| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Endocrine Sciences Research Group, Department of Medicine, University of Manchester (N.T., W.R.R., J.R.E.D.), M13 9PT Manchester; School of Biological Sciences, University of Liverpool (M.R.H.W., C.D.W.), L69 7ZB Liverpool, United Kingdom
Address all correspondence and requests for reprints to: Julian R. E. Davis, Endocrine Sciences Research Group, Department of Medicine, Stopford Building, University of Manchester, M13 9PT, United Kingdom. E-mail: julian.davis{at}man.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Rat pituitary GH3 cells were transfected by lipofection with a luciferase reporter gene linked to 5000 bp from the human PRL gene 5'-flanking region. A series of stably transfected cell clones were generated, and one of these was chosen for detailed study on the basis of appropriate regulation of high-level luciferase expression by a series of known stimuli including TRH, forskolin, the calcium channel agonist Bay K8644, and basic fibroblast growth factor (bFGF). These cells were subjected to direct imaging of luciferase activity using a Hamamatsu photon-counting camera linked to a Zeiss Axiovert microscope with an Argus-50 image processor. Cells were exposed to 1 mM luciferin, and images were integrated over 30-min periods for up to 72 h. The total photon count over a given field settled to steady levels within 10 h and then remained constant for over 55 h. Addition of forskolin, TRH, or bFGF increased the total photon count of fields of 20–100 cells by 2- to 4-fold consistent with previous data from transient expression assays using the human PRL promoter. Individual cells, on the other hand, showed marked marked temporal and spatial heterogeneity and variability of luciferase expression when studied at 3-h intervals. Unstimulated cells showed variable luciferase expression with up to 40-fold excursions in photon counts per single cell area within 12-h periods. Stimulation of cells with either TRH, forskolin, or bFGF resulted in smooth increases in photon output over fields of 20–100 cells, but again individual cell responses differed widely, with some cells showing slow progressive rises in photon output, others showing phasic or transient responses, and yet others showing no response.
In conclusion, we found a surprising degree of heterogeneity and temporal variability in the level of gene expression in individual living pituitary tumor cells over long periods of time, with markedly divergent responses to hormonal or intracellular stimulation. The use of stably transfected clonal cell lines with extended periods of reporter gene imaging offers a valuable insight into control of gene expression in living cells in real time.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Clonal stable cell lines containing the luciferase gene under the
control of either viral (2) or eukaryotic (3, 4) promoters produce
readily detectable amounts of light using such CCD cameras, and the
relatively short half-life of luciferase activity (
1 h) makes it a
suitable reporter gene for dynamic studies of the kinetics of gene
expression in living cells. These recent studies indicated a surprising
degree of heterogeneity in both the basal level of reporter gene
activation and also the responsiveness to activators and inhibitors of
gene expression (2). Previous data on pituitary GC cells stably
transfected with the luciferase reporter gene under the control of the
GH promoter indicate that approximately 60% of the cells demonstrated
activation of the GH promoter, implying that some cells were
transcriptionally silent at given times (3). Similar variability in
reporter gene expression was found in microinjected pituitary
lactotroph cells over 24-h periods (5).
The human PRL (hPRL) gene has been well characterized, and comprises two independent promoter regions, a proximal 5000-bp region directing pituitary-specific expression (6), and a separate upstream region directing extrapituitary expression (7). The hPRL pituitary-specific regulatory region extends over 5000 bp upstream from the transcriptional start site, and is responsible for multihormonal regulation of the gene (6), similar to that shown for the rat PRL promoter. We previously confirmed the involvement of the pituitary transcription factor Pit-1, and in particular we characterized its role in hPRL activation by TRH, epidermal growth factor, calcium, and cAMP (6, 8, 9), using rat pituitary GH3 cells in which signaling systems have been well delineated.
In this study, we utilized the well-characterized lactotroph system to study the regulation of hPRL promoter activity in living individual cells over extended periods. We used repeated imaging of luciferase expression at 3-h intervals to demonstrate dramatic heterogeneity and short-term variability of expression in both resting and stimulated cells.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Transfection and cell line selection
Rat pituitary GH3 cells were transfected by lipofection (Tfx-50,
Promega) with the hPRL-5000-Luc+ construct together with a pSV2-neo
vector (Clontech, Palo Alto, CA), and maintained in DMEM (ICN, Costa
Mesa, CA) supplemented with 10% FCS. Stable transfectant clones were
selected and recloned using G418 (500 µg/ml; GIBCO-BRL, Paisley,
UK).
A series of cloned cell lines were then tested to ensure adequate level of resting (unstimulated) luciferase activity and appropriate hormonal inducibility, because this may differ according to site of integration of the transcriptional unit within the cell’s chromosomes. Cells were treated with 300 nM Bayer K8644 (Calbiochem, Nottingham, UK), 1000 nM TRH (Sigma, Poole, Dorset, UK), or 5 µM forskolin (Sigma) for 2–8 h, and cell lysates tested for luciferase activity using a Berthold Lumat LB 9501 luminometer (St. Albans, UK). One of the cell lines (termed phPRL-5000-Luc+D44, hereafter D44) was selected for further detailed study on the basis of reproducible 2- to 5-fold inductions in luciferase activity by each of the above stimuli, similar to the results seen using transient expression assays (data not shown). Basal and inducible luciferase activity in this line remained stable over 12 weeks in continuous cell culture.
Single cell imaging
Continuously cultured D44 cells were then subjected to direct
imaging of luciferase activity. The cells were seeded on 35-mm dishes
with coverslips attached at the base, and cultured for 1 or 2 days
before imaging. The dishes were set on the 37 C heated stage of the
microscope (Zeiss Axiovert-135TV, Vehuyn, Garden City, UK) in 5%
CO2/95% air in a completely dark room. Images were
obtained using either a 10x, 0.5NA dry, or 40x, 1.3NA oil immersion
objective. Brightfield images were obtained using differential
interference contrast (DIC). Beetle luciferin (potassium salt; Promega)
was added to the cells (final concentration, 1 mM). Three
minutes after adding luciferin, the first image (0 h) was captured
using a photon-counting CCD camera (Hamamatsu Photonics, Enfield, UK).
All images were integrated for 30 min. Images were taken sequentially
every 3 h unless otherwise indicated. Both slice and
center-of-gravity images were captured. The slice images were used for
image display, whereas the center-of-gravity images were used for
quantitative analysis. Images were processed using an Argus 50 image
processor (Hamamatsu Photonics). Luminescence images were generated
using a smoothing filter in the Hamamatsu Argus-50 image processing
software, with an offset of -1 counts, and DIC images were generated
using a sharpening filter.
Quantitation of photon count data
Quantitative data were obtained using the area analysis
functions in the Argus-50 software. For the data shown in Fig. 2
, circular areas of 10 pixel diameter, i.e. 79 pixels, were
placed over the position of single cells, and the counts recorded
throughout the image series. The average background for an area of this
size was 0.76 counts per area per 30 min. Estimates of the variance of
the background counts were obtained, and showed only minimal deviations
from this background level (SD ± 1 count). An additional
source of noise that must be taken into account in this quantification
of photon-limited luminescence images is photon shot noise. This
statistical variation in photon arrival may be estimated by the square
root of the number of photons counted. A crude estimate of the likely
variance of the photon counts between images may thus be calculated, in
which variance = 
(total count) +
(SDbackground)2. This confidence
interval is represented by the gray envelopes associated with the
graphs of photon counts from individual cell areas in Fig. 2. In Figs. 3 and 4, the count from each cell was measured using different-sized
pixel areas. The SD of the background was determined
experimentally from areas of these images in which there were no cells.
The background estimate for each cell was therefore calculated
according to the pixel area sampled, and this is similarly shown as a
shaded envelope for each graph of individual cell data in Figs. 3 and 4.
|
|
|
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
and 2
–4),
and in subsequent experiments, test substances were therefore added at
least 10 h after luciferin to ensure a stable baseline.
|
). Individual cells, or in some cases small clumps of 2–3 cells, were
selected from overall fields of 20–40 cells and tracked throughout the
imaging period for up to 80 h after addition of luciferin,
i.e. for up to 70 h after the achievement of a stable
baseline photon output.
Individual cell areas, or cell-free background areas, were imaged for
30-min periods at 3-h intervals to generate quantitative data. Resting,
unstimulated cells showed marked heterogeneity and variability of
luciferase expression when studied at 3-h intervals over 55 h of
observation, with up to 40-fold excursions in photon counts per area
within 12-h periods (Fig. 2). More than 90% of cells studied exhibited
high level luciferase activity immediately after luciferin addition,
but subsequently some of these cells exhibited phasic oscillations or
pulses in photon output, whereas other cells (that were capable of
expressing luciferase during the initial burst of activity after
luciferin addition) remained effectively silent throughout the
observation period.
Whereas induction of hPRL promoter activity with any of the three
stimuli (forskolin, TRH, or bFGF) led to smooth sustained overall
luciferase activity in any given field, this concealed widely different
responses when individual cells were analyzed. An experiment showing
the response to bFGF is illustrated in Fig. 1B, and examples of
responses to forskolin and TRH are shown graphically in Figs. 3 and 4.
From these quantitative data, it is clear that some cell areas
displayed steady progressive rises in light emission, whereas others
showed phasic or transient responses, and yet others showed no response
to stimulation despite high initial luciferase activity
(e.g. Fig. 4G). The degree of response of any
given cell during the experiment was independent of the height of the
initial surge of light output after luciferin. Although the overall
increase in photon output from a field of 20 cells might be only 3- to
4-fold, individual cells displayed excursions of photon output as high
as 50-fold during these studies.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The most important novel finding presented by these data is the dynamic nature of gene expression during long-term continuous observations in both resting unstimulated cells and following hormonal stimulation or activation of intracellular signaling pathways. This poses interesting questions about the nature of stochastic phenomena controlling the rate of gene transcription in a given cell from hour to hour, or probably from minute to minute, and it is likely that the temporal nature of transcriptional integration of numerous stimuli will be increasingly accessible to future study using such approaches.
A number of reasons may be considered for both the spatial and temporal heterogeneity that we have observed, including cell-to-cell contact or paracrine factors, cell cycle effects, and variable activation of intracellular signaling systems and expression of transcription factors. In fact it has been known for several years that cells in the anterior pituitary gland, and in pituitary tumors in man, display marked heterogeneity in peptide hormone content and receptor expression, and this heterogeneity extends to clonal cell lines (10, 11). Hofland and co-workers showed highly variable GH secretion rates and messenger RNA content among individual cells from human pituitary adenomas (12, 13). Our present data indicate that this probably reflects widely differing and fluctuating rates of gene transcription. However the additional significance of the present work is in the unique new ability to analyze this cell-to-cell heterogeneity in gene expression at frequent intervals, and hence, show that it is not a static phenomenon but changes dramatically over time. Individual somatotrophic cells show varying amplitude of calcium transients (14), and it will be of interest to determine whether heterogeneity of intracellular signaling relates to the variable gene expression that we have observed.
Cell-to-cell contact may have an important effect on gene expression,
which may be underestimated in cell culture systems, given the intimate
spatial relationship between lactotrophs and gonadotrophic cells in the
intact anterior pituitary gland. Recent data have confirmed that
cell-cell contact among primary pituitary cell cultures affects gene
expression, but that the effect depends on the pituitary cell type,
with PRL-Luc expression being suppressed by adjacent lactotrophs but
not by nonlactotrophic cells (15). In the present studies, we noted
that cells may move over time, and in some cases formed cell-cell dyads
during extended periods of observation. Interestingly, we observed a
number of instances in which the formation of cell-cell dyads was
associated with extinction of PRL-Luc expression in both of the
contiguous lactotrophic cells (e.g. Fig. 1b, mid-left of
panels at 24.5–36.5 h and other data not shown). Peptide hormone
promoter activity may alter according to stage of cell cycle, and this
may be a relevant factor in a clonal cell line with a doubling time of
40–50 h. Recent work demonstrated marked heterogeneity also in primary
cultures of pituitary lactotroph cells, in which cell proliferation is
probably limited (5), so this is probably not the only factor.
Nonetheless the stable cell line that we describe now offers an
important opportunity to analyze interrelationships between cell cycle
and gene expression in a well-delineated cell model system that may be
of general importance to a number of endocrine tissues.
Previous studies had not shown that the addition of luciferin had a significant effect on the level of stable light emission. However, none of the previous studies with transfected mammalian cells were performed over the very long consecutive imaging periods reported here. In transgenic plants expressing luciferase, the addition of luciferin has been shown to have a dramatic effect on the luciferase-directed luminescence in vivo (16, 17) and similar results have been shown in single heart cells microinjected with high levels of luciferase enzyme (18). It has been suggested that this decrease in luminescence following luciferin addition may be caused by inactivation by the luciferase metabolite oxyluciferin (18). Our results here show that a similar effect can occur in transfected mammalian cells, although we do not know whether this effect will prove to be general to all cell types.
In conclusion, we describe here for the first time dynamic and heterogeneous physiological changes in PRL promoter activation, recorded continuously in individual living pituitary tumor cells over periods of several days. The establishment of the stable cell system will allow future analysis of the significance and mechanisms of this aspect of gene regulation. The low copy number of the reporter gene construct in stably transfected cell lines offers the advantage over microinjected cells or transiently transfected cells of representing transcriptional phenomena as faithfully as possible similar to that of the endogenous genes, without possible distortion by factors such as competition for limiting amounts of transcription factors by very high copy numbers of episomal promoter-reporter constructs. Thus, despite the well-known cautions that apply to work with tumor cell lines, the system that we describe has important merits in terms of a tool for studying transcriptional phenomena in relation to other aspects of cell biology such as cell cycle and intracellular signaling. Systematic exploration of the role of paracrine phenomena and cell-cell contact will be greatly aided by the use of such cell systems using additional fluorescent or bioluminescent reporter genes that offer the possibility of studying activation of more than one gene promoter simultaneously.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Supported by Biotechnology and Biological Sciences Research
Council.
3 Supported by Pharmacia and Upjohn.
Received August 11, 1997.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Semprini, S. Friedrichsen, C. V. Harper, J. R. McNeilly, A. D. Adamson, D. G. Spiller, N. Kotelevtseva, G. Brooker, D. G. Brownstein, A. S. McNeilly, et al. Real-Time Visualization of Human Prolactin Alternate Promoter Usage in Vivo Using a Double-Transgenic Rat Model Mol. Endocrinol., April 1, 2009; 23(4): 529 - 538. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A McMaster, T Chambers, Q-J Meng, S Grundy, A S I Loudon, R Donn, and D W Ray Real-time analysis of gene regulation by glucocorticoid hormones J. Endocrinol., May 1, 2008; 197(2): 205 - 211. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. D. Adamson, S. Friedrichsen, S. Semprini, C. V. Harper, J. J. Mullins, M. R. H. White, and J. R. E. Davis Human Prolactin Gene Promoter Regulation by Estrogen: Convergence with Tumor Necrosis Factor-{alpha} Signaling Endocrinology, February 1, 2008; 149(2): 687 - 694. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. T. French, W. H. Hanneman, L. S. Chubb, R. E. Billings, and M. E. Andersen Induction of CYP1A1 in Primary Rat Hepatocytes by 3,3',4,4',5-Pentachlorobiphenyl: Evidence for a Switch Circuit Element Toxicol. Sci., April 1, 2004; 78(2): 276 - 286. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. JOHNSTON, F. R. A. CAGAMPANG, J. A. STIRLAND, A.-J. F. CARR, M. R. H. WHITE, J. R. E. DAVIS, and A. S. I. LOUDON Evidence for an endogenous per1- and ICER-independent seasonal timer in the hamster pituitary gland FASEB J, May 1, 2003; 17(8): 810 - 815. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. J. Norris, J. A. Stirland, D. W. McFerran, Z. C. Seymour, D. G. Spiller, A. S. I. Loudon, M. R. H. White, and J. R. E. Davis Dynamic Patterns of Growth Hormone Gene Transcription in Individual Living Pituitary Cells Mol. Endocrinol., February 1, 2003; 17(2): 193 - 202. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Vazquez-Martinez, G. M. Leclerc, M. E. Wierman, and F. R. Boockfor Episodic Activation of the Rat GnRH Promoter: Role of the Homeoprotein Oct-1 Mol. Endocrinol., September 1, 2002; 16(9): 2093 - 2100. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. L. Shorte, G. M. Leclerc, R. Vazquez-Martinez, D. C. Leaumont, W. J. Faught, L. S. Frawley, and F. R. Boockfor PRL Gene Expression in Individual Living Mammotropes Displays Distinct Functional Pulses That Oscillate in a Noncircadian Temporal Pattern Endocrinology, March 1, 2002; 143(3): 1126 - 1133. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Vazquez-Martinez, S. L. Shorte, W. J. Faught, D. C. Leaumont, L. S. Frawley, and F. R. Boockfor Pulsatile Exocytosis Is Functionally Associated with GnRH Gene Expression in Immortalized GnRH-Expressing Cells Endocrinology, December 1, 2001; 142(12): 5364 - 5370. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. W. McFerran, J. A. Stirland, A. J. Norris, R. A. Khan, N. Takasuka, Z. C. Seymour, M. S. Gill, W. R. Robertson, A. S. I. Loudon, J. R. E. Davis, et al. Persistent Synchronized Oscillations in Prolactin Gene Promoter Activity in Living Pituitary Cells Endocrinology, July 1, 2001; 142(7): 3255 - 3260. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. R. E. Davis, J. McVerry, G. A. Lincoln, S. Windeatt, P. R. Lowenstein, M. G. Castro, and A. S. McNeilly Cell Type-Specific Adenoviral Transgene Expression in the Intact Ovine Pituitary Gland after Stereotaxic Delivery: An in VivoSystem for Long-Term Multiple Parameter Evaluation of Human Pituitary Gene Therapy Endocrinology, February 1, 2001; 142(2): 795 - 801. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Hume Probability in transcriptional regulation and its implications for leukocyte differentiation and inducible gene expression Blood, October 1, 2000; 96(7): 2323 - 2328. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. D. Southgate, S. Windeatt, J. Smith-Arica, C. A. Gerdes, M. J. Perone, I. Morris, J. R. E. Davis, D. Klatzmann, P. R. Lowenstein, and M. G. Castro Transcriptional Targeting to Anterior Pituitary Lactotrophic Cells Using Recombinant Adenovirus Vectors in Vitro and in Vivo in Normal and Estrogen/Sulpiride-Induced Hyperplasic Anterior Pituitaries Endocrinology, September 1, 2000; 141(9): 3493 - 3505. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Faria, C. D. Wood, L. Perrouault, J. S. Nelson, A. Winter, M. R. H. White, C. Helene, and C. Giovannangeli Targeted inhibition of transcription elongation in cells mediated by triplex-forming oligonucleotides PNAS, April 11, 2000; 97(8): 3862 - 3867. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. T. Willard, M. D. Amstutz, E. J. Abraham, J. P. Castano, D. C. Leaumont, W. J. Faught, and L. S. Frawley Simultaneous indirect activity measurements of GH and PRL genes in the same, living mammosomatotrope Am J Physiol Endocrinol Metab, December 1, 1999; 277(6): E1150 - E1153. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Villalobos, W. J. Faught, and L. S. Frawley Dynamics of Stimulus-Expression Coupling as Revealed by Monitoring of Prolactin Promoter-Driven Reporter Activity in Individual, Living Mammotropes Mol. Endocrinol., October 1, 1999; 13(10): 1718 - 1727. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
