| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Interactions Cellulaires Neuroendocrines, UMR 6544, Centre National de la Recherche Scientifique (P.J., A.S., G.G., A.J.Z., A.E.); Laboratoire de Transfert d’Oncologie Biologique-Assistance Publique Hopitaux de Marseille (F.F., L’H.O.); and Unit 501, INSERM (M.G.), Institut Fédératif Jean Roche, Faculté de Médecine Nord, 13916 Marseille Cedex 20, France; and Biomeasure, Inc. (M.D.C., J.P.M.), Milford, Massachussetts 01757
Address all correspondence and requests for reprints to: Dr. Philippe Jaquet, Interactions Cellulaires Neuroendocrines, UMR 6544, Centre National de la Recherche Scientifique, Institut Fédératif Jean Roche, Faculté de Médecine Nord, boulevard Pierre Dramard, 13916 Marseille Cedex 20, France. E-mail: jaquet.p{at}jean-roche.univ.mrs.fr
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In cells cultured from the 10 mixed adenomas that secreted both GH and PRL, RT-PCR analysis revealed a consistent coexpression of SSTR5, SSTR2, and SSTR1 mRNA. In all cases SRIF14, SRIF28, and the SSTR5-preferential analog, BIM-23268, significantly suppressed PRL secretion, with a mean maximal inhibition of 48 ± 4%. In contrast, the SSTR2-preferential analogs, BIM-23197 and octreotide, were effective in suppressing PRL in only 6 of 10 cases. In cells cultured from adenomas taken from patients partially responsive to the SRIF analog, octreotide, partial additivity in suppressing both GH and PRL secretion was observed when the SSTR2- and SSTR5-preferring analogs, BIM-23197 and BIM-23268, were tested in combination. Our data show a highly variable ratio of the SSTR2 and SSTR5 transcripts, according to tumors. The SSTR2-preferring compound consistently inhibits GH release, whereas the SSTR5-preferring compound is the main inhibitor of PRL secretion. When both drugs are combined, the partial additivity observed in mixed GH- plus PRL-secreting adenomas may be of interest in the therapeutic approach of such tumors.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Our data reveal a similar pattern of SSTR2 and SSTR5 mRNA expression in both pure and mixed adenomas. Despite its lower expression, SSTR2 mRNA expression correlated significantly with the GH-suppressive effects of SRIF14, SRIF28, and BIM-23197. SSTR1 transcripts were only expressed in tumors secreting PRL. The SSTR5-preferential compound, BIM-23268, is the dominant inhibitor of PRL in the mixed adenomas. Therefore, due to the different inhibitory actions regulated by the two receptor subtypes, a combination of both preferential compounds may allow a better control of hormones hypersecretion in these mixed adenomas.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The present study was approved by the ethics committee of the
University of Aix-Marseilles (Marseilles, France) and was undertaken
after obtaining informed consent from each patient. Fifteen acromegalic
patients (eight women and seven men), aged 18–66 yr, presenting with a
macroadenoma were studied. Their endocrine status and the
neuroradiological characterization of the tumors were documented before
any treatment. Basal GH levels were expressed as the mean of
consecutive measurements obtained hourly between 0800–1300 h. SRIF
agonist sensitivity was assessed by an acute test using a single sc
200-µg injection of octreotide (Sandostatin, Novartis,
Basel, Switzerland). Blood samples were withdrawn hourly before
and for 6 h after octreotide treatment to measure GH variations.
Sensitivity to octreotide was expressed as the percent decrease in GH
from basal to the mean GH value 2–6 h after octreotide injection. The
basal insulin-like growth factor I (IGF-I) value was evaluated under
fasting conditions between 0800–0900 h. Magnetic resonance imaging
revealed adenomas with a maximal 11- to 48-mm diameter either enclosed
in the pituitary fossa (n = 8) or with an extension toward the
adjacent structures, mainly the cavernous sinus (n = 7). All
patients underwent transsphenoidal surgery. The clinical endocrine and
tumoral status of each patient is summarized in Tables 1
and 2,
which are divided according to classification as pure GH-secreting
adenomas or mixed GH+PRL adenomas. The definition of the later subgroup
was not based upon basal plasma PRL values, but was established
according to the 24-h release of both GH and PRL on day 3 of cell
culture, as shown in Table 2.
|
|
GH and PRL were measured using commercial immunoradiometric kits (Immunotech, Marseilles, France). Normal GH values ranged from 0.05–2.4 µg/L; normal PRL values ranged from 1–24 µg/L in women and from 1–17 µg/L in men. After an ethanol-acid extraction, the plasma IGF-I assay was performed using the IGF-I RIA kit from Nichols Institute Diagnostics (San Juan Capistrano, CA). The normal ranges, according to sex and age, were established by our laboratory.
Detection of SSTR subtypes
Total RNA was extracted from 30–60 mg tissue from each tumor using the SV total RNA isolation system (Promega Corp., Lyon, France). The RNA samples were subsequently treated with 30 U ribonuclease-free deoxyribonuclease I (Roche Molecular Biochemicals, Mannheim, Germany) to prevent any contamination by genomic DNA. Total RNA was reverse transcribed into complementary DNA (cDNA) using 1 µg hexamers (Pharmacia Biotech, Orsay, France) and Moloney murine leukemia virus reverse transcriptase, as described by the manufacturer.
The 5'-exonuclease (Taq Man) assay, which produces a direct
proportional readout for the progression of PCR reactions, was used
(10). Amplification of cDNA derived from 50–150 ng total RNA was
performed in a 50-µL reaction volume with a buffer consisting of 10
mmol/L Tris-HCl (pH 8.3; 25 C); 50 mmol/L KCl; 10 mmol/L
ethylenediamine tetraacetate; 5 mmol/L MgCl2 in
the presence of 200 µmol/L deoxy (d)-ATP, dCTP, and dGTP; 400
µmol/L dUTP; 1 µmol/L of each primer; 200 nmol/L probe; 1 U Amp
Erase UNG (Perkin-Elmer Corp., Paris, France); and 1.25 U AmpliTaq Gold
polymerase (Perkin-Elmer Corp.). The probe comprised
20–30 nucleotides with 5'-end substitution with a fluorophore and a
quencher substitution at the 3'-end. The synthetic SSTR cDNA primers
used in the PCR reaction were 19- or 20-mers as follows: SSTR1: sense,
1411–1433; antisense, 1511–1492; probe, 1442–1463; SSTR2: sense,
10–29; antisense, 109–91; probe, 58–32; SSTR3, sense, 1188–1206;
antisense, 1254–1236; probe, 1209–1234; SSTR4: sense, 1282–1301;
antisense, 1362–1343; probe, 1331–1301; and SSTR5: sense, 1103–1119;
antisense, 1156–1139; probe, 1137–1121. The annealing-extension
temperatures were: for SSTR1, 66 C; for SSTR2, 56 C; for SSTR3, 70 C;
for SSTR4, 66 C; and for SSTR5, 70 C. Forty cycles of two-step PCR
reaction-annealing extension at specified temperatures for 30 s
and denaturation at 95 C for 20 s,were performed on a ABI Prism
7700 sequence detection apparatus (Perkin-Elmer Corp.).
For quantitation of data, SSTR mRNA levels were, in the same reaction,
normalized to the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA
levels. The control GAPDH primers was as follows: sense, 222–240;
antisense, 322–303; and probe, 277–301. For each measurement, three
independent RT-PCR analyses were performed. To produce standard curves
for each SSTR mRNA and GAPDH mRNA, RNAs were produced by in
vitro transcription from linearized templates corresponding to
SSTRs and GAPDH cDNA constructs using T7 or T3 polymerases as
previously described (11). The synthesized RNAs were reverse
transcribed to cDNA for each parameter as described above. Using the
fluorogenic probes for SSTR receptors and GAPDH with the experimental
conditions defined above, we obtained a linear relationship between the
RNA concentration (previously transcribed into cDNA) and the
fluorescent signal (
RQ) for SSTR and GAPDH RNAs in 1–250 pg DNA
target. For each unknown sample, we determined the RQ values for
both genes, and the results were expressed as picograms of SSTR per pg
GAPDH.
Cell culture studies
A portion of each tumor obtained at surgery was dissociated by
mechanical and enzymatic methods. Depending on the tumor, 3–30 x
106 isolated cells were obtained. Tumoral cells
were initially cultured in DMEM supplemented with 10% FCS for 3 days.
On day 3, the cells were washed and plated in multiwell culture dishes
coated with extracellular matrix from bovine corneal endothelial cells.
When they were attached to the matrix on days 4–6, depending on the
culture, the medium was removed and replaced with serum-free DMEM
supplemented with antibiotics, insulin, transferrin, and selenium as
previously described (12). The effects of various doses of the SRIF
agonists, SRIF14, SRIF28, octreotide, BIM-23268, and BIM-23197, on
inhibition of GH and PRL release were measured over an 8-h period
between days 5–7 of culture. Each drug concentration was tested in
quadruplicate culture wells (Costar 3524, Brumath,
France). SRIF14 and SRIF28 were purchased from Sigma
(Saint-Quentin Fallavier, France). Octreotide was supplied by Novartis
(Basel, Switzerland). BIM-23268 and BIM-23197, respectively
preferential for SSTR5 and SSTR2 subtypes (13), were provided by
Biomeasure, Inc. (Milford, MA). The human SSTR subtype affinities
(IC50; nanomoles per L) of each compound are
summarized in Table 3. The native SRIFs and SRIF analogs were dissolved
in 0.01 mol/L acetic acid containing 0.1% purified serum albumin
(Life Technologies, Inc., Cergy Pontoise, France). The
drugs were stored at -80 C as 10-3 mol/L
solutions. Fresh working solutions were prepared from a new aliquot for
each experiment.
|
In situ hybridization was performed in 2 mixed GH+PRL adenomas, characterized as either sensitive (A6) or resistant (A15) to the SRIF agonists based upon in vivo data and cell culture studies. Sections 12 µm thick were cut in a cryostat microtome at -20 C. They were thaw-mounted onto twice gelatin-coated slides, dried on a slide warmer, and kept at -80 C. The sections were hybridized simultaneously with the SSTR2, SSTR5, PRL, and GH probes, as previously described (14). The SSTR2 probe was a 706-bp fragment of the human SSTR2 cDNA (bases 320-1025) cloned into Bluescript and labeled with [35S]UTP (SA, 1200 Ci/mmol; New England Nuclear, Paris, France) using either T3 (antisense probe) or T7 (sense probe) RNA polymerase. The SSTR5 probe was a 936-bp fragment of the human SSTR5 cDNA (bases 81–1016) cloned into Bluescript and labeled with [35S]UTP (SA, 1200 Ci/mmol; New England Nuclear) using either T3 (antisense probe) or T7 (sense probe) RNA polymerase. The PRL probe was a 588-bp fragment of the human PRL cDNA (bases 94–681) cloned into pPRC script and labeled with digoxigenin-UTP (Roche Molecular Biochemicals, Meylan, France) using T3 (antisense probe) or T7 (sense probe) RNA polymerase. The GH probe was a 558-bp fragment of the human GH cDNA (bases 56–611) cloned into pPRC script and labeled with streptavidin-UTP (Roche Molecular Biochemicals) using either T3 (antisense probe) or T7 (sense probe) RNA polymerase. PRL and GH hybridization signals were revealed using fluorescein isothiocyanate- or rhodamine-tyramide, respectively. Slides were subsequently dipped into nuclear emulsion (1:1 water: K5, Ilford, Saint-Priest, France) and exposed for 2 months for SSTR5 mRNAs and for 4 months for SSTR2 mRNAs. Sections triple hybridized with the corresponding sense riboprobes served as controls. Specimens were viewed under a Leitz DMDR microscope (Wetzlar, Germany) equipped with a planachromatic fluorite 40/0.70 objective. Ten systematic random fields were sampled for each tumor (15), captured with a color CCD camera and digitized through a Neotech Image Grabber (Neotech, Hampshire, UK) slotted on a microcomputer. Each sampled field was captured 4 times: once under epifluorescence with the barrier filter centered at 520 nm (fluorescein isothiocyanate), once similarly with the filter barrier at 600 nm (rhodamine), once under Koehler’s illumination transmitted darkfield for detecting the radioautography silver grains, and once under phase contrast to assess the total number of cells in the field (including the unlabeled cells). The corresponding 4 digitized images were overlaid with Photoshop software (Adobe System, Inc. Mountain View, CA). Cells were identified individually and computed according to their label characteristics, i.e. unlabeled cells; single labeled cells for GH, PRL, SSTR2, or SSTR5 mRNAs; dual labeled cells for the messengers of 1 hormone and SSTR2 or SSTR5 mRNA; and triple labeled cells for both hormones and either SSTR2 or SSTR5 mRNA. The onus of counting was undertaken by a single investigator. Countings were expressed as the ratio of the percentage of the class to the total number of cells computed in the 10 microscopic fields representing each tumor.
Statistics
The results are presented as the mean ± SEM. Statistical significance between two unpaired groups was determined by the Mann-Whitney test. To measure the strength of association between the pairs of variables, without specifying dependencies, Spearman order correlations were run. P < 0.05 was considered significant for all tests.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
RT-PCR quantitative analysis was performed on tumor fragments of
the 15 adenomas. The individual patterns of SSTR2 and SSTR5 mRNA
expression are presented in Fig. 1. Based on
hormone release by cell cultures prepared from these tumors, the
present series included 5 pure GH-secreting adenomas and 10 mixed
GH+PRL adenomas. This latter group included 5 tumors from
hyperprolactinemic patients (plasma PRL, 27–83 µg/L) and 5 tumors
from normoprolactinemic patients but, nevertheless, with significant
PRL release (0.3–3.2 µg/day·106 cells)
in vitro, as indicated in Table 2
. RT-PCR analysis
demonstrated the presence of both SSTR2 and SSTR5 mRNA in all tumors.
The mean level of SSTR2 mRNA expression was 100 ± 30 pg/pg GAPDH,
whereas the mean level of SSTR5 mRNA was 1052 ± 406 pg/pg GAPDH
in the series. The individual pattern of expression was highly
variable, depending on the tumor. In particular, 4 tumors (2 pure GH
and 2 mixed GH+PRL adenomas) expressed SSTR5 mRNA levels over 1000
pg/pg GAPDH. No tumor expressed only SSTR2 or SSTR5 mRNA alone,
although the expression of SSTR5 mRNA was largely dominant in 2 tumors
(A4 and A13). Although trends were observed, the expression level of
SSTR2 and SSTR5 was not significantly different between the 5 pure GH
and the 10 mixed GH+PRL adenomas. In particular, SSTR5 mRNA expression,
previously shown to be correlated to the inhibition of PRL release in
prolactinomas (9), was greater, although not significantly so, in the
pure GH-secreting adenomas (1874 ± 888 pg/pg GAPDH for pure
GH-secreting adenomas vs. 640 ± 390 pg/pg GAPDH for
GH+PRL adenomas). The SSTR1 mRNA, which has been reported to be highly
expressed in pure prolactinomas (8, 9), was poorly expressed among the
15 tumors studied (30 ± 18 pg/pg GAPDH). In the majority of
tumors, the level of SSTR1 expression ranged from 1–16 pg/pg GAPDH,
and it was mainly expressed in the mixed GH+PRL tumors. In only 3 of 10
GH+PRL tumors (A6, 7, and 13) was SSTR1 mRNA expression high, ranging
from 43–278 pg/pg GAPDH. SSTR3 mRNA expression was highly variable,
being either absent (n = 7) or at a mean expression level of
271 ± 104 pg/pg GAPDH in 8 tumors (A2, 3, 4, 9, 10, 12, 14, and
15). Finally, SSTR4 mRNA expression was not observed in any tumor.
|
In the cell culture studies, the effects of SRIF14 and SRIF28 on
GH and PRL secretion were measured after an 8-h incubation period.
Previous experiments (not shown) demonstrated that maximal inhibition
of GH, by nanomolar concentrations of these drugs, occurred after
4 h of incubation. Because in some tumors, the basal release of
either GH or PRL was low during a 4-h incubation (1–5 ng/medium in
control conditions), an 8-h incubation period was adapted to achieve
accurate measurements of hormone release from all cultures. As shown in
Fig. 2a, the maximal inhibition of GH release
by SRIF14, compared with controls (medium alone), ranged from 29–59%
among the 5 pure GH-producing adenoma cell cultures and from 25–51%
among the 10 mixed GH+PRL adenoma cell cultures. Under the same
conditions, SRIF28-induced similar inhibition of GH release in 14 of 15
cases. The mean inhibitory effect of SRIF14 on GH release was 41
± 3% and was not different from the 39 ± 4% inhibitory effect
of SRIF28 at nanomolar concentrations.
|
). The mean
suppressive effect on PRL secretion was positively correlated to the
mean maximal GH suppression in the same 10 cultures studies (mean GH
suppression by SRIF14 and SRIF28, 47 ± 3% and 38 ± 5%,
respectively; P < 0.02). Correlation between SSTR subtype mRNA expression and inhibition of hormone release (GH and PRL)
The degree of GH inhibition by SRIF14 and SRIF28 and the levels of
SSTR2 mRNA expression were highly correlated among the 15 tumors
analyzed (P < 0.003), as illustrated for SRIF14 in
Fig. 3a. In the same tumors, despite a much
higher expression of SSTR5 transcripts, there was no correlation with
the ability of the 2 SRIFs to inhibit GH release, as illustrated for
SRIF14 (Fig. 3b). Among the limited number of mixed tumors that
cosecreted GH and PRL, no correlation could be demonstrated between the
inhibition of PRL release and the expression of either SSTR2 or SSTR5
mRNAs.
|
The absence of correlation between SSTR5 mRNA expression and the
inhibition of GH release by either SRIF14 or SRIF28 prompted us to
analyze more precisely the responses of adenoma cells to compounds
preferential for the SSTR2 and SSTR5 subtypes. A dose-response
inhibition of GH release was examined with
10-12–10-8 mol/L concentrations of SRIF14;
of the SSTR2-preferring analog, BIM-23197; of the SSTR5-preferring
compound, BIM-23268; and of octreotide, the analog used in
vivo (Tables 1 and 2). The correlation between GH inhibition by
octreotide in vivo and in vitro (with nanomolar
concentrations) was highly significant (r = 0.74;
P < 0.008) in the 15 cases. Among the 15 adenoma cell
cultures, 2 patterns of responses to SSTR2- and 5-preferential analogs
were observed. In cultures from 7 of 15 tumors (Fig. 4a), the mean
dose-response inhibitions of GH release by BIM-23197 and BIM-23268 were
identical (EC50, 0.02 ± 0.01 and 0.03
± 0.01 nmol/L, respectively). Superimposable results were achieved
with SRIF14 in these studies. Octreotide realized a similar
dose-response inhibition of GH at slightly higher concentrations
(EC50, 0.06 ± 0.06 nmol/L) than with
BIM-23197 or SRIF14. These 7 cases (A1, 3, 4, 6, 12, 14, and 15) were
from both GH only and mixed GH+PRL adenomas. In the others 8 of 15
cultures from the remaining 8 tumors (Fig. 4b), superimposable
dose-response inhibition for GH occurred in the presence of BIM-23197
and SRIF14 with a similar EC50 (0.02 ± 0.01
nmol/L). In contrast to the previously discussed tumors, the SSTR5
agonist produced a dose-dependant inhibition of GH release at 30-fold
higher concentrations (EC50, 0.6 ± 0.2
nmol/L) than required by either SRIF14 or BIM-23197. In the same
tumors, octreotide produced a dose-response inhibition for GH with an
EC50 of 0.1 ± 0.04 nmol/L. The discrepancy
between the results obtained with BIM-23197 and BIM-23268 can be
explained on the basis of the binding characteristics of BIM-23268
(Table 3), which, at high concentrations, behaves as a weak SSTR2
agonist. Thus, these results reveal 1 class of tumors that are equally
responsive to SSTR2- and SSTR5-preferential agonists and a second class
of tumors in which the GH-suppressive effect is only mediated
through the SSTR2 subtype.
|
Dose-response inhibition of PRL secretion by SRIF14, BIM-23197,
BIM-23268, and octreotide was also studied in all cultures of the 10
mixed (GH-PRL) tumors. As already shown (Fig. 2) SRIF14 and SRIF28, at
nanomolar concentrations, produced a significant inhibition of PRL
release from all tumors. In all 10 tumors (Fig. 5), SRIF14 and the SSTR5 preferential analog,
BIM-23268, induced a superimposable dose-related suppression of PRL
(EC50, 0.09 ± 0.08 and 0.08 ± 0.06
nmol/L, respectively). The dose-related inhibition of PRL by the
SSTR2-preferring compound, BIM-23197 (as well as by octreotide), showed
2 markedly distinct patterns. In 6 of 10 tumors (A6, 7, 8, 12, 14, and
15) BIM-23197 and octreotide induced a dose-related suppression of PRL
parallel to that of SRIF14, with 5-fold higher concentrations for
octreotide (Fig. 5a). In the remaining 4 tumors (Fig. 5b), BIM-23197
and octreotide were unable to significantly suppress PRL release at any
concentration. Thus, as observed for GH inhibition, 2 patterns of PRL
suppression were revealed by the preferential SSTR2 and 5 compounds. In
6 tumors, PRL suppression was equally achieved through activation of
either SSTR5 or SSTR2 subtype, whereas in the remaining 4 tumors PRL
inhibition was only mediated by the SSTR5 subtype.
|
The previous data, analyzed as the mean inhibitory effects of
BIM-23197 or BIM-23268 on either GH release in the 15 tumors or PRL
release from the 10 of 15 mixed adenomas, do not fully reveal the
heterogeneous range of responses to the SSTR2 and SSTR5 preferential
compounds encountered among individual tumors. From the analysis of the
individual dose-response inhibition of GH alone, 2 types of responses
were identified in the GH-only secreting adenomas. Examples of the
responses to the preferential SSTRs analogs are presented in Fig. 6. Among the 5 GH-only secreting adenomas, 3
showed similar GH suppression with either SSTR2- or SSTR5-preferring
compounds, as exemplified by A1 (Fig. 6a), whereas in the remaining 2
GH-only secreting adenomas, GH was only suppressed by the
SSTR2-preferential compound, BIM-23197 (A2, Fig. 6b). In these latter 2
tumors, GH suppression obtained with higher concentrations of BIM-23268
was interpreted as being due to the weak SSTR2 agonist activity of the
compound. Among the 10 mixed adenoma cell cultures, 4 of 10 showed
identical suppression of GH and PRL release with both the SSTR2- and
the SSTR5-preferential compounds. In one such case (A6, Fig. 6c), the
EC50 was 0.005 nmol/L for both GH and PRL
suppression. In another 4 of 10 mixed adenomas a differential pattern
of GH and PRL was observed in response to BIM-23197 and BIM-23268, as
exemplified by A11 (Fig. 6d). The GH component of this tumor was highly
sensitive to the inhibitory effect of the SSTR2-preferential agonist.
Conversely, the PRL component was sensitive only to the inhibitory
effects of the SSTR5-preferential agonist, BIM-23268. In this latter
subgroup, such specific inhibitory effects of BIM-23197 and BIM-23268
on GH and PRL secretion argue for the coexistence of tumor subclones
secreting either GH alone or PRL (or GH and PRL) alone. Finally, in 2
of 10 mixed adenomas the pattern of GH and PRL response to the two
analogs was halfway between the 2 preceding examples, as GH suppression
was highly sensitive to BIM-23197 only, whereas PRL was equally
suppressed by BIM-23197 and BIM-23268. Thus, these individual data
provide a more realistic overview of the functional heterogeneity among
acromegalic tumors with regard to their sensitivity to the two SRIF
analogs preferential for the SSTR2 and SSTR5 subtypes.
|
The possibility of additivity of BIM-23197 and BIM-23268 was
examined in three mixed GH+PRL tumors (A13, 14, and 15), which
(in vivo studies) presented with a 18–40% partial
inhibition of GH release after octreotide acute testing (Table 2 and
Fig. 7). In these experiments the
effects of BIM-23197 and BIM-23268, either alone or in combination, as
well as of octreotide on GH and PRL secretion were examined using
10-12-10-9 mol/L of each
compound (and of equimolar concentrations when BIM-23197 and BIM-23268
were combined). The mean maximal GH suppression achieved by BIM-23197
plus BIM-23268 was 34 ± 6%, which was not statistically
different from that obtained with BIM-23197 alone (25 ± 3%) or
BIM-23268 alone (22 ± 7%). The maximal GH suppression produced
by octreotide (14 + 7%) was significantly lower than that obtained
with the combination of BIM-23197 and BIM-23268 (P <
0.02). These data indicate a lack of synergism between both compounds
at maximally effective concentrations. Nevertheless, at submaximal
concentration (10-10 mol/L), BIM-23197 plus
BIM-23268 achieved a better GH and PRL inhibition than that obtained
with either drug alone (P < 0.05). These data show
that, in those tumors characterized as partial responders to
octreotide, the combination of BIM-23197 and BIM-23268 is clearly
additive within the dose-response range. In such tumors, the
combination of SSTR2- and SSTR5-preferential compounds allowed
comparable inhibition of GH and PRL with concentrations 10–30 times
lower than those required when each drug was tested alone.
|
)
The cellular localization of SSTR2 and SSTR5 mRNA was
analyzed on fragments of mixed tumors characterized as either sensitive
(A6) or partially responsive (A15) to SRIF agonists (Table 2). In both
tumors, hybridization with both PRL mRNA and GH mRNA antisense probes
allowed identification of distinct hormonal cell phenotypes. About 18%
of cells in both tumors were not labeled for either GH or PRL mRNAs.
The cells expressing PRL mRNA only were extremely rare (4.7–1.3%,
respectively, in each tumor). Based on the presence of cells labeled
with either GH mRNA alone or with both GH and PRL mRNAs, both tumors
were found to be bimorphous. In A6, the GH transcript alone was
identified in 40% of the cells, whereas both GH and PRL transcripts
were present in 41% of the cells. In A15, 64% of the cells expressed
GH mRNA alone, whereas 15% expressed both GH and PRL transcripts.
SSTR5 mRNA expression was observed in most of the cells from both
tumors in both the pure GH cells and the GH+PRL cells. The mean
percentages of cells labeled with the SSTR5 probe were 85% and 87%
for A6 and A15 tumors, respectively. With regard to SSTR2 mRNA
labeling, a marked difference was observed between the two tumors. In
the SRIF agonist-sensitive adenoma (A6), the SSTR2 transcript was
expressed in 66–60% of the GH alone and GH+PRL cells. In the
partially sensitive tumor (A15), SSTR2 mRNA expression was much lower.
Only 29–30% of the GH alone and GH+PRL cells expressed the SSTR2
transcript. Interestingly the percentage of SSTR2 mRNA labeling (63%
and 30%) in these two tumors correlated well with the 51% and 25%
maximal inhibitory effects of SRIF14 on GH release observed in the cell
culture studies of cells from the same adenomas (Fig. 2).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
-cells, and SSTR5 is quasiselectively
expressed in the 
-cells. Such cellular specificities for the
different SSTR subtypes indeed underlines the need for SSTR-specific
ligands to target hormone suppression. The expression of SSTR2 mRNA was quantitatively correlated to the degree of GH release inhibition by both SRIF14 and SRIF28 in our study. No difference in the inhibition of GH release was observed with either native SRIF in most cases. The dominant expression of SSTR5 mRNA, however, was not correlated to either GH or PRL suppression by any of the compounds tested. Whether the observed high expression of SSTR5 mRNA correlates with the presence of SSTR5 protein is unknown. The expression of SSTR mRNAs has been correlated to the receptor proteins in some studies (18, 28), but not in others (29). Further studies, using specific antibodies raised against the different SSTR subtypes, are necessary to investigate whether there are either transcriptional defects or abnormal coupling mechanisms that may explain SSTR5 nonfunctionality in some GH-secreting tumors (30). The presence of the SSTR5 cannot be explained on the basis of its role in suppressing PRL secretion, as equivalent SSTR5 mRNA expressions were found in pure GH and mixed GH+PRL tumors.
Although SRIF14 and SRIF28 were equipotent in hormone suppression, the
SSTR2 and SSTR5 preferential analogs identified heterogeneous
suppressive effects on GH and PRL secretion. In the past, the subtype
selectivity of some peptide SRIF analogs has been questioned (31). More
recently, cyclic and linear SRIF analogs have been developed and
present with a preferential selectivity for SSTR2, such as BIM-23197
(32), or for SSTR5, such as BIM-23268 (33). This latter compound has a
40-fold selectivity in affinity for SSTR5 over SSTR2 when tested for
binding in CHO-1 cells transfected with the human receptor subtypes
(33). Classification of the tumors on the basis of their hormonal
response patterns to the SSTR2- and SSTR5-preferring analogs revealed
distinct phenotypes that can be correlated with prior morpho-functional
subclassifications of somatotroph adenomas (34). In somatotroph
adenomas, the tumor cells present either as GH only, densely or
sparsely granulated cells or as mixed GH+PRL adenoma cells (35). In the
latter case, according to double labeling immunogold electron
microscopy, the tumors appear either as monomorphous (GH and PRL
granules colocalized in the same mammosomatotroph cell) or as
bimorphous tumors composed of a mixture of both GH only and PRL only
adenomas cells. In initial studies, the GH only adenomas were reported
to represent 64% of the tumors (35). A more recent immunohistochemical
analysis of 69 acromegalic tumors revealed PRL- as well as
-subunit-immunopositive cells in 90% of somatotroph tumors (36).
The preponderance of mixed GH+PRL tumors was also encountered in our
in vitro study, which detected an associated mild PRL
secretion in 66% of the adenoma cell cultures. Temporal analysis of
the somatolactotrope lineage during fetal development displays several
different morphological steps in somatomamotroph cell commitment (37)
that are superimposable on the different ultrastructural aspects of the
somatotroph adenomas. The evolving human fetal pituitary initially
displays well differentiated somatotroph cells at 8–9 weeks. A second
population of mammosomatotroph cells, appearing at 12 weeks gestation,
develops during the second trimester of pregnancy. Finally, the pure
lactotroph cells differentiate from week 23 to the term of gestation.
This developmental progression is corroborated by the experimental
model of transgenic mice bearing the thymidine kinase obliteration
system (38). In keeping with the hypothesis that adenomas could occur
from cells at any stage of somatolactotroph development, our analysis
of the variable inhibitory effects of the SSTR-preferential compounds
suggests an evolutive pattern for the differentiation of SSTR subtype
functionality. The progenitor cell somatotroph adenoma could bear both
SSTR2 and SSTR5 functional subtypes. These two SSTRs subtypes remain
functional for GH and PRL inhibition in the mammosomatotroph
bisecreting adenoma. This concept is supported by the results of a
previous study of human fetal pituitary cells, collected at weeks
23–25 (i.e. when both somatotroph and somatomammotroph
cells are well differentiated), in which both BIM-23197 and BIM-23268
were shown to equally inhibit GH release (13). The third tumoral cell
phenotype arises from the somatomammotroph lineage in which there is
somatostatinergic regulation with preferential GH suppression through
SSTR2 and preferential PRL suppression through SSTR5. The fourth
somatotroph tumor phenotype is characterized by its SSTR2-only
regulation of GH suppression. It has been previously shown that the
suppression of PRL in pure prolactinomas was mediated mainly via the
SSTR5 subtype (8, 9), whereas the SSTR2-preferential agonists only
occasionally could mediate PRL suppression (9). Such arguments favor an
ultimate differentiation in which some lactotrophs lose the SSTR2
regulatory component, as opposed to the terminal differentiation of
somatotrophs, in which the SSTR5 regulatory component is lost. Such a
differentiation pathway from progenitor cells, as proposed in the
theories of stem cell regulation (39), implies a default of
differentiation executed by the stem cell under factors influencing the
pituitary cell commitment. This was already documented for the GHRH
receptor (40), which plays an important role in the differentiation of
somatotrophs. The suggestion of an evolutive pattern of expression of
SSTRs subtypes is not surprising, as it has already been documented
during the ontogenesis of the fetal rat brain, which initially
expresses SSTR2 and SSTR3 in different areas, whereas SSTR4-SSTR5
emerge only around birth (41).
Finally, and possibly most importantly from a therapeutic perspective, our study as well as previous cell culture studies of acromegalic tumors (8, 42, 43, 44, 45) underline the variable suppressive effects of SRIF agonists on GH secretion, according to the individual nature of each tumor. In the adenoma cells taken from acromegalic patients that were partially responsive to the inhibitory effects of octreotide, our hybridization studies revealed a clear loss of SSTR2 transcript expression in the majority of the adenoma cells. Such findings underline the key role of the SSTR2 subtype in the inhibitory regulation of GH release. Identical observations of a patchy, heterogeneous distribution of SSTR2 in GH-secreting tumors, poorly responsive to octreotide administration, have been previously noticed by receptor autoradiography in two tumors (4). However, it should be noted that in addition to the SSTR2 level of expression, other mechanisms are involved in the sensitivity of GH-secreting adenomas to SRIF analogs. In fact, despite a better octreotide sensitivity of gsp+ vs. gsp- adenomas, SSTR2 mRNA is not different in the two categories of tumors (46). In the GH-secreting tumors partially responsive to octreotide or lanreotide, can we expect an additive suppressive effect on GH secretion when the SSTR2 and SSTR5 analogs are combined? Indeed, in a previous report of three cultured octreotide- or lanreotide-resistant adenomas (8), the combination of both BIM-23197 and BIM-23268 was found to significantly improve the inhibition of GH. Such additivity was nevertheless partial, as an only a 10% greater inhibitory effect was obtained by the combination of both compounds compared with the maximal GH inhibition achieved by one of the drugs alone. In our three cases, combining SSTR2- and SSTR5-preferential analogs did not produce a significant synergistic effect in suppressing GH. Nevertheless, at submaximal concentrations, a partial additivity of BIM-23197 and BIM-23268 on GH suppression was observed. Such data can be explained by the fact that such a drug combination results in binding affinities of 0.19 and 0.42 nmol/L, respectively, for the SSTR2 and SSTR5 subtypes, which are similar to those of native SRIF28. Furthermore, in our studies, both BIM-23197 and 23268 produced a greater maximal PRL suppression than that obtained using octreotide. Due to the variable tumor cell phenotypes and to the variable distribution of SSTR2 and SSTR5 transcripts, the combination of both analogs may mimic the effects of the native SRIF more accurately than octreotide or lanreotide, and could allow better control of hormone hypersecretion in acromegalic tumors.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received June 25, 1999.
Revised October 8, 1999.
Accepted October 15, 1999.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Acromegaly Study Group. 1990 Long term effect
of incremental doses of the somatostatin analog SMS 201–995 in 58
acromegalic patients. J Clin Endocrinol Metab. 71:391–397.[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  T. Florio, F. Barbieri, R. Spaziante, G. Zona, L. J Hofland, P. M van Koetsveld, R. A Feelders, G. K Stalla, M. Theodoropoulou, M. D Culler, et al. Efficacy of a dopamine-somatostatin chimeric molecule, BIM-23A760, in the control of cell growth from primary cultures of human non-functioning pituitary adenomas: a multi-center study Endocr. Relat. Cancer, June 1, 2008; 15(2): 583 - 596. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Plockinger, S. Albrecht, C. Mawrin, W. Saeger, M. Buchfelder, S. Petersenn, and S. Schulz Selective Loss of Somatostatin Receptor 2 in Octreotide-Resistant Growth Hormone-Secreting Adenomas J. Clin. Endocrinol. Metab., April 1, 2008; 93(4): 1203 - 1210. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. L. Fougner, J. Bollerslev, F. Latif, J. K. Hald, T. Lund, J. Ramm-Pettersen, and J. P. Berg Low Levels of Raf Kinase Inhibitory Protein in Growth Hormone-Secreting Pituitary Adenomas Correlate with Poor Response to Octreotide Treatment J. Clin. Endocrinol. Metab., April 1, 2008; 93(4): 1211 - 1216. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. F Taboada, R. M Luque, L. V. Neto, E. d. O Machado, B. C Sbaffi, R. C Domingues, J. B Marcondes, L. M C Chimelli, R. Fontes, P. Niemeyer, et al. Quantitative analysis of somatostatin receptor subtypes (1-5) gene expression levels in somatotropinomas and correlation to in vivo hormonal and tumor volume responses to treatment with octreotide LAR Eur. J. Endocrinol., March 1, 2008; 158(3): 295 - 303. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Resmini, P. Dadati, J.-L. Ravetti, G. Zona, R. Spaziante, A. Saveanu, P. Jaquet, M. D. Culler, F. Bianchi, A. Rebora, et al. Rapid Pituitary Tumor Shrinkage with Dissociation between Antiproliferative and Antisecretory Effects of a Long-Acting Octreotide in an Acromegalic Patient J. Clin. Endocrinol. Metab., May 1, 2007; 92(5): 1592 - 1599. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Sharif, L. Gendron, J. Wowchuk, P. Sarret, J. Mazella, A. Beaudet, and T. Stroh Coexpression of Somatostatin Receptor Subtype 5 Affects Internalization and Trafficking of Somatostatin Receptor Subtype 2 Endocrinology, May 1, 2007; 148(5): 2095 - 2105. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. van der Hoek, S. W J Lamberts, and L. J Hofland Preclinical and clinical experiences with the role of somatostatin receptors in the treatment of pituitary adenomas Eur. J. Endocrinol., April 1, 2007; 156(suppl_1): S45 - S51. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. F Taboada, R. M Luque, W. Bastos, R. F C Guimaraes, J. B Marcondes, L. M C Chimelli, R. Fontes, P. J P Mata, P. N. Filho, D. P Carvalho, et al. Quantitative analysis of somatostatin receptor subtype (SSTR1-5) gene expression levels in somatotropinomas and non-functioning pituitary adenomas Eur. J. Endocrinol., January 1, 2007; 156(1): 65 - 74. [Abstract] [Full Text] [PDF]
|
|
|
|
