Vol. 1, 129-136, January 1995 Clinical Cancer Research 129 3 L. M. Wakefield, unpublished data. Transforming Growth Factor- 31 Circulates in Normal Human Plasma and Is Unchanged in Advanced Metastatic Breast Cancer Lalage M. Wakefield,1 John J. Letterlo, Theresa Chen, David Danielpour, Robin S. H. Allison, Lee H. Pai, Andrea M. Denicoff, Marianne H. Noone, Kenneth H. Cowan, Joyce A. O’Shaughnessy, and Michael B. Sporn Laboratory of Chemoprevention [L. M. W., J. J. L., T. C., D. D., R. S. H. A., M. B. S.] and Medicine Branch [L H. P., A. M. D., M. H. N., K. H. C., J. A. 0’S.], National Cancer Institute, Bethesda, Maryland 20892-5055 ABSTRACT A method has been developed to determine true plasma transforming growth factor 1 (TGF- 3) levels by using the platelet a granule-specific marker, platelet factor 4, to cor- rect for the TGF- 3 contributed by platelets degranulated ex vivo. TGF-f3 levels were measured on acid-ethanol extracts of human plasma using isoform-specific sandwich enzyme- linked immunosorbent assays. Normal human subjects had 41 ± 2.0 ng/ml TGF- 31 (range, 2.0-12.0; n 42), <0.2 ng/ml TGF- 32, and <0.1 ng/ml TGF- 3 in their plasma. There were no significant changes with age or with hor- monal status, but any given individual showed fluctuations of up to 3-fold in measured plasma TGF-f levels due to unknown factors. Of 28 patients with advanced metastatic breast cancer, 2 had greatly elevated TGF- 31 levels, while the rest were in the normal range. The presence of physio- logically significant levels of TGF- 31 in the plasmas of normal human subjects may indicate previously unsus- pected endocrine roles for this peptide, while TGF- 2 and TGF- 33 appear to act only in a local autocrine/paracrine fashion. INTRODUCTION The TGF- 3s2 are multifunctional cytokines whose proper- ties include being highly potent inhibitors of epithelial cell growth (1). The presence of one or more isoforms in essentially every epithelial tissue probably reflects a critical role in regu- lating normal epithelial homeostasis. TGF431 null mice show hyperproliferation in the basal layer of the epidermis (2), and Received 7/19/94; accepted 9/2/94. 1 To whom requests for reprints should be addressed, at Laboratory of Chemoprevention, National Cancer Institute, Building 41, Room Bi 11 1, 41 LIBRARY DR MSC 5055, Bethesda, MD 20892-5055. 2 The abbreviations used are: TGF- 3, transforming growth factor 3; PF4, platelet factor 4; 3Th, 3-thmombogIobulin; SELISA, sandwich enzyme-linked immunosombent assay. epidermal tumors derived from TGF- 31 null keratinocytes show more rapid malignant progression than their wild-type counter- parts, suggesting that dysregulation of TGF- 31 function may enhance the carcinogenic process (3). Similarly, colon carci- noma cells transfected with antisense TGF- 31 show a more aggressive phenotype when grown in nude mice than do their untransfected counterparts (4). The observation that the anti- estrogen tamoxifen increases TGF- 3 isoform expression in breast cancer cells in vitro and locally in vivo in human breast tumors is also consistent with a potential role for TGF- 3s in preventing tumor progression (5, 6). In later phases of tumor progression, TGF- 3 overexpmes- sion may actually enhance tumorigenesis through paracnine stimulation of the tumor stroma and inhibition of the immune surveillance system. Transfection of a rat prostatic carcinoma cell line with TGF- 3i resulted in larger, more metastatic tumors (7). Furthermore, three immunohistochemical studies (8-10) have shown a correlation between increased staining for TGF- 31 in human mammary carcinomas and disease progres- sion, although it is not clear whether this is causally related to progression or represents a failed homeostatic response. Clearly, changes in local TGF- 3 levels may be important during carcinogenesis in a number of organ systems, with potentially different effects at different stages in the process. We therefore wished to develop a method for accurate determination of cir- culating TGF43 levels to evaluate their potential usefulness as prognostic indicators, as disease markers, or as intermediate biomarkers of chemopreventive or chemotherapeutic efficacy. Human platelets contain large amounts of TGF- 31 in the a granules (- 2500 molecules/platelet (1 i). Thus measurements of circulating TGF- 3 levels must be made on plasma, not serum. Since it is extremely difficult to prepare plasma without any platelet degranulation, it is not clear whether previously deter- mined levels of TGF- 3 in the plasma of normal subjects might not have been derived from platelets degranulated ex vivo in the course of sample handling (12-14). To circumvent this problem, we have developed a method to correct for the platelet contri- bution by measuring levels of the specific platelet a granule protein, PF4. We now present our results for normal controls and for patients with advanced metastatic breast cancer. MATERIALS AND METHODS Collection of Blood and Preparation of Plasma Plasma TGF43 levels were determined in 42 apparently healthy subjects, ranging in age from 20 to 60 years, and included 17 males, 14 premenopausal females, 6 pregnant fe- Research. on December 11, 2021. © 1995 American Association for Cancer clincancerres.aacrjournals.org Downloaded from