryonic neurons resemble more closely those of the central rather than peripheral nervous systems (31) . Consistent with this observation is the gradual decline in the level of the transcript (PROML1) that encodes the AC133/CD133 antigen, expressed on human CNS stem cells (Fig. 3Citation ; Ref. 32 ). Transcripts for other proteins typically expressed in CNS stem cells were either not expressed such as the RNA-binding protein musashi (MSI1) and the SRY-related high-mobility group box containing transcription factors SOX1 and SOX3, or no changes in expression were detected during differentiation such as the intermediate filament vimentin (VIM) and the SH2/PTB-containing SHC-transforming protein adaptor (SHC1). However, an induction of expression, albeit late, of two basic helix-loop-helix transcription factors, ASCL1 and NSCL2 being early regulators of the development of CNS and neural crest cells (33 , 34) , was detected (Fig. 4)Citation . The transcription factor PAX6 has been reported to have a role in the specification of neural identity (35) , and an up-regulation of this factor was observed during RA-induced differentiation of NT2/D1 cells (Fig. 1)Citation . Also of note is the induction of neurofilament transcripts that are highly specific markers for CNS neurons (Ref. 9 ; Fig. 4Citation ) and the down-regulation of a neurofilament found in fetal brain (INA/NF66; Ref. 36 ). Microtubule-associated proteins typically expressed in neuronal dendrites as well as neural-specific tubulins were also induced (Fig. 4)Citation . Examination of expression of transcripts coding for different regions of the brain revealed a sequential activation of HOXA2 and ZIC1 typically observed in hindbrain development and the generation of cerebellar granule cells (37) . On the other hand, no changes in expression of other genes specific for anterior hindbrain, midbrain, diencephalon, or forebrain identity such as GBX2 and PAX8 (transcripts present); or PAX2, PAX5, EN2, DLX2, EMX1, and EMX2 (transcripts absent) were detected. One notable exception was increased expression of FOXG1B/HBF1 at day 14, a forkhead-related transcription factor expressed in early forebrain development (Fig. 4)Citation . The rapid and consistent induction of HOX gene clusters in the absence of any induction of the more anterior CNS patterning genes suggest that RA-induced differentiation of NT2/D1 EC cells might preferentially induce early neuronal precursor phenotypes with regional characteristics of the posterior hindbrain and other posterior CNS regions.



View larger version (47K):
[in this window]
[in a new window]
 
Fig. 4. Expression patterns of neural-associated transcripts during NT2/D1 RA-induced differentiation. The expression patterns of transcripts associated with neural lineage differentiation and function are shown. Each data point represents the unnormalized mean of the log-transformed f values for the triplicate experiments.

 
Transplantation of NT2/D1 neurons into murine brains has suggested that these differentiated cells display plasticity in the development of specific subtypes of neurons (38) . In the present study, no specific neurotransmitter subtypes could be characterized based on the expression patterns of transcripts specific for neurotransmitter synthesis or transport. Long-term differentiation studies will have to address whether the early presence of hindbrain-specific regional identity markers, in the absence of any mid- or forebrain specific makers, might restrict the neurotransmitter subtypes or function available for these neurons. A number of genes associated with vesicular trafficking necessary for neuropeptide/neurotransmitter release were up-regulated within the time period of this study (Fig. 4)Citation . Whereas RA treatment of NT2/D1 cells predominantly induces the differentiation of cells along a neuronal lineage, it is recognized that other non-neuronal lineages are also induced at a lower frequency (2) . The non-neuronal cells observed at day 14 have been reported not to be persistent undifferentiated cells, although they were variably positive for vimentin and cytokeratin (9) . The identity of these non-neuronal cells remains largely unknown; however, examination of the expression patterns of genes associated with the glial lineage has revealed some evidence for the presence of oligodendroglia, with no evidence of the induction or presence of astroglia markers (Fig. 4)Citation . Additional culturing of differentiated NT2/D1 results in a population of cells comprising ~99% neurons (38) .

Concluding Remarks.
In this study, a pluripotential EC cell line, NT2/D1, was induced to undergo differentiation by RA along a somatic lineage of a predominantly neuronal phenotype, and the patterns of expression of >12,000 transcripts associated with the induction of this differentiation program were assayed by oligonucleotide microarray hybridization. Such an extensive analysis revealed patterns of expression of genes consistent with the slowing of cell growth, loss of a stem cell phenotype, and the induction of lineage markers. The molecular mechanisms by which the changes in specific gene transcript levels were effected are largely unknown. At the transcriptional level, regulation could occur directly by RA, by virtue of the presence of a retinoic acid response element in a gene promoter or enhancer responsive to different members of the RA nuclear receptor family, or changes in the composition of activator/repressor/cofactor complexes at regulatory sites possibly including those found to be altered in this study. Alternatively, changes in the steady state levels of transcripts could also arise by changes in the stability of the transcripts, possibly affected in some cases by RNA-binding proteins such as those induced in this analysis. The functional relevance of the changes in transcript levels to the induction of differentiation and lineage commitment observed in this study require confirmation through function-based assays.

Human EC cell lines provide an excellent model system in which the molecular mechanisms governing cell fate and lineage decisions normally occurring in ES cells can be investigated. Comparison of expression profiles of the differentiation program in this study with other differentiation programs induced by other morphogens and in other EC cell lines may well identify candidates ranging from master regulators of differentiation to molecular inducers of specific differentiation programs. Whereas the possibility exists for the use of human ES and germ cell cultures in similar studies, their availability in culture and the controversial nature of such studies has thus far limited their overall potential. On the other hand, human EC cell lines offer an unlimited supply of cells with various capacities for pluripotentiality and, thus, represent a powerful model system in which to identify genes having functions during human development.


    Materials and Methods
 TOP
 Abstract
 Introduction
 Results and Discussion
 Materials and Methods
 References
 
Cell Culture.
Cultures of NT2/D1 cells were maintained in DMEM as described previously (13) . Cells were seeded at 2.5 x 106/10 cm plate and 24 h later were treated with 10 µM RA (dissolved in DMSO; Sigma). Parallel mock-treated plates served as controls. Prior studies by us and others (8 , 13) have shown no detectable effect of the comparable concentration of DMSO (0.1%) on the morphology, cell doubling, or expression of select mRNAs in these cells. Cultures treated continuously with RA were collected at 6, 12, 24, 48, and 72 h for RNA extraction. A mock-treated culture collected at 24 h served as the control for these treatment times. For later collection times (6, 10, and 14 days), RA- and mock-treated cultures were reseeded every 3–4 days and collected no less than 24 h after reseeding (13) . A mock-treated day 14 culture served as the control for these collection times. Three independent replicates of RA- and mock-treatment of cultures were performed.

RNA Isolation and Array Hybridization.
Total RNA was extracted from the collected cell pellets using an RNeasy kit (Qiagen), and the quality was assessed by spectrophotometry and agarose gel electrophoresis (13) . First- and second-strand cDNA synthesis was performed using a cDNA synthesis kit (Invitrogen), and biotinylated cRNA was prepared according to manufacturer’s recommendations (Enzo). The cRNA was fragmented and 15 µg hybridized to Human Genome 95A oligonucleotide arrays (Affymetrix) according to the manufacturer’s recommended conditions. Using a specially designed fluidics station (Affymetrix), the arrays were then subjected to a series of rinses and stained before scanning in a confocal scanner (Affymetrix).

Expression Data Analysis.
Comparison expression analyses were performed using the Affymetrix Microarray Suite 4.0. An absolute analysis was performed on each of the untreated cultures to serve as baseline for the respective comparison analyses. Such analyses were performed for each of the individual triplicate experiments. A change in the level of expression of a transcript was considered only when the following criteria were satisfied: (a) it was designated as an increase or decrease by the custom software; (b) the calculated fold change (f) was >=2; (c) for a designated increase the transcript was present in the RA-treated culture and for a decrease was present in the mock-treated culture; (d) it was not designated as an increase or decrease at >=2-fold between the mock-treated cultures of the triplicates; and (e) it fulfilled the above criteria in all three of the individual experiments. Transcripts that exhibited a change in expression for at least one collection time were submitted to k-means clustering.

For clustering purposes, the f values as calculated by the custom software for each changed transcript at each time point and for each replicate were first transformed on to a continuous scale, such that f' = f if f >= 1 and f' = 1/f if f <= -1, and then log transformed to give x = log(f'). For each transcript at each time point, the mean of the log-transformed values (f') was determined as x(replicate 1) + x(replicate 2) + x(replicate 3)/3. The variances of the three values for each transcript at each time point was calculated and used in the clustering algorithm in the computation of the cluster means. Clustering was performed on the n-dimensional vectors where n is the number of time points, using the mean log-transformed values for each transcript. Before clustering, the vectors were normalized to the unit sphere so that transcripts clustered according to their pattern of expression over time rather than the absolute level of expression at each time point. K-means clustering was then performed on the normalized vectors, with the number of clusters fixed at 20 (selected based on exploratory studies to yield a manageable number of coherent clusters). The distances between transcripts were assessed using Pearson’s correlation coefficient. For each iteration of the k-means clustering algorithm, the mean vector for each vector was updated using the mean of the transcript vectors assigned to that cluster, weighted by the inverse of the variances of the mean vector for each transcript.

All of the expression and clustering data are available (see supporting information). Public databases were used to functionally categorize the transcripts for which a change in expression level was identified. UniGene cluster assignments were confirmed for each transcript with those provided in the Affymetrix EASI Database (Version 2.31). The OMIM database was used to identify the given gene nomenclature and other synonyms, and along with searching PubMed (MEDLINE), a gene was categorized according to its reported function

Northern Hybridization.
Northern hybridization of total RNA was performed as described previously (13) using the following probes: CCND1 (entire coding region cDNA, gift of A. Arnold, University of Connecticut School of Medicine, Farmington, CT), MYCN (1-kb genomic probe containing 0.4-kb coding sequence, gift of M. Schwab, German Cancer Research Center, Heidelberg, Germany), PAX6 (2 kb cDNA, gift of M. Ladanyi, Memorial Sloan-Kettering Cancer Center, New York, NY), CDKN1A (entire cDNA, gift of B. Vogelstein, The Johns Hopkins University, Baltimore, MD), NEFL (1.5-kb cDNA, purchased from American Type Culture Collection), and ANXA1 (1.44 kb cDNA, purchased from American Type Culture Collection). Filters were sequentially hybridized and then hybridized finally with an ACTB probe to control for loading.


    Acknowledgments
 
We thank Vladan Miljkovic for dedicated handling of the hybridization, washing, and scanning of the arrays, and Riccardo Dalla-Favera for use of the Columbia University Institute of Cancer Genetics, Affymetrix Hybridization Facility. We also thank Richard Rifkind of the Sloan-Kettering Institute for enthusiastic support.


    Footnotes
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported in part by grants from the NIH, Byrne Fund, and Lance Armstrong Foundation. Back

2 To whom requests for reprints should be addressed, at Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021. Phone: (212) 639-8121; Fax: (212) 717-3541; E-mail: chagantr{at}mskcc.org Back

3 The abbreviations used are: ES, embryonal stem; EC, embryonal carcinoma; NT2/D1, NTera2/CloneD1; RA, all-trans-retinoic acid; CNS, central nervous system; BMP, bone morphogenetic protein. Back

4 J. Houldsworth and R. S. K. Chaganti, unpublished observations. Back

Received for publication 1/28/02. Revision received 4/22/02. Accepted for publication 5/ 6/02.


    References
 TOP
 Abstract
 Introduction
 Results and Discussion
 Materials and Methods
 References
 

  1. Chaganti R. S. K., Houldsworth J. Genetics and biology of adult human male germ cell tumors. Cancer Res., 60: 1475-1482, 2000.[Abstract/Free Full Text]
  2. Andrews P. W. Teratocarcinomas and human embryology: pluripotent human EC cell lines. A review. APMIS, 106: 158-168, 1998.[Medline]
  3. Damjanov I., Horvat B., Gibas Z. Retinoic acid-induced differentiation of the developmentally pluripotent human germ cell tumor-derived cell line. NCCIT. Lab. Investig., 68: 220-232, 1993.
  4. Suzuki N., Yamada T., Matsuoka K., Hiraoka N., Iwamaru Y., Hata J. Functional expressions of fms and M-CSF during trophoectodermal differentiation of human embryonal carcinoma cells. Placenta, 20: 203-211, 1999.[Medline]
  5. Andrews P. W., Nudelman E., Hakomori S-I., Fenderson B. A. Different patterns of glycolipid antigens are expressed following differentiation of TERA-2 human embryonal carcinoma cells induced by retinoic acid, hexamethylene bisacetamide (HMBA) or bromodeoxyuridine (BudR). Differentiation, 43: 131-138, 1990.[Medline]
  6. Roach S., Schmid W., Pera M. F. Hepatocyte transcription factor expression in human embryonal carcinoma and yolk sac carcinoma cell lines: expression of HNF-3{alpha} in models of early endodermal cell differentiation. Exp. Cell Res., 215: 189-198, 1994.[Medline]
  7. Andrews P. W. Human teratocarcinomas. Biochim. Biophys. Acta, 948: 17-36, 1988.[Medline]
  8. Andrews P. W. Retinoic acid induces neuronal differentiation of a cloned human embryonal carcinoma cell line in vitro. Dev. Biol., 103: 285-293, 1984.[Medline]
  9. Lee V. M., Andrews P. W. Differentiation of NTERA-2 clonal human embryonal carcinoma cells into neurons involves the induction of all three neurofilament proteins. J. Neurosci., 6: 514-521, 1986.[Abstract]
  10. Hong C. S., Caromile L., Nomata Y., Mori H., Bredesen D. E., Koo E. H. Contrasting role of presenilin-1 and presenilin-2 in neuronal differentiation in vitro. J. Neurosci., 19: 637-643, 1999.[Abstract/Free Full Text]
  11. Andrews P. W., Damjanov I., Berends J., Krumpf S., Zappavigna V., Mavilio F., Sampath K. Inhibition of proliferation and induction of differentiation of pluripotent human embryonal carcinoma cells by osteogenic protein-1 (or bone morphogenic protein-7). Lab. Investig., 71: 243-251, 1994.[Medline]
  12. Caricasole A., Ward-Van Oostwaard D., Zeinstra L., Van Den Eijnden-Van Raaij A., Mummery C. Bone morphogenetic proteins (BMPs) induce epithelial differentiation of NT2D1 human embryonal carcinoma cells. Int. J. Dev. Biol., 44: 443-450, 2000.[Medline]
  13. Houldsworth J., Reuter V. E., Bosl G. J., Chaganti R. S. K. ID expression varies with lineage during differentiation of pluripotential male germ cell tumor cell lines. Cell Tissue Res., 303: 371-379, 2001.[Medline]
  14. Tsukui T., Capdevila J., Tamura K., Ruiz-Lozano P., Rodriguez-Esteban C., Yonei-Tamura S., Magallon J., Chandraratna R. A., Chien K., Blumberg B., Evans R. M., Belmonte J. C. Multiple left-right asymmetry defects in Shh(-/-) mutant mice unveil a convergence of the Shh and retinoic acid pathways in the control of Lefty-1. Proc. Natl. Acad. Sci. USA, 96: 11376-11381, 1999.[Abstract/Free Full Text]
  15. Wasiak S., Lohnes D. Retinoic acid affects left-right patterning. Dev. Biol., 215: 332-342, 1999.[Medline]
  16. Nagata Y., Todokoro K. Activation of helix-loop-helix proteins Id1. Id2 and Id3 during neural differentiation. Biochem. Biophys. Res. Commun., 199: 1355-1362, 1994.[Medline]
  17. Roose J., Huls G., van Beest M., Moerer P., van der Horn K., Goldshmeding R., Logtenberg T., Clevers H. Synergy between tumor suppressor APC and the ß-catenin-Tcf4 target Tcf1. Science (Wash. DC), 285: 1923-1926, 1999.[Abstract/Free Full Text]
  18. Elder G. A., Liang Z., Lee N., Friedrich V. L., Jr., Lazzarini R. A. Novel DNA binding proteins participate in the regulation of human neurofilament H gene expression. Brain Res. Mol. Brain Res., 15: 85-98, 1992.[Medline]
  19. Mark M., Ghyselinck N. B., Wendling O., Dupe V., Mascrez B., Kastner P., Chambon P. A. Genetic dissection of the retinoid signaling pathway in the mouse. Proc. Nutr. Soc., 58: 609-613, 1999.[Medline]
  20. Wolf G. Cellular retinoic acid-binding protein II: a coactivator of the of the transactivation by the retinoic acid receptor complex RAR. RXR. Nutr. Rev., 58: 151-153, 2000.[Medline]
  21. Horne M. C., Donaldson K. L., Goolsby G. L., Tran D., Mulheisen M., Hell J. W., Wahl A. F. Cyclin G2 is up-regulated during growth inhibition and B cell antigen receptor-mediated cell cycle arrest. J. Biol. Chem., 272: 12650-12661, 1997.[Abstract/Free Full Text]
  22. Spinella M. J., Freemantle S. J., Secula D., Chang J. H., Christie A. J., Dmitrovsky E. Retinoic acid promotes ubiquitination and proteolysis of cyclin D1 during induced tumor cell differentiation. J. Biol. Chem., 274: 22013-22018, 1999.[Abstract/Free Full Text]
  23. Guardavaccaro D., Corrente G., Covone F., Micheli L., D’Agnano I., Starace G., Caruso M., Tirone F. Arrest of G(1)-S progression by the p53-inducible gene PC3 is Rb dependent and relies on the inhibition of cyclin D1 transcription. Mol. Cell. Biol., 20: 1797-1815, 2000.[Abstract/Free Full Text]
  24. Tirone F. The gene PC3(TIS21/BTG2), prototype member of the PC3/BTG/TOB family: regulator in control of cell growth, differentiation, and DNA repair?. J. Cell Physiol., 187: 155-165, 2001.[Medline]
  25. Yoshida Y., Tanaka S., Umemori H., Minowa O., Osui M., Ikematsu N., Hosoda E., Imamura T., Kuno J., Yamashita T., Miyazono K., Nada M., Noda T., Yamamoto T. Negative regulation of BMP/Smad signaling by Tob in osteoblasts. Cell, 103: 1085-1097, 2000.[Medline]
  26. Brehm A., Ovitt C. E., Scholer H. R. Oct-4: more than just a POUerful marker of the mammalian germline?. APMIS, 106: 114-126, 1998.[Medline]
  27. Wakeman J. A., Walsh J., Andrews P. W. Human Wnt-13 is developmentally regulated during differentiation of NTERA-2 pluripotent human embryonal carcinoma cells. Oncogene, 17: 179-186, 1998.[Medline]
  28. Simeone A., Acampora D., Nigro V., Faiello A., D’Esposito M., Stornaiuolo A., Mavilio F., Boncinelli E. Differential regulation by retinoic acid of the homeobox genes of the four HOX loci in human embryonal carcinoma cells. Mech. Dev., 33: 215-228, 1991.[Medline]
  29. Akasaka T., van Lohuizen M., van Der Lugt N., Mizutani-Koseki Y., Kanno M., Taniguchi M., Vidal M., Berns A., Koseki H. Mice doubly deficient for the polycomb group genes Mel18 and Bmi1 reveal synergy and requirement for maintenance but not initiation of Hox gene expression. Development (Camb.), 128: 1587-1597, 2001.[Abstract]
  30. Bertolino E., Reimund B., Wildt-Perinic D., Clerc R. G. A novel homeobox protein which recognizes a TGT core and functionally interferes with a retinoid-responsive motif. J. Biol. Chem., 52: 31178-31188, 1995.
  31. Pleasure S. J., Lee V. M. NTera 2 cells: a human cell line which displays characteristics expected of a human committed neuronal progenitor cell. J. Neurosci. Res., 35: 585-602, 1993.[Medline]
  32. Uchida N., Buck D. W., He D., Reitsma M. J., Masek M., Phan T. V., Tsukamoto A. S., Gage F. H., Weissman I. L. Direct isolation of human central nervous system stem cells. Proc. Natl. Acad. Sci. USA, 97: 14720-14725, 2000.[Abstract/Free Full Text]
  33. Lipkowitz S., Gobel V., Varterasian M. L., Nakahara K., Tchorz K., Kirsch I. R. A comparative structural characterization of the human NSCL-1 and NSCL-2 genes. Two basic helix-loop-helix genes expressed in the developing nervous system. J. Biol. Chem., 267: 21065-21071, 1992.[Abstract/Free Full Text]
  34. Nieto M., Schurmans C., Britz O., Guillemot F. Neural bHLH genes control the neuronal versus glial fate decision in cortical progenitors. Neuron, 29: 401-413, 2001.[Medline]
  35. Mastick G. S., Andrews G. L. Pax6 regulates the identity of embryonic diencephalic neurons. Mol. Cell Neurosci., 17: 190-207, 2001.[Medline]
  36. Chien C. L., Mason C. A., Liem R. K. {alpha}-Internexin is the only neuronal intermediate filament expressed in developing cerebellar granule neurons. J. Neurobiol., 29: 304-318, 1996.[Medline]
  37. Wingate R. J., Hatten M. E. The role of rhombic lip in avian cerebellum development. Development (Camb.), 126: 4395-4404, 1999.[Abstract]
  38. Trojanowski J. Q., Mantione J. R., Hong Lee J., Seid D. P., You T., Inge L. J., Lee V. M. Neurons derived from a human teratocarcinoma cell line establish molecular and structural polarity following transplantation into the rodent brain. Exp. Neurol., 122: 283-294, 1993.[Medline]



This article has been cited by other articles:


Home page
Genes Dev.Home page
S. S. Dhar, S.-H. Lee, P.-Y. Kan, P. Voigt, L. Ma, X. Shi, D. Reinberg, and M. G. Lee
Trans-tail regulation of MLL4-catalyzed H3K4 methylation by H4R3 symmetric dimethylation is mediated by a tandem PHD of MLL4
Genes & Dev., December 15, 2012; 26(24): 2749 - 2762.
[Abstract] [Full Text] [PDF]


Home page
JCOHome page
J. E. Korkola, J. Houldsworth, D. R. Feldman, A. B. Olshen, L.-X. Qin, S. Patil, V. E. Reuter, G. J. Bosl, and R.S.K. Chaganti
Identification and Validation of a Gene Expression Signature That Predicts Outcome in Adult Men With Germ Cell Tumors
J. Clin. Oncol., November 1, 2009; 27(31): 5240 - 5247.
[Abstract] [Full Text] [PDF]


Home page
Genes Dev.Home page
A. P. Bracken, N. Dietrich, D. Pasini, K. H. Hansen, and K. Helin
Genome-wide mapping of Polycomb target genes unravels their roles in cell fate transitions
Genes & Dev., May 1, 2006; 20(9): 1123 - 1136.
[Abstract] [Full Text] [PDF]


Home page
Cancer Res.Home page
J. E. Korkola, J. Houldsworth, R. S.V. Chadalavada, A. B. Olshen, D. Dobrzynski, V. E. Reuter, G. J. Bosl, and R.S.K. Chaganti
Down-Regulation of Stem Cell Genes, Including Those in a 200-kb Gene Cluster at 12p13.31, Is Associated with In vivo Differentiation of Human Male Germ Cell Tumors
Cancer Res., January 15, 2006; 66(2): 820 - 827.
[Abstract] [Full Text] [PDF]


Home page
BloodHome page
J. Houldsworth, A. B. Olshen, G. Cattoretti, G. B. Donnelly, J. Teruya-Feldstein, J. Qin, N. Palanisamy, Y. Shen, K. Dyomina, M. Petlakh, et al.
Relationship between REL amplification, REL function, and clinical and biologic features in diffuse large B-cell lymphomas
Blood, March 1, 2004; 103(5): 1862 - 1868.
[Abstract] [Full Text] [PDF]


This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Houldsworth, J.
Right arrow Articles by Chaganti, R. S. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Houldsworth, J.
Right arrow Articles by Chaganti, R. S. K.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cell Growth & Differentiation