Cellular and Molecular Neurobiology, Vol. 12, No. 5, 1992
Immunoreactive Growth Hormone Production by Human Lymphocyte Cell Lines Ting-Lin Kao, 1 Scott C. Supowit, ~ E. Aubrey Thompson, 1 and Walter J. Meyer, 1111'2 Received February I, 1992; accepted March 20, 1992 KEY WORDS: growth hormone; lymphocytes; cell culture; polymerase chain reaction; pituitary.
SUMMARY 1. Two human lymphocyte cell lines, a T-cell line and a B-cell line, were shown to produce and secrete immunoreactive growth hormone (irGH). The irGH molecules secreted by the two cell lines appeared to be de novo synthesized and their molecular size was similar to that of pituitary GH as well as irGH secreted by peripheral blood lymphocytes. 2. Affinity-purified irGH molecules had human growth hormone (hGH)-like mitogenic activity on Nb2 cells. These findings indicate that the irGH molecules produced by H9 and IM9 were similar to hGH in structure. 3. However, the irGH messages could not be amplified by polymerase chain reaction (PCR) primers which had been demonstrated to be able to amplify reverse-transcribed hGH messenger RNA successfully, suggesting that the lymphocyte-derived irGH and pituitary hGH are not exactly identical molecules. 4. We conclude that the H9 and IM9 cells produce a growth hormone-related molecule whose structure is different from that in the anterior pituitary.
INTRODUCTION In recent years, numerous bidirectional interactions between the immune system and the neuroendocrine system have been described. Alterations in the neuroen1Department of Human Biological Chemistry and Genetics, University of Texas Medical Branch, Galveston, Texas 77550. 2 To whom correspondence should be addressed at Departments of Psychiatry & Behavioral Sciences and Human Biological Chemistry & Genetics, University of Texas Medical Branch, Galveston, Texas 77550. 483 0272-4340/92/1000-0483506.50/0(~ 1992 Plenum Publishing Corporation
484
Kao, Supowit, Thompson, and Meyer
docrine system affect the immune system, and the reverse is also true. One of the proposed mechanisms for the intersystem communication is the production of neuropeptides by the cells of the immune system. Several of the lymphocytederived pituitary hormones have been characterized, such as ACTH, endorphin, thyrotropin (TSH), chorionic gonadotropin, luteinizing hormone, and growth hormone (Smith et al., 1983, 1986; Emanuele et al., 1990; Weigent and Blalock, 1990; Harbour-McMenamin et al., 1986). Since all these hormones can also function as immunomodulators, it is proposed that cells in the immune system are able to produce neuroendocrine hormones, which can in turn have autocrine or paracrine effects on the immunocytes themselves. In the past two decades, increasing evidence has pointed to the theory that growth hormone (GH) not only is an essential neuroendocrine hormone but also plays an important role in the immune system. G H has been found to have influences on the thymus, the lymphocytes, and the macrophages, both in vitro and in vivo (Kelley, 1990). Recently, production of immunoreactive growth hormone (irGH) by human peripheral blood lymphocytes (PBL) has been reported (Weigent et al., 1988; Hattori et al., 1990). Characterization of lymphocyte-derived irGH molecules indicate that they have molecular weight and bioactivity similar to their pituitary counterpart. Quantitation using a highly sensitive immunoassay showed that PBL secreted irGH at picogram level (Hattori et al., 1990). Furthermore, irGH molecules have been shown to be able to stimulate lymphocyte proliferation, suggesting that irGH may play an autocrine/paracrine role in lymphocyte replications (Weigent et al., 1991). The use of primary cell cultures for study of irGH production results in an inability to obtain a large, homogeneous cell population and adds difficulty to studies on regulation of irGH secretion. Development of model systems using lymphocyte cell lines could be one way to solve the problem. Similar systems have been developed for TSH and prolactin (Harbour et al., 1989; DiMattia et al., 1988). Previous study suggested that both T and B lymphocytes secrete irGH (Hattori et al., 1990). In the present study, we report identification of a T-cell and a B-cell lymphocyte cell lines which produce irGH molecules as well as their characterization.
MATERIALS A N D METHODS Cell Cultures
IM9 cells were obtained from the laboratory of Dr E. Brad Thompson, and H9 cells from Dr Miles W. Cloyd, both at the University of Texas Medical Branch in Galveston. The cells were cultured in RPMI 1640 supplemented with either 5% fetal calf serum (GCS) or insulin-transferrin-sodium selenite medium supplement (Sigma). Nb2 cells were provided by Dr Li Yu-Lee of Baylor College of Medicine. GH3 cells were provided by Dr Scott Supowit of University of Texas Medical Branch in Galveston.
Immunoreactive Growth Hormone Production
485
Reagents Human growth hormone (hGH) was kindly donated by the National Pituitary Agency or purchased from Eli Lilly. Rabbit antisera to hGH were donated by the National Pituitary Agency or purchased from Chemicon. Purified monoclonal antibody to hGH was obtained from Medix Biochemica, Finland. DNA primers were synthesized by the UTMB D N A synthesis laboratory or Genosys.
Indirect Immunofluorescence The assay was performed as described previously (Blalock and Smith, 1980) with the modification that bovine serum albumin (BSA) was employed as a blocker of nonspecific binding. Cells were air-dried to glass coverslips and fixed with absolute ethanol. After preincubation in phosphate-buffered saline (PBS)0.1% BSA for 45 rain, the coverslips were stained first with either polyclonal rabbit antiserum to hGH (Chemicon, 1 : 20) or normal rabbit serum (1 : 100) and then with fluorescein-conjugated goat anti-rabbit IgG (Cappell Laboratories, 1:15). The stained coverslips were mounted and read at 100× magnification under a fluorescence microscope.
Reverse Hemolytic Plaque Assay This was performed as described previously with several minor modifications (Neill and Frawley, 1983; P. F. Smith et al., 1984). Sheep red blood cells (SRBC) were covalently coupled to staphylococcal protein A (SPA; Sigma) as described previously (Mishell and Shiigi, 1980). The protein A-coupled SRBC suspension was mixed with an equal volume of lymphocytes resuspended in RPMI-0.1% BSA-1000/~g/ml penicillin/streptomycin at a concentration of 106/ml, anti-hGH antiserum (Chemicon, 1:50), and guinea pig complement (Cordis Labs, 1:20). The resulting mixture was infused into poly-tAysine-coated Cunningham slide chambers. The plaques were counted after overnight incubation.
Intrinsic Labeling and Affinity Chromatography Cells were resuspended at a concentration of 106/ml in low-amino acid Selectamine medium. At 24 hr, the culture was supplemented with 2 #Ci/ml 3H-amino acid mixture (ICN or Amersham). The supernatant was harvested at 48hr. Activated affinity chromatography units (Nalgene, U12) were bound to purified monoclonal GH antibody (Medix Biochemica, clone code 5802) according to the manufacturer's specification. Cell culture supernatant concentrates were passed 20 times over a PBS-equilibrated anti-hGH unit, and the bound portion was eluted with 0.1 M glyeine.
Sodium Dodecyl Sulfate (SDS)-Polyacrylamide Gel Electrophoresis Affinity-purified, intrinsically radiolabeled irGH was reduced by boiling and electrophoresed on 15% reducing polyacrylamide gels. After electrophoresis, the
486
Kao, Supowit, Thompson, and Meyer
lanes were sliced into 24 0.25-cm fractions. The fractions were individually solubilized with 0.5 ml of Solusol (National Diagnostics) and radioactivity per fraction was monitored in a E-counter.
Immunoradiometric Assay The cells were grown in Dulbecco's modified Eagle medium ( D M E M ) - 5 % FCS at 106/ml cells. When the cells were grown in serum-free DMEM medium, 5% FCS was added at the end of incubation. The culture supernatant was dialyzed against 1% PBS, dispensed into 15-ml centrifuge tubes, lyophilized, and reconstituted in 1/50 of the original volume in water. Fresh D M E M - 5 % FCS media containing hGH were dialyzed, lyophilized, and reconstituted in the same way to set up the standard curve. The concentrated samples were assayed for GH immunoreactivity using the Nichols Institute hGH Immunoassay. The detection limit of the modified assay was 1 pg/ml.
Bioassay The G H bioassay was based on the proliferative response of Nb2 node rat lymphoma cells to lactogenic hormones (Tanaka et al., 1980). Nb2 cells were grown in Fischer's medium for leukemic mice (GIBCO) supplemented with 10% horse serum, 10% FCS, and 10-4M mercaptoethanol. For the assay, Nb2 cells were washed twice in serum-free Fischer's medium and then resuspended in maintenance medium, which was the growth medium without FCS, for 22-24 hr. At the end of incubation, the cells were washed twice, resuspended in the maintenance medium at 2 × 105 ml, and added to a 96-well microplate. Pituitary hGH or afffinity-purified irGH samples at different dilutions were then added to the wells. To test the blocking effects of anti-hGH antibodies, 0.1vol of monoclonal antibodies (Medix Biochemica, 1 mg/ml) was incubated with the irGH samples for 1-1.5 hr before addition to the Nb2 cells. The cells were counted under a microscope using a hemocytometer after 68-70 hr of incubation.
RNA Isolation and Dot-Blot Hybridization Cytoplasmic R N A was extracted from the lymphocytes using a method described by Gough (1988). Total R N A was transferred onto 0.45-/~m nitrocellulose filters using a Bio-Rad dot-blot apparatus. To obtain cDNA probes, plasmid DNA was purified from HB 101 bacteria carrying the plasmids (PBR322) which contain 800-bp H i n d l I I inserts of either rat or human G H cDNA. A 500-bp fragment within the coding region was excised with P v u I I . The bacteria were provided by Dr Weigent of the University of Alabama, who transformed the bacteria using plasmids donated by Dr Denoto of the University of California at San Francisco. Hybridization was carried out according to a standard protocol (Wanl et al., 1981). The blot was incubated with prehybridization buffer (50% formamide, 5× Denhardt's solution, 0.1% SDS, 100/~g/ml of salmon sperm DNA, 5 × SSPE) at 42°C for 1-2 hr. For hybridization a p32 nick-translated probe with a specific activity of 1-5 x 108 cpm//~g was added to the hybridization buffer
ImmunoreactiveGrowth HormoneProduction
487
(prehybridization buffer with 10% dextran sulfate added) at 1-5 x 106 cpm/ml. After overnight incubation, the blot was washed extensively and autoradiographed with an intensifying screen for a week.
Reverse Transcription and Polymerase Chain Reaction (PCR) Total RNA prepared from lymphocytes or pXGH5-transfected cells was reverse transcribed using a modification of a standard protocol (Innis et al., 1990). The following reagents were assembled in a final volume of 20/zl: 1 mM of each of the four dNTP's, 1 unit//~l of RNasin, 0.2/zg of oligo(dT); 5/tg of RNA sample, 200 units of MoMuLV reverse transcriptase, I x MoMuLV RT buffer, and 10 mM DTT. For negative controls, all reagants were included except the MoMuLV reverse transcriptase. The tubes were incubated at 37°C for 1.5 hr, heated at 95-100°C for 5 min, and quick-chilled on ice. Then the transcribed DNA was precipitated twice and resuspended in 20/~1 of water. The PCR mixture was set up in a final volume of 100/~1:200/~mol of each dNTP, 40 pmol of each of the two primers, 20 #1 of the template, l x PCR buffer (Perkin Elmer Cetus), and 2.5 units of Taq DNA polymerase. The reaction mixtures were then run for 35 cycles, with a thermal cycle profile of denaturation at 92°C for 1 min, annealing at 45°C for 2 min, and extension at 72°C for 3 min. The resulting PCR mixtures were analyzed on 2% agarose gels.
Southern Blotting and Hybridization After electrophoresis of PCR samples, the DNA was denatured by alkaline and the treated gel was capillary-transferred onto a 0.45-~m nitrocellulose membrane overnight. The blot was incubated with prehybridization buffer (50% formamide, 6x SSPE, 5x Denhardt's solution, 0.5% SDS, 100/~g/ml denatured, fragmented salmon sperm DNA) at 42°C for 1-2 hr, For hybridization, 32p random-primed, full-length hGH eDNA probe was denatured with NaOH and added directly to the prehybridization buffer at 106 cpm/ml. After 8 to 10 hr of incubation, the blot was washed extensively and autoradiographed with an intensifying screen for 0.5-2 hr.
RESULTS Identification of Cell Lines Which Produce irGH In order to identify cell lines which produce irGH, 10 T-cell and B-cell lines were screened by indirect immunofluorescence. The 10 cell lines were IM9, C7, Hut78, M4, CLL155, CLL156, HB34, TIB190, $49, and BCL. Of the 10 cell lines screened, IM9 cells have shown the best positive staining (Fig. 1). Other cell lines did not stain significantly above the normal rabbit serum background. To show that the IM9 cells were capable of secreting irGH molecules as well as producing them, the reverse hemolytic plaque assay was employed. Representative photomicrographs of the assay are shown in Fig. 2. When no complement was added to the mixture, microscopic analysis revealed lymphocytes in close proximity to the sheep erythrocytes without specific plaque formation (Fig. 2A).
488
Kao, Supowit, Thompson, and Meyer
Fig. 1. Immunofluorescent detection of irGH production in IM9 cells. Top: IM9 ceils stained with anti-GH antiserum. Bottom: IM9 cells stained with normal rabbit serum.
Immunoreactive Growth Hormone Production
Fig. 2. Photomicrographs of lysed area around IM9 cells obtained by GH-specific reverse hemolytic plaque assay after overnight incubation. (A) Complement control at 8x; (B-D) IM9 cells at (B) 8x, (C) 32x, and (D) 100x.
489
49t1
Kao, Supowit, Thompson, and Meyer
Leaving out anti-hGH antiserum produced similar effects. With both complement and antibody, specific plaques or zones of clear hemolysis surrounding lymphocytes were observed and indicated that irGH were released from IM9 cells (Figs. 2B-D). However, the number of plaques observed was relatively low (32 + 13 plaques per 50,000 lymphocytes added), suggesting that the amount of irGH secreted may be only a small fraction of total irGH molecules produced. Attempts have been made to quantitate the level of irGH secretion by IM9 cells. Unfortunately, even though the irGH secretion level is higher than the medium background as measured in the immunoassay, it is still below the detection limit (1 pg/ml) and cannot be accurately measured (data not shown). Nevertheless, screening of a few cell lines not previously screened by immunofluorescence helped us to identify H9, a T-cell line which secreted irGH at a level measurable by our immunoassay. Figure 3 shows the time course of irGH secretion from H9 cells. H9 cells adapted to serum-free medium secrete irGH at a level of 11.3 pg/107 cells at 48 hr, which was comparable to that secreted by cells grown in FCS-containing medium (14.1 pg/107 cells).
Cytoplasmic GH mRNA Expression Since we could not quantitate irGH secretion from IM9 cells using our immunoassay, we attempted to measure its G H R N A expression by dot-blot analysis as an alternative means for possible quantitation studies. Both rat and human GH cDNA probes were employed. The insert was digested with PvulI, which cut a fragment entirely within the translated coding region. GH3 cells, a rat pituitary cell line, were used as the positive control; and L ceils, a fibroblast cell line, were used as the negative control. Figure 4 shows the results of dot-blot analysis. 30
24 o9
Ld ©
6
12
24
48
hours
Fig. 3. Kinetics of irGH secretion by the H9 cells, measured by double-antibody immunoassay. Incubated H9 cell culture supernatants were harvested at various time points and assayed. Error bars represent standard deviations of the means.
Immunoreactive Growth Hormone Production
491
GH3 L IM9
GH3 L IM9 i~iiii~ iiii~ ~
~.~~ ~ ~ ~ ~ ~
~ ~i~ ~ii~i!~!!~i~i!
Fig. 4. Dot-blot of I M 9 R N A hybridized with e i t h e r r G H o r h G H D N A p r o b e s . T o p : R a t p r o b e . G H 3 R N A , 2 a n d 1 ktg; L-cell R N A , 10 a n d 5 ~tg; I M 9 R N A , 40 a n d 2 0 # g . B o t t o m : H u m a n p r o b e . G H 3 R N A - - 2 , 1, a n d 0 . 5 / ~ g : L - c e l l R N A - - 1 0 , 5, a n d 2.5 # g ; I M 9 R N A - - 4 0 , 20, a n d 10/~g.
492
Kao, Supowit, Thompson, and Meyer
De Novo Synthesis of Lymphocyte-Derived irGH
To assure that the irGH molecules secreted by IM9 and H9 cells were synthesized de novo, the cultures were intrinsically labeled with tritiated amino acids. In typical experiments, radiolabeled amino acids (2/tCi/ml) were added to 50 ml of lymphocyte cell cultures and incubated for 24 hr. The supernatant was then passed through an affinity chromatography unit coupled with a purified
5400
.~
IM9
hGH
4540
3680
2820
1960
1100 5
10
15
20
Fraction iXkrnber 1200
H9
hGH 10:30
860
690
520
350
5
10
15
20
Fraction NLrnber
Fig. 5. SDS-PAGE of 3H-labeled irGH molecules eluted from affinity chromatography unit. MW standards: 92,000, 66,000, 45,000, 31,000, 21,000, and 14,000. Arrows point to where pituitary hGH runs on the gel. Top, IM9; bottom, H9.
Immunoreactive Growth Hormone Production
493
monoclonal a n t i - h G H antibody. T h e affinity column bound to 1 - 2 × 104 cpm (0.005-0.01% of the total cpm added). These data strongly suggest that the i r G H molecules were synthesized de novo by the lymphocytes. The radiolabeled, affinity-purified i r G H molecules were electrophoresed through a S D S - p o l y a c r y m i d e gel under reducing condition and the resulting gel cut into small slices to be counted on the fi-counter for determination of electrophoretic profile. Both protein molecular weight standards and pituitary h G H were also run on the same gel. Figure 5 shows the electrophoretic profile of i r G H molecules produced by H9 and IM9 cells. T h e two arrows point to where pituitary h G H migrates on the gel and should represent, respectively, the 21,500 and 20,000 variants of h G H . For both H9 and IM9, the main i r G H band ran between the two h G H bands. T h e r e are also some higher molecular weight bands present, which may represent aggregates of G H molecules (Stolar et al., 1984).
Mitogenic Activity of Lymphocyte-Derived irGH Molecules Nb2 is a rat l y m p h o m a cell line whose growth is dose dependent on h G H and is widely used for bioassay of lactogenic hormones. W e employed the Nb2 bioassay to test the bioactivity of i r G H molecules derived from H9 and IM9 cells. As shown in Table I, when affinity-purified i r G H molecules were added to stationary Nb2 cells, the growth of these cells was stimulated in a dose-dependent fashion and this mitogenic effect could be blocked with a n t i - h G H antibodies. The observation that the bioactivity level of i r G H derived from IM9 cells seemed to be lower than that from H9 cells is consistent with the relative amounts detected in the immunoassay.
Polymerase Chain Reaction Considering the extremely low level of i r G H production by the lymphocyte cell lines, we a t t e m p t e d to amplify the i r G H message using the PCR. The entire
Table I. Nb2 Bioassay of Affinity-Purified Culture Supernatants from H9 and IM Cells
Treatment Medium only HGH
H9 supernatant, affinity purified
IM9 supernatant, affinity purified
hGH concentration (ng/ml) or sample dilution
Number of Nb2 cells × 104/4 ml
0 0.01 0.02 0.05 0.01 1 :2 1 : 2 + anti-hGH 1:4 1:4 + anti-hGH 1:2 1:2 + anti-hGH 1:4 1:4 + anti-hGH
24.8 + 3.7 31.6+2.7 36.2 + 1.1 48.1 + 6.4 62.0 + 9.7 38.5 + 2.3 27.0 + 1.4 34.2 + 0.8 25.8+2.5 31.5+0.9 20.5+2.0 26.8 ± 5.1 21.8 ± 1.3
494
Kao, Supowit, Thompson, and Meyer
hGH cDNA, which was approximately 800 bp long, was to be amplified in two separate segments using two sets of PCR primers. The first set of primers, with sequences identical to nucleotides sense 41-60 (primer a) and antisense 421-440 (primer b), were targeted to amplify a 400-bp fragment at the 5' end of the hGH cDNA, and the second set of primers, with sequences identical to nucleotides sense 421-440 (primer c) and antisense 764-780 (primer d), were targeted to amplified a 360-bp fragment at the 3' end of the hGH cDNA. Total RNA isolated from the H9 and IM9 cells was reverse-transcribed and amplified. To assure that the reverse transcription and PCR reaction were functional, L cells were transfected with pXGH5, a plasmid containing the human hGH DNA, and RNA isolated from successfully transfected cells was used as a positive control in reverse transcription and PCR. Since the plasmid used for transfection contained the full-length (2-kb) hGH genomic DNA, there was no possibility of falsepositives resulting from the amplification of plasmid cDNA. For negative controls, reverse transcriptase was left out of each sample to be reversetranscribed and amplified. The results of the PCR are shown in Fig. 6. Amplifications from cDNA and the transfected cell lines gave fragments of the expected sizes for both the 5' end and the 3' end. No 5' end fragment of the
cDNA1 2 3 4 5 6
.f I 2 3 4 56 cDNA
Fig. 6. Amplifications of reverse-transcribed RNA from H9 and IM9 cells by polymerase chain reaction. The first set of primers is sense primer (bases 41-64) and antisense primer (bases 421-440) targeted to amplify a 400-bp fragment at the 5' end of the hGH cDNA, and the second set of primers is sense primer (bases 421-440) and antisense primer (bases 764-780) targeted to amplify a 360-bp fragment at the 3' end of the hGH cDNA. Left, 5'-end amplification; right, 3' end. For both photos, Lane 1--transfected cell line, reverse-transcribed RNA; Lane 2--transfected cell line, not reverse transcribed; Lane 3---H9, reverse-transcribed RNA; Lane 4---H9 RNA; Lane 5--IM9, reversetranscribed RNA; Lane 6----IM9 RNA. Lanes amplified from hGH cDNA are labeled "cDNA." The first lane in the 3'-end photo is the 500-bp company kit control, marked with an arrow.
Immunoreactive Growth Hormone Production
123456 Fig. 7. Southern hybridization of the amplified 3'-end DNA with h G H cDNA probe. The lanes are numbered as in Fig. 6. Lane 1--transfected cell line, reverse-transcribed RNA; Lane 2--transfected cell line RNA; Lane 3--H9 reverse-transcribed RNA; Lane 4--H9 RNA; Lane 5--IM9 reverse-transcribed RNA; Lane 6--IM9 RNA.
495
496
Kao, Supowit, Thompson, and Meyer
expected size was observed in the lymphocyte lanes, but for the 3' end, a band with almost the expected size was shown in the lymphocyte lanes and was absent in the negative control lanes. Consequently, the gel for the 3' end amplification was transferred to a nitrocellulose membrane and hybridized with 32p-labeled, full-length hGH cDNA. Figure 7 shows the results of Southern hybridization. The band resulted from amplification of the transfected cell line hybridized as expected; however, none of the lymphocyte lanes showed any visible band. The most likely explanation is that the irGH message could not be amplified from the two chosen set of primers, and the appearance of a band of expected size in the lymphocyte lanes of the 3' end gel was a coincidence resulting from nonspecific amplification of some unrelated DNA by the primers.
DISCUSSION
In this paper we report the identification of two lymphocyte cell lines, one T cell and one B cell, which produced irGH. The irGH molecules were characterized using protein chemistry and molecular biology techniques. Although we characterized irGH molecules produced by neoplastic cell lines, our data are in agreement with previous reports on irGH produced by human peripheral blood lymphocytes (Weigent et al., 1988; Hattori et al., 1990). Briefly, we have confirmed these observations that lymphocytes produced irGH which was similar to pituitary hGH in size, bioactivity, and message reactivity with G H cDNA. In addition, Hattori et al. (1990) have reported that both T and B cells were capable of producing irGH, and our identification of an irGH-producing B-ceU and an irGH-producing T-cell line seem to support the observation. Attempts were made to amplify the irGH RNA with a polymerase chain reaction. Even though the 3'-end amplification consistently gave a band of the correct size, the fact that a visible ethidium bromide-stained band could not be hybridized in a Southern blot, while the neighboring positive control bands of similar intensity readily hybridized (band visible in audiograph within I h of exposure), clearly indicates that the amplified band shares very little homology, if any, with the hGH cDNA sequence. Taken together with the fact that the 5'-end fragment amplification was not successful, we may conclude that h G H messenger RNA could not be amplified from the lymphocyte cell lines with the sets of primers we employed. It is most likely that the irGH molecules produced by H9 and IM9 cells were dissimilar to pituitary hGH in the PCR primer region(s). Possibly, these irGH molecules were very similar, but not exactly identical, to their pituitary counterparts. This possibility could be the reason for the extremely low level of immunoreactivity and bioactivity. No literature up to today has definitively proved or disproved that lymphocyte-derived irGH and pituitaryderived hGH are exactly identical. Our study strongly suggested that dissimilarity existed between the two species of molecules. There were other neuropeptidelike immunoreactivities which were identified in lymphocytes and were shown to be not exactly identical to their counterparts in the neuroendocrine system. Examples include immunoreactive corticotropin releasing hormone and prolactin
Immunoreactive Growth Hormone Production
497
(Staphanous et al., 1990; Hiestand et al., 1986). It is possible that i r G H also belongs to this group. The secretion of i r G H by H9 cells has been quantitated by a highly sensitive immunoassay. The major advantage of studying a cell line model system over a peripheral blood lymphocyte system is that the regulation of i r G H secretion by various pharmacological regulators can be observed without the need to consider possible interferences such as individual donor differences, mixed T- and B-cell populations, or the metabolic and endocrinological states on the donor at the time blood was drawn. Using peripheral blood lymphocytes for their study, Hattori et al. (1990) have been unable to observe any effect of G H R H or somatostatin analogue on secretion of irGH. Preliminary studies in our laboratory have also revealed a regulatory mechanism which appeared to be different from that in the pituitary (unpublished results). Since growth hormone has a variety of effects on the immune system, the possibility that human lymphocyte-derived i r G H has autocrine/paracrine effects on the lymphocytes themselves needs to be considered. In rats, it has been demonstrated that antisense oligonucleotides to rat growth hormone message can decrease the production of i r G H by rat spleen lymphocytes as well as inhibit the proliferation of the lymphocytes themselves (Weigent et al., 1991). The facts that both IM9 and human peripheral blood lymphocytes possess growth hormone receptors (Eshet et al., 1975; Kiess and Butenandt, 1985; Lesniak et al., 1974)) and h G H can stimulate proliferation of normal and transformed human lymphocytes (Suzuki et al., 1990; Mercola et al., 1981) suggest that similar growth promoting effects may be observed for H9- and IM9-derived i r G H molecules. Such effects are currently under study in our laboratory.
ACKNOWLEDGMENTS This project was partially supported by a grant from the George and Mary Josephine H a m m a n Foundation, the James W. McLaughlin Fellowship Fund, and N I M H Grant D H H S RO1 MH43279-02. The authors wish to thank Ms Nita Brannon and Ms Judy Van Over for help in preparing the manuscript.
REFERENCES Blalock, J. E., and Smith, E. M. (1980). Human leukocyte interferon: Structure and biological relatedness to adrenocorticotropin hormone and endorphins. Proc. Natl. Acad. Sci. USA 77:5972-5974. DiMattia, G. E., Gellersen, B., Bohnet, H. G., and Friesen, H. G. (1988). A human Blymphoblastoid cell line produces prolactin. Endocrinology 122:2508-2517. Emanuele, N. V., Emanuele, M. A., Teutler, J., Kirstein, L., Azad, N., and Lawrence, A. M. (1990). Rat spleen lymphocytes contain an immunoreactive and bioactive luteinizing hormone releasing hormone. Endocrinology 126:2482-2486. Eshet, R., Manheimer, S., Chobsieng, P., and Laron, Z. (1975). Human growth hormone receptors in human circulating lymphocytes. Horm. Metab. Res. 7:352-353. Gough, N. M. (1988). Rapid and quantitative preparation of cytoplasmicRNA from small number of cells. Anal. Biochem. 73:93-95.
498
Kao, Supowit, Thompson, and Meyer
Harbour, D. V., Kruger, T. E., Coppenhaver, D., Smith, E. M., and Meyer, W. J. (1989). Differential expression and regulation of thyrotropin (TSH) in T cell lines. Mol. Cell. Endocrinol. 64:229-241. Harbour-McMenamin, D., Smith, E. M., and Blalock, J. E. (1986). Production of lymphocyte derived chorionic gonadotropin in a mixed lymphocyte reaction (MCR). Proc. Natl. Acad. Sci. USA 83:6834-6838. Hattori, N., Shimatsu, A., Sugita, M., Kumagi, S., and Immura, H. (1990). Immunoreactive growth hormone (GH) secretion by human lymphocytes: Augmented release by exogenous GH. Biochem. Biophys. Res. Commun. 168:396-401. Hiestand, P. C., Mekler, P., Noramann, R., Grieder, A., and Permmongkol, C. (1986). Prolactin as a modulator of lymphocyte responsiveness provides a possible mechanism of action for cyclosporine. Proc. Natl. Acad. Sci. USA 83:2599-2603. InniS, M. A., Gelland, D. H., Snindky, and White, T. J. (eds) (1990). PCR Protocols: Guide to Methods and Application, Academic Press, New York. Kelley, K. W. (1990). The role of growth hormone in modulation of the immune response. Ann. N.Y. Acad. Sci. 594:95-103. Kiess, W., and Butenandt, O. (1985). Specific growth hormone receptors on human peripheral mononuclear cells: Reexpression, identification, and characterization. J. Clin. Endocrinol. Metab. 60:740-746. Lesniak, M. A., Gorden, P., Roth, J., and Gavin, J. R. (1974). Binding of 125I-human growth hormone to specific receptors in human cultured lymphocytes. J. Biol. Chem. 249:1661-1667. Mercola, K. E., Cline, M. J., and Golde, D. W. (1981). Growth hormone stimulation of normal and leukemic human T-lymphocyte proliferation in vitro. Blood 58:337-340. Mishell, B. B., and Shiigi, S. M. (eds) (1980). Selected Methods in Cellular Immunology, W. H. Freeman, San Francisco. Neill, J. D., and Frawley, L. S. (1983). Detection of hormone release from individual cells in mixed populations using a reverse hemolytic plaque assay. Endocrinology 112:1135-1137. Smith, E. M., Phan, M., Coppenhaver, D., Kruger, T. E., and Blalock, J. E. (1983). Human lymphocyte production of immunoreactive thyrotropin. Proc. Natl. Acad. Sci. USA 80:60106013. Smith, E. M., Morrill, A. C., Meyer, W. J., and Blalock, J. E. (1986). Corticotropin releasing factor induction of leukocyte-derived immunoreactive ACTH and endorphins. Nature 322:881-882. Smith, P. F., Frawley, L. S. and NeiU, J. D. (1984). Detection of LH release from individual pituitary cells by the reverse hemolytic plaque assay: Estrogen increases the fraction of gonadotropes responding to GnRH. Endocrinology 115:2484-2486. Staphanous, A., Jessop, D. S., Knight, R. A., and Lightman, S. L. (1990). Corticotropin-releasing factor-like immunoreactivity and mRNA in human leukocytes. Brain Behav. lmmun. 4:67-73. Stolar, M. W., Amburn, K., and Baumann, G. (1984). Plasma "big" and "big-big" growth hormone (GH) in man: An oligomeric series composed of structurally diverse GH monomers. J. Clin. Endocrinol. Metab. 59:212-218. Suzuki, K., Suzuki, S., Saito, Y., Ikebuchi, H., and Terao, T. (1990). Human growth hormonestimulated growth of human cultured lymphocytes (IM9) and its inhibition by phorbol diesterase through down-regulation of the hormone receptors. J. Biol. Chem. 265:11,320-11,327. Tanaka, T., Shiu, R. P. C., and Gout, P. W. (1980). A new sensitive and specific bioassay for lactogenic hormones: measurement of prolactin and growth hormone in human serum. J. Clin. EndocrinoL Metab. 51:1058-1063. Wanl, G. M., Ong, E., Meinkoth, J., Franco, R., and Baringa, M. (1981). Method for the Diazotized Paper Solid Supports, Schleicher and Schuell, Keene. Weigent, D. A., and Blalock, J. E. (1980). Immunoreactive growth hormone-releasing hormone in rat leukocytes. J. Neuroimmunol. 29:1-13. Weigent, D. A., Baxter, J. B., Wean, W. E., Smith, L. R., Bost, K. L., and Blalock, J. E. (1988). Production of immunoreactive growth hormone by mononuclear leukocytes. FASEB J. 2: 2812-2818. Weigent, D. A., Blalock, J. E., and LeBoeuf, R. D. (1991). An antisense oligodeoxynucleotide to growth hormone messenger ribonucleic acid inhibits lymphocyte proliferation. Endocrinology 128:2053-2057.
Immunoreactive Growth Hormone Production by Human Lymphocyte Cell Lines Ting-Lin Kao, 1 Scott C. Supowit, ~ E. Aubrey Thompson, 1 and Walter J. Meyer, 1111'2 Received February I, 1992; accepted March 20, 1992 KEY WORDS: growth hormone; lymphocytes; cell culture; polymerase chain reaction; pituitary.
SUMMARY 1. Two human lymphocyte cell lines, a T-cell line and a B-cell line, were shown to produce and secrete immunoreactive growth hormone (irGH). The irGH molecules secreted by the two cell lines appeared to be de novo synthesized and their molecular size was similar to that of pituitary GH as well as irGH secreted by peripheral blood lymphocytes. 2. Affinity-purified irGH molecules had human growth hormone (hGH)-like mitogenic activity on Nb2 cells. These findings indicate that the irGH molecules produced by H9 and IM9 were similar to hGH in structure. 3. However, the irGH messages could not be amplified by polymerase chain reaction (PCR) primers which had been demonstrated to be able to amplify reverse-transcribed hGH messenger RNA successfully, suggesting that the lymphocyte-derived irGH and pituitary hGH are not exactly identical molecules. 4. We conclude that the H9 and IM9 cells produce a growth hormone-related molecule whose structure is different from that in the anterior pituitary.
INTRODUCTION In recent years, numerous bidirectional interactions between the immune system and the neuroendocrine system have been described. Alterations in the neuroen1Department of Human Biological Chemistry and Genetics, University of Texas Medical Branch, Galveston, Texas 77550. 2 To whom correspondence should be addressed at Departments of Psychiatry & Behavioral Sciences and Human Biological Chemistry & Genetics, University of Texas Medical Branch, Galveston, Texas 77550. 483 0272-4340/92/1000-0483506.50/0(~ 1992 Plenum Publishing Corporation
484
Kao, Supowit, Thompson, and Meyer
docrine system affect the immune system, and the reverse is also true. One of the proposed mechanisms for the intersystem communication is the production of neuropeptides by the cells of the immune system. Several of the lymphocytederived pituitary hormones have been characterized, such as ACTH, endorphin, thyrotropin (TSH), chorionic gonadotropin, luteinizing hormone, and growth hormone (Smith et al., 1983, 1986; Emanuele et al., 1990; Weigent and Blalock, 1990; Harbour-McMenamin et al., 1986). Since all these hormones can also function as immunomodulators, it is proposed that cells in the immune system are able to produce neuroendocrine hormones, which can in turn have autocrine or paracrine effects on the immunocytes themselves. In the past two decades, increasing evidence has pointed to the theory that growth hormone (GH) not only is an essential neuroendocrine hormone but also plays an important role in the immune system. G H has been found to have influences on the thymus, the lymphocytes, and the macrophages, both in vitro and in vivo (Kelley, 1990). Recently, production of immunoreactive growth hormone (irGH) by human peripheral blood lymphocytes (PBL) has been reported (Weigent et al., 1988; Hattori et al., 1990). Characterization of lymphocyte-derived irGH molecules indicate that they have molecular weight and bioactivity similar to their pituitary counterpart. Quantitation using a highly sensitive immunoassay showed that PBL secreted irGH at picogram level (Hattori et al., 1990). Furthermore, irGH molecules have been shown to be able to stimulate lymphocyte proliferation, suggesting that irGH may play an autocrine/paracrine role in lymphocyte replications (Weigent et al., 1991). The use of primary cell cultures for study of irGH production results in an inability to obtain a large, homogeneous cell population and adds difficulty to studies on regulation of irGH secretion. Development of model systems using lymphocyte cell lines could be one way to solve the problem. Similar systems have been developed for TSH and prolactin (Harbour et al., 1989; DiMattia et al., 1988). Previous study suggested that both T and B lymphocytes secrete irGH (Hattori et al., 1990). In the present study, we report identification of a T-cell and a B-cell lymphocyte cell lines which produce irGH molecules as well as their characterization.
MATERIALS A N D METHODS Cell Cultures
IM9 cells were obtained from the laboratory of Dr E. Brad Thompson, and H9 cells from Dr Miles W. Cloyd, both at the University of Texas Medical Branch in Galveston. The cells were cultured in RPMI 1640 supplemented with either 5% fetal calf serum (GCS) or insulin-transferrin-sodium selenite medium supplement (Sigma). Nb2 cells were provided by Dr Li Yu-Lee of Baylor College of Medicine. GH3 cells were provided by Dr Scott Supowit of University of Texas Medical Branch in Galveston.
Immunoreactive Growth Hormone Production
485
Reagents Human growth hormone (hGH) was kindly donated by the National Pituitary Agency or purchased from Eli Lilly. Rabbit antisera to hGH were donated by the National Pituitary Agency or purchased from Chemicon. Purified monoclonal antibody to hGH was obtained from Medix Biochemica, Finland. DNA primers were synthesized by the UTMB D N A synthesis laboratory or Genosys.
Indirect Immunofluorescence The assay was performed as described previously (Blalock and Smith, 1980) with the modification that bovine serum albumin (BSA) was employed as a blocker of nonspecific binding. Cells were air-dried to glass coverslips and fixed with absolute ethanol. After preincubation in phosphate-buffered saline (PBS)0.1% BSA for 45 rain, the coverslips were stained first with either polyclonal rabbit antiserum to hGH (Chemicon, 1 : 20) or normal rabbit serum (1 : 100) and then with fluorescein-conjugated goat anti-rabbit IgG (Cappell Laboratories, 1:15). The stained coverslips were mounted and read at 100× magnification under a fluorescence microscope.
Reverse Hemolytic Plaque Assay This was performed as described previously with several minor modifications (Neill and Frawley, 1983; P. F. Smith et al., 1984). Sheep red blood cells (SRBC) were covalently coupled to staphylococcal protein A (SPA; Sigma) as described previously (Mishell and Shiigi, 1980). The protein A-coupled SRBC suspension was mixed with an equal volume of lymphocytes resuspended in RPMI-0.1% BSA-1000/~g/ml penicillin/streptomycin at a concentration of 106/ml, anti-hGH antiserum (Chemicon, 1:50), and guinea pig complement (Cordis Labs, 1:20). The resulting mixture was infused into poly-tAysine-coated Cunningham slide chambers. The plaques were counted after overnight incubation.
Intrinsic Labeling and Affinity Chromatography Cells were resuspended at a concentration of 106/ml in low-amino acid Selectamine medium. At 24 hr, the culture was supplemented with 2 #Ci/ml 3H-amino acid mixture (ICN or Amersham). The supernatant was harvested at 48hr. Activated affinity chromatography units (Nalgene, U12) were bound to purified monoclonal GH antibody (Medix Biochemica, clone code 5802) according to the manufacturer's specification. Cell culture supernatant concentrates were passed 20 times over a PBS-equilibrated anti-hGH unit, and the bound portion was eluted with 0.1 M glyeine.
Sodium Dodecyl Sulfate (SDS)-Polyacrylamide Gel Electrophoresis Affinity-purified, intrinsically radiolabeled irGH was reduced by boiling and electrophoresed on 15% reducing polyacrylamide gels. After electrophoresis, the
486
Kao, Supowit, Thompson, and Meyer
lanes were sliced into 24 0.25-cm fractions. The fractions were individually solubilized with 0.5 ml of Solusol (National Diagnostics) and radioactivity per fraction was monitored in a E-counter.
Immunoradiometric Assay The cells were grown in Dulbecco's modified Eagle medium ( D M E M ) - 5 % FCS at 106/ml cells. When the cells were grown in serum-free DMEM medium, 5% FCS was added at the end of incubation. The culture supernatant was dialyzed against 1% PBS, dispensed into 15-ml centrifuge tubes, lyophilized, and reconstituted in 1/50 of the original volume in water. Fresh D M E M - 5 % FCS media containing hGH were dialyzed, lyophilized, and reconstituted in the same way to set up the standard curve. The concentrated samples were assayed for GH immunoreactivity using the Nichols Institute hGH Immunoassay. The detection limit of the modified assay was 1 pg/ml.
Bioassay The G H bioassay was based on the proliferative response of Nb2 node rat lymphoma cells to lactogenic hormones (Tanaka et al., 1980). Nb2 cells were grown in Fischer's medium for leukemic mice (GIBCO) supplemented with 10% horse serum, 10% FCS, and 10-4M mercaptoethanol. For the assay, Nb2 cells were washed twice in serum-free Fischer's medium and then resuspended in maintenance medium, which was the growth medium without FCS, for 22-24 hr. At the end of incubation, the cells were washed twice, resuspended in the maintenance medium at 2 × 105 ml, and added to a 96-well microplate. Pituitary hGH or afffinity-purified irGH samples at different dilutions were then added to the wells. To test the blocking effects of anti-hGH antibodies, 0.1vol of monoclonal antibodies (Medix Biochemica, 1 mg/ml) was incubated with the irGH samples for 1-1.5 hr before addition to the Nb2 cells. The cells were counted under a microscope using a hemocytometer after 68-70 hr of incubation.
RNA Isolation and Dot-Blot Hybridization Cytoplasmic R N A was extracted from the lymphocytes using a method described by Gough (1988). Total R N A was transferred onto 0.45-/~m nitrocellulose filters using a Bio-Rad dot-blot apparatus. To obtain cDNA probes, plasmid DNA was purified from HB 101 bacteria carrying the plasmids (PBR322) which contain 800-bp H i n d l I I inserts of either rat or human G H cDNA. A 500-bp fragment within the coding region was excised with P v u I I . The bacteria were provided by Dr Weigent of the University of Alabama, who transformed the bacteria using plasmids donated by Dr Denoto of the University of California at San Francisco. Hybridization was carried out according to a standard protocol (Wanl et al., 1981). The blot was incubated with prehybridization buffer (50% formamide, 5× Denhardt's solution, 0.1% SDS, 100/~g/ml of salmon sperm DNA, 5 × SSPE) at 42°C for 1-2 hr. For hybridization a p32 nick-translated probe with a specific activity of 1-5 x 108 cpm//~g was added to the hybridization buffer
ImmunoreactiveGrowth HormoneProduction
487
(prehybridization buffer with 10% dextran sulfate added) at 1-5 x 106 cpm/ml. After overnight incubation, the blot was washed extensively and autoradiographed with an intensifying screen for a week.
Reverse Transcription and Polymerase Chain Reaction (PCR) Total RNA prepared from lymphocytes or pXGH5-transfected cells was reverse transcribed using a modification of a standard protocol (Innis et al., 1990). The following reagents were assembled in a final volume of 20/zl: 1 mM of each of the four dNTP's, 1 unit//~l of RNasin, 0.2/zg of oligo(dT); 5/tg of RNA sample, 200 units of MoMuLV reverse transcriptase, I x MoMuLV RT buffer, and 10 mM DTT. For negative controls, all reagants were included except the MoMuLV reverse transcriptase. The tubes were incubated at 37°C for 1.5 hr, heated at 95-100°C for 5 min, and quick-chilled on ice. Then the transcribed DNA was precipitated twice and resuspended in 20/~1 of water. The PCR mixture was set up in a final volume of 100/~1:200/~mol of each dNTP, 40 pmol of each of the two primers, 20 #1 of the template, l x PCR buffer (Perkin Elmer Cetus), and 2.5 units of Taq DNA polymerase. The reaction mixtures were then run for 35 cycles, with a thermal cycle profile of denaturation at 92°C for 1 min, annealing at 45°C for 2 min, and extension at 72°C for 3 min. The resulting PCR mixtures were analyzed on 2% agarose gels.
Southern Blotting and Hybridization After electrophoresis of PCR samples, the DNA was denatured by alkaline and the treated gel was capillary-transferred onto a 0.45-~m nitrocellulose membrane overnight. The blot was incubated with prehybridization buffer (50% formamide, 6x SSPE, 5x Denhardt's solution, 0.5% SDS, 100/~g/ml denatured, fragmented salmon sperm DNA) at 42°C for 1-2 hr, For hybridization, 32p random-primed, full-length hGH eDNA probe was denatured with NaOH and added directly to the prehybridization buffer at 106 cpm/ml. After 8 to 10 hr of incubation, the blot was washed extensively and autoradiographed with an intensifying screen for 0.5-2 hr.
RESULTS Identification of Cell Lines Which Produce irGH In order to identify cell lines which produce irGH, 10 T-cell and B-cell lines were screened by indirect immunofluorescence. The 10 cell lines were IM9, C7, Hut78, M4, CLL155, CLL156, HB34, TIB190, $49, and BCL. Of the 10 cell lines screened, IM9 cells have shown the best positive staining (Fig. 1). Other cell lines did not stain significantly above the normal rabbit serum background. To show that the IM9 cells were capable of secreting irGH molecules as well as producing them, the reverse hemolytic plaque assay was employed. Representative photomicrographs of the assay are shown in Fig. 2. When no complement was added to the mixture, microscopic analysis revealed lymphocytes in close proximity to the sheep erythrocytes without specific plaque formation (Fig. 2A).
488
Kao, Supowit, Thompson, and Meyer
Fig. 1. Immunofluorescent detection of irGH production in IM9 cells. Top: IM9 ceils stained with anti-GH antiserum. Bottom: IM9 cells stained with normal rabbit serum.
Immunoreactive Growth Hormone Production
Fig. 2. Photomicrographs of lysed area around IM9 cells obtained by GH-specific reverse hemolytic plaque assay after overnight incubation. (A) Complement control at 8x; (B-D) IM9 cells at (B) 8x, (C) 32x, and (D) 100x.
489
49t1
Kao, Supowit, Thompson, and Meyer
Leaving out anti-hGH antiserum produced similar effects. With both complement and antibody, specific plaques or zones of clear hemolysis surrounding lymphocytes were observed and indicated that irGH were released from IM9 cells (Figs. 2B-D). However, the number of plaques observed was relatively low (32 + 13 plaques per 50,000 lymphocytes added), suggesting that the amount of irGH secreted may be only a small fraction of total irGH molecules produced. Attempts have been made to quantitate the level of irGH secretion by IM9 cells. Unfortunately, even though the irGH secretion level is higher than the medium background as measured in the immunoassay, it is still below the detection limit (1 pg/ml) and cannot be accurately measured (data not shown). Nevertheless, screening of a few cell lines not previously screened by immunofluorescence helped us to identify H9, a T-cell line which secreted irGH at a level measurable by our immunoassay. Figure 3 shows the time course of irGH secretion from H9 cells. H9 cells adapted to serum-free medium secrete irGH at a level of 11.3 pg/107 cells at 48 hr, which was comparable to that secreted by cells grown in FCS-containing medium (14.1 pg/107 cells).
Cytoplasmic GH mRNA Expression Since we could not quantitate irGH secretion from IM9 cells using our immunoassay, we attempted to measure its G H R N A expression by dot-blot analysis as an alternative means for possible quantitation studies. Both rat and human GH cDNA probes were employed. The insert was digested with PvulI, which cut a fragment entirely within the translated coding region. GH3 cells, a rat pituitary cell line, were used as the positive control; and L ceils, a fibroblast cell line, were used as the negative control. Figure 4 shows the results of dot-blot analysis. 30
24 o9
Ld ©
6
12
24
48
hours
Fig. 3. Kinetics of irGH secretion by the H9 cells, measured by double-antibody immunoassay. Incubated H9 cell culture supernatants were harvested at various time points and assayed. Error bars represent standard deviations of the means.
Immunoreactive Growth Hormone Production
491
GH3 L IM9
GH3 L IM9 i~iiii~ iiii~ ~
~.~~ ~ ~ ~ ~ ~
~ ~i~ ~ii~i!~!!~i~i!
Fig. 4. Dot-blot of I M 9 R N A hybridized with e i t h e r r G H o r h G H D N A p r o b e s . T o p : R a t p r o b e . G H 3 R N A , 2 a n d 1 ktg; L-cell R N A , 10 a n d 5 ~tg; I M 9 R N A , 40 a n d 2 0 # g . B o t t o m : H u m a n p r o b e . G H 3 R N A - - 2 , 1, a n d 0 . 5 / ~ g : L - c e l l R N A - - 1 0 , 5, a n d 2.5 # g ; I M 9 R N A - - 4 0 , 20, a n d 10/~g.
492
Kao, Supowit, Thompson, and Meyer
De Novo Synthesis of Lymphocyte-Derived irGH
To assure that the irGH molecules secreted by IM9 and H9 cells were synthesized de novo, the cultures were intrinsically labeled with tritiated amino acids. In typical experiments, radiolabeled amino acids (2/tCi/ml) were added to 50 ml of lymphocyte cell cultures and incubated for 24 hr. The supernatant was then passed through an affinity chromatography unit coupled with a purified
5400
.~
IM9
hGH
4540
3680
2820
1960
1100 5
10
15
20
Fraction iXkrnber 1200
H9
hGH 10:30
860
690
520
350
5
10
15
20
Fraction NLrnber
Fig. 5. SDS-PAGE of 3H-labeled irGH molecules eluted from affinity chromatography unit. MW standards: 92,000, 66,000, 45,000, 31,000, 21,000, and 14,000. Arrows point to where pituitary hGH runs on the gel. Top, IM9; bottom, H9.
Immunoreactive Growth Hormone Production
493
monoclonal a n t i - h G H antibody. T h e affinity column bound to 1 - 2 × 104 cpm (0.005-0.01% of the total cpm added). These data strongly suggest that the i r G H molecules were synthesized de novo by the lymphocytes. The radiolabeled, affinity-purified i r G H molecules were electrophoresed through a S D S - p o l y a c r y m i d e gel under reducing condition and the resulting gel cut into small slices to be counted on the fi-counter for determination of electrophoretic profile. Both protein molecular weight standards and pituitary h G H were also run on the same gel. Figure 5 shows the electrophoretic profile of i r G H molecules produced by H9 and IM9 cells. T h e two arrows point to where pituitary h G H migrates on the gel and should represent, respectively, the 21,500 and 20,000 variants of h G H . For both H9 and IM9, the main i r G H band ran between the two h G H bands. T h e r e are also some higher molecular weight bands present, which may represent aggregates of G H molecules (Stolar et al., 1984).
Mitogenic Activity of Lymphocyte-Derived irGH Molecules Nb2 is a rat l y m p h o m a cell line whose growth is dose dependent on h G H and is widely used for bioassay of lactogenic hormones. W e employed the Nb2 bioassay to test the bioactivity of i r G H molecules derived from H9 and IM9 cells. As shown in Table I, when affinity-purified i r G H molecules were added to stationary Nb2 cells, the growth of these cells was stimulated in a dose-dependent fashion and this mitogenic effect could be blocked with a n t i - h G H antibodies. The observation that the bioactivity level of i r G H derived from IM9 cells seemed to be lower than that from H9 cells is consistent with the relative amounts detected in the immunoassay.
Polymerase Chain Reaction Considering the extremely low level of i r G H production by the lymphocyte cell lines, we a t t e m p t e d to amplify the i r G H message using the PCR. The entire
Table I. Nb2 Bioassay of Affinity-Purified Culture Supernatants from H9 and IM Cells
Treatment Medium only HGH
H9 supernatant, affinity purified
IM9 supernatant, affinity purified
hGH concentration (ng/ml) or sample dilution
Number of Nb2 cells × 104/4 ml
0 0.01 0.02 0.05 0.01 1 :2 1 : 2 + anti-hGH 1:4 1:4 + anti-hGH 1:2 1:2 + anti-hGH 1:4 1:4 + anti-hGH
24.8 + 3.7 31.6+2.7 36.2 + 1.1 48.1 + 6.4 62.0 + 9.7 38.5 + 2.3 27.0 + 1.4 34.2 + 0.8 25.8+2.5 31.5+0.9 20.5+2.0 26.8 ± 5.1 21.8 ± 1.3
494
Kao, Supowit, Thompson, and Meyer
hGH cDNA, which was approximately 800 bp long, was to be amplified in two separate segments using two sets of PCR primers. The first set of primers, with sequences identical to nucleotides sense 41-60 (primer a) and antisense 421-440 (primer b), were targeted to amplify a 400-bp fragment at the 5' end of the hGH cDNA, and the second set of primers, with sequences identical to nucleotides sense 421-440 (primer c) and antisense 764-780 (primer d), were targeted to amplified a 360-bp fragment at the 3' end of the hGH cDNA. Total RNA isolated from the H9 and IM9 cells was reverse-transcribed and amplified. To assure that the reverse transcription and PCR reaction were functional, L cells were transfected with pXGH5, a plasmid containing the human hGH DNA, and RNA isolated from successfully transfected cells was used as a positive control in reverse transcription and PCR. Since the plasmid used for transfection contained the full-length (2-kb) hGH genomic DNA, there was no possibility of falsepositives resulting from the amplification of plasmid cDNA. For negative controls, reverse transcriptase was left out of each sample to be reversetranscribed and amplified. The results of the PCR are shown in Fig. 6. Amplifications from cDNA and the transfected cell lines gave fragments of the expected sizes for both the 5' end and the 3' end. No 5' end fragment of the
cDNA1 2 3 4 5 6
.f I 2 3 4 56 cDNA
Fig. 6. Amplifications of reverse-transcribed RNA from H9 and IM9 cells by polymerase chain reaction. The first set of primers is sense primer (bases 41-64) and antisense primer (bases 421-440) targeted to amplify a 400-bp fragment at the 5' end of the hGH cDNA, and the second set of primers is sense primer (bases 421-440) and antisense primer (bases 764-780) targeted to amplify a 360-bp fragment at the 3' end of the hGH cDNA. Left, 5'-end amplification; right, 3' end. For both photos, Lane 1--transfected cell line, reverse-transcribed RNA; Lane 2--transfected cell line, not reverse transcribed; Lane 3---H9, reverse-transcribed RNA; Lane 4---H9 RNA; Lane 5--IM9, reversetranscribed RNA; Lane 6----IM9 RNA. Lanes amplified from hGH cDNA are labeled "cDNA." The first lane in the 3'-end photo is the 500-bp company kit control, marked with an arrow.
Immunoreactive Growth Hormone Production
123456 Fig. 7. Southern hybridization of the amplified 3'-end DNA with h G H cDNA probe. The lanes are numbered as in Fig. 6. Lane 1--transfected cell line, reverse-transcribed RNA; Lane 2--transfected cell line RNA; Lane 3--H9 reverse-transcribed RNA; Lane 4--H9 RNA; Lane 5--IM9 reverse-transcribed RNA; Lane 6--IM9 RNA.
495
496
Kao, Supowit, Thompson, and Meyer
expected size was observed in the lymphocyte lanes, but for the 3' end, a band with almost the expected size was shown in the lymphocyte lanes and was absent in the negative control lanes. Consequently, the gel for the 3' end amplification was transferred to a nitrocellulose membrane and hybridized with 32p-labeled, full-length hGH cDNA. Figure 7 shows the results of Southern hybridization. The band resulted from amplification of the transfected cell line hybridized as expected; however, none of the lymphocyte lanes showed any visible band. The most likely explanation is that the irGH message could not be amplified from the two chosen set of primers, and the appearance of a band of expected size in the lymphocyte lanes of the 3' end gel was a coincidence resulting from nonspecific amplification of some unrelated DNA by the primers.
DISCUSSION
In this paper we report the identification of two lymphocyte cell lines, one T cell and one B cell, which produced irGH. The irGH molecules were characterized using protein chemistry and molecular biology techniques. Although we characterized irGH molecules produced by neoplastic cell lines, our data are in agreement with previous reports on irGH produced by human peripheral blood lymphocytes (Weigent et al., 1988; Hattori et al., 1990). Briefly, we have confirmed these observations that lymphocytes produced irGH which was similar to pituitary hGH in size, bioactivity, and message reactivity with G H cDNA. In addition, Hattori et al. (1990) have reported that both T and B cells were capable of producing irGH, and our identification of an irGH-producing B-ceU and an irGH-producing T-cell line seem to support the observation. Attempts were made to amplify the irGH RNA with a polymerase chain reaction. Even though the 3'-end amplification consistently gave a band of the correct size, the fact that a visible ethidium bromide-stained band could not be hybridized in a Southern blot, while the neighboring positive control bands of similar intensity readily hybridized (band visible in audiograph within I h of exposure), clearly indicates that the amplified band shares very little homology, if any, with the hGH cDNA sequence. Taken together with the fact that the 5'-end fragment amplification was not successful, we may conclude that h G H messenger RNA could not be amplified from the lymphocyte cell lines with the sets of primers we employed. It is most likely that the irGH molecules produced by H9 and IM9 cells were dissimilar to pituitary hGH in the PCR primer region(s). Possibly, these irGH molecules were very similar, but not exactly identical, to their pituitary counterparts. This possibility could be the reason for the extremely low level of immunoreactivity and bioactivity. No literature up to today has definitively proved or disproved that lymphocyte-derived irGH and pituitaryderived hGH are exactly identical. Our study strongly suggested that dissimilarity existed between the two species of molecules. There were other neuropeptidelike immunoreactivities which were identified in lymphocytes and were shown to be not exactly identical to their counterparts in the neuroendocrine system. Examples include immunoreactive corticotropin releasing hormone and prolactin
Immunoreactive Growth Hormone Production
497
(Staphanous et al., 1990; Hiestand et al., 1986). It is possible that i r G H also belongs to this group. The secretion of i r G H by H9 cells has been quantitated by a highly sensitive immunoassay. The major advantage of studying a cell line model system over a peripheral blood lymphocyte system is that the regulation of i r G H secretion by various pharmacological regulators can be observed without the need to consider possible interferences such as individual donor differences, mixed T- and B-cell populations, or the metabolic and endocrinological states on the donor at the time blood was drawn. Using peripheral blood lymphocytes for their study, Hattori et al. (1990) have been unable to observe any effect of G H R H or somatostatin analogue on secretion of irGH. Preliminary studies in our laboratory have also revealed a regulatory mechanism which appeared to be different from that in the pituitary (unpublished results). Since growth hormone has a variety of effects on the immune system, the possibility that human lymphocyte-derived i r G H has autocrine/paracrine effects on the lymphocytes themselves needs to be considered. In rats, it has been demonstrated that antisense oligonucleotides to rat growth hormone message can decrease the production of i r G H by rat spleen lymphocytes as well as inhibit the proliferation of the lymphocytes themselves (Weigent et al., 1991). The facts that both IM9 and human peripheral blood lymphocytes possess growth hormone receptors (Eshet et al., 1975; Kiess and Butenandt, 1985; Lesniak et al., 1974)) and h G H can stimulate proliferation of normal and transformed human lymphocytes (Suzuki et al., 1990; Mercola et al., 1981) suggest that similar growth promoting effects may be observed for H9- and IM9-derived i r G H molecules. Such effects are currently under study in our laboratory.
ACKNOWLEDGMENTS This project was partially supported by a grant from the George and Mary Josephine H a m m a n Foundation, the James W. McLaughlin Fellowship Fund, and N I M H Grant D H H S RO1 MH43279-02. The authors wish to thank Ms Nita Brannon and Ms Judy Van Over for help in preparing the manuscript.
REFERENCES Blalock, J. E., and Smith, E. M. (1980). Human leukocyte interferon: Structure and biological relatedness to adrenocorticotropin hormone and endorphins. Proc. Natl. Acad. Sci. USA 77:5972-5974. DiMattia, G. E., Gellersen, B., Bohnet, H. G., and Friesen, H. G. (1988). A human Blymphoblastoid cell line produces prolactin. Endocrinology 122:2508-2517. Emanuele, N. V., Emanuele, M. A., Teutler, J., Kirstein, L., Azad, N., and Lawrence, A. M. (1990). Rat spleen lymphocytes contain an immunoreactive and bioactive luteinizing hormone releasing hormone. Endocrinology 126:2482-2486. Eshet, R., Manheimer, S., Chobsieng, P., and Laron, Z. (1975). Human growth hormone receptors in human circulating lymphocytes. Horm. Metab. Res. 7:352-353. Gough, N. M. (1988). Rapid and quantitative preparation of cytoplasmicRNA from small number of cells. Anal. Biochem. 73:93-95.
498
Kao, Supowit, Thompson, and Meyer
Harbour, D. V., Kruger, T. E., Coppenhaver, D., Smith, E. M., and Meyer, W. J. (1989). Differential expression and regulation of thyrotropin (TSH) in T cell lines. Mol. Cell. Endocrinol. 64:229-241. Harbour-McMenamin, D., Smith, E. M., and Blalock, J. E. (1986). Production of lymphocyte derived chorionic gonadotropin in a mixed lymphocyte reaction (MCR). Proc. Natl. Acad. Sci. USA 83:6834-6838. Hattori, N., Shimatsu, A., Sugita, M., Kumagi, S., and Immura, H. (1990). Immunoreactive growth hormone (GH) secretion by human lymphocytes: Augmented release by exogenous GH. Biochem. Biophys. Res. Commun. 168:396-401. Hiestand, P. C., Mekler, P., Noramann, R., Grieder, A., and Permmongkol, C. (1986). Prolactin as a modulator of lymphocyte responsiveness provides a possible mechanism of action for cyclosporine. Proc. Natl. Acad. Sci. USA 83:2599-2603. InniS, M. A., Gelland, D. H., Snindky, and White, T. J. (eds) (1990). PCR Protocols: Guide to Methods and Application, Academic Press, New York. Kelley, K. W. (1990). The role of growth hormone in modulation of the immune response. Ann. N.Y. Acad. Sci. 594:95-103. Kiess, W., and Butenandt, O. (1985). Specific growth hormone receptors on human peripheral mononuclear cells: Reexpression, identification, and characterization. J. Clin. Endocrinol. Metab. 60:740-746. Lesniak, M. A., Gorden, P., Roth, J., and Gavin, J. R. (1974). Binding of 125I-human growth hormone to specific receptors in human cultured lymphocytes. J. Biol. Chem. 249:1661-1667. Mercola, K. E., Cline, M. J., and Golde, D. W. (1981). Growth hormone stimulation of normal and leukemic human T-lymphocyte proliferation in vitro. Blood 58:337-340. Mishell, B. B., and Shiigi, S. M. (eds) (1980). Selected Methods in Cellular Immunology, W. H. Freeman, San Francisco. Neill, J. D., and Frawley, L. S. (1983). Detection of hormone release from individual cells in mixed populations using a reverse hemolytic plaque assay. Endocrinology 112:1135-1137. Smith, E. M., Phan, M., Coppenhaver, D., Kruger, T. E., and Blalock, J. E. (1983). Human lymphocyte production of immunoreactive thyrotropin. Proc. Natl. Acad. Sci. USA 80:60106013. Smith, E. M., Morrill, A. C., Meyer, W. J., and Blalock, J. E. (1986). Corticotropin releasing factor induction of leukocyte-derived immunoreactive ACTH and endorphins. Nature 322:881-882. Smith, P. F., Frawley, L. S. and NeiU, J. D. (1984). Detection of LH release from individual pituitary cells by the reverse hemolytic plaque assay: Estrogen increases the fraction of gonadotropes responding to GnRH. Endocrinology 115:2484-2486. Staphanous, A., Jessop, D. S., Knight, R. A., and Lightman, S. L. (1990). Corticotropin-releasing factor-like immunoreactivity and mRNA in human leukocytes. Brain Behav. lmmun. 4:67-73. Stolar, M. W., Amburn, K., and Baumann, G. (1984). Plasma "big" and "big-big" growth hormone (GH) in man: An oligomeric series composed of structurally diverse GH monomers. J. Clin. Endocrinol. Metab. 59:212-218. Suzuki, K., Suzuki, S., Saito, Y., Ikebuchi, H., and Terao, T. (1990). Human growth hormonestimulated growth of human cultured lymphocytes (IM9) and its inhibition by phorbol diesterase through down-regulation of the hormone receptors. J. Biol. Chem. 265:11,320-11,327. Tanaka, T., Shiu, R. P. C., and Gout, P. W. (1980). A new sensitive and specific bioassay for lactogenic hormones: measurement of prolactin and growth hormone in human serum. J. Clin. EndocrinoL Metab. 51:1058-1063. Wanl, G. M., Ong, E., Meinkoth, J., Franco, R., and Baringa, M. (1981). Method for the Diazotized Paper Solid Supports, Schleicher and Schuell, Keene. Weigent, D. A., and Blalock, J. E. (1980). Immunoreactive growth hormone-releasing hormone in rat leukocytes. J. Neuroimmunol. 29:1-13. Weigent, D. A., Baxter, J. B., Wean, W. E., Smith, L. R., Bost, K. L., and Blalock, J. E. (1988). Production of immunoreactive growth hormone by mononuclear leukocytes. FASEB J. 2: 2812-2818. Weigent, D. A., Blalock, J. E., and LeBoeuf, R. D. (1991). An antisense oligodeoxynucleotide to growth hormone messenger ribonucleic acid inhibits lymphocyte proliferation. Endocrinology 128:2053-2057.
