| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Biology and Genetics |
Department of Cell Biology and Cancer Center [C. M. J. v. d. M., M. X. L., D. P-H., B. C., J. L. S., A. J. v. W., G. S. S.] and Information Services [D. S. L.], University of Massachusetts Medical School, Worcester, Massachusetts 01655
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Cell cycle control during G1 phase is mediated in part by interactions between retinoblastoma-related tumor suppressor proteins (p105, p130, and p107; Ref. 1 ) and the E2 promoter binding factor (E2F) class of transcription factors (E2F1 to 6; Ref. 2 ), which are regulated by the actions of distinct cyclins (cyclins D, E, and A) and cdks (cdk4/6 and cdk2; Ref. 3, 4, 5 ). Consistent with the important role of E2F in late G1, many cell growth-regulated genes have E2F binding sites. However, genes for other cell cycle-regulated proteins whose expression is up-regulated immediately after G1 (e.g., histones H1, H2A, H2B, H3, and H4) do not have E2F sites (6) . Thus, there may be multiple gene-regulatory programs involved in the control of E2F-dependent and -independent genes during the G1 and S phases of the cell cycle.
During S-phase when the genome is replicated, newly synthesized DNA is organized into nucleosomes through interactions with histone proteins. This functional coupling is reflected by the tight linkage between the timing of DNA synthesis and histone gene expression (6 , 7) . The elevation of nucleotide pools, necessary to support DNA synthesis, and histone-dependent packaging of nascent DNA into nucleosomes are interdependent processes. These processes may very well be sequentially executed, because accumulation of histone mRNA and synthesis of histone proteins are only necessary after the appearance of nascent DNA.
To obtain a genome-wide understanding of the temporal pattern of gene expression in relation to DNA synthesis, we assessed RNA populations in cells synchronized at the G1-S-phase transition. Two principal findings of our study are that: (a) subsets of genes are successively expressed during the DNA synthesis phase of the cell cycle; (b) there is a core group of cell growth-related genes that is required for cell cycle progression in both normal and tumor cells.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Isolation of mRNA and Hybridization Analyses.
Total cellular RNA was isolated with Trizol (Life Technologies, Inc.) and with RNeasy kits (Qiagen, Valencia, CA) according to the manufacturers’ instructions. RNA was quantitated by spectrophotometry (DU640B; Beckman Coulter, Fullerton, CA), and the quality of the RNA was assessed by electrophoresis in a 1% agarose gel. Size-fractionated RNA was transferred to a Hybond-N+ nylon membrane (Amersham Pharmacia Biotech, Piscataway, NJ) by capillary transfer overnight. Using [32P]dCTP (NEN) and Rediprime II (Amersham Pharmacia Biotech), DNA fragments encoding histone H4 (H4,3–2 SK+) and GAPDH (phGAPDH) were labeled, purified, and hybridized to the RNA blots at 42°C. For microarray analyses, total RNA (5 µg) was reverse transcribed into double-stranded cDNA (SuperScript Choice system; Life Technologies, Inc.) with the T7-dT24 primer (Genset Corp., La Jolla, CA). After phenol-chloroform extraction, the cDNA was transcribed into biotinylated cRNA (Enzo, Farmingdale, NY), which was purified (RNeasy; Qiagen) and fragmented according to the manufacturer’s instructions (Affymetrix Expression Analysis Technical manual rev. 4, Santa Clara, CA). We used test chips and the HU6800 GeneChip probe arrays and followed the standard hybridization procedures with the control oligonucleotide B2 (Genset). The arrays were washed in the Affymetrix GeneChip Fluidics Station 400 and scanned using the Affymetrix GeneChip array scanner.
Bioinformatic Analyses of Gene Expression.
The signals from each microarray were analyzed, normalized, and converted to a numerical output using the Affymetrix GeneChip software. The data generated by the different arrays were globally scaled to the target average intensity of 1500. The average expression value for each gene across the arrays was used to normalize the mRNA intensities. Adjusted data were subjected to further analysis using Xcluster.4
Expression data from two HeLa cell synchronies were analyzed with Xcluster using log scaling and Pearson’s correlation coefficients to assess the similarity of expression profiles during the time course, and data were grouped using the k-means (k = 10) and self-organizing map options of Xcluster. Data were further analyzed by functional groupings, for which we used both Microsoft Excel and Microsoft Access. The HeLa cell microarray data were compared with data generated from normal human fibroblasts.5
Several criteria were used in this comparison: (a) we only included genes that exhibited a relative level of expression above a threshold (average difference value, >500); (b) expression patterns with values that fluctuated <1.5-fold during the time course were considered constitutive; and (c) to be considered similar between the two cell types, the peaks or nadirs of the gene expression pattern should occur within a timeframe of 2 h. For additional information, see our website.
|
RESULTS AND DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
. We found that the RNA expression profiles for histone H4, which is a well-established and highly specific S-phase-related parameter, give quantitatively comparable results with both methods. The distribution of cells in the cell cycle was monitored by FACS analyses (Fig. 2A)
and reveals that the fraction of cells in S-phase is proportional to histone gene expression. The FACS analyses also demonstrate that timepoints chosen reflect enrichment for cells in early S-phase (1.5 h), mid S-phase (3 h), late S-phase (4.5–6 h), G2-M phases (6–9 h), and early G1 phase (9 h).
|
|
. Analysis of the microarray data indicates that 
40% of the genes (2,846) have signal intensities that are significantly above background. On the basis of the temporal pattern of expression during this time course, each gene was assigned to a cluster by using the k-means clustering algorithm (k = 10; Fig. 2B
). The clusters were organized according to their temporal expression pattern, with lower cluster numbers assigned to groups of genes with the highest expression levels in early S-phase and higher cluster numbers assigned to groups of genes that peak in G2-M to G1 phase. We found that many clusters are enriched in genes that correspond with specific, cell cycle stage-related functions (representative genes are indicated on the right of Fig. 2B
).
The correlation between temporal clusters and functional groupings was further assessed by scoring the frequency of occurrence of distinct groups of genes in each cluster (Table 1)
. We found several groups of genes involved in, for example, signal transduction, protein synthesis, and protein degradation, that are more prevalent in certain clusters. Inspection of the temporal expression pattern for each cluster suggests that there are substages during S-phase. For example, cluster 2 represents genes that are predominantly expressed in early S-phase, whereas genes in cluster 6 are expressed at highest levels in mid S-phase (Fig. 3)
. About half of the expressed genes exhibit temporally regulated expression patterns, in contrast with genes encoding proteins such as cytoplasmic ß-actin and most proteasomal proteins, which show barely any fluctuation. Overall, a succession of gene expression occurs over time, with particular genes or groups of genes appearing at particular stages of the cell cycle phases.
|
|
, with DHFR expression increasing before the up-regulation of histone H4. This finding suggests that there are distinct substages in S-phase to accommodate biological requirements for the duplication of the genome.
Induction of Genes Involved in DNA Replication and Early Response Genes in Early S-Phase.
To further elucidate temporal transitions during S-phase, we examined the genes that are expressed at the G1-S boundary and the subsequent entry into S-phase. We addressed whether genes whose expression is up-regulated at early S-phase are involved in DNA replication and/or in reinitiating proliferation. We found an increase in transcripts for the proteins MCM6, ORC1, ORC2, and the kinases cdc6 and cdc7 (Fig. 2B
, clusters 1 and 3; Fig. 3
). These proteins all participate in the initial stages of DNA replication, which encompass recognition of the origin of replication and assembly of an initiation complex (10)
. The subsequent initiation of DNA synthesis is mediated by DNA primase and DNA polymerase 
, followed by DNA polymerase 
(11)
. The corresponding gene expression patterns for these DNA replication enzymes (shown in Fig. 4A
) are comparable with those for genes involved in origin recognition. Enzymes involved in generating the cellular supply of nucleotides are also expressed in early S-phase (e.g., DHFR in Fig. 2B
). Taken together, many enzymes and proteins involved in DNA synthesis are present in early S-phase.
|
) in the group of genes that are maximally expressed at 0–1.5 h after release (e.g., cluster 2) as well as in gene groups that are expressed at high levels in subsequent stages. We assessed which specific genes exhibit high expression in early S-phase. Genes present at release or up-regulated immediately thereafter included a number of early response genes (ATF-3, SRF, ERF-2, IEX-1, and HSP-70), transcriptional regulatory factors (SWI/SNF, ATRX, and others; Ref. 12
), cytokines (MDNCF and BSF-2), nuclear orphan receptors (LXR-, TR3, Ear-3, and MINOR), and kinases (e.g., c-fes and cdk10; Fig. 3
). Other genes that are induced in early S-phase include those encoding cell surface proteins, such as cytokine receptors (e.g., frizzled) and transport-related proteins (e.g., rab13 and TWIK; Table 1
). We also note that there is a moderate increase in metabolic proteins [e.g., cyclooxygenase (Cox-2), and acid sphingomyelinase] when the cell cycle proceeds (Fig. 2B)
. Therefore, in this early stage of S-phase, expression of members from several signal transduction pathways as well as expression of metabolic enzymes is prominent.
Several mRNAs were present at high levels at the G1-S boundary but exhibited a transient decrease in expression during S-phase. Among these were envoplakin (a cytoskeletal linker protein), the growth arrest-specific gene (gas1), the transcription coregulators Smad1 and Smad6, and BARD1 (Fig. 3)
. The temporal expression of these genes may be G1 or G1-S transition specific or correlated with the synchronization procedure. Thus, there are multiple gene-regulatory programs that are initiated in early S-phase, and only a subset directly supports the onset and progression of DNA replication.
Maximal Expression of Genes Involved in DNA Repair and Histone Genes Is Linked to Maximal DNA Synthesis.
At mid S-phase, we found a second category of genes that are maximally expressed when the DNA synthesis rate is highest, which defines this substage of S-phase. DNA replication may be coupled to mechanisms that monitor the integrity of nascent DNA. Consistent with this concept, we found the highest levels of transcripts for proteins involved in DNA repair (e.g., mismatch repair gene) when cells progress toward mid S-phase (Fig. 2B
, clusters 3, 4, and 5). The coordinate up-regulation of histone RNA levels at 4.5–6 h into S-phase (Fig. 2B
, cluster 6) accommodates the high demand for core histones (i.e., H2A, H2B, H3, and H4; Refs. 7
and 9
), which are involved in the formation of chromatin as DNA is synthesized. Preceding the up-regulation of these two classes of genes, we find high expression for several gene regulators involved in chromatin assembly [such as nuclear protein, ataxia-telangiectasia locus (NPAT) and the chromatin assembly factor p150 (CAF); Fig. 2B
, cluster 3]. Thus, when cells exhibit peak rates of DNA synthesis, we found maximal expression of transcripts for proteins that monitor and restore newly replicated DNA and chromatin.
E2F-dependent versus E2F-independent Gene Regulation.
Our studies indicate that there are temporal differences in RNA expression of proteins involved in DNA synthesis versus transcripts encoding histone proteins. The first group of transcripts is present at relatively high levels at the G1-S boundary, whereas the second group does not peak until at least 1.5 hours into S-phase (Figs. 4, A and B)
. This difference in timing is at least in part regulated through transcriptional mechanisms, as reflected by particular promoter characteristics for the different groups. The predominant regulatory elements that govern G1-S-phase-specific expression and corresponding transcription factors have been identified previously. Of these recognition sequences, the E2F site is one element critical for the passage through the R point and G1-S transition, and many genes involved in DNA replication are E2F responsive (2
, 13
, 14)
. Promoters of these genes have at least one E2F site as well as one or more auxiliary elements (e.g., Sp1 sites) but in general lack TATA and CCAAT boxes. It is remarkable that histone gene promoters almost without exception lack E2F sites. Instead, promoters of the histone gene family are controlled by TATA and CCAAT boxes, Sp1 sites, and histone subtype-specific elements (e.g., binding sites for IRF, Oct1, CDP, and HiNF P; Refs. 6
and 9
).
Distinctions in promoter organization and gene expression profiles of DNA replicative genes predict that the cognate transcription factors may also have differences in expression. The gene expression patterns of transcription factors that are either known or presumed to interact with G1-S-phase-related promoter elements (E2F and Sp1 sites, CCAAT and TATA boxes, and others) were clustered by Pearson’s correlation and analyzed with k-means (Fig. 4C)
. The factors we analyzed appear to separate into three temporally distinct groups. Three of these genes (i.e., E2F-2, an E2F-like protein, and IRF-5) have expression patterns resembling those of several DNA polymerases (Fig. 4, A and C)
. A second group of factors [i.e., Sp1, basic transcription element binding protein (BTEB2), and down-regulator of transcription 1 (DR1); Refs. 15
and 16
] is regulated in parallel with the histone gene subtypes. Hence, our data suggest that the expression of at least a subset of transcription factors present in S-phase correlates with the expression of potential downstream target genes.
Transcripts Encoding Protein-modifying Factors and Nuclear Components Are Temporally Linked with Mitosis.
After completion of S-phase, cells prepare for mitotic division, which requires condensation, separation, and redistribution of the chromosomes, as well as major turnover and/or posttranslational modification of proteins. The mitotic phase is preceded by the G2 checkpoint, which is essential to monitor correct duplication of genetic information, accurate structural formation, and competency for division and completion of the cell cycle (17)
. We found maximal expression for several transcriptional repressors as cells progressed from mid S-phase to early G1 (e.g., DR1, C/EBP
, TGIF, and CTCF; Figs. 3
and 4C
), which may be related to the dramatic decrease in RNA synthesis that occurs during mitosis (18)
. We found that transcripts encoding proteins involved in ubiquitination and degradation are present at later stages of the cell cycle (Table 1
, cluster 7). For example, maximal transcript levels for several genes encoding proteins involved in protein degradation, such as small SUMO and the cullin proteins Hs-cul-1 and Hs-cul-4A, were found at 4.5–6 h after release (Fig. 2B)
. The increased expression of these protein-inactivating factors indicates that cells have completed S-phase and are preparing for cell division. Similarly, other enzymes mediating protein modification are predominantly present in the same temporal cluster (Table 1
, cluster 7; e.g., phosphatases and kinases). RNA encoding proteins known to be involved in mitotic regulation or mitosis are successively expressed with maximal expression at mid S-phase [e.g., kinesin-like spindle protein (HKSP) and ß-tubulin], late S-phase (e.g., cyclin F and PIN1), or the G2-M phases (e.g., p55cdc and cyclin B; Figs. 2B
and 3
). We conclude that several division-specific proteins show an increase in RNA expression during the mitotic period.
The redistribution of chromatin and synthesis of the nuclear envelope in the daughter cells mark the final stage of mitosis and beginning of G1 (20)
. During these phases, we observed a striking rise in nuclear proteins (Table 1
, clusters 8 and 9). Representative genes expressed during this transition encoded proteins involved in nuclear trafficking (e.g., RCC1, RanBP1 and 2, importin ß, and nucleopore protein), nuclear structure (e.g., the nuclear matrix protein matrin 3 and lamins A and B), and nucleolar organization (e.g., the proliferating cell nucleolar antigen; Fig. 3
). We found a significant increase in the majority of RNA-processing factors and translation initiation factors in early G1. For example, we found that most heterogeneous nuclear ribonucleoproteins, small nuclear ribonucleoprotein particles, the nuclear cap binding protein, and translation initiation factors show peak expression at 9 h (Fig. 2B)
. In summary, our gene profiling studies have revealed the sequential expression during S-phase of RNA encoding proteins such as growth signal-responsive proteins, proteins involved in DNA synthesis, and proteins involved in chromatin assembly (e.g., histones), followed by maximal expression in the later cell cycle phases of proteins involved in cell division and nuclear structure (Fig. 5)
.
|
. In early S-phase, we found high expression of cytokine receptors involved in tumor proliferation (21
, 22)
, as well as a number of oncoproteins (23, 24, 25, 26)
. These expression patterns parallel the increased expression of transcripts encoding other cell signaling-related proteins that are maximally expressed in early S-phase (see Fig. 3
). At mid to late S-phase, we observed peak expression for several genes encoding proteins with mitogenic properties that are involved in metastasis (27)
, maintenance of the transformed phenotype (28)
, or induction of proliferation (29
, 30)
. We also found increased expression of genes encoding several tumor suppressor proteins or proteins with antiproliferative properties. Thus, although many proteins with oncogenic properties are expressed from mid to late S-phase, a similar peak of expression occurs for numerous growth inhibitors and tumor suppressors (31, 32, 33, 34, 35, 36, 37, 38)
. From late S-phase to G2-M, we found enhanced expression of several genes encoding proteins involved in calcium and inositol signaling. Moreover, some of these proteins have also been associated with negative effects on cell growth and/or involvement in apoptosis (39, 40, 41)
. Hence, we have characterized expression of signal transduction molecules in early S-phase, proteins involved in cell growth and tumor-related proteins from mid to late S-phase, and potential signal transduction attenuators during G2-M.
|
Comparison of Normal and Tumor Cells Revealed a Core Group of Cell Growth-related Genes.
Gene expression analysis for synchronized normal fibroblasts has been presented (19)
using the same type of microarrays. To evaluate the similarities and differences in gene expression in normal and tumor cells, we analyzed the global expression data from the two studies. The results of this comparison reveal that two-thirds (i.e., 1,837) of the total number of genes expressed in HeLa cells are also expressed in normal fibroblasts and vice versa. A large number of genes exhibit higher expression levels in the cancer cell and often lose their cell cycle-related pattern. This group mostly included transcription factors, structural proteins, and several proteins involved in the process of chromosomal segregation. We scored the occurrence of genes in several functional groups and compiled the results (Table 2)
. Two groups (Receptors and Transport) include many genes that exhibit cell type-specific expression and may represent phenotypic markers. In contrast, genes in five groups (i.e., Cell cycle, DNA replication, DNA repair, RNA/protein synthesis, and Degradation) are expressed in both normal and tumor cells. Thus, despite significant phenotypic differences in the cell types, there is a striking conservation in the expression of multiple functional classes of genes. For additional information, see our website.
|
In contrast to these genes that are not expressed, the group of cell cycle-regulated genes identified in the diploid fibroblast cells contains 141 (64%) genes that were also expressed in the synchronized HeLa cells. Table 3
provides an overview of the similarities and differences in the expression patterns of cell cycle-specific genes identified in the two studies. The data reveal that the expression patterns for many genes (45 of 141) are different, and most of these genes are constitutively expressed in the tumor cell line (for example, transcription factors and genes involved in mitosis; see Fig. 7
). Only a limited number of genes exhibited temporal expression patterns in the cancer cells that were fluctuating yet different from those in the normal cell type. Among these genes were the tumor repressor p21, the transcription factor serum response factor, the protein spectrin (involved in cell shape), and metabolic proteins such as ATP synthase (Table 3
and Fig. 7
). Thus, although a number of genes show alterations in the temporal regulation of RNA levels, the expression of the majority of this subgroup of genes is no longer cell cycle regulated.
|
|
|
|
and Fig. 7
). Hence, a relatively small fraction (37%) of the genes that are cell cycle regulated in the diploid fibroblast cell type are also expressed and similarly modulated during the tumor cell cycle. These genes may constitute a core set of regulatory genes fundamental to cell cycle progression. In conclusion, in this study we have characterized the sequential expression of genes that regulate or mediate genome replication. We identified groups of genes that are maximally expressed at specific substages during S-phase and G2. These genes encode cell growth and/or cancer-related proteins that may play specific roles at critical cell cycle transitions. We propose that the classically defined S and G2 phases of the cell cycle can be subdivided into distinct substages based on diagnostic gene expression profiles. The second principal finding that emerged from our studies is the identification of a core set of genes that is similarly regulated in both normal and tumor cells and may support the fundamental mechanisms required for cell cycle progression.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This work was supported by NIH Grant GM32010. 
2 To whom requests for reprints should be addressed, at Department of Cell Biology, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA 01655.
3 The abbreviations used are: cdk, cyclin-dependent kinase; FACS, fluorescence-activated cell sorter; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; DHFR, dihydrofolate reductase.
4 Internet address: http://genome-www.stanford.edu/sherlock/cluster.html.
5 Internet address: http://www.salk.edu/docs/labs/chipdata/.
6 Internet address: http://www.salk.edu/docs/labs/chipdata/primary.html.
7 For additional information, see our web site: http://www.umassmed.edu/cellbio/labs/steinlab.
Received 1/ 2/02. Accepted 4/ 2/02.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
-primase complex couples DNA replication, cell-cycle progression and DNA-damage response. Trends Biochem. Sci., 22: 424-427, 1997.[Medline]
signaling pathway. J. Biol. Chem., 276: 9640-9648, 2001.This article has been cited by other articles:
|
|
 |
|
  K. Mochizuki, A. Nishiyama, M. K. Jang, A. Dey, A. Ghosh, T. Tamura, H. Natsume, H. Yao, and K. Ozato The Bromodomain Protein Brd4 Stimulates G1 Gene Transcription and Promotes Progression to S Phase J. Biol. Chem., April 4, 2008; 283(14): 9040 - 9048. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. W. Young, M. Q. Hassan, X.-Q. Yang, M. Galindo, A. Javed, S. K. Zaidi, P. Furcinitti, D. Lapointe, M. Montecino, J. B. Lian, et al. Mitotic retention of gene expression patterns by the cell fate-determining transcription factor Runx2 PNAS, February 27, 2007; 104(9): 3189 - 3194. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. E. Polo, S. E. Theocharis, J. Klijanienko, A. Savignoni, B. Asselain, P. Vielh, and G. Almouzni Chromatin Assembly Factor-1, a Marker of Clinical Value to Distinguish Quiescent from Proliferating Cells Cancer Res., April 1, 2004; 64(7): 2371 - 2381. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Wang, C. Shiels, P. Sasieni, P. J. Wu, S. A. Islam, P. S. Freemont, and D. Sheer Promyelocytic leukemia nuclear bodies associate with transcriptionally active genomic regions J. Cell Biol., February 16, 2004; 164(4): 515 - 526. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Zhao, F. Granberg, L. Elfineh, U. Pettersson, and C. Svensson Strategic Attack on Host Cell Gene Expression during Adenovirus Infection J. Virol., October 15, 2003; 77(20): 11006 - 11015. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Pratap, M. Galindo, S. K. Zaidi, D. Vradii, B. M. Bhat, J. A. Robinson, J.-Y. Choi, T. Komori, J. L. Stein, J. B. Lian, et al. Cell Growth Regulatory Role of Runx2 during Proliferative Expansion of Preosteoblasts Cancer Res., September 1, 2003; 63(17): 5357 - 5362. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Menges, L. Hennig, W. Gruissem, and J. A. H. Murray Cell Cycle-regulated Gene Expression in Arabidopsis J. Biol. Chem., October 25, 2002; 277(44): 41987 - 42002. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK< |
] representing samples generated from log-phase cells. Numbers on the left indicate different clusters. Shown on the right are particular genes found in different clusters. The yellow color indicates a high level of expression, black represents no change, and blue indicates the lowest expression as shown on the bottom. The transition of yellow from the left top to the right bottom of the bin reflects succession in maximal expression levels.;datamatics;(v)
