| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Biochemistry, University of Texas Health Science Center, San Antonio, Texas 78229-3900
Address all correspondence and requests for reprints to: Dr. Martin L. Adamo, Department of Biochemistry, MSC7760, University of Texas Health Science Center, 7703 Floyd Curl Drive, San Antonio, Texas 78229-3900. E-mail: adamo{at}biochem.uthscsa.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
One variable that could mediate autonomous cellular regulation of gene expression is cell density. It is now established that cell density can modulate the level of messenger RNA (mRNA) expression in a variety of systems, including the myelin gene in immortalized Schwann cells (3), the fibronectin gene in NIH-3T3 cells (4), and the phosphoenolpyruvate carboxykinase (PEPCK) gene in H4IIEC3 rat hepatoma cells (5). Cell-cell contact is also involved in the regulation of gene expression. It was reported that glucocorticoid receptor transcriptional activity was regulated by glial/neuronal interaction (6), which may be involved in the cell contact-dependent glucocorticoid induction of glutamine synthetase in Müller glial cells (7). Transcription of the fibroblast matrix metalloproteinase-9 gene in primary rat fibroblasts was activated when they were plated on a monolayer of 2.8 cells, a transformed rat embryo cell line (8).
There have been several reports that IGF system components are regulated by cell density. The IGF-I mRNA level and secretion decreased markedly as a function of cellular confluence in rat vascular smooth muscle cells (9, 10). However, the IGF-I peptide level per unit of cellular protein in conditioned medium (CM) was higher when hepatocytes were plated at increasing cell density (11). In the C6 cell line, the IGF-I mRNA level was higher in confluent than in rapidly growing cultures (12). The IGF-I receptor number was reported to be higher in sparse human fibroblast cultures than in dense cultures (13). Other components of the IGF system, such as IGF-II and IGF-binding proteins, are also regulated by cell density (14, 15, 16, 17, 18, 19). All of these observations suggest that cell density is a very important regulator of IGF system gene expression.
IGF-I action is also regulated by cell density in a cell type-specific manner (20, 21, 22, 23). It was reported that IGF-I has a higher stimulatory effect on DNA synthesis than epidermal growth factor, fibroblast growth factors, and cholera toxin at high cell density in proximal tubule epithelial cells (24). Thus, characterization of the mechanisms by which cell density influences IGF-I gene expression will be useful for understanding the physiological functions of IGF-I in cell culture models.
In this study we have shown that cell density has a cell type- and gene type-specific effect on IGF-I gene expression. We demonstrate that the induction of IGF-I mRNA in C6 cells by cell density is at least partially due to transcriptional activation. Furthermore, we have begun to localize the putative cell density response element(s) in the IGF-I proximal exon 1 promoter region. Our data also suggest that the stimulation of IGF-I mRNA in C6 cells by cell density is not due to the release of soluble peptidic factors into conditioned medium.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Solution hybridization/RNase protection assays
A pGEM4Z plasmid containing a 464-bp rat IGF-I gene fragment was
linearized and used to synthesize a 32P-labeled
antisense RNA probe (26). In ribonuclease protection assays (RPAs),
hybridization of this probe to exon 1 containing IGF-I mRNAs results in
a protected band of 238 bases, whereas hybridization to exon 2
containing mRNAs results in bands of 290 bases and a doublet of
approximately 305–309 bases. These bands result from hybridization to
mRNAs transcribed from the different exon 2 start sites (26). To detect
rat IGF-I receptor mRNA, a 32P-labeled antisense
RNA probe was synthesized from a linearized pGEM3 plasmid containing a
portion of rat IGF-I receptor complementary DNA (cDNA) (27).
Hybridization of this probe to rat RNA resulted in a protected band of
265 bases.
A pGEM3Z plasmid containing a human IGF-I cDNA was linearized and used to synthesize a 32P-labeled antisense RNA probe. Hybridization of this probe to human RNA resulted in a band of 329 bases corresponding to exon 2-containing mRNA and a band of 418 bases corresponding to exon 1-containing mRNA (28). To detect human IGF-I receptor mRNA, a 394-base 32P-labeled antisense probe was used, and a 379-base protected band was detected by RPA (29). Both human IGF-I and IGF-I receptor plasmids were gifts from Dr. Charles Roberts, Oregon Health Sciences University (Portland, OR). Linearized plasmids containing rat or human ß-actin cDNA insert were obtained from Ambion, Inc. (Austin, TX) and were used to synthesize rat or human ß-actin antisense RNA probes. The rat ß-actin antisense probe resulted in a 126-base protected band on RPA, whereas the human ß-actin antisense probe resulted in a 245-base protected band on RPA.
Antisense RNA probes were synthesized using either the MaxiScript kit from Ambion, Inc., or using the protocol described by Adamo et al. (25), with reagents from Promega Corp. (Madison, WI) and Ambion, Inc. Solution hybridization/RPAs were conducted using either the RPA II kit from Ambion, Inc., or the protocol described by Adamo et al. (25), with reagents supplied by Ambion, Inc. The gels were dried, autoradiographed, and then subjected to phosphorimage analysis of protected band intensity using the phosphorimage system from Molecular Dynamics, Inc. (Sunnyvale, CA).
Promoter constructs and transient transfections
The following constructs were used in transient transfections:
pGL2-Basic (promoterless luciferase vector; Promega Corp.), pGL2-Control (simian virus 40 enhancer/promoter
luciferase vector; Promega Corp.), and rat IGF-I exon 1
promoter fragments -1500/+319, -1000/+319, -500/+319, -500/+282,
-500/+192, -250/+319, -250/+282, -250/+192, +75/+319, +75/+282, and
+75/+192 cloned into pGL2-Basic. All of the IGF-I promoter fragments
were generated by PCR amplification of these regions using a sense
primer with a KpnI restriction site and an antisense primer
with a HindIII restriction site, as described in Wang
et al. (30). The PCR products were ligated into the same
sites of pGL2-Basic.
Transient transfection was performed using the Lipofectamine Plus system in Opti-MEM medium (Life Technologies, Inc./BRL, Gaithersburg, MD). Three hours after transfection, Opti-MEM medium was replaced with Ham’s F-12 medium containing 0.1% BSA in place of 10% FBS. Twenty-four hours later, lysates were prepared and assayed for luciferase enzyme activity using reagents and protocol supplied by Promega Corp., with chemiluminescence measurements performed on a model IL-A911 semiautomatic luminometer from Tropix (Bedford, MA). Protein concentration was assayed on the lysates using the method of Bradford (31).
Statistical analysis
Data in Figures 5
and 6 are the mean + SEM for
three separate experiments. Statistical differences between means were
determined using one-way ANOVA in the SIMSTAT3 package (Normand
Peladeau, Provalis Research, Montréal, Canada).
|
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
), at which time
cell density was close to 4 million cells/60-mm plate. C6 cells express
mostly exon 1 containing IGF-I mRNA transcripts (Fig. 1A). IGF-I mRNA
was increased by over 100-fold when the cells reached confluence and
increased another 2- to 3-fold as cell density increased from 4 million
to 6 million cells/60-mm plate (Fig. 1, A and B). Then, although the
cell number was still increasing, IGF-I mRNA remained constant (Fig. 1, A and B). The levels of IGF-I receptor and ß-actin mRNA also
increased under these conditions. However, the inductions were much
lower than for IGF-I mRNA. Furthermore, the levels of 28S and 18S
ribosomal RNA, as measured by ethidium bromide staining of
agarose/formaldehyde gels of equal microgram amounts of total RNA, were
not altered by cell density (data not shown). The experiments were
repeated three times, and similar inductions were observed (data not
shown). These results indicate that cell density has a gene-specific
influence on gene expression in C6 cells, particularly regulating IGF-I
mRNA levels.
|
).
IGF-I mRNA that was transcribed from the exon 1 promoter increased
about 5-fold, and IGF-I mRNA that was transcribed from exon 2 promoter
increased about 4-fold when cells grew from days 1–4 (Fig. 2, A and
B). After day 4, IGF-I mRNA levels were not changed with increasing
cell number. IGF-I receptor mRNA transcripts also increased about
2-fold under these conditions. Interestingly, ß-actin mRNA
transcripts were induced about 2.2-fold from days 1–4, but then
slightly decreased on the following days.
|
, both exon 1-
and exon 2-containing IGF-I mRNAs were present in SK-N-MC cells,
although the majority of IGF-I transcripts contained exon 1. Both mRNAs
were stimulated about 4-fold in the first 4 days of growth and slightly
increased during the following days of culture. Cellular confluence was
reached on day 7 (Fig. 3C). IGF-I receptor mRNA was also induced by
cell density, but to a lesser extent (Fig. 3, A and B). Similar to that
in GH3 cells, the ß-actin mRNA level increased
about 2.7-fold from days 1–5, but then slightly decreased on the
following days (Fig. 3, A and B).
|
). In this study
only the exon 2-containing IGF-I mRNA level was quantified, as shown in
Fig. 4B. IGF-I, IGF-I receptor, and ß-actin mRNA levels were all
relatively constant as cells became more and more confluent (Fig. 4).
For GH3, SK-N-MC, and OVCAR-3 cells, experiments
were repeated three times, and similar results were obtained (data not
shown). In all of these experiments 18S and 28S ribosomal RNA levels
were equivalent for equal amount of total RNA from the sparse and dense
cultures (data not shown). These results suggest that in
GH3 cells, SK-N-MC cells, and OVCAR-3 cells, the
regulation of IGF-I mRNA expression by cell density is distinct from
that in C6 cells, which indicates that cell density has a cell
type-specific effect on IGF-I gene expression.
|
). This
result was used as a control for the specific effect of cell density on
the IGF-I promoter.
The construct containing 1.5 kb of 5'-flanking region and 319 bp of
exon 1 sequence stimulated luciferase activity by 2.7-fold over
pGL2-Basic in low density cultures and by 4.9-fold over pGL2-Basic in
high density cultures, i.e. a 1.8-fold stimulation by cell
density (Fig. 5A; P < 0.05). 5' Deletions to -1000,
-500, and -250 led to increased promoter activity in low density
cultures. Further deletion to +75 resulted in a slight decrease in
promoter activity in low density cultures. Cell density had a
significant stimulatory effect on the activities of all of these
promoter constructs (Fig. 5A; P < 0.05).
Therefore, 3' deletion was also conducted. The +75/+282 construct
stimulated luciferase activity by 20-fold over pGL2-Basic in low
density cultures and by 51-fold over pGL2-Basic in high density
cultures, i.e. a 2.5-fold stimulation by cell density (Fig. 5A; P < 0.001). These data suggest that a cell density
response element(s) is probably located between +75 and +282 of the
IGF-I exon 1 promoter region.
To further determine the location of the cell density response element,
constructs -500/+282, -500/+192, -250/+282, and -250/+192 were
tested in C6 cells. All four constructs stimulated significant promoter
activity over pGL2-Basic in both low density and high density cultures
(Fig. 5B; P < 0.001). The constructs -500/+282 and
-250/+282 stimulated luciferase activity by 13- and 4.4-fold,
respectively, over pGL2-Basic in low density cultures and by 29- and
18-fold over pGL2-Basic in high density cultures, i.e. 2-
and 4.2-fold stimulations by cell density, respectively (Fig. 5B;
P < 0.001). However, 3' deletion to +192 led to loss
of the stimulatory effect of cell density (Fig. 5B). These data
suggested that the +192 to +282 region of the proximal exon 1 promoter
may contain a cell density response element(s). Parallel plates,
transfected with the pGL2-Basic plasmid, were harvested for total RNA
extraction. A 4- to 5-fold induction of IGF-I mRNA transcripts by
cellular confluence was observed in these experiments (data not
shown).
As the +192 to +282 region may mediate the cell density response in C6
cells, it was important to determine whether this region is also
effective in other cell lines in which IGF-I mRNA was induced by cell
density. IGF-I exon 1 promoter constructs -250/+282 and -250/+192,
which were responsive or unresponsive, respectively, to cell density in
C6 cells, were transfected into SK-N-MC cells. The promoter activities
of these constructs were not significantly different between the dense
and sparse cultures (Fig. 5C; P > 0.05). Under the
same conditions, endogenous IGF-I mRNA from parallel plates,
transfected with pGL2-Basic plasmid, was 2-fold higher in the dense
cultures than in the sparse cultures (data not shown).
Soluble peptidic factor(s) plays a minor role in the induction of
IGF-I mRNA by cell density
To understand whether soluble factors released into CM are
involved in the induction of IGF-I mRNA in C6 cells, CM of sparse
cultures was exchanged with that of dense cultures in the presence of a
cocktail of protease inhibitors, as described in Fig. 6. Control plates in which CM was not
exchanged were also treated with the same concentrations of protease
inhibitors. As shown in Fig. 6, IGF-I mRNA levels were not altered
after the medium exchange (P > 0.05). This indicates
that soluble peptidic factors in CM probably do not mediate the
induction of IGF-I mRNA by cell density in C6 cells.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In contrast to the results we obtained from these tumor cell lines, Giannella-Neto et al. reported that IGF-I mRNA transcripts decreased markedly as a function of cellular confluence in vascular smooth muscle cells (10), which is consistent with the rationale that in contact-inhibited cells, expression of the autocrine mitogen IGF-I should be down-regulated as cells grow from sparse cultures to confluent cultures. However, the IGF-I peptide level was higher in dense culture of primary hepatocytes (11). IGF-I produced by hepatocytes is the major source of circulating IGF-I, and IGF-I does not function as a hepatic growth factor, as mature hepatocytes lack IGF-I receptor (reviewed in Ref. 2). The expression level of IGF-I in adult liver is greatest among the tissues. Therefore, the regulation of IGF-I in response to cellular confluence in hepatocytes may be different from that in other primary cell cultures.
Expression of basic fibroblast growth factor is inhibited by cell contact in human astrocytes (35). In contrast, tumorigenic glioma cells that are no longer susceptible to contact inhibition of growth express basic fibroblast growth factor constitutively. All four tumor cell lines used in this study continued to proliferate even after they reached confluence. This lack of contact inhibition may result in the high level of IGF-I mRNA expression in the confluent cultures of these tumor cells, which may increase the tumorigenic capacity of these cell lines. In support of this idea is the observation that inhibition of IGF-I expression using an antisense RNA technique in C6 cells prevented tumorigenesis when the cells were injected into nude mice (36).
The low level of IGF-I mRNA in rapidly growing C6 cells suggests that IGF-I may not mainly act as a mitogen during the early growth of these cells. IGF-I is known to be capable of preventing apoptosis in a number of cell types (37, 38, 39). IGF-I receptor plays a dominant role in inhibiting programmed cell death or apoptosis (40). Thus, it is possible that autocrine IGF-I is primarily a survival factor, rather than a mitogen, in C6 cells. In GH3, SK-N-MC, and OVCAR-3 cells, IGF-I may act as both a mitogen and a survival factor. However, as we did not examine IGF-I peptide production in this study, we can only speculate that the expression pattern of the IGF-I peptide level regulated by cell density would follow the profile of IGF-I mRNA. Studies are currently underway to determine whether IGF-I acts in an autocrine manner to regulate cellular proliferation and survival in these tumor cells.
The levels of mRNA encoded by other cell density-regulated genes have been reported to be induced 2- to 20-fold by increasing cell density (3, 4, 5, 18). In our study the induction of IGF-I mRNA transcripts in C6 cells in response to cellular confluence is more than 200-fold, which is much higher than the cell density induction of IGF-I mRNA in C6 cells reported by Straus and Burke (12). This may be due to the different ranges of cell density used by the two groups. Nonetheless, the present study suggests that IGF-I gene expression is one of the most sensitive to the influence of cell density. Due to the low transfection efficiency in very dense cultures and inaccurate measurement of specific luciferase activity in very sparse cultures, we cannot use the extreme conditions of cell density that were used in the assays of endogenous mRNA expression during cell growth for the transient transfection assays. Under the conditions used for the transient transfection experiments in this study, the endogenous IGF-I mRNA was 4- to 5-fold higher in the dense cultures compared with that in the sparse cultures, whereas IGF-I promoter activity was increased 2- to 4-fold. Therefore, the majority of the induction of IGF-I mRNA is comparable to the increase in promoter activity, suggesting that transcription is probably a major mode of regulation. However, the possible effect of cell density on IGF-I mRNA stability cannot be excluded. It is technically difficult to assess whether cellular confluence affects the IGF-I mRNA half-life because it would be necessary to perform a time-course experiment to study mRNA decay during a time period in which cell density would change. We do not know the mechanism by which increasing cell density stimulates both exon 1- and exon 2-containing IGF-I mRNAs in GH3 and SK-N-MC cells. We showed that the IGF-I exon 1 promoter sequence that was responsive to cell density in C6 cells was not responsive in SK-N-MC cells. Therefore, this sequence does not contain a cell density response region in SK-N-MC cells, making it unlikely to be responsible for the increases in both exon 1- and exon 2-containing mRNAs in SK-N-MC cells.
In C6 cells, deletion of the +192 to +282 region reduced IGF-I exon 1
promoter activity in both high density and low density cultures, which
suggests that this region is required for the maximum promoter
activity. The promoter region from +192 to +282 includes the 3'-end of
region III, all of region IV, and the 5'-end of region V, which are
detected by deoxyribonuclease I footprinting using rat liver nuclear
extract (41). Regions III and V are metabolically regulated during
insulinopenic diabetes. The 5'-end of the promoter region between +192
and +282 also includes a C/EBP
-binding site, which is involved in
the stimulation of IGF-I promoter in a cAMP-dependent pathway in rat
osteoblasts (42). Parts of rat liver footprints HS3D and HS3E (43) also
overlap with the +192 to +282 region. Our transient transfection
experiments with cells of different densities suggest that the +192 to
+282 region may also contain a putative cell density response
element(s) in C6 cells. Therefore, this region probably serves as an
anchor for the binding of multiple basic and regulatory transcription
factors.
We are also interested in the signal that stimulates IGF-I mRNA transcripts as C6 cells proliferate to confluence. This signal may be produced by a soluble factor(s) released into CM as cell number increases. However, in the presence of a cocktail of protease inhibitors, exchange of CM from low density and high density cultures had a minor effect on the expression level of IGF-I mRNA. This observation argues against the hypothesis that a soluble peptidic factor(s) in CM plays an important role in this event. Therefore, it is more likely that cell-cell and/or cell-matrix contacts are responsible for the induction of IGF-I mRNA transcripts in C6 cells. However, the possibility that other labile factors, e.g. lipids, in the CM were degraded or inactivated before they were able to stimulate IGF-I should not be excluded. In this study we showed that IGF-I mRNA transcripts were much less inducible in GH3 and SK-N-MC cells and were relatively constant in OVCAR-3 cells during cell growth. These data suggest that the signaling pathway(s) involved in the regulation of IGF-I mRNA by cell density in C6 cells may be altered in GH3, SK-N-MC, and OVCAR-3 cells. This alteration may represent a cell type-specific regulation of IGF-I gene expression.
In conclusion, our results demonstrate that cell density has a cell type- and gene type-specific effect on the expression of IGF-I mRNA transcripts. This effect is at least partially due to transcriptional activation. The IGF-I exon 1 proximal promoter region from +192 to +282 may contain the cell density response element(s). A soluble peptidic factor(s) in CM plays a minor role in this induction. Identification of the signaling pathway(s) from cell surface to nucleus, which is involved in this induction, will illustrate a new aspect of how the IGF-I gene is regulated in a cell type-specific manner and may be a model for regulation of IGF-I gene expression during development, remodeling programs, and tumorigenesis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received November 3, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
activates insulin-like
growth factor-I gene transcription in osteoblasts. Identification of a
novel cyclic AMP signaling pathway in bone. J Biol Chem 272:31793–31800This article has been cited by other articles:
 |
|
 |
|
  S. Gao, M. McGarry, T. Ferrier, B. Pallante, B. Gasparrini, J. Fletcher, L. Harkness, P. De Sousa, J. McWhir, and I. Wilmut Effect of Cell Confluence on Production of Cloned Mice Using an Inbred Embryonic Stem Cell Line Biol Reprod, February 1, 2003; 68(2): 595 - 603. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Chacko and M. L. Adamo Double-Stranded RNA Decreases IGF-I Gene Expression in a Protein Kinase R-Dependent, but Type I Interferon-Independent, Mechanism in C6 Rat Glioma Cells Endocrinology, February 1, 2002; 143(2): 525 - 534. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Wang, X. Ma, L.-C. C. Yeh, and M. L. Adamo Differential Regulation of IGF-Binding Protein Gene Expression by cAMP in Rat C6 Glioma Cells Endocrinology, September 1, 2001; 142(9): 3917 - 3925. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Wang and M. L. Adamo Cyclic Adenosine 3',5'-Monophosphate Inhibits Insulin-Like Growth Factor I Gene Expression in Rat Glioma Cell Lines: Evidence for Regulation of Transcription and Messenger Ribonucleic Acid Stability Endocrinology, July 1, 2001; 142(7): 3041 - 3050. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. S. Chacko and M. L. Adamo Double-Stranded Ribonucleic Acid Decreases C6 Rat Glioma Cell Numbers: Effects on Insulin-Like Growth Factor I Gene Expression and Action Endocrinology, October 1, 2000; 141(10): 3546 - 3555. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
represents 16F-I mRNA level, 
represents 16F-I
receptor mRNA level, and 
represents ß-actin mRNA level. C,
Representative photographs of parallel plates. The number under
the right corner of each photograph is the cell number
(millions) counted in that plate.
