Regulatory Peptides, 37 (1992) 213-226
213
© 1992 Elsevier Science Publishers B.V. All rights reserved. 0167-0115/92/$05.00 REGPEP 01138
Vasoactive intestinal peptide: autocrine growth factor in neuroblastoma M. Sue O'Dorisio, Daniel J. Fleshman, Stephen J. Q u a l m a n and T h o m a s M. O'Dorisio Departments of Pediatrics, Pediatric Pathology and Internal Medicine, The Ohio State University, College of Medicine, Columbus, OH (U.S.A.)
(Received 25 July 1991; revisedversion received25 October 1991; accepted 28 October 1991) K e y words: Vasoactive intestinal peptide; Neuroblastoma; Tumor
Summary Neuroblastoma is the most common solid tumor of children less than 5 years of age; yet the biology of this tumor is poorly understood. Neuroblastoma tumors are derived from neural crest precursors; they synthesize both adrenergic and peptidergic neurotransmitters. This study determined VIP receptor expression in primary neuroblastoma tumors prior to chemotherapy. The VIP receptor was expressed in 12 of 15 neuroblastoma tumors as determined by direct binding studies (KD = 1.3-12.4 nM) and VIP-mediated stimulation of adenylate cyclase. The VIP stimulation index for adenylate cyclase in the primary tumor was inversely correlated with the VIP content of the tumor, suggesting that VIP regulates its own receptor expression. Similar observations were made in vitro by comparison of two human neuroblastoma cell lines, IMR32 and SKNSH. Both cell lines were demonstrated to express specific, high affinity VIP receptors (K D = 4 nM and 2.5 nM for IMR32 and SKNSH, respectively). IMR32 cells contained very low levels of VIP (0.6 pg VIP/106 cells). Exogenous VIP stimulated adenylate cyclase 22-fold over basal activity and VIP inhibited proliferation of IMR32 cells by 49% in 6-day cultures. On the other hand, SKNSH cells synthesized high levels of VIP (6.3 pg/106 cells), metabolized VIP rapidly and demonstrated a low level of VIP-mediated stimulation of adenylate cyclase; their proliferation rate was minimally inhibited by exogenous VIP. These observations help validate the hypothesis that VIP serves as an autocrine growth factor in neuroblastoma.
Correspondence: M. Sue O'Dorisio, Department of Pediatrics, The Ohio State University,700 Children's Drive, Columbus, OH 43205, U.S.A.
214 Introduction
Overall, the long term survival of children with malignancies has improved rapidly over the past 30 years [ 1]; however, neuroblastoma remains a poorly controlled and little understood malignancy [2,3]. Diagnosis late in the disease course is common, with more than 70~o of tumors in older children being advanced in stage at presentation [4]. Although a 5 year survival for advanced disease varies from 44~o to 65~o depending on the staging system [5], little impact on long-term survival has occurred in the past three decades [ 1]. Early analysis of ongoing clinical trials suggests that aggressive chemotherapy alone is as effective as chemotherapy combined with autologous bone marrow transplantation [6,7]. Although both modes of therapy have prolonged survival slightly, late relapses are common. Thus, an ultimate impact on cure rates in neuroblastoma will require a better understanding of the biology of neuroblastoma cells. Neuroblastoma tumors are composed primarily of cells resembling neuroblasts and more mature glial or Schwann type cells which are thought to derive from a common neural crest precursor cell [8]. The neuroblasts actively synthesize transmitters of the sympathetic nervous system, including epinephrine and norepinephrine and express ~-adrenergic receptors [ 8,9]. More recent evidence has documented synthesis of several peptide neurotransmitters including neuropeptide Y, substance P, somatostatin and vasoactive intestinal peptide in human neuroblastoma cell lines and in primary tumors [ 10,11,44]. VIP is a 28 amino acid peptide originally isolated from porcine intestine [12]; the VIP gene was originally cloned from a human neuroblastoma cell line [13], and watery diarrhea syndrome has been associated with neuroblastoma [ 14], suggesting that some neuroblastomas may express this gene in vivo. VIP gene expression has also been observed in adrenal chromaffin cells which derive from the same neural crest precursor cell as sympathetic neurons [ 15-17 ]. The actions of VIP as a neurotransmitter or neuromodulator are mediated via binding of VIP to high affinity receptors on the respective target cells with activation of adenylate cyclase and protein kinase A [18]. VIP appears to have growth factor potential in developing nervous tissue, but whether this is a direct effect or the result of VIP-mediated release of other known growth factors such as nerve growth factor and epidermal growth factor is not known [ 19,20]. VIP has been proposed as an autocrine growth factor in neuroblastoma based on recent observations that VIP induces differentiation of neuroblastoma cell lines [21] and that some neuroblastoma cell lines both secrete VIP and express the VIP receptor [ 10]. The hypothesis that VIP is an autocrine growth factor has not yet been tested in primary human neuroblastoma tumor tissue. The present study was undertaken to evaluate whether any human neuroblastoma tumors express both VIP and VIP receptors in situ and to assess the effects of VIP on proliferation of neuroblastoma tumor cell lines.
215 Materials and Methods
Tumor procurement Neuroblastoma tumor tissue was obtained from the Cooperative Human Tissue Network. All patients were untreated prior to surgery. Immediately after removal of the tumors, sections were taken for histological examination; the remaining tissue was placed on ice and frozen at - 80 °C within 60 min. VIP receptor binding Membranes were prepared by our modification of the technique of Kaslow and coworkers [22,23]. Briefly, tissue suspended at 10 mg/ml (or 50" 106 cultured cells/ml) in buffer A (20 mM Hepes, 2 mM MgCI2, 5 mM EDTA, 1 mM 2-mercaptoethanol, 150 mM NaC1, 50 #g/ml phenylmethylsulfonylfluoride (PMSF) (pH 7.4) was disrupted by Brinkman polytron followed by centrifugation at 750 g for 5 min. The pellet was resuspended in 1/2 the original volume of buffer and again disrupted by polytron for 30 s. After a second 750 g spin, the two supernatant fractions were combined and centrifuged for 20 rain at 48,000 g. The resulting particulate fraction was washed in the original volume of buffer A and centrifuged 20 rain at 48,000 g. The particulate membrane fraction was resuspended in buffer A and stored at - 80 °C until use. Binding studies were performed as described by O'Dorisio et al. [24] using 100-200 #g membrane protein in buffer A. Competition experiments for estimation of receptor affinity (KD) and number (Bm~x) utilized 50 pM [125I]VIP (specific activity 2000 Ci/mmol, New England Nuclear) and increasing concentrations of unlabeUed VIP in a total volume of 0.5 ml. Reactions were started by the addition of membrane and the vials then kept in a shaking controlled temperature waterbath. The reaction was stopped by layering triplicate 0.15 ml aliquots of binding mixture over GF/C filters presoaked in 0.3 % polyethylenimine; f'dters were washed three times in 3 ml buffer A containing 0.2~o bovine serum albumin (BSA). Radioactivity bound to filters was quantified in a Beckman Gamma 5500. Bmax and K D were determined by non-linear computer assisted fit of a single-site model using the equation of Akera and Cheng [25]. Adenylate cyclase assay Enzymatic activity was assayed on tissue homogenates as previously described [26]. Briefly, the reaction was initiated by addition of 25-50 #g homogenate protein in buffer A to assay media containing 1 mM ATP, 2 mM MgCI2, 10 mM phosphocreatine, 3 U creatine kinase and 100 mM Tris-HCl (pH 7.7) (all reagents purchased from Sigma) in a total volume of 0.1 ml for 30 rain at 37 ° C. The reaction was terminated by addition of 0.1 ml of 0.1 M sodium acetate (pH 4.0), followed by heating at 90 °C for 3 rain; cAMP generated in the reaction was quantified by radioimmunoassay using specific antisera generated in our laboratories [27]. Culture of lMR32 and SKNSH cell lines Two human neuroblastoma cell lines, IMR-32 [28] and SKNSH [29], were obtained from American Type Culture Collection, Rockville, MD and grown in 75 cmz flask (Coming) in culture media consisting of minimal essential medium (MEM) with 15
216 heat-inactivated fetal bovine serum (FBS), 100 U/ml penicillin, 100 #g/ml streptomycin and 12 #l/ml nonessential amino acids. All components were purchased from Gibco, Grand Island, NY. Cells were incubated in a humidified atmosphere of 95~o air and 5~o COz at 37 °C. Cells were subcultured once a week. In proliferation experiments, cells were removed from the plate by trypsinization and enumerated on a Coulter counter.
Extraction of cellsfor peptide radioimmunoassay Cultured cells were suspended in buffer A at a concentration of 50.106 cells/ml buffer and disrupted with a Brinkman polytron. Homogenates were mixed (1 : 1) with 50~o EtOH-0.05 M acetic acid and extracted 18 h at 4 °C followed by dilution with an equal volume of 0.05 M acetic acid and centrifugation 20 min at 10,000 g. The resulting supernatant was lyophilized. The residue was dissolved in 2 ml of 0.05 M acetic acid for radioimmunoassay. VIP was quantified according to the method of O'Dorisio [30]. Synthetic porcine VIP antibody ( # 674, CalBioChem, San Diego, CA) was used. This antibody recognizes at least two forms of VIP and is N-terminally directed. Intra and interassay coefficients of variation are 5 and 17%, respectively. No cross reactivity is demonstrated with peptide histidine isoleucine, secretin, glucagon, growth hormone releasing factor or gastric inhibitory peptide. Highly purified porcine [125I]VIP was purchased from New England Nuclear (Boston, MA) and porcine synthetic VIP antigen from Peninsula Laboratories (San Carlos, CA).
Results
Competitive binding studies were performed on plasma membranes prepared from primary neuroblastoma tumors removed at diagnosis and before chemotherapy. Specific binding of [~25I]VIP was observed in 9 of 15 membrane preparations; the VIP receptor expressed in these tumors appears to be of high affinity with affinity constants (KD) ranging from 1.3-12.4 nM (Fig. 1 and Table I). Three tumors ( # 1389, # 1181 and # 3332) each bound more than 0.5 ~ of the total added [ ~25I]VIP with estimated K D values > 200 nM (Table I); these tumors were considered negative for VIP receptor expression. The presence or absence of VIP receptors did not correlate with stage of disease (P > 0.2). Further analysis of these VIP binding sites was performed to determine if the receptors were linked to a signal transduction pathway. Basal adenylate cyclase activity as well as VIP and guanine nucleotide stimulated activities were quantified in each tumor with the results shown in Table II. The basal adenylate cyclase activity was highly correlated with VIP levels; basal activity and VIP levels appeared to increase concomittantly (Fig. 2A, r -- 0.8, P < 0.01). Guanine nucleotide stimulation of adenylate cyclase was observed in each tumor, indicating that the G-protein-adenylate cyclase complex is intact in these tumors (Table II). Eight of the tumors including three tumors in which insufficient tissue was available for direct binding studies ( # 2998, # 3192 and # 2264 in Table I) demonstrated VIP-
217 Competitive Binding of VIP t o NeuPoblastomo Tumor 15000
tO000
5000 .x
II .
it
,..,
.
iO
-log
\
o
X
9 8 7 [VIP]. N
6
.
.
Fig. 1. Competitive binding of [m25I]VIP to neuroblastoma tumor membranes. Tumor tissue was placed in buffer A (10 mg/ml) and disrupted by polytron. Plasma membranes were prepared and competitive binding performed as described in Materials and Methods using 150/~g membrane protein per point. Binding affinities are as listed in Table I. TABLE I VIP receptor expression in neuroblastoma Tumor No.
Stage
mo F/U"
Receptor binding K D (nM)
107GN b 0864 2001 1389 1181 3332 3014 2998 3192 3129 4277 4386 1452 3023 2264
II II I III I II II III IV II IV III III III III
26 (wod) 8 (wod) 30 (wod) 12 (D) 16 (wod) 50 (wod) 52 (wod) 29 (wod) 9 (wod) 52 (wod) 12 (D) 17 (D) 34 (wod) 32 (wod) 33 (D)
7.4 9.6 6.5 c c c 4.0 n.d. n.d. 4.7 3.0 12.4 1.3 2.4 n.d.
Bm.~ (nM) 1.1 1.04 1.6 _ 1.3
1.4 0.2 1.3 1.3 1.7
a wed, without disease; D, dead. b Ganglioneuroma component of composite neuroblastoma tumor. c No detectable specific binding. n.d., not done due to insufficient tumor tissue. Tumor tissue was homogenized in buffer A and plasma membrane prepared by differential centrifugation. Competitive binding was performed as described in Fig. 1.
218 T A B L E II VIP-mediated stimulation of adenylate cyclase in neuroblastoma tumor Tumor
VIP (pg/~g D N A )
Adenylate cyclase (pmol cAMP/mg protein per min) basal
107GN a 0864 2001 1389 1181 3332 3014 2998 3192 3129 4277 4386 1452 3023 2264
150.5 34.0 23.4 10.0 9.5 2.6 1.2 1.2 0.6 0.5 0.2 0.0 0.0 0.0 0.0
42.0 5.9 20.7 0.29 15.3 3.4 6.4 1.2 0.36 5.4 2.8 7.1 4.3 1.6 0.94
+ 7.3 + 0.6 +_ 0.2 +3.2 + 0.6 + 0.9 + 0.2 + 0.08 + 1.7 + 0.2 -+ 1.5 + hl + 0.3 + 0.28
1 # M VIP
10 # M G p p N H p
28.2 6.4 17.2 0.25 14.8 2.4 9.5 3.3 1.4 15.8 3.3 11.5 9.0 2.6 2.1
201.8 17.2 48 0.69 53.3 9.8 22.9 9.1 2.6 12.7 6.6 12.4 18.0 8.4 4.0
_+ 5.6 + 0.7 + 0.04 + 1.4 +0.4 + 0.9* + 0.1" + 0.3* _+6.4* + 0.5 + 0.6* -+2.0* + 0.6* + 0.1"
_+ 42.0* + 1.0" + 0.12' +9.3* + 1.4' + 3.8* + 1.9" + 0.7* +4.2* _+ 1.4' + 3.2* +0.9* + 0.8* + 1.0"
Tumor tissue was homogenized in buffer A (10 mg tissue/ml buffer) and adenylate cyclase activity quantified as described in Materials and Methods using 20-40 #g protein per assay. Values are mean + S.D. for triplicate assays. * Statistically different from basal, P < 0.05. a Ganglioneuroma foci of composite neuroblastoma tumor.
mediated activation of adenylate cyclase (Table II), thus confirming the expression of VIP receptors in these tumors. Together the binding studies and adenylate cyclase studies demonstrated the presence of VIP receptors in 12 of 15 tumors (Tables I and II). The three tumors which did not demonstrate high affinity binding of VIP ( # 1389, # 1181 and # 3332 in Table I), also failed to demonstrate VIP-mediated activation of adenylate cyclase (Table II). Four tumors, # 107GN, #0864, #2001 and #4277, expressed high affinity VIP receptors with affinity constants of 7.4, 9.6, 6.5 and 3.0 nM, respectively, but did not show VIP stimulation of adenylate cyclase (Tables I and II). The receptors in these cells are apparently uncoupled from the adenylate cyclase complex, suggesting that the high VIP content in these tumors induces down regulation of the receptor from the signal transduction mechanism [31,32]. VIP did not stimulate adenylate cyclase in the homogenate of any tumor which contained endogenous VIP > 2 pg/#g DNA (Table II). The VIP content of the tumors showed a high inverse correlation with VIP-mediated stimulation of adenylate cyclase in each of the tumors studied (Fig. 2B, P < 0.005 using Fisher's exact test). A similar correlation between receptor expression, signal transduction and VIP levels can be seen in neuroblastoma cell lines. Both the SKNSH and IMR32 cell lines express high affinity VIP receptors (Table III). IMR32 cells have very low or non-detectable VIP levels, but exhibit a VIP-adenylate cyclase stimulation index of 22, whereas SKNSH
219 Relationship between VIP and Basal Adenylate Cyclase Activity 50 A 40 4( 30
"S T,,
20
.•..
1C
0.1
•
-
1
1-o
Ioo
VIP (Pg/Ng DNA) Relationship between VIP and Adenylate Cyclase Stimulation Index
B 4
c M
3
_a 2 E
0
.
0.1
.
.
.
.
.
.
.
.
'
.
1
.
.
.
.
.
.
.
.
10
.
.
,
.....
i
100
VIP (pg/~Jg DNA)
Fig. 2. Effect of tumor VIP content on basal VIP and stimulated adenylate cyclase activity in neuroblastoma tumors. Tumor tissue was disrupted as described in Fig. 1. Cell free protein (10-20 #gin) was incubated with adenylate cyclase media containing Mg-ATP _+1 #M VIP as described in Materials and Methods. The cAMP generated in the reaction was quantified by radioimmunoassay. A second aliquot of the cell-free protein in buffer A was extracted for quantification of VIP content by radioimmunoassay as described in Materials and Methods. (A) Basal activity was measured in the absence of VIP. Data were fit by 2nd order polynomial for basal adenylate cyclase activity regressed on log VIP content. (B) Adenylate cyclase activity was measured in the presence of 1 #M VIP. The adenylate cyclase stimulation index is the ratio of VIP-stimulated : basal adenylate cyclase activity using the values shown in Table II. Data were analyzed using Fisher's exact test.
cells have higher VIP levels and a much lower VIP-adenylate cyclase stimulation index (Table III). Earlier studies by Muller et al. [ 10] demonstrated higher levels of VIP in more differentiated SKNSH cell lines compared to neuroblast-like subcultures. The VIP-mediated stimulation of adenylate cyclase observed in SKNSH cultures may reflect this heterogeneity of cell type which includes both neuroblast-like and Schwanoma-like cells [33,34]. The rate of VIP metabolism was examined in neuroblastoma cell lines. Both IMR32 and SKNSH cells were incubated in the presence of 1 #M VIP and the amount of VIP remaining in the culture was quantified by RIA. IMR32 cells appear to degrade (or metabolize) VIP quite slowly; VIP remained immunologically recognizable for 96 h in
220 TABLE III VIP and VIP receptors in neuroblastoma cell lines KD
SKNSH IMR32
(nM)
pg VIP/106 cells
2.5 + 1.0 4.0 _+ 2.1
Adenylate cyclase SI a
6.3 + 2.5 0.6 + 1.10"
VIP
Forsk
5.3 + 1.1 22.2 _+ 3.8
6.3 + 2.2 8.2 + 0.8
* Values not detectable in RIA considered as 0. a Adenylate cyclase stimulation index in the ratio of VIP-stimulated: basal activity. Cells were subcultured as described in Materials and Methods, allowed to grow to confluency (7 days) and washed two times with serum free media. Competitive binding at 17 °C was performed 30 min using 150 #g membrane protein. VIP was extracted and quantified by RIA as described in Materials and Methods. Adenylate cyclase activity assayed 30 min at 37 °C using 10-20 #g cell free protein. Values are mean _+ S.D. for three cultures of each cell type.
VIP Metobol ism/Degradot i on in Neuroblastomo Cell Lines 105(
"~---~''~'~
lo 4
~
1 03 o_ >
102
lo 1 0
1
2
3
4
5
6
Time External VIP Inoubation -0-
SKNSH
+IMR-32
[days]
~
Media Alone
Detection of VIP Fragments By Radioimmunoassoy 1O0
8O 70
0
60
[] O
5o 40
O 0
30
20 10
[] O
B
~o
~oo
~ooo
~oooo
%o000
pmol VlP
•
VIP
0
VIP 4_
0
VlPlo_
Fig. 3. (A) Metabolism of VIP by neuroblastoma cell lines in culture. S K N S H and IMR32 cells were grown in culture as described in Materials and Methods. VIP (I #M) was added to each well (containing 1.105 cells or media alone) on day 0. Cells and media were harvested at indicated times and extracted for quantification of VIP as described in Materials and Methods. (B) Radioimmunoassay ofVIP. VIP standard curve with VIP4_28 and VIPlo_2s run as unknowns. Peptides were purchased from Bachem.
221 TABLE IV Interaction of VIP and VIP fragments with VIP receptor on neuroblastoma cells
VIP
VIP4_28 VIPlo_28
KD (nM)
Adenylate cyclase SI a
2.9 + 1.2 1.5 + 0.8 110 + 79
3.8 + 0.6 1.l + 0.l 0.96 + 0.03
a Adenylate cyclase stimulation index is the ratio of VIP-stimulated: basal activity. SKNSH cells were subcultured as described in Materials and Methods, allowed to grow to confluency (7 days) and washed twice with serum free media. Competitive binding was performed for 30 min at 17 °C using 150/~g cell free protein. Adenylate cyclase was measured as described in Materials and Methods. All peptides were added at 1 #M concentrations. Values are mean + S.D. (n = 3).
the culture with a d e c a y r a t e e q u i v a l e n t to the d e c a y rate in m e d i a a l o n e (Fig. 3A). E x o g e n o u s V I P w a s rapidly lost f r o m S K N S H m e t a b o l i z e d in the rlrst 24 h. (Fig. 3A).
cultures w i t h 99~o h y d r o l y z e d o r
S e v e r a l proteolytic f r a g m e n t s o f V I P h a v e b e e n d e m o n s t r a t e d in cell cultures [ 3 5 - 3 7 ] . T w o o f t h e s e f r a g m e n t s , VIP4_28 a n d VIPlo_28 w e r e tested in the r a d i o i m m u n o a s s a y
SKNSH
0
A i
i
i
i
i
IMR32
B
Days of culture
Fig. 4. Effect o f exogenous VIP on proliferation ofneuroblastoma cell lines in culture. S K N S H and I M R 3 2
cells were grown in tissue culture in the presence of no VIP (e), 1 #M VIP ( I ) , or 10 #M VIP (A). Peptide was added on day 0 and at the indicated times, cells were released from the plates by addition of trypsin and cells enumerated by Coulter counter. Viability was > 90% for all cultures. (A) SKNSH cells and (B) IMR32 cells; values shown are mean + S.D. for three independent experiments.
222 (Fig. 3B) and examined for receptor affinity as well as for their ability to induce signal transduction. VIP and the peptide fragment VIP4_2s are recognized with equal potency in this assay, but the carboxy-terminal fragment VIPIo_2s is not measurable except at very high concentrations (Fig. 3B). Smaller carboxy-terminal fragments such as VIP17_28 and VIP23_28 although known degradation products [35-37] would not be measurable in this assay [38]. VIP4_28 binds to the VIP receptor, but does not activate adenylate cyclase; VIP1o_2s shows very low affinity for the VIP receptor and has no demonstrable against activity in the adenylate cyclase assay (Table IV). These results suggest (Fig. 3 and Table IV) that SKNSH cells degrade VIP rapidly to fragments smaller than VIP~o_2s, which are not effective ligands for the VIP receptor. Thus, effects of VIP on proliferation of neuroblastoma cells appear to be dependent on an intact VIPl_28 molecule as has been shown in intestine, heart and lung [39-41]. Further experiments were conducted to determine if VIP affects the viability and/or rate of growth of the two neuroblastoma tumor cell lines. 1 #M VIP added to the media on day one had no significant effect on the viability or proliferation rate of SKNSH cells, but inhibited the proliferation of IMR32 cells by 30~o over a 6 day culture period (P < 0.02) (Fig. 4). At higher concentrations (10 #M) VIP inhibited the proliferation of SKNSH by 20~ (P < 0.05) and inhibited IMR32 proliferation by 49~o (P < 0.01) (Fig. 4). The morphology of IMR32 cells incubated with VIP for six days showed less confluent cultures with increased neurite extension (Fig. 5). These results correlate well with the recent observation of Pence and Shorter that VIP induces differentiation in IMR32 cells in culture [21].
Fig. 5. Effect of exogenous VIP on morphology ofIMR32 neuroblastoma cells in culture. IMR32 cells were grown under identical conditions as in Fig. 4 and photographed under phase contrast microscopy (40 x ) on day 6. Viability was >90% for all cultures. A, no peptide; B, 1 #M VIP; C, 10/~M VIE
223
Discussion
These studies have provided the first evidence that human neuroblastoma tumors both synthesize (or concentrate) VIP and express VIP receptors in vivo, thus fulfilling a basic requirement for autocrine growth regulation. In addition, endogenous tumor VIP levels appear to regulate the ability of the VIP receptor to transduce intracellular signals via the adenylate cyclase complex, with high VIP levels resulting in uncoupling of the receptor-Gs-cyclase pathway.
224 Detection of VIP in tumor tissue by RIA is not absolute proof of VIP gene expression by the malignant cells; however, uptake of VIP from the circulation is unlikely because the levels of VIP in serum are extremely low except in individuals with VIP secreting tumors [38]. Likewise, release of VIP by peptidergic neurons in close proximity to the tumor is unlikely to account for the high levels observed in these tumors. In addition, at least three human neuroblastoma cell lines ( S K N S H , IMR32, NB-OK-1) synthesize VIP [ 11 ], (Table III). The current data also might be explained by a paracrine action of VIP, i.e., a subpopulation of tumor cells which synthesize and secrete VIP which binds to VIP receptors on adjacent tumor cells. VIP decreases the proliferation rate of a neuroblastoma cell line which does not hydrolyze VIP rapidly, but has little effect in a cell line which appears to degrade VIP rapidly. This observation supports the hypothesis that VIP can serve as an autocrine growth factor in human neuroblastoma and suggests several steps at which an aberration of the normal VIP regulation of proliferation or differentiation might contribute to malignant transformation. Overexpression or amplification of a protease gene might account for the rapid hydrolysis of VIP in S K N S H cells and thus inhibit normal VIP effects on neuronal cell differentiation. Similarly, mutations in the normal VIP gene might result in lower VIP gene expression and a failure in the differentiation pathway. Mutations in the VIP receptor gene might prevent the normal VIP mediated signal transduction and also result in deregulation of a normal proliferation/differentiation pathway. The recent cloning of both the VIP [42] and VIP receptor [43] genes will allow for direct testing of these hypotheses.
Acknowledgements This work was supported by grants from the National Cancer Institute (RO1-CA41997 to M.S.O.) and The Ohio State University Comprehensive Cancer Center (S.J.Q.). The authors wish to thank Jadwiga Labanowska, Brent Howe, Laura Campolito and Monica Summers for expert technical assistance; John Hayes for advice on the statistical analysis of data and Debbie Fine for preparation of the manuscript.
References 1 Birch, J.M., Marsden, H.B., Morris-Jones, P.H., Pearson, D. and Blair, V., Improvements in survival from childhood cancer: results of a population based surveyover 30 years, Br. Med. J., 296 (1988) 1372. 2 Woods, W.G. and Tuchman, M., Neuroblastoma: the case for screening infants in North America, Pediatrics, 79 (1987) 869-873. 3 Kushner, B.H. and Cheung, N.-K.V., Treatment of neuroblastoma, Cancer, 2 (1988) 1-11. 4 Hayes, F.A. and Smith, E.I., Neuroblastoma. In P.A. Pizzo and D.G. Poplack (Eds.), Principles and Practice of Pediatric Oncology,J.B. Lippincott, Philadelphia, pp. 607-622, 1989. 5 Evans,A.E., D'Angio,G.J., Sather, H.N., De Lorimier,A.A., Dalton, A., Ungerleider, R.S., Finklestein, J.Z. and Hammond, G.D., A comparison of four staging systems for localized and regional neuroblastoma: a report from the Children's Cancer Study Group, J. Clin. Oncol., 8 (1991) 678-688. 6 Philip, T., Chauvin, F. and Michon, J., A pilot study of double A.B.M.T. in advanced neuroblastoma. In K. Dicke, G. Spitzer and S. Jagannath (Eds.), International Conferenceon Autologous Bone Marrow Transplantation, Houston, The University of Texas, pp. 799-805, 1989.
225 70'Leary, M., Sather, H., Ramsay, N., Harris, R., Strandjord, S., Romansky, S., Haase, G., Shimada, H., Novak, L., Siegel, S., Harm, H.W., Brodeur, G., Secger, R. and Hammond, D., Intensive chemotherapy for poor prognosis neuroblastoma, Prec. Am. Ass. Cancer Res., 31 (1990) 201 (abstract). 8 Ciecarone, V., Spengler, B.A., Meyers, M.B., Biedler, J.L. and Ross, R.A., Phenotype diversification in human neuroblastoma cells: expression ofdistinct neural crest lineages, Cancer Res., 49 (1989) 219-225. 9 Baron, B.M. and Siegel, B.W., Alpha-2-adrenergic and muscarinic cholinergic receptors have opposing actions on cyclic AMP levels in SK-N-SH human neuroblastoma cells, J. Neurochem., 53 (1989) 602-609. 10 Muller, J.M., Lolait, S.J., Yu, V.C., Sadee, W. and Waschek, J.A., Functional vasoactive intestinal polypeptidc (VIP) receptors in human neuroblastoma subclones that contain VIP precursor mRNA and release VIP-like substances, J. Biol. Chem., 264 (1989) 3647-3650. 11 Hoshino, M., Yanaihara, C., Ogino, K., Iguchi, K., Sate, H., Suzuki, T. and Yanaihara, N., Production of VIP- and PHM (human PHI)-related peptides in human neuroblastoma cells, Peptides, 5 (1984) 155-160. 12 Said, S.I. and Mutt, V., Polypeptide with broad biological activity: isolation from small intestine, Science, 169 (1970) 1217-1218. 13 Itoh, N., Obata, K., Yanaihara, N. and Okamoto, H., Human preprovasoactiv¢ intestinal peptide contains a novel, PHI-27-1ike peptide. PHM 27, Nature, 304 (1983) 547-549. 14 Funato, M., Fujimura, M., Shimada, S. et al., Rapid changes of serum vasoactive intestinal peptide after removel of ganglioneuroblastoma with watery diarrhea-hypokalemia-achlorhydria syndrome in a child, J. Ped. Gastroent. Nutr., 1 (1982) 131-135. 15 Anderson, D.J. and Axel, R., A bipotential neuroendocrine precursor whose choice of cell fate is determined by NGF and glucocorticoids, Cell, 47 (1986) 1079-1090. 16 Tischler, A.S., Dayal, Y., Balogh, K., Cohen, R.B., Connolly, J.L and Tallberg, K., The distribution of immunoreactive chromogranins, S-100 protein, and vasoactive intestinal peptide in compound tumors of the adrenal medulla, Hum. Pathol., 18 (1987) 909-917. 17 Eiden, L.E. and Hotchkiss, A.J., Cyclic adenosine monophosphate regulates vasoactive intestinal polypeptide and enkephalin biosynthesis in cultured bovine chromaffin cells, Neuropeptides, 4 (1983) 1-9. 18 Tseng, J. and O'Dorisio, M.S.,Mechanism ofvasoactiveintestinal peptide (VlP)-mediatedimmunoregulation. In E.J. Goetzl and N.H. Spector (Eds.), Neuroimmune Networks: Physiology and Diseases, Alan R. Liss, New York, pp. 105-111, 1989. 19 Gozes, I. and Brenneman, D.E., VIP: molecular biology and neurobiological function, Mol. Neurobiol., 3 (1989) 201-236. 20 Brenneman, D.E., Nieol, T., Warren, D. and Bowers, LM., Vasoactive intestinal peptide: a neurotrophic releasing agent and an astroglial mitogen, J. Neurosci. Res., 25 (1990) 386-394. 21 Pence, J.C. and Shorter, N.A., In vitro differentiation of human neuroblastoma cells caused by vasoacrive intestinal peptide, Cancer Res., 50 (1990) 5177-5183. 22 O'Dorisio, M.S., O'Dorisio, T.M., Wood, C.L., Beattie, M.S., Bresnahan, J.C. and Campolito, L.B., Characterization of VIP receptors in nervous and immune systems, Ann. NY Acad. Sci., 527 (1988) 257-265. 23 Kaslow, H.R., Farfel, Z., Johnson, G.L. and Bourne, H.R., Adenylate cyclase assembled in vitro: cholera toxin substrates determine different patterns of regulation by isoproterenol and guanosine 5'-triphosphate, Mol. Pharmacol., 15 (1979) 472-83. 24 O'Dorisio, M.S., Shannon, B.T., Fleshman, D.J. and Campolito, L.B., Identification of high affinity receptors for vasoactive intestinal peptide on human lymphocytes of B cell lineage, J. Immnnol., 142 (1989) 3533-3536. 25 Akera, T. and Cheng, V., A simple method for the determination of affinity and binding site concentration in receptor binding studies, Biochim. Biophys. Acta, 4790 (1979) 412-423. 26 O'Dorisio, M.S., Hermina, N., Balcerzak, S.P. and O'Dorisio, T.M., Vasoactive intestinal polypeptide stimulation of adenylate cyclase in purified human leukoeytes, J. Immunol., 127 (1981) 2551-2554. 27 O'Dorisio, M.S., Vandenbark, G.R. and Lo Buglio, A.F., Human monocyte killing ofS. a u r e u s : modulation by agonists of cAMP and cGMP, Infect. Immun., 26 (1979) 604-611. 28 Tumilowicz, J.J., Nichols, W.W., Cholon, J.J. and Greene, A.E., Definition of a continuous human cell line derived from neuroblastoma, Cancer Res., 30 (1970) 2110-2118.
226 29 Biedler, J.I., Nelson, I. and Spengler, B.A., Morphology and growth, tumorgenicity and cytogenetics of human neuroblastoma cells in continuous culture, Cancer Res., 33 (1973) 2643-2652. 30 O'Dorisio, T.M., Mekhjian, H.S., Ellison, E.C., O'Dorisio, M.S., Gaginella, T.S. and Woltering, E.A., Role ofpeptide radioimmunoassay in understandingpeptide-peptide interactions and clinical expression of gastroenteropancreatic endocrine tumors, Am. J. Med., 82 (1987) 60-67. 31 Sibley, D.R., Peters, J.R., Nambi, P., Caron, M.G. and Letkowitz, R.J., Desensitization of turkey erythrocyte adenylate cyclase, J. Biol. Chem., 259 (1984) 9742-9748. 32 Iyengar, R., Bhat, M.K., Riser, M.E. and Birnbaumer, L., Receptor-specific desensitization of the $49 lymphoma cell adenylyl cyclase, J. Biol. Chem., 256 (1981) 4810-4899. 33 Rettig, W.J., Spengler, B.A., Chesa, P.G., Old, L.J. and Biedler, J.L., Coordinate changes in neuronal phenotype and surface antigen expression in human neuroblastoma cell variants, Cancer Res., 47 (1987) 1383-1389. 34 Ross, R.A., Spengier, B.A. and Biedler, J.L., Coordinate morphological and biochemical interconversion of human neuroblastoma cells, J. Nat. Cancer Inst., 71 (1983) 741-747. 35 Goetzl, E.J., Kodama, K.T., Turck, C.W., Schiogolev, S.A. and Sreedharan, S.P., Unique pattern of cleavage of vasoactive intestinal peptide by human lymphocytes, Immunology, 66 (1989) 554-558. 36 Misbin, R.I., Wolfe, M.M., Morris, P., Buynitzky, S.J. and McGuigan J.E., Uptake of vasoactive intestinal peptide by rat liver, Am. J. Physiol., 243 (1982) G103-G111. 37 Goetzl, E.J., Sreedharan, S.P. and Turck, C.W., Structurally distinctive vasoactive intestinal peptides from rat basophilic leukemia cells, J. Biol. Chem., 263 (1988) 9083-9086. 38 O'Dorisio, T.M., Howe, B.A., Hill, D.S., Obermeyer, J. and Cataland, S., Vasoactive intestinal peptide (VIP) and neurotensin radioimmunoassays, J. Clin. Immunoassay, 10 (1987) 85-90. 39 Couvineau, A., Rouyer-Fessard, C., Fournier, A., St. Pierre, S., Pipkorn, R. and Laburthe, M., Structural requirements for VIP interaction with specific receptors in human and rat intestinal membranes: effect of nine partial sequences, Biochem. Biophys. Res. Commun., 121 (1984) 493-498. 40 O'Donnell,M., Garippa, R., O'Neill, N., Bolin, D. and Cottrell, J., Structure-activitystudies ofvasoactive intestinal polypeptide, J. Biol. Chem. 266 (1991) 6389-6392. 41 Fournier, A., Saunders, J.K. and St. Pierre, S., Synthesis, conformational studies and biological activities of VIP and related fragments, Peptides, 5 (1984) 169-177. 42 Tsukada, T., Horovitch, S.J., Montiminy, M.R., Mandel, G. and Goodman, R.H., Structure of the human vasoactive intestinal polypeptide gene, DNA, 4 (1985) 293-300. 43 Sreedharan, S.P., Robichon, A., Peterson, K.E. and Goetzl, E., Cloning and expression of the human vasoactive intestinal peptide receptor, Proc. Natl. Acad. Sci. USA, 88 (1991) 4986-4990. 44 Qualman, S.F., O'Dorisio, M.S., Fleshman, D.F., Shimada, H. and O'Dorisio, T.M., Neuroblastoma: correlation of neuropeptide and oncogene expression with other prognostic factors, Cancer, in press.
213
© 1992 Elsevier Science Publishers B.V. All rights reserved. 0167-0115/92/$05.00 REGPEP 01138
Vasoactive intestinal peptide: autocrine growth factor in neuroblastoma M. Sue O'Dorisio, Daniel J. Fleshman, Stephen J. Q u a l m a n and T h o m a s M. O'Dorisio Departments of Pediatrics, Pediatric Pathology and Internal Medicine, The Ohio State University, College of Medicine, Columbus, OH (U.S.A.)
(Received 25 July 1991; revisedversion received25 October 1991; accepted 28 October 1991) K e y words: Vasoactive intestinal peptide; Neuroblastoma; Tumor
Summary Neuroblastoma is the most common solid tumor of children less than 5 years of age; yet the biology of this tumor is poorly understood. Neuroblastoma tumors are derived from neural crest precursors; they synthesize both adrenergic and peptidergic neurotransmitters. This study determined VIP receptor expression in primary neuroblastoma tumors prior to chemotherapy. The VIP receptor was expressed in 12 of 15 neuroblastoma tumors as determined by direct binding studies (KD = 1.3-12.4 nM) and VIP-mediated stimulation of adenylate cyclase. The VIP stimulation index for adenylate cyclase in the primary tumor was inversely correlated with the VIP content of the tumor, suggesting that VIP regulates its own receptor expression. Similar observations were made in vitro by comparison of two human neuroblastoma cell lines, IMR32 and SKNSH. Both cell lines were demonstrated to express specific, high affinity VIP receptors (K D = 4 nM and 2.5 nM for IMR32 and SKNSH, respectively). IMR32 cells contained very low levels of VIP (0.6 pg VIP/106 cells). Exogenous VIP stimulated adenylate cyclase 22-fold over basal activity and VIP inhibited proliferation of IMR32 cells by 49% in 6-day cultures. On the other hand, SKNSH cells synthesized high levels of VIP (6.3 pg/106 cells), metabolized VIP rapidly and demonstrated a low level of VIP-mediated stimulation of adenylate cyclase; their proliferation rate was minimally inhibited by exogenous VIP. These observations help validate the hypothesis that VIP serves as an autocrine growth factor in neuroblastoma.
Correspondence: M. Sue O'Dorisio, Department of Pediatrics, The Ohio State University,700 Children's Drive, Columbus, OH 43205, U.S.A.
214 Introduction
Overall, the long term survival of children with malignancies has improved rapidly over the past 30 years [ 1]; however, neuroblastoma remains a poorly controlled and little understood malignancy [2,3]. Diagnosis late in the disease course is common, with more than 70~o of tumors in older children being advanced in stage at presentation [4]. Although a 5 year survival for advanced disease varies from 44~o to 65~o depending on the staging system [5], little impact on long-term survival has occurred in the past three decades [ 1]. Early analysis of ongoing clinical trials suggests that aggressive chemotherapy alone is as effective as chemotherapy combined with autologous bone marrow transplantation [6,7]. Although both modes of therapy have prolonged survival slightly, late relapses are common. Thus, an ultimate impact on cure rates in neuroblastoma will require a better understanding of the biology of neuroblastoma cells. Neuroblastoma tumors are composed primarily of cells resembling neuroblasts and more mature glial or Schwann type cells which are thought to derive from a common neural crest precursor cell [8]. The neuroblasts actively synthesize transmitters of the sympathetic nervous system, including epinephrine and norepinephrine and express ~-adrenergic receptors [ 8,9]. More recent evidence has documented synthesis of several peptide neurotransmitters including neuropeptide Y, substance P, somatostatin and vasoactive intestinal peptide in human neuroblastoma cell lines and in primary tumors [ 10,11,44]. VIP is a 28 amino acid peptide originally isolated from porcine intestine [12]; the VIP gene was originally cloned from a human neuroblastoma cell line [13], and watery diarrhea syndrome has been associated with neuroblastoma [ 14], suggesting that some neuroblastomas may express this gene in vivo. VIP gene expression has also been observed in adrenal chromaffin cells which derive from the same neural crest precursor cell as sympathetic neurons [ 15-17 ]. The actions of VIP as a neurotransmitter or neuromodulator are mediated via binding of VIP to high affinity receptors on the respective target cells with activation of adenylate cyclase and protein kinase A [18]. VIP appears to have growth factor potential in developing nervous tissue, but whether this is a direct effect or the result of VIP-mediated release of other known growth factors such as nerve growth factor and epidermal growth factor is not known [ 19,20]. VIP has been proposed as an autocrine growth factor in neuroblastoma based on recent observations that VIP induces differentiation of neuroblastoma cell lines [21] and that some neuroblastoma cell lines both secrete VIP and express the VIP receptor [ 10]. The hypothesis that VIP is an autocrine growth factor has not yet been tested in primary human neuroblastoma tumor tissue. The present study was undertaken to evaluate whether any human neuroblastoma tumors express both VIP and VIP receptors in situ and to assess the effects of VIP on proliferation of neuroblastoma tumor cell lines.
215 Materials and Methods
Tumor procurement Neuroblastoma tumor tissue was obtained from the Cooperative Human Tissue Network. All patients were untreated prior to surgery. Immediately after removal of the tumors, sections were taken for histological examination; the remaining tissue was placed on ice and frozen at - 80 °C within 60 min. VIP receptor binding Membranes were prepared by our modification of the technique of Kaslow and coworkers [22,23]. Briefly, tissue suspended at 10 mg/ml (or 50" 106 cultured cells/ml) in buffer A (20 mM Hepes, 2 mM MgCI2, 5 mM EDTA, 1 mM 2-mercaptoethanol, 150 mM NaC1, 50 #g/ml phenylmethylsulfonylfluoride (PMSF) (pH 7.4) was disrupted by Brinkman polytron followed by centrifugation at 750 g for 5 min. The pellet was resuspended in 1/2 the original volume of buffer and again disrupted by polytron for 30 s. After a second 750 g spin, the two supernatant fractions were combined and centrifuged for 20 rain at 48,000 g. The resulting particulate fraction was washed in the original volume of buffer A and centrifuged 20 rain at 48,000 g. The particulate membrane fraction was resuspended in buffer A and stored at - 80 °C until use. Binding studies were performed as described by O'Dorisio et al. [24] using 100-200 #g membrane protein in buffer A. Competition experiments for estimation of receptor affinity (KD) and number (Bm~x) utilized 50 pM [125I]VIP (specific activity 2000 Ci/mmol, New England Nuclear) and increasing concentrations of unlabeUed VIP in a total volume of 0.5 ml. Reactions were started by the addition of membrane and the vials then kept in a shaking controlled temperature waterbath. The reaction was stopped by layering triplicate 0.15 ml aliquots of binding mixture over GF/C filters presoaked in 0.3 % polyethylenimine; f'dters were washed three times in 3 ml buffer A containing 0.2~o bovine serum albumin (BSA). Radioactivity bound to filters was quantified in a Beckman Gamma 5500. Bmax and K D were determined by non-linear computer assisted fit of a single-site model using the equation of Akera and Cheng [25]. Adenylate cyclase assay Enzymatic activity was assayed on tissue homogenates as previously described [26]. Briefly, the reaction was initiated by addition of 25-50 #g homogenate protein in buffer A to assay media containing 1 mM ATP, 2 mM MgCI2, 10 mM phosphocreatine, 3 U creatine kinase and 100 mM Tris-HCl (pH 7.7) (all reagents purchased from Sigma) in a total volume of 0.1 ml for 30 rain at 37 ° C. The reaction was terminated by addition of 0.1 ml of 0.1 M sodium acetate (pH 4.0), followed by heating at 90 °C for 3 rain; cAMP generated in the reaction was quantified by radioimmunoassay using specific antisera generated in our laboratories [27]. Culture of lMR32 and SKNSH cell lines Two human neuroblastoma cell lines, IMR-32 [28] and SKNSH [29], were obtained from American Type Culture Collection, Rockville, MD and grown in 75 cmz flask (Coming) in culture media consisting of minimal essential medium (MEM) with 15
216 heat-inactivated fetal bovine serum (FBS), 100 U/ml penicillin, 100 #g/ml streptomycin and 12 #l/ml nonessential amino acids. All components were purchased from Gibco, Grand Island, NY. Cells were incubated in a humidified atmosphere of 95~o air and 5~o COz at 37 °C. Cells were subcultured once a week. In proliferation experiments, cells were removed from the plate by trypsinization and enumerated on a Coulter counter.
Extraction of cellsfor peptide radioimmunoassay Cultured cells were suspended in buffer A at a concentration of 50.106 cells/ml buffer and disrupted with a Brinkman polytron. Homogenates were mixed (1 : 1) with 50~o EtOH-0.05 M acetic acid and extracted 18 h at 4 °C followed by dilution with an equal volume of 0.05 M acetic acid and centrifugation 20 min at 10,000 g. The resulting supernatant was lyophilized. The residue was dissolved in 2 ml of 0.05 M acetic acid for radioimmunoassay. VIP was quantified according to the method of O'Dorisio [30]. Synthetic porcine VIP antibody ( # 674, CalBioChem, San Diego, CA) was used. This antibody recognizes at least two forms of VIP and is N-terminally directed. Intra and interassay coefficients of variation are 5 and 17%, respectively. No cross reactivity is demonstrated with peptide histidine isoleucine, secretin, glucagon, growth hormone releasing factor or gastric inhibitory peptide. Highly purified porcine [125I]VIP was purchased from New England Nuclear (Boston, MA) and porcine synthetic VIP antigen from Peninsula Laboratories (San Carlos, CA).
Results
Competitive binding studies were performed on plasma membranes prepared from primary neuroblastoma tumors removed at diagnosis and before chemotherapy. Specific binding of [~25I]VIP was observed in 9 of 15 membrane preparations; the VIP receptor expressed in these tumors appears to be of high affinity with affinity constants (KD) ranging from 1.3-12.4 nM (Fig. 1 and Table I). Three tumors ( # 1389, # 1181 and # 3332) each bound more than 0.5 ~ of the total added [ ~25I]VIP with estimated K D values > 200 nM (Table I); these tumors were considered negative for VIP receptor expression. The presence or absence of VIP receptors did not correlate with stage of disease (P > 0.2). Further analysis of these VIP binding sites was performed to determine if the receptors were linked to a signal transduction pathway. Basal adenylate cyclase activity as well as VIP and guanine nucleotide stimulated activities were quantified in each tumor with the results shown in Table II. The basal adenylate cyclase activity was highly correlated with VIP levels; basal activity and VIP levels appeared to increase concomittantly (Fig. 2A, r -- 0.8, P < 0.01). Guanine nucleotide stimulation of adenylate cyclase was observed in each tumor, indicating that the G-protein-adenylate cyclase complex is intact in these tumors (Table II). Eight of the tumors including three tumors in which insufficient tissue was available for direct binding studies ( # 2998, # 3192 and # 2264 in Table I) demonstrated VIP-
217 Competitive Binding of VIP t o NeuPoblastomo Tumor 15000
tO000
5000 .x
II .
it
,..,
.
iO
-log
\
o
X
9 8 7 [VIP]. N
6
.
.
Fig. 1. Competitive binding of [m25I]VIP to neuroblastoma tumor membranes. Tumor tissue was placed in buffer A (10 mg/ml) and disrupted by polytron. Plasma membranes were prepared and competitive binding performed as described in Materials and Methods using 150/~g membrane protein per point. Binding affinities are as listed in Table I. TABLE I VIP receptor expression in neuroblastoma Tumor No.
Stage
mo F/U"
Receptor binding K D (nM)
107GN b 0864 2001 1389 1181 3332 3014 2998 3192 3129 4277 4386 1452 3023 2264
II II I III I II II III IV II IV III III III III
26 (wod) 8 (wod) 30 (wod) 12 (D) 16 (wod) 50 (wod) 52 (wod) 29 (wod) 9 (wod) 52 (wod) 12 (D) 17 (D) 34 (wod) 32 (wod) 33 (D)
7.4 9.6 6.5 c c c 4.0 n.d. n.d. 4.7 3.0 12.4 1.3 2.4 n.d.
Bm.~ (nM) 1.1 1.04 1.6 _ 1.3
1.4 0.2 1.3 1.3 1.7
a wed, without disease; D, dead. b Ganglioneuroma component of composite neuroblastoma tumor. c No detectable specific binding. n.d., not done due to insufficient tumor tissue. Tumor tissue was homogenized in buffer A and plasma membrane prepared by differential centrifugation. Competitive binding was performed as described in Fig. 1.
218 T A B L E II VIP-mediated stimulation of adenylate cyclase in neuroblastoma tumor Tumor
VIP (pg/~g D N A )
Adenylate cyclase (pmol cAMP/mg protein per min) basal
107GN a 0864 2001 1389 1181 3332 3014 2998 3192 3129 4277 4386 1452 3023 2264
150.5 34.0 23.4 10.0 9.5 2.6 1.2 1.2 0.6 0.5 0.2 0.0 0.0 0.0 0.0
42.0 5.9 20.7 0.29 15.3 3.4 6.4 1.2 0.36 5.4 2.8 7.1 4.3 1.6 0.94
+ 7.3 + 0.6 +_ 0.2 +3.2 + 0.6 + 0.9 + 0.2 + 0.08 + 1.7 + 0.2 -+ 1.5 + hl + 0.3 + 0.28
1 # M VIP
10 # M G p p N H p
28.2 6.4 17.2 0.25 14.8 2.4 9.5 3.3 1.4 15.8 3.3 11.5 9.0 2.6 2.1
201.8 17.2 48 0.69 53.3 9.8 22.9 9.1 2.6 12.7 6.6 12.4 18.0 8.4 4.0
_+ 5.6 + 0.7 + 0.04 + 1.4 +0.4 + 0.9* + 0.1" + 0.3* _+6.4* + 0.5 + 0.6* -+2.0* + 0.6* + 0.1"
_+ 42.0* + 1.0" + 0.12' +9.3* + 1.4' + 3.8* + 1.9" + 0.7* +4.2* _+ 1.4' + 3.2* +0.9* + 0.8* + 1.0"
Tumor tissue was homogenized in buffer A (10 mg tissue/ml buffer) and adenylate cyclase activity quantified as described in Materials and Methods using 20-40 #g protein per assay. Values are mean + S.D. for triplicate assays. * Statistically different from basal, P < 0.05. a Ganglioneuroma foci of composite neuroblastoma tumor.
mediated activation of adenylate cyclase (Table II), thus confirming the expression of VIP receptors in these tumors. Together the binding studies and adenylate cyclase studies demonstrated the presence of VIP receptors in 12 of 15 tumors (Tables I and II). The three tumors which did not demonstrate high affinity binding of VIP ( # 1389, # 1181 and # 3332 in Table I), also failed to demonstrate VIP-mediated activation of adenylate cyclase (Table II). Four tumors, # 107GN, #0864, #2001 and #4277, expressed high affinity VIP receptors with affinity constants of 7.4, 9.6, 6.5 and 3.0 nM, respectively, but did not show VIP stimulation of adenylate cyclase (Tables I and II). The receptors in these cells are apparently uncoupled from the adenylate cyclase complex, suggesting that the high VIP content in these tumors induces down regulation of the receptor from the signal transduction mechanism [31,32]. VIP did not stimulate adenylate cyclase in the homogenate of any tumor which contained endogenous VIP > 2 pg/#g DNA (Table II). The VIP content of the tumors showed a high inverse correlation with VIP-mediated stimulation of adenylate cyclase in each of the tumors studied (Fig. 2B, P < 0.005 using Fisher's exact test). A similar correlation between receptor expression, signal transduction and VIP levels can be seen in neuroblastoma cell lines. Both the SKNSH and IMR32 cell lines express high affinity VIP receptors (Table III). IMR32 cells have very low or non-detectable VIP levels, but exhibit a VIP-adenylate cyclase stimulation index of 22, whereas SKNSH
219 Relationship between VIP and Basal Adenylate Cyclase Activity 50 A 40 4( 30
"S T,,
20
.•..
1C
0.1
•
-
1
1-o
Ioo
VIP (Pg/Ng DNA) Relationship between VIP and Adenylate Cyclase Stimulation Index
B 4
c M
3
_a 2 E
0
.
0.1
.
.
.
.
.
.
.
.
'
.
1
.
.
.
.
.
.
.
.
10
.
.
,
.....
i
100
VIP (pg/~Jg DNA)
Fig. 2. Effect of tumor VIP content on basal VIP and stimulated adenylate cyclase activity in neuroblastoma tumors. Tumor tissue was disrupted as described in Fig. 1. Cell free protein (10-20 #gin) was incubated with adenylate cyclase media containing Mg-ATP _+1 #M VIP as described in Materials and Methods. The cAMP generated in the reaction was quantified by radioimmunoassay. A second aliquot of the cell-free protein in buffer A was extracted for quantification of VIP content by radioimmunoassay as described in Materials and Methods. (A) Basal activity was measured in the absence of VIP. Data were fit by 2nd order polynomial for basal adenylate cyclase activity regressed on log VIP content. (B) Adenylate cyclase activity was measured in the presence of 1 #M VIP. The adenylate cyclase stimulation index is the ratio of VIP-stimulated : basal adenylate cyclase activity using the values shown in Table II. Data were analyzed using Fisher's exact test.
cells have higher VIP levels and a much lower VIP-adenylate cyclase stimulation index (Table III). Earlier studies by Muller et al. [ 10] demonstrated higher levels of VIP in more differentiated SKNSH cell lines compared to neuroblast-like subcultures. The VIP-mediated stimulation of adenylate cyclase observed in SKNSH cultures may reflect this heterogeneity of cell type which includes both neuroblast-like and Schwanoma-like cells [33,34]. The rate of VIP metabolism was examined in neuroblastoma cell lines. Both IMR32 and SKNSH cells were incubated in the presence of 1 #M VIP and the amount of VIP remaining in the culture was quantified by RIA. IMR32 cells appear to degrade (or metabolize) VIP quite slowly; VIP remained immunologically recognizable for 96 h in
220 TABLE III VIP and VIP receptors in neuroblastoma cell lines KD
SKNSH IMR32
(nM)
pg VIP/106 cells
2.5 + 1.0 4.0 _+ 2.1
Adenylate cyclase SI a
6.3 + 2.5 0.6 + 1.10"
VIP
Forsk
5.3 + 1.1 22.2 _+ 3.8
6.3 + 2.2 8.2 + 0.8
* Values not detectable in RIA considered as 0. a Adenylate cyclase stimulation index in the ratio of VIP-stimulated: basal activity. Cells were subcultured as described in Materials and Methods, allowed to grow to confluency (7 days) and washed two times with serum free media. Competitive binding at 17 °C was performed 30 min using 150 #g membrane protein. VIP was extracted and quantified by RIA as described in Materials and Methods. Adenylate cyclase activity assayed 30 min at 37 °C using 10-20 #g cell free protein. Values are mean _+ S.D. for three cultures of each cell type.
VIP Metobol ism/Degradot i on in Neuroblastomo Cell Lines 105(
"~---~''~'~
lo 4
~
1 03 o_ >
102
lo 1 0
1
2
3
4
5
6
Time External VIP Inoubation -0-
SKNSH
+IMR-32
[days]
~
Media Alone
Detection of VIP Fragments By Radioimmunoassoy 1O0
8O 70
0
60
[] O
5o 40
O 0
30
20 10
[] O
B
~o
~oo
~ooo
~oooo
%o000
pmol VlP
•
VIP
0
VIP 4_
0
VlPlo_
Fig. 3. (A) Metabolism of VIP by neuroblastoma cell lines in culture. S K N S H and IMR32 cells were grown in culture as described in Materials and Methods. VIP (I #M) was added to each well (containing 1.105 cells or media alone) on day 0. Cells and media were harvested at indicated times and extracted for quantification of VIP as described in Materials and Methods. (B) Radioimmunoassay ofVIP. VIP standard curve with VIP4_28 and VIPlo_2s run as unknowns. Peptides were purchased from Bachem.
221 TABLE IV Interaction of VIP and VIP fragments with VIP receptor on neuroblastoma cells
VIP
VIP4_28 VIPlo_28
KD (nM)
Adenylate cyclase SI a
2.9 + 1.2 1.5 + 0.8 110 + 79
3.8 + 0.6 1.l + 0.l 0.96 + 0.03
a Adenylate cyclase stimulation index is the ratio of VIP-stimulated: basal activity. SKNSH cells were subcultured as described in Materials and Methods, allowed to grow to confluency (7 days) and washed twice with serum free media. Competitive binding was performed for 30 min at 17 °C using 150/~g cell free protein. Adenylate cyclase was measured as described in Materials and Methods. All peptides were added at 1 #M concentrations. Values are mean + S.D. (n = 3).
the culture with a d e c a y r a t e e q u i v a l e n t to the d e c a y rate in m e d i a a l o n e (Fig. 3A). E x o g e n o u s V I P w a s rapidly lost f r o m S K N S H m e t a b o l i z e d in the rlrst 24 h. (Fig. 3A).
cultures w i t h 99~o h y d r o l y z e d o r
S e v e r a l proteolytic f r a g m e n t s o f V I P h a v e b e e n d e m o n s t r a t e d in cell cultures [ 3 5 - 3 7 ] . T w o o f t h e s e f r a g m e n t s , VIP4_28 a n d VIPlo_28 w e r e tested in the r a d i o i m m u n o a s s a y
SKNSH
0
A i
i
i
i
i
IMR32
B
Days of culture
Fig. 4. Effect o f exogenous VIP on proliferation ofneuroblastoma cell lines in culture. S K N S H and I M R 3 2
cells were grown in tissue culture in the presence of no VIP (e), 1 #M VIP ( I ) , or 10 #M VIP (A). Peptide was added on day 0 and at the indicated times, cells were released from the plates by addition of trypsin and cells enumerated by Coulter counter. Viability was > 90% for all cultures. (A) SKNSH cells and (B) IMR32 cells; values shown are mean + S.D. for three independent experiments.
222 (Fig. 3B) and examined for receptor affinity as well as for their ability to induce signal transduction. VIP and the peptide fragment VIP4_2s are recognized with equal potency in this assay, but the carboxy-terminal fragment VIPIo_2s is not measurable except at very high concentrations (Fig. 3B). Smaller carboxy-terminal fragments such as VIP17_28 and VIP23_28 although known degradation products [35-37] would not be measurable in this assay [38]. VIP4_28 binds to the VIP receptor, but does not activate adenylate cyclase; VIP1o_2s shows very low affinity for the VIP receptor and has no demonstrable against activity in the adenylate cyclase assay (Table IV). These results suggest (Fig. 3 and Table IV) that SKNSH cells degrade VIP rapidly to fragments smaller than VIP~o_2s, which are not effective ligands for the VIP receptor. Thus, effects of VIP on proliferation of neuroblastoma cells appear to be dependent on an intact VIPl_28 molecule as has been shown in intestine, heart and lung [39-41]. Further experiments were conducted to determine if VIP affects the viability and/or rate of growth of the two neuroblastoma tumor cell lines. 1 #M VIP added to the media on day one had no significant effect on the viability or proliferation rate of SKNSH cells, but inhibited the proliferation of IMR32 cells by 30~o over a 6 day culture period (P < 0.02) (Fig. 4). At higher concentrations (10 #M) VIP inhibited the proliferation of SKNSH by 20~ (P < 0.05) and inhibited IMR32 proliferation by 49~o (P < 0.01) (Fig. 4). The morphology of IMR32 cells incubated with VIP for six days showed less confluent cultures with increased neurite extension (Fig. 5). These results correlate well with the recent observation of Pence and Shorter that VIP induces differentiation in IMR32 cells in culture [21].
Fig. 5. Effect of exogenous VIP on morphology ofIMR32 neuroblastoma cells in culture. IMR32 cells were grown under identical conditions as in Fig. 4 and photographed under phase contrast microscopy (40 x ) on day 6. Viability was >90% for all cultures. A, no peptide; B, 1 #M VIP; C, 10/~M VIE
223
Discussion
These studies have provided the first evidence that human neuroblastoma tumors both synthesize (or concentrate) VIP and express VIP receptors in vivo, thus fulfilling a basic requirement for autocrine growth regulation. In addition, endogenous tumor VIP levels appear to regulate the ability of the VIP receptor to transduce intracellular signals via the adenylate cyclase complex, with high VIP levels resulting in uncoupling of the receptor-Gs-cyclase pathway.
224 Detection of VIP in tumor tissue by RIA is not absolute proof of VIP gene expression by the malignant cells; however, uptake of VIP from the circulation is unlikely because the levels of VIP in serum are extremely low except in individuals with VIP secreting tumors [38]. Likewise, release of VIP by peptidergic neurons in close proximity to the tumor is unlikely to account for the high levels observed in these tumors. In addition, at least three human neuroblastoma cell lines ( S K N S H , IMR32, NB-OK-1) synthesize VIP [ 11 ], (Table III). The current data also might be explained by a paracrine action of VIP, i.e., a subpopulation of tumor cells which synthesize and secrete VIP which binds to VIP receptors on adjacent tumor cells. VIP decreases the proliferation rate of a neuroblastoma cell line which does not hydrolyze VIP rapidly, but has little effect in a cell line which appears to degrade VIP rapidly. This observation supports the hypothesis that VIP can serve as an autocrine growth factor in human neuroblastoma and suggests several steps at which an aberration of the normal VIP regulation of proliferation or differentiation might contribute to malignant transformation. Overexpression or amplification of a protease gene might account for the rapid hydrolysis of VIP in S K N S H cells and thus inhibit normal VIP effects on neuronal cell differentiation. Similarly, mutations in the normal VIP gene might result in lower VIP gene expression and a failure in the differentiation pathway. Mutations in the VIP receptor gene might prevent the normal VIP mediated signal transduction and also result in deregulation of a normal proliferation/differentiation pathway. The recent cloning of both the VIP [42] and VIP receptor [43] genes will allow for direct testing of these hypotheses.
Acknowledgements This work was supported by grants from the National Cancer Institute (RO1-CA41997 to M.S.O.) and The Ohio State University Comprehensive Cancer Center (S.J.Q.). The authors wish to thank Jadwiga Labanowska, Brent Howe, Laura Campolito and Monica Summers for expert technical assistance; John Hayes for advice on the statistical analysis of data and Debbie Fine for preparation of the manuscript.
References 1 Birch, J.M., Marsden, H.B., Morris-Jones, P.H., Pearson, D. and Blair, V., Improvements in survival from childhood cancer: results of a population based surveyover 30 years, Br. Med. J., 296 (1988) 1372. 2 Woods, W.G. and Tuchman, M., Neuroblastoma: the case for screening infants in North America, Pediatrics, 79 (1987) 869-873. 3 Kushner, B.H. and Cheung, N.-K.V., Treatment of neuroblastoma, Cancer, 2 (1988) 1-11. 4 Hayes, F.A. and Smith, E.I., Neuroblastoma. In P.A. Pizzo and D.G. Poplack (Eds.), Principles and Practice of Pediatric Oncology,J.B. Lippincott, Philadelphia, pp. 607-622, 1989. 5 Evans,A.E., D'Angio,G.J., Sather, H.N., De Lorimier,A.A., Dalton, A., Ungerleider, R.S., Finklestein, J.Z. and Hammond, G.D., A comparison of four staging systems for localized and regional neuroblastoma: a report from the Children's Cancer Study Group, J. Clin. Oncol., 8 (1991) 678-688. 6 Philip, T., Chauvin, F. and Michon, J., A pilot study of double A.B.M.T. in advanced neuroblastoma. In K. Dicke, G. Spitzer and S. Jagannath (Eds.), International Conferenceon Autologous Bone Marrow Transplantation, Houston, The University of Texas, pp. 799-805, 1989.
225 70'Leary, M., Sather, H., Ramsay, N., Harris, R., Strandjord, S., Romansky, S., Haase, G., Shimada, H., Novak, L., Siegel, S., Harm, H.W., Brodeur, G., Secger, R. and Hammond, D., Intensive chemotherapy for poor prognosis neuroblastoma, Prec. Am. Ass. Cancer Res., 31 (1990) 201 (abstract). 8 Ciecarone, V., Spengler, B.A., Meyers, M.B., Biedler, J.L. and Ross, R.A., Phenotype diversification in human neuroblastoma cells: expression ofdistinct neural crest lineages, Cancer Res., 49 (1989) 219-225. 9 Baron, B.M. and Siegel, B.W., Alpha-2-adrenergic and muscarinic cholinergic receptors have opposing actions on cyclic AMP levels in SK-N-SH human neuroblastoma cells, J. Neurochem., 53 (1989) 602-609. 10 Muller, J.M., Lolait, S.J., Yu, V.C., Sadee, W. and Waschek, J.A., Functional vasoactive intestinal polypeptidc (VIP) receptors in human neuroblastoma subclones that contain VIP precursor mRNA and release VIP-like substances, J. Biol. Chem., 264 (1989) 3647-3650. 11 Hoshino, M., Yanaihara, C., Ogino, K., Iguchi, K., Sate, H., Suzuki, T. and Yanaihara, N., Production of VIP- and PHM (human PHI)-related peptides in human neuroblastoma cells, Peptides, 5 (1984) 155-160. 12 Said, S.I. and Mutt, V., Polypeptide with broad biological activity: isolation from small intestine, Science, 169 (1970) 1217-1218. 13 Itoh, N., Obata, K., Yanaihara, N. and Okamoto, H., Human preprovasoactiv¢ intestinal peptide contains a novel, PHI-27-1ike peptide. PHM 27, Nature, 304 (1983) 547-549. 14 Funato, M., Fujimura, M., Shimada, S. et al., Rapid changes of serum vasoactive intestinal peptide after removel of ganglioneuroblastoma with watery diarrhea-hypokalemia-achlorhydria syndrome in a child, J. Ped. Gastroent. Nutr., 1 (1982) 131-135. 15 Anderson, D.J. and Axel, R., A bipotential neuroendocrine precursor whose choice of cell fate is determined by NGF and glucocorticoids, Cell, 47 (1986) 1079-1090. 16 Tischler, A.S., Dayal, Y., Balogh, K., Cohen, R.B., Connolly, J.L and Tallberg, K., The distribution of immunoreactive chromogranins, S-100 protein, and vasoactive intestinal peptide in compound tumors of the adrenal medulla, Hum. Pathol., 18 (1987) 909-917. 17 Eiden, L.E. and Hotchkiss, A.J., Cyclic adenosine monophosphate regulates vasoactive intestinal polypeptide and enkephalin biosynthesis in cultured bovine chromaffin cells, Neuropeptides, 4 (1983) 1-9. 18 Tseng, J. and O'Dorisio, M.S.,Mechanism ofvasoactiveintestinal peptide (VlP)-mediatedimmunoregulation. In E.J. Goetzl and N.H. Spector (Eds.), Neuroimmune Networks: Physiology and Diseases, Alan R. Liss, New York, pp. 105-111, 1989. 19 Gozes, I. and Brenneman, D.E., VIP: molecular biology and neurobiological function, Mol. Neurobiol., 3 (1989) 201-236. 20 Brenneman, D.E., Nieol, T., Warren, D. and Bowers, LM., Vasoactive intestinal peptide: a neurotrophic releasing agent and an astroglial mitogen, J. Neurosci. Res., 25 (1990) 386-394. 21 Pence, J.C. and Shorter, N.A., In vitro differentiation of human neuroblastoma cells caused by vasoacrive intestinal peptide, Cancer Res., 50 (1990) 5177-5183. 22 O'Dorisio, M.S., O'Dorisio, T.M., Wood, C.L., Beattie, M.S., Bresnahan, J.C. and Campolito, L.B., Characterization of VIP receptors in nervous and immune systems, Ann. NY Acad. Sci., 527 (1988) 257-265. 23 Kaslow, H.R., Farfel, Z., Johnson, G.L. and Bourne, H.R., Adenylate cyclase assembled in vitro: cholera toxin substrates determine different patterns of regulation by isoproterenol and guanosine 5'-triphosphate, Mol. Pharmacol., 15 (1979) 472-83. 24 O'Dorisio, M.S., Shannon, B.T., Fleshman, D.J. and Campolito, L.B., Identification of high affinity receptors for vasoactive intestinal peptide on human lymphocytes of B cell lineage, J. Immnnol., 142 (1989) 3533-3536. 25 Akera, T. and Cheng, V., A simple method for the determination of affinity and binding site concentration in receptor binding studies, Biochim. Biophys. Acta, 4790 (1979) 412-423. 26 O'Dorisio, M.S., Hermina, N., Balcerzak, S.P. and O'Dorisio, T.M., Vasoactive intestinal polypeptide stimulation of adenylate cyclase in purified human leukoeytes, J. Immunol., 127 (1981) 2551-2554. 27 O'Dorisio, M.S., Vandenbark, G.R. and Lo Buglio, A.F., Human monocyte killing ofS. a u r e u s : modulation by agonists of cAMP and cGMP, Infect. Immun., 26 (1979) 604-611. 28 Tumilowicz, J.J., Nichols, W.W., Cholon, J.J. and Greene, A.E., Definition of a continuous human cell line derived from neuroblastoma, Cancer Res., 30 (1970) 2110-2118.
226 29 Biedler, J.I., Nelson, I. and Spengler, B.A., Morphology and growth, tumorgenicity and cytogenetics of human neuroblastoma cells in continuous culture, Cancer Res., 33 (1973) 2643-2652. 30 O'Dorisio, T.M., Mekhjian, H.S., Ellison, E.C., O'Dorisio, M.S., Gaginella, T.S. and Woltering, E.A., Role ofpeptide radioimmunoassay in understandingpeptide-peptide interactions and clinical expression of gastroenteropancreatic endocrine tumors, Am. J. Med., 82 (1987) 60-67. 31 Sibley, D.R., Peters, J.R., Nambi, P., Caron, M.G. and Letkowitz, R.J., Desensitization of turkey erythrocyte adenylate cyclase, J. Biol. Chem., 259 (1984) 9742-9748. 32 Iyengar, R., Bhat, M.K., Riser, M.E. and Birnbaumer, L., Receptor-specific desensitization of the $49 lymphoma cell adenylyl cyclase, J. Biol. Chem., 256 (1981) 4810-4899. 33 Rettig, W.J., Spengler, B.A., Chesa, P.G., Old, L.J. and Biedler, J.L., Coordinate changes in neuronal phenotype and surface antigen expression in human neuroblastoma cell variants, Cancer Res., 47 (1987) 1383-1389. 34 Ross, R.A., Spengier, B.A. and Biedler, J.L., Coordinate morphological and biochemical interconversion of human neuroblastoma cells, J. Nat. Cancer Inst., 71 (1983) 741-747. 35 Goetzl, E.J., Kodama, K.T., Turck, C.W., Schiogolev, S.A. and Sreedharan, S.P., Unique pattern of cleavage of vasoactive intestinal peptide by human lymphocytes, Immunology, 66 (1989) 554-558. 36 Misbin, R.I., Wolfe, M.M., Morris, P., Buynitzky, S.J. and McGuigan J.E., Uptake of vasoactive intestinal peptide by rat liver, Am. J. Physiol., 243 (1982) G103-G111. 37 Goetzl, E.J., Sreedharan, S.P. and Turck, C.W., Structurally distinctive vasoactive intestinal peptides from rat basophilic leukemia cells, J. Biol. Chem., 263 (1988) 9083-9086. 38 O'Dorisio, T.M., Howe, B.A., Hill, D.S., Obermeyer, J. and Cataland, S., Vasoactive intestinal peptide (VIP) and neurotensin radioimmunoassays, J. Clin. Immunoassay, 10 (1987) 85-90. 39 Couvineau, A., Rouyer-Fessard, C., Fournier, A., St. Pierre, S., Pipkorn, R. and Laburthe, M., Structural requirements for VIP interaction with specific receptors in human and rat intestinal membranes: effect of nine partial sequences, Biochem. Biophys. Res. Commun., 121 (1984) 493-498. 40 O'Donnell,M., Garippa, R., O'Neill, N., Bolin, D. and Cottrell, J., Structure-activitystudies ofvasoactive intestinal polypeptide, J. Biol. Chem. 266 (1991) 6389-6392. 41 Fournier, A., Saunders, J.K. and St. Pierre, S., Synthesis, conformational studies and biological activities of VIP and related fragments, Peptides, 5 (1984) 169-177. 42 Tsukada, T., Horovitch, S.J., Montiminy, M.R., Mandel, G. and Goodman, R.H., Structure of the human vasoactive intestinal polypeptide gene, DNA, 4 (1985) 293-300. 43 Sreedharan, S.P., Robichon, A., Peterson, K.E. and Goetzl, E., Cloning and expression of the human vasoactive intestinal peptide receptor, Proc. Natl. Acad. Sci. USA, 88 (1991) 4986-4990. 44 Qualman, S.F., O'Dorisio, M.S., Fleshman, D.F., Shimada, H. and O'Dorisio, T.M., Neuroblastoma: correlation of neuropeptide and oncogene expression with other prognostic factors, Cancer, in press.
