Prolactin synthesized and secreted by human ... - Europe PMC

21 downloads 0 Views 1MB Size Report
Feb 25, 1992 - 2.0o agarose gel electrophoresis, transferred by capillary transfer to a Hybond-N ... 4-chloro-3-indolyl phosphate p-toluidine salt. Hormonal ...
Proc. Natl. Acad. Sci. USA Vol. 89, pp. 7713-7716, August 1992

Medical Sciences

Prolactin synthesized and secreted by human peripheral blood mononuclear cells: An autocrine growth factor for lymphoproliferation (endocrine/immune system)

P. SABHARWAL*, R. GLASER*t, W. LAFUSEt, S. VARMA*, Q. Liut, S. ARKINSf, R. KOOIJMAN§, L. KUTZt, K. W. KELLEYt, AND W. B. MALARKEY*¶ Departments of *Medicine and tMedical Microbiology and Immunology, The Ohio State University Medical Center, Columbus, OH 43210; tLaboratory of Immunophysiology, Department of Animal Sciences, University of Illinois, Urbana, IL 61801; and §Universiteitskliniek voor Kinderen en Jeugdigen, Utrecht, The Netherlands

Communicated by Leo Paquette, April 21, 1992 (received for review February 25, 1992)

ABSTRACT Prolactin has been shown to have an immunoregulatory role in the rodent immune response. A prolactinlike molecule has also been found in mouse splenocytes and a human B-lymphoblastoid cell line. We have evaluated whether human peripheral blood mononuclear cells (PBMCs) synthesize and/or secrete prolactin. We used the polymerase chain reaction (PCR) to generate a 276-base-pair prolactin product from human PBMCs, and Southern blot analysis confirmed that it was related to prolactin. Western blotting using a polyclonal antibody to prolactin indicated that cell extracts prepared from human PBMCs contained a high molecular mass (60-kDa) immunoreactive prolactin. To determine whether this PBMC prolactin was being secreted, we developed a highly sensitive and specific hormonal enzyme-linked immunoplaque assay. With this assay, we were able to detect human prolactin secretion from concanavalin A (Con A)- or phytohemagglutinin-stimulated PBMCs but not from unstimulated PBMCs. We next sought to determine whether this secreted prolactin could function as an autocrine growth factor in lymphoproliferation. We observed that anti-human prolactin antiserum significanly inhibited human PBMC proliferation in response to Con A or phytohemagglutinin. We conclude that a prolactin-like molecule is synthesized and secreted by human PBMCs and that it functions in an autocrine manner as a growth factor for lymphoproliferation.

mononuclear cells (PBMCs) synthesize and secrete a prolactin-like molecule that functions in an autocrine loop as a growth factor for lymphoproliferation.

MATERIALS AND METHODS Oligonucleotide Primer Design. The sequences and positions of the two 21-mer primers used for PCR amplification of human prolactin cDNA are shown in Fig. 1 (10). The 5' primer is located within exon 3 of the human prolactin gene and the 3' primer spans exons 4 and 5. The 3' primer was chosen so that any contaminating genomic DNA would not be amplified in the PCR. These primers amplify a 276-basepair (bp) fragment spanning exons 3-5. cDNA Synthesis and PCR Amplifications. One microliter (-0.5 Zg) of cellular RNA, obtained from human decidua or Ficoll/Hypaque (Pharmacia LKB) isolated PBMCs (11) obtained from healthy subjects, was used for each cDNA strand-synthesis reaction. Total cellular RNA was isolated by a single-step method using thiocyanate/phenol/chloroform extraction. The cDNA reaction mixture (20 AlI) contained 1.25 mM dNTPs (United States Biochemical), 1.6 ,ug of random hexamer [pd(N)6; Pharmacia LKB), and 200 units of Moloney murine leukemia virus reverse transcriptase (GIBCO/BRL). The reaction was carried out at 37°C for 1 hr and the enzyme was inactivated by heating at 95°C for 5 min. Ten microliters of the cDNA reaction mixture was then added to a PCR mixture (final volume, 100 ,l) containing Taq polymerase (1.65 units; Perkin-Elmer/Cetus) 5' and 3' primers (0.2 1.M), and PCR buffer (0.5 mM MgCl2/50 mM KCI/10 mM Tris-HCl, pH 8.3/0.001% gelatin). Reaction mixtures were subjected to 30 cycles of PCR (each consisting of denaturation at 94°C for 75 sec, annealing at 55°C for 90 sec, and extension at 72°C for 90 sec. In-gel amplification was then performed in 1.5% remelted low-melting-point agarose (GIBCO/BRL). Ten microliters of PCR products (=1 ng/ul) was subjected to PCR amplification as described above. Double-distilled water served as a negative control for reverse transcription and amplification. Southern Blotting. Southern transfer and hybridization were performed to confirm that the PBMC cDNA was specifically amplified. A human prolactin cDNA clone (PhPRL) containing the complete coding sequence was obtained from American Type Culture Collection) (ATCC no. 31721). The cDNA insert was excised by Pst I digestion and purified from agarose gels by Geneclean (Midwest Scientific, Valley Park, MO). The PCR products were separated by

Recent studies suggest an immunoregulatory role for prolactin in rodents. In animals, hypophysectomy results in cessation of the growth of the thymus gland (1), decreased antibody titers against sheep red blood cells, and depressed delayed hypersensitivity reaction to chlorodinitrobenzene (2). Bromocryptine-induced hypoprolactinemia in mice injected with Listeria monocytogenes increases mortality that is associated with impaired lymphocyte proliferation and decreased production of macrophage-activating factors by T lymphocytes (3). Further, a prolactin-like molecule is secreted following Con A stimulation of murine lymphocytes (4), and a prolactin-like mRNA as well as a secreted product have been detected in human B-lymphoblastoid cell lines (5, 6). In contrast, Clevenger et al. (7) could not demonstrate prolactin-specific mRNA or prolactin secretion following interleukin 2 stimulation of a mouse T-lymphocyte line. Several investigators (7-9) using rodent lymphoid cell lines or splenocytes have found that a prolactin-like protein is required for lymphocyte mitogenesis. The relevance of these observations to human cellular immunity has not been explored. Here we document that human peripheral blood

Abbreviation: PBMC, peripheral blood mononuclear cell. ITo whom reprint requests should be addressed at: Ohio State University Medical Center, Room N-1106, Doan Hall, 410 West Tenth Avenue, Columbus, OH 43210.

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. 7713

7714

Proc. Natl. Acad Sci. USA 89 (1992)

Medical Sciences: Sabharwal et al. 5' PRIME: 5'-GGG-TTC-ATT-ACC-AAG-GCC-ATC-3' Intron A

Intron B

! EXON2

T

5'

Intron

Intron a

IEXO3

3L PRIMEL: 3'-CAG-TCG-GTC-CAA-GTA-GGA-CTT-5' FIG. 1. Diagram of human prolactin gene and PCR primers.

2.0o agarose gel electrophoresis, transferred by capillary transfer to a Hybond-N nylon membrane (Amersham), and crosslinked to the membrane by UV irradiation. The membrane was prehybridized in hybridization buffer [50%o (vol/ vol) deionized formamide/0.25 M sodium phosphate/pH 7.2/0.25 M NaCl/1 mM EDTA/7% (wt/vol) SDS/7% (wt/ vol) PEG] at 420C for 15 min and then hybridized to a prolactin-specific cDNA probe (2 x 106 cpm/ml) in the same hybridization buffer at 42TC overnight. The probe was 32p_ labeled by the random primer method (GIBCO/BRL) and denatured by the addition of 0.15 M NaOH. The membrane was washed by incubation with 300 mM NaCl/30 mM sodium citrate, pH 7.0) at 22TC for 5 min followed by 0.25 M sodium phosphate, pH 7.2/1% SDS/1 mM EDTA at 65TC for 1.5 hr and then exposed to Kodak X-AR film at -70°C for 6 hr. Restriction Endonuclease Digestion. Samples of human placenta and PBMC cDNA obtained by reverse transcription and PCR amplification were precipitated, resuspended in 15 ,ul of double-distilled water, and incubated for 1 hr at 37°C with 10 units of Rsa I (GIBCO/BRL). Restriction products, together with undigested samples and a 123-base-pair molecular size marker (GIBCO/BRL), were electrophoresed in a 2% agarose gel and stained with ethidium bromide. Western Blot Analysis. Cell extracts were prepared by lysis in 50 mM TrisHCl, pH 8.0/150 mM NaCl/1% (vol/vol) Triton X-100/0.02% (wt/wt) NaN3 containing phenylmethanesulfonyl fluoride (100 jig/ml) and aprotinin (1 jug/ ml). Samples were dissolved in 63 mM Tris HCl, pH 6.8/4%

A

53

r

;

8 9 1 C;G

B

SDS/10%o glycerol/5% dithiothreitol by heating for 5 min at 100°C. Samples were subjected to SDS/polyacrylamide gel electrophoresis using a 4% stacking gel and 14% separating gel (1.0 mm x 130 mm x 160 mm). Proteins were transferred from gels to nitrocellulose membranes in a Bio-Rad Transblot apparatus (0.1 M Trizma base) at 0.2 mA per gel for 4.0 hr at 4°C. Immunostaining was achieved with an amplified alkaline phosphatase immunoblot assay kit (Bio-Rad). Antiserum dilutions were as follows: anti-human prolactin [National Institute of Diabetes and Digestive and Kidney Diseases; crossreactivity < 0.001% for growth hormone (GH), thyroidstimulating hormone (TSH), follicle-stimulating hormone (FSH), luteinizing hormone (LH), lipotropic hormone, and placental lactogen], 1:75,000; goat anti-rabbit IgG conjugated with biotin, 1:3000. Prolactin-like molecules were visualized by reaction with nitroblue tetrazolium chloride and 5-bromo4-chloro-3-indolyl phosphate p-toluidine salt. Hormonal Enzyme-Linked Immunopaque Assay. Prolactin was measured by a highly sensitive hormonal enzyme-linked immunoplaque assay method (12) developed in our laboratory. In brief, the nitrocellulose membrane filter bottoms of microtiter plates were first coated with a monoclonal antibody (clone 94194) to human prolactin (1:1000) (Fitzgerald Industries, Chemsford, MA); crossreactivity < 0.01% for GH, TSH, FSH, and LH), and then freshly prepared PBMCs (105 in 100 std) were plated on the nitrocellulose membranes and incubated for 48 hr. Thereafter, cells were removed and the plates were washed twice with 100 Al of phosphatebuffered saline (PBS). A second antibody, goat anti-human prolactin antiserum (1:1000) (BioSpacific, Emerysville, CA; crossreactivity < 0.01% for GH, TSH, FSH, and LH) was added to the wells and incubated at 37°C for 4 hr. The wells were washed eight times with 200 j of PBS/0.05% Tween 20. Horseradish peroxidase-conjugated rabbit anti-goat IgG (1:800) (BioSpacific) was added to the wells and incubated for A B

C D

E F

_W4W

G H -60 kDa

24(E

- 26 kDa

FIG. 2. (A) PCR-amplifled prolactin cDNA derived from human placenta and human PBMCs visualized by ethidium. bromide staining after electrophoresis in 2% agarose. Lane 1, no cDNA template; lane 2, 123-bp molecular size marker; lane 3, placenta; lane 4, PBMCs from donor 1; lane 5, PBMCs from donor 2; lane 6, PBMCs (donor 2) stimulated with Con A for 48 hr; lanes 7-10, Rsa I digestion of the PCR-amplified cDNA shown in lanes 3-6, respectively. (B) Southern blot of PCR-amplified prolactin products hybridized with a randomprimer-labeled prolactin cDNA. Lane 1, placenta; lane 2, PBMCs from donor 1; lane 3, PBMCs from donor 2; lane 4, PBMCs (donor 2) stimulated with Con A for 48 hr.

FIG. 3. Immunoblot showing immunoreactive prolactin-like molecule from human PBMCs and the Burkitt lymphoma cell line AG-876. Lanes A and B, placenta; lanes C and D, AG-876; lanes E and F, human PBMCs stimulated with Con A for 48 hr; lanes G and H, human PBMCs. Note the 60-kDa prolactin-like molecule from extracts of resting and Con A-treated PBMCs as well as AG-876 cells.

Medical Sciences: Sabharwal et al.

Proc. Natl. Acad. Sci. USA 89 (1992)

7715

FIG. 4. Photomicrographs of hormonal enzyme-linked immunoplaque assay showing that unstimulated PBMCs secrete no prolactin (A) whereas human PBMCs stimulated by the mitogen Con A (10 jlg/ml) form specific prolactin plaques. (B) Similar prolactin plaques were observed when PBMCs were stimulated by phytohemagglutinin (2.5 pg/ml). These findings were repeated using PBMCs from 10 healthy individuals.

1 hr at 370C. The wells were washed eight times with 200 Al of PBS/0.05% Tween 20 and then twice with 200 A.l of Tris-buffered saline. Prolactin-specific plaques were visualized by reaction with 4-chloro-1-naphthol. The violet plaques on the dried filter membranes in individual wells were photographed at x50 magnification, using a microscope (Wild Heerburgg, Switzerland) x 10 with a zoom objective. Measurement of [3HlThymidine Incorporation. Peripheral blood lymphocytes (106 per ml) from five subjects were cultured in RPMI-1640 supplemented with 1% Nutridoma SP (Boehringer Mannheim) and 1% fetal bovine serum. Samples were run in triplicate in 100-l. volumes in 96-well tissue culture plates. The cells were treated with either Con A (0.3-40 pg/ml) or phytohemagglutinin (0.3-40 pg/ml) plus anti-human prolactin antiserum (1:10,000). To evaluate DNA synthesis, 0.5 gCi of [3H]thymidine (6.7 Ci/mmol; ICN; 1 Ci = 37 GBq) was added to each well and 8 hr later the cells were harvested onto glass-fiber filters with a PHD cell harvester (Cambridge Technology, Watertown, MA). The filters were air-dried and radioactivity was measured by liquid scintillation spectroscopy (Betaplate scintillation counter; LKB).

RESULTS Evidence for Prolactin mRNA in Human PBMCs. To determine whether prolactin mRNA could be expressed in human PBMCs, we attempted to measure steady-state levels of prolactin mRNA in resting and Con A-stimulated human PBMCs by Northern blot analysis. Despite repeated attempts, prolactin mRNA was not found in resting or Con A-stimulated PBMCs grown in chemically defined (Nutridoma SP) RPMI 1640 medium containing only 1% fetal bovine serum. We then attempted to identify prolactin mRNA by PCR. Prolactin gene sequences were selectively amplified by reverse transcription followed by PCR. Total cellular RNA derived from placenta and PBMCs were used as a template to synthesize first-strand cDNA (Fig. 2A). This first-strand cDNA was then subjected to PCR with 5' and 3' primers complementary to exon 3 and exon 5, respectively (Fig. 1). Prolactin PCR products of the appropriate size (276 bp) were generated from both human placenta and PBMC cDNA (Fig. 2A). Southern blot analysis confirmed that the PCR products were related to prolactin (Fig. 2B). If the PCR products originated from prolactin cDNA, enzymatic restriction with Rsa I should yield two fragments of 116 and 160 bp. The results (Fig. 2A) showed that two fragments of correct size were generated with Rsa I-digested PCR products. A similar band pattern was evident in both placenta and PBMC cDNA cleaved with Rsa I. Also, a similar PCR product was generated from an Epstein-Barr virus genome-positive Burkitt

lymphoma B-lymphoblastoid cell line (AG-876) (data not shown). Evidence for a Prolactin-Like Molecule in Human PBMCs and a Human B-Lymphoblastoid Cell Line. We found a 60-kDa immunoreactive prolactin product in Western blots from extracts of resting human PBMCs and PBMCs treated with Con A for 48 hr to induce proliferation (Fig. 3). Activation of lymphocytes by Con A increased the amount of prolactin in cell extracts (Fig. 3). The Epstein-Barr virus genome-positive Burkitt lymphoma B-lymphoblastoid cell line AG-876 was found to express the same 60-kDa prolactin product (Fig. 3). Inactive PBMCs did not secrete detectable prolactin when evaluated by the enzyme-linked immunoplaque assay (Fig. 4). However, when activated with Con A and photohemagglutinin, the cells secreted a prolactin-like peptide into the culture medium. Secreted Prolactin Functions as an Autocrine Growth Factor. We explored the possibility that this human prolactin-like product secreted by PBMCs could function as a lymphocyte comitogen. When human prolactin antiserum was added to phytohemagglutinin- or Con A-stimulated PBMCs, [3H]thymidine incorporation fell by more than 45% (Figs. 5 and 6). We conclude from these data that the prolactin synthesized 14 7

12-

10l 0

x

E

c0 0

1

10

Phytohemagglutinin, ,u g/ml FIG. 5. Effect of rabbit anti-human prolactin antiserum on [3H]thymidine incorporation into DNA in phytohemagglutinin-

stimulated (E) human PBMCs. Nonspecific IgG (-) or specific anti-human prolactin antiserum (*) was added to cultures at a final dilution of 1:10,000. Data points represent mean + SEM from four experiments. Note the inhibition of mitogenesis in wells containing anti-prolactin antibodies.

Proc. Natl. Acad Sci. USA 89 (1992)

Medical Sciences: Sabharwal et al.

7716 10-

6

10 x E

0

2.4.

2-

0.-

0.1

,..i1

10

100

Con A, ,ug/ml

FIG. 6. Effect of rabbit anti-human prolactin antiserum on [3H]thymidine incorporation into DNA in Con A-stimulated (s) human PBMCs. Nonspecific rabbit IgG (*) or specific anti-human prolactin antiserum (m) was added to cultures at a final dilution of 1:10,000. Data points represent mean + SEM from four experiments. Note the inhibition of mitogenesis in wells containing anti-prolactin antibodies.

by human PBMCs was interacting in an autocrine fashion following stimulation by these mitogens.

DISCUSSION In this study we provide evidence that human PBMC synthesize, store, and secrete prolactin that acts as a comitogen in lymphoproliferation. The active synthesis of prolactin by human PBMCs was demonstrated by our ability to generate an appropriate 276-bp prolactin PCR product from human PBMC cDNA, which was similar to the prolactin PCR product generated from human placental cDNA. Northern blots of both unstimulated and Con A-stimulated PBMCs were not sensitive enough to reveal the presence of any specific prolactin mRNA, but prolactin mRNA was present in human PBMCs, since it served as a template for the cDNA in our PCR study. Cell extracts and supernatants prepared from both unstimulated and mitogen-stimulated human PBMCs contained a protein that was antigenically similar to pituitary-derived prolactin. After reduction with dithiothreitol this prolactin molecule derived from human PBMCs had an approximate molecular mass of 60 kDa in SDS/polyacrylamide gels. The Burkitt lymphoma cell line AG-876 also synthesized a 60-kDa form of prolactin and contained human prolactin mRNA detectable by PCR. These data are consistent with studies that demonstrated a prolactin-like protein in a B-lymphoma cell line (5) and mouse splenocyte lysates (13, 14), with a molecular mass of 48 kDa or 22 kDa. Multiple molecular weight forms of pituitary and serum prolactin in humans have been described: a predominant 22-kDa (little prolactin) form with lesser amounts of a 48-kDa

(big prolactin) and a 150-kDa (big big prolactin) species (15). In this study we found only a 60-kDa prolactin species in the Burkitt lymphoma cell line and in human PBMCs. Further experiments will be needed to determine the sequence homology of the 48-kDa prolactin species from human pituitary and the 60-kDa species we found in human PBMCs. Using a highly sensitive and specific hormonal enzymelinked immunoplaque assay, we detected human prolactin secretion by PBMCs only after stimulation with mitogens. This finding suggests that prolactin is synthesized and stored in resting human PBMCs and that activation of the cells is required to induce prolactin secretion. Similar findings have been noted with mouse splenocytes, which require Con A stimulation for prolactin secretion (5). We were able to demonstrate that this secreted 60-kDa prolactin was of physiological importance, as antiserum to human prolactin inhibited DNA synthesis in Con A- and phytohemagglutinin-stimulated lymphocytes. Since there was no human prolactin available to the PBMCs in the culture medium, this finding suggests that the 60-kDa prolactin-like protein was secreted by the cells, bound to prolactin receptors, and then migrated to the nucleus to serve as a comitogen and autocrine regulator of cell growth. These findings support the concept that locally generated neuropeptides can modulate immune function. We thank Michelle Vermillion for assistance in manuscript preparation. This work was supported in part by National Institute of Mental Health Grants MH40787 and MN44660, by National Institutes of Health Clinical Research Center Grant M01-RR-0034, by National Institutes of Health Grant AG06246 to K.W.K., and by The Ohio State University Comprehensive Cancer Center Core Grant CA16058. 1. Smith, P. E. (1980) Anat. Rec. 47, 119-126. 2. Nagy, E. & Berczi, I. (1978) Acta Endocrinol. 89, 530-537. 3. Bernton, E., Meltzer, M. & Holaday, J. W. (1988) Science 239, 401-404. 4. Montgomery, D. W., Shah, G. N., Zukoski, C. F., Buckley, A. R., Pacholczyk, C. F. & Russell, D. H. (1987) Biochem. Biophys. Res. Commun. 145, 692-698. 5. DiMattia, G. E., Gellerson, B., Bohnet, H. G. & Friesen, H. G. (1988) Endocrinology 122, 2508-2517. 6. Boglia, L. A., Cruz, D. & Shaw, J. E. (1991) Endocrinology 128, 2266-2272. 7. Clevenger, C. V., Russell, D. H., Appasamy, P. M. & Prystowsky, M. B. (1990) Proc. Nadl. Acad. Sci. USA 87, 64606464. 8. Hartmann, D., Holaday, J. W. & Bernton, E. W. (1989) FASEB J. 3, 2194-2202. 9. Clevenger, C. V., Altmann, S. W. & Prystowsky, M. B. (1991) Science 253, 77-79. 10. Truong, A. T., Duez, C., Belayew, A., Renard, A., Pictet, R., Bell, G. I. & Martial, J. A. (1984) EMBO J. 3, 429-437. 11. Glaser, R., Rice, J., Sheridan, J., Fertel, R., Stout, J., Speicher, C., Pinsky, D., Kotur, M., Post, A., Beck, M. & Glaser, J. K. (1987) Brain Behav. Immunol. 1, 17-22. 12. Varma, S., Sabharwal, P., Sheridan, J. & Malarkey, W. B. (1992) Endocr. Soc. 74, 456 (abstr.). 13. Kenner, J. R., Holaday, J. W., Bernton, E. W. & Smith, P. F. (1990) Prog. NeuroEndocrinImmunol. 3, 188-195. 14. Montgomery, D. W., LeFevre, J. A., Ulrich, E. D., Adamson, C. R. & Zukoski, C. F. (1990) Endocrinology 127, 2601-2603. 15. Kiefer, K. A. & Malarkey, W. B. (1978) J. Clin. Endocrinol. Metab. 46, 119-124.