Species differences in the expression of transforming growth factor-alpha (TGF-α) in the submandibular gland and pancreas

lnternanonal Journal ofPancreatology, vol. 22. no. 2, 111-119, October 1997 �9 Copyright 1997 by Humana Press Inc. All rights of any nature whatsoever reserved. 0169-4197/97/22:111-119/$10.25

Species Differences in the Expression of Transforming Growth Factor-Alpha (TGF- )

in the Submandibular Gland and Pancreas

Yoshito Ikematsu, Parviz M. Pour,* and Katherine Kazakoff The UNMC/Eppley Cancer Center, University of Nebraska Medical Center, Omaha, NE

Summary

Conclusion. Significant differences exist in the expression of transforming growth factor-alpha (TGF- c0 in the submandibular glands (SMG) and the pancreas of different species and among cell components in the same species.

Background. Our previous studies have shown marked differences in the expression of TGF-~ in the pancreas of humans and Syrian hamsters. To examine whether these differences also exist in other species, we examined the expression of TGF-~ in the pancreas of mouse, rat, Syrian hamster, guinea pig, rabbit, pig, dog, and monkey. We included the SMG of these species for comparison.

Materials and Methods. The formalin-fixed tissues of these species (n = 3) were investigated by immunohistochemistry using a monoclonal antibody to TGF-~. The SMG of rat, mouse, hamster, rabbit, pig, dog, and monkey were examined by RT-PCR to assure the specificity of the antibody.

Results. Remarkable species differences were found in the expression of this peptide in both the SMG and the pancreas. In the SMG, the expression varied in different cell components, even in the same tissue of the species. Although excretory and secretory ducts of the SMG of most species reacted with the antibody, intercalated ducts were immunoreactive only in mouse and guinea pig. Acinar cells were either weakly positive or nonimmunoreactive. In the pancreas of most species, the cells of the large and medium-sized ducts expressed TGF-~, whereas centroacinar cells of only rat and dog reacted with the antibody. Marked differences were found in the expression of TGF-~ in islet cells and in its spatial distribution. Differences were also found in the immunoreactivity of mesenchymal and neural cells among the species.

Key Words: Pancreas; submandibular gland; TGF-~ expression; immunohistochemistry.

Introduction

Transforming growth factor alpha (TGF-c~) is a 50-amino-acid polypeptide that is structurally simi- lar to epidermal growth factor (1) and binds to the

Received August 7, 1996; Revised March 17, 1997; Accepted April 4, 1997.

*Author to whom all correspondence and reprint requests should be addressed: The UNMC/Eppley Cancer Center, University of Nebraska Medical Center, 600 South 42nd Street, Omaha, NE 68198-6805.

same cell- surface receptors as epidermal growth factor (2,3). Originally found in the culture superna- tant of transformed fibroblasts (4,5), this polypeptide is overexpressed in many tumors. Based on the assumption that TGF-c~ produced by tumor cells acts as an autocrine growth factor (6-9), it is believed that overexpression of this polypeptide gives tumor cells a growth advantage. The expression of TGF-c~ has been demonstrated in many adult and fetal human tissues, including digestive tract, pancreas, kidney,

111

112 Ikematsu, Pour, and Kazakoff

thyroid gland, adrenal gland, brain, skin, mammary gland, genital tissue, and placenta (6,7,10-13). Whether or not the expression of the polypeptide in these tissues is comparable to that in cancer cells deriving from them has not been well studied.

In our previous immunohistochemical study, we found high levels of TGF-o~ in the normal pancreas and duodenal mucosa of some individuals, but con- sistently high levels in ductal/ductular cells in individuals with chronic pancreatitis to the same extent as in pancreatic ductal adenocarcinomas (13). How- ever, the induced pancreatic ductal cancer in Syrian hamsters, which in many aspects resembles the human disease, did not overexpress this growth factor. In this comparative study, we also found that pancreatic cells of humans and hamsters express TGF-o~ differently. A higher expression of this growth factor was found in human pancreatic acinar and ductal cells than in the hamster pancreas. Strik- ingly, unlike the random distribution of TGF-ot in human islet-cells, a strong reactivity of the antibody was found only in the c~ cells of the hamster pancreas. This finding indicates species differences in the expression of this ligand. Because there is no information on species differences in the expression of TGF-a, we investigated the pattern of TGF-c~ expression in the pancreas of the mouse, rat, hamster, guinea pig, rabbit, pig, dog, and monkey. The submandibular glands of these animals were also used as control tissue because this gland is known to be a major source of epidermal growth factor production (10,14,15).

Materials and Methods

Animals The following animals were used: Swiss mouse,

MRC-Wistar-derived rat, and Syrian golden hamster (all 8-10 wk old and all from the Eppley colony); guinea pig (Hartley, 8-wk old); dog (mongrel, middle age); pig (Yucatan, 2-yr old); and monkey (Rhesus, 7-yr old).

Antibody

Monoclonal anti-TGF-~ (Ab-2, clone 213-4.4, IgG2a) was purchased from Oncogene Science Inc. (Uniondale, NY). According to the manufacturer's information, this antibody reacts with denatured and native TGF-ot of human and rat origin but shows no

crossreactivity with human or mouse EGF. The reactivity of this antibody has been shown by us and oth- ers (7) to correlate with the mRNA expression of the ligand.

RNA Isolation and Enzymatic Amplification of mRNA of TGF- a Total RNA from the salivary gland of mouse,

Syrian golden hamster, rat, rabbit, dog, pig, and monkey was isolated using Tri-Reagent (Molecular Research Center, Cincinnati, OH). Hamster abdominal rectus muscle and kidney were also used as a negative and positive control, respectively. GAPDH was amplified as an internal control of mRNA. Each sample of mRNA was then amplified by reverse transcriptase (RT). Briefly, total RNA samples (2 ~tg) were incubated with 50 mM Tris-HC1, pH 8.3, 75 mM KC1, 3 mM MgC12, 4 mM DTT, 1 [tg oligo(dT)12q8 primer, 4 x 0.5 ~tL deoxynucleotides (10 mM), 10 U ribonuclease inhibitor, and 200 U Superscript TM

RNase H-reverse transcriptase (Gibco-BRL, Gaithers- burg, MD) at 42~ for 60 min. The products were amplified by PCR. The PCR primers of TGF-a are as follows: the 5' primer, TC-3, is a 22-mer 5'-GGCCTT GGAGAACAGCACGTCC-3' from nucleotide position 97 to 118, the 3' primer, TC-2, is a 20-mer 3'-GCTCTTCGGGTCGCGGGAGG-5' from nucleotide position 454 to 473 (16). Primers of GAPDH are as follows: the 5' primer, p7, is a 20-mer 5'-GCA TCCTGGGCTACACTGAG-3', the 3' primer p8 is a 20-mer 5'-CACCACCCTGTTGCTGTAGC- 3' (17). Primers were synthesized as described (16) by a model 394 DNA/RNA Synthesizer (Applied Bio- systems, Foster City, CA). PCR core reagents were purchased from Perkin-Elmer-Cetus (Norwalk, CT). PCR was performed with 5 ~tL of RT products in a mixture containing 50 mM KC1, 1.5 mM MgC12, 10 mM Tris-HC1, pH 8.0, 4 • 0.5 [tL deoxynucleotides (10 mM), 0.3 ~L Taq-polymerase (5 ~/uL), 2 x 1 ~L primers (1 ~tg/~L) in a total volume of 50 ~tL. The PCR cycle was varied as follows: 3 min at 95~ then 40 times the following cycle: 1 min at 95~ 1 min at 66~ (60~ for GAPDH), 1 min at 72~ The last cycle was followed by 7 min at 72~ PCR products were analyzed in 3% agarose gel and visualized by ethidium bromide staining. PCR products of hamster, mouse, and rat were purified by a gel-extraction kit from Qiagen (Chatsworth, CA). DNA sequencing with a fluorescent detection method was carried out

International Journal of Pancreatology Volume 22, 1997

RRH Species Difference in TGF-ot Expression 113

using the DyeDeoxy Terminator fluorescent sequencing kit (Applied Biosystems) on an Applied Biosystems model 373A automatedDNA sequencer (17). Sequence analysis was accomplished using the sequence analysis program from Applied Biosystems and was compared with the previously reported sequences (16).

Tissue Preparation and Histology

The pancreas of each species was removed and immediately placed in Bouin's solution, fixed overnight, and processed for histology according to con- ventional methods.

Immunohistochemistry

Paraffin-embedded tissues were cut into serial sections, each 4 ~t thick. One of these sections was stained with hematoxylin and eosin, and the rest were processed by immunoperoxidase procedures with a Vectastain ABC kit (Vector Laboratories, Burlin- game, CA) as described earlier (13). The optimal concentration of TGF-(x was found to be 2.5 ~tg/mL based on the manufacturer's recommendation and our pilot studies using the antibody. We incubated the specimens with the antibody overnight at 4~ The slides were pretreated with saponin (0.05% for 30 rain at room temperature), as recommended by the manufacturer. Control slides were treated as follows: Tissues were incubated with PBS instead of the primary antibody, and tissues were incubated with UPC-10 (isotype IgG1) or MOPC-21 (IgG 2) instead of the primary monoclonal antibody. To minimize the technical bias in the staining, we kept the incuba- tion time and temperature constant and processed as many slides as possible at the same time. In some cases, we repeated the staining process for confirma- tion. The reactivity and staining intensity of pancreatic cells were compared with that of human nerve fibers, which express TGF-c~.

Scoring The reactivity of the cells with the antibody was

given in arbitrary units: -, +, ++, and +++, when the staining intensity was judged to be negative, weak, moderate or strong, respectively. The cellular localization of the reactive product was classified as a cytoplasmic, Golgi, or glycocalyx pattern. Cytoplas- mic staining was defined as staining of the whole cytoplasm. If the stained coarse granules were confined to the projected region of the Golgi apparatus, a Golgi pattern of reactivity was used. The glycocalyx

TGI:

506

GAP

506

Fig. 1. PCR product analysis. Ethidium bromide- stained agarose gel (3%) of PCR products. Top, TGF-~; bottom, GAPDH. M = 1 kb molecular-weight marker elec- trophoresed at the same time. Monkey SMG (A), pig SMG (B), dog SMG (C), rabbit SMG (D), rat SMG (E), mouse SMG (F), hamster SMG (G), hamster abdominal rectus muscle (H), and hamster kidney (I).

pattern was defined as staining of the luminal surface of the cells.

Results

RT-PCR Analysis

RT-PCR of TGF-c~ and GAPDH is shown in Fig. 1. TGF-~ mRNA was visualized in the SMG of monkey (Fig. 1A), dog (Fig. 1C), rat (Fig. 1E), mouse (Fig. 1F), and hamster (Fig. 1G). Expected molecular size was 374 bp in TGF-~ and 162 bp in GAPDH.

Immunohistochemistry The reactivity of the pancreata and SMGs of each

species with the antibody is summarized in Tables 1 and 2. In the mouse, the reactivity of pancreatic acinar cells was very weak and was confined to the basal cytoplasm. No other pancreatic cells were stained. The cell membranes of peripancreatic fat cells were strongly immunoreactive. In the SMG, the excretory and granulated ducts stained with a moderate intensity, whereas the reactivity of intercalated ductal cells was strong (Fig. 2). The reactivity was generally cytoplasmic, except in the excretory ducts, which



Table 1 The Patterns of TGF-~ Expression in the Pancreas

Species Large ducts Medium-sized ducts Terminal ductules Centroacinar cells Islet Acini

Mouse a . . . . + Rat + + - +++ - /+ ++ Hamster + + - - +++ + Guinea pig +++ ++ ++ - +/++ + Rabbit + + - /+++ - - +/++ pig + . . . . +++ Dog - - - +++ - - Monkey +++ +++ +++ - ++ - /+

a_, no staining; +, weak staining; ++, moderate staining; +++ strong staining;/, or.

Table 2 The Reactivity of Submandibular Glands of Different Species with Anti-TGF-er

Species Excretory ducts Secretory ducts Intercalated ducts Acinar cells

Mouse ++a ++ +++ _

R a t - - / + + + - - /+ Hamster ++ ++ - - /+ Guinea pig +/+++ +/+++ ++ - Rabbit ++ ++ - - /+ pig + + - - Dog +/++ - - + Monkey +++ +++ - - /+

a_, no staining; +, weak staining; ++, moderate staining; +++ strong staining;/, or.

v~v!,

Fig. 2. Submandibular gland (SMG) of mouse. Reac- tivity of excretory ducts (arrows) and of intercalated ducts (arrowheads) with anti-TGF-~. Anti-TGF-cr antibody and ABC method were used in all the figures (Bar = 100 ~tm).

showed a stronger staining of the apical cy top lasm

than the remaining cytoplasm.

In the rat, pancreat ic ductal and islet cells were unstained, whereas acinar cells were stained weakly and centroacinar cells were stained strongly. A very

weak staining was seen in smooth muscle and endothelial cells, and a modera te staining was seen in the membranes of fat cells. In the SMG, the secretory, but not the excretory ducts, were immunoreac t ive .

The reactivity varied in different regions of the ducts

as did the staining intensity (Fig. 3). Only a few aci-

nar cells were stained weakly or moderately .

The reactivity of the hamster pancreat ic cells with

both antibodies was the same as reported earlier (13). A strong cytoplasmic staining of cells in the islet

per iphery corresponding to the cr cells (13), a weak staining of acinar cells, and a very weak staining of

ductal cells were seen (Fig. 4). Nerve fibers, smooth musc le cells, endothel ia l cells, and fat cells also

stained weakly. In the SMG, the excre tory ducts

stained strongly, whereas the secretory ducts stained


Fig. 3. SMG of rat. A weak staining of the excretory ducts (arrow) but a strong immunoreactivity of some secretory ductal cells (arrowheads) with anti-TGF-~. A few acinar cells show weak reactivity (•

AI

�9 j

I

Fig. 4. Pancreas of Syrian hamster. A strong reactivity of islet cells in the islet periphery. A weak staining of acinar cells and a weaker reactivity of ductal cells (arrows) with anti-TGF-~. ABC method (Bar = 250 p.m).

Fig. 5. (A) Pancreas of guinea pig. Moderate-to-strong staining of large and medium-sized ductal cells (arrows) and a weaker staining of the small ducts (arrowheads). Islets (I) are weakly stained. (B) SMG of guinea pig. Excretory and secretory ductal cells are stained with different intensity. Acinar cells (A) are unstained (Bar = 100 ~tm).

randomly and moderately to strongly. The staining intensity varied from one area to another.

The pancreas of guinea pig showed a strong staining of large ducts, moderate staining of small ducts, and a very weak or unrecognizable stain-

ing of the acinar cells with anti-TGF-c~ (Fig. 5A). The staining of the islet cells differed. It was weak in the central portion, but moderate in the periphery of the islets (Fig. 5A). A strong staining was found in smooth muscle cells, the pancreat ic

International Journal of Pancreatology 115 Volume 22, 1997


Fig. 6. Pancreas of rabbit. Scattered strong staining of terminal ductules (arrowheads) and a weak staining of acinar cells (A) with anti-TGF-a. Islet cells (I) are unstained (Bar = 100 p.m).

Fig. 7. Pancreas of pig. A moderate-to-strong staining of acinar cells (A) and a weak staining of islet cells (I). In the islet, the capillary cells are strongly stained (•

fibrous capsule, and nerve fibers, and a moderate staining was found in endothelial cells and fat cells. In the SMG, the staining was confined to the excretory and secretory ducts, where the reactivity and the staining intensity varied considerably from one area to another (Fig. 5B).

In the pig pancreas, a weak staining of ductal cells and a moderate staining of acinar cells was found. In acinar cells, the reactivity was stronger in the basal portion of the cell than in the apical portion. Islet cells were unstained. The staining of endothelial cells and nerve fibers was very weak and that of fat cell membranes was moderate. In the SMG a weak staining of excretory and secretory ducts, nerve fibers and ganglia was found. No other cells of this gland were stained.

In the rabbit pancreas, the acinar cells were stained moderately to strongly. The staining was diffuse cytoplasmic and was stronger in the basal cell zone. In some areas a few or many ductular cells were strongly stained, whereas the staining of large ducts was weak and that of the islets was negative (Fig. 6).

The pancreas of the pig showed a strong staining of acinar cells, whereas the ductal sys- tem and islets were unstained. A strong stain-

ing was also seen in vascular endothelial cells of islets (Fig. 7).

In the dog, centroacinar cells of the pancreas showed a strong reactivity to the antibody. Other cells, including acinar cells, ductal cells, vascular endothelium, and nerve fibers were unstained or showed a very weak reactivity (Fig. 8A). In the SMG, the excretory ducts showed granular staining of the cytoplasm, mostly in the luminal cell zone and less frequently in the basal cell portion. A stronger staining was seen in a few cells at the base of the epithelium (Fig. 8B). These reactive cells seemed to represent the myoepithelial cells. The basal portion of some acinar cells were stained weakly and diffusely.

In the monkey, the reactivity of acinar cells with the antibody was weak, whereas ducts and terminal ductular cells were strongly stained (Fig. 9A). Almost all islet cells were stained with moderate intensity. The reactivity of vascular and endothelial cells was strong. In the SMG, the excretory and secretory ducts were stained moderately and diffusely, whereas the staining of acinar cells was patchy, diffuse, and mostly in the basal cell portion (Fig. 9B). Nerve fibers, vascular wall cells, and smooth muscle cells were stained strongly.


RRH Species Difference in TGF-a Expression 117

A

~ v

B

Fig. 8. (A) Pancreas of dog. The immunoreactivity is confined to centroacinar cells. Acinar cells, ducts (arrows), and islets (I) are unstained (x250). (B) SMG of dog. Granu- lar staining of excretory ducts. Stronger staining of paraductal cells (arrowheads). Other structures, including vessels (V) are unstained (Bar = 100 ~tm).

Fig. 9. (A) Pancreas of monkey. Moderate staining of islet cells (I), small ducts (arrowheads) and ductules, and centroacinar cells (small arrowheads), and stronger staining of large ducts (arrows). (B) SMG of monkey. A strong staining of excretory and secretory ducts and a random weak staining of acinar cells (A) (Bar = 100 gin).

Discussion

Immunohistochemically, considerable differences were found in the expression of TGF-~ and its cellular distribution in the SMGs and pancreata of eight species. TGF-~ mRNA was demonstrated

by RT-PCR in all but the pig and rabbit, although the signals of GAPDH appeared intact. However, because the immunohistochemistry confirmed the presence of TGF-~ in the SMG of both pig and rabbit, the negative RT-PCR results could be a result of inappropriate primer design. The primers used in this



study were originally designed for hamster (16), which might not be comparable with that in pig and rabbit. Considering the correlation between immunohistochemistry and RT-PCR in the other five species, it appears that the antibody recognizes a common epitope in all eight species. It is also possible that in pig and rabbit the protein is derived from external sources. The results showed that findings in these eight species also differ from the same human tissues (12,13). Differences in the reactivity of anti-TGF-~ with the islets of these species was the most striking. Only human (12,13), guinea pig, monkey, and hamster pancreatic islets express TGF-~. Yet, among these species, the level of expression and the cellular localization of the peptide varied considerably. In the normal human pancreas, TGF-a was randomly dis- tributed among the islet cells, whereas in the hamster, the expression was restricted to the ~ cells, and in the guinea pig, the expression was moderate and corresponded to the expression of epidermal growth factor in this species (19). Reasons for these remarkable differences are obscure.

The differences in the level and the site of expression of this peptide among the species may reflect different functions of the same cells in different species or be the result of differences in the nutritional status or age. The latter possibility is supported by the finding of a greater expression of TGF-~ in the duodenal epithelium of human fetal tissue than in the adult mucosa (10), and the first appearance of EGFR in the SMG of pigs after 7 d of age (20). These results indicate that the growth factors are important in the developing tissues. A higher expression of TGF-c~ in the ductal cells of patients with chronic pancreatitis also points to the involvement of this growth factor in regenerative and differentiation processes. A weak or lack of expression of TGF-o~ in the pancreatic acinar cells of all species may reflect the minor role of these peptides in the functional regulation of these cells, whereas the type 13 transforming growth factor (TGF-131) has been shown to influence pancreatic acinar cell growth (21). However, the reason for the lack of TGF-~ in the pancreatic islets of some species and its exceptionally high expression in the hamster pancreas is obscure. Because the hamster is thus far the only species that, in response to certain carcino- gens, develops pancreatic ductal cancer comparable to human tumors, the overexpression of TGF-o~ in the islets of this species may play a role, in that TGF-c~ acts

in a paracrine mode as a growth promoting factor for the development of tumors, most of which arise within or in the immediate area of islets (22-24).

The presence of a high level of TGF-~x in pancreatic ductal carcinomas has been thought to be an important event in malignancy (6, 7). However, the high level of this growth factor in the normal tissue of various species, and the low incidence of SMG tumors expressing EGFR (25) conflict with this assumption. Examina- tion of larger samples from normal tissues of different species and tissues with benign and malignant dis- eases can help us to understand the role of these growth factors in development, differentiation, and malignancy. The study also calls for caution in relating findings in animal models to humans.

References

1 Derynck R, Goeddel DV, Ullrich A, Gutterman JU, Wil- liams RD, Bringman TS, Berger WH. Synthesis of mes- senger RNAs for transforming growth factors alpha and beta and the epidermal growth factor receptor by human tumors. Cancer Res 1987; 47: 707-712.

2 Derynck R, Roberts AB, Winkler ME, Chen KY, Goeddel DV. Human transforming growth factor-a: precursor structure and expression in E. Coli. Cell 1984; 38: 287-297.

3 Massague J. Epidermal growth factor-like transforming growth factor. II. Interaction with epidermal growth actor receptors in human placenta membranes and A431 cells. JBiol Chem 1983; 258: 13,614-13,620.

4 Delarco JE, Todaro GI. Growth factor from murine sarcoma virus-transformed cells. Proc Natl Acad Sci USA 1978; 75: 4001-4005.

5 Roberts AB, Lamb LC, Newton DL, Sporn MB, Delarco JD, Todaro GJ. Transforming growth factors: isolation of polypeptides from virally and chemically transformed cells by acid/ethanol extraction. Proc Natl Acad Sci USA 1980; 77: 3494-3498.

6 Korc M, Chandrasekar B, Yamanaka Y, Friess H, Buchgler M, Beger HG. Overexpression of the epidermal growth factor receptor in human pancreatic cancer is associated with concomitant increases in the levels of epidermal growth factor and transforming growth factor alpha. J Clin Invest 1992; 90: 1352-1360.

7 Smith JJ, Derynck R, Korc M. Production of transforming growth factor a in human pancreatic cancer cells: evidence for a superagonist autocrine cycle. Proc NatlAcad Sci USA 1987; 84: 7567-7570.

8 Yoshida K, Yasui W, Ito H, Tahara E. Growth factors in progression of human esophageal and gastric carcinomas. Exp Pathol 1990; 40: 291-300.

9 Yoshida K, Kyo E, Tsujino T, Sano T, Niimoto M, Tahara E. Expression of epidermal growth factor, transforming growth factor a and their receptor genes in human gastric carcinoma; implication for autocrine growth. Jpn J Cancer Res 1990; 81: 43-51.


RRH Species Difference in TGF-a Expression 119

10 Miettinen PJ. Transforming growth factor alpha and epidermato growth factor expression in human fetal gastrointestinal tract. Pediatr Res 1993; 33:481-486.

11 Lei ZM, Rao CV. Expression of epidermal growth factor and transforming growth factor-alpha, in human fallopian tubes. Endocrinology 1992; 131: 947-957.

12 Yasui W, Ji ZQ, Kuniyasu H, Ayhan A, Yokozaki H, Ito H, Tahara E. Expression of transforming growth factor alpha in human tissues: immunohistochemical study and Northern blot analysis. Virchow's Archiv A Pathol Anat 1992; 421: 513-519.

13 Tomioka T, Toshkov I, Kazakoff K, Andren-Sandberg A, Takahashi T, Buchler M, Friess H, Vaughn R, Pour PM. Cellular and subcellular localization of transforming growth factor-a and epidermal growth factor receptor in normal and diseased human and hamster pancreas. Teratogenesis, Carcinogenesis, and Mutagenesis 1995; 15: 231-250.

14 Dubiel B, Mytar B, Tarnawski A, Zembala M, Stachura J. Epidermal growth factor (EGF) expression in human salivary glands. An immunohistochemical study. J Physiol Pharmacol 1992; 43: 21-32.

15 Kajikawa K, Yasui W, Sumiyoshi HI, Yoshida K, Naka- yama H, Ayhan A, Yokozaki H, Ito H, Tahara E. Expression of epidermal growth factor in human tissues. Immuno- histochemical and biochemical analysis. Virchow's Archiv PatholAnat Histopathol 1991; 418: 27-32.

16 Chiang, T, McBride, J, Chou, MY, Nishimura, I, Wong DTW. Molecular cloning of the complementary DNA encoding for the hamster TGF-cz mature peptide. Carcino- genesis 1991; 12: 529-532.

17 Visser CJ, Wouterson RA, Bruggink AH, van-Garderen- Hoetmer A, Seifert-Bock I, Tilanus MG, de-Weger RA.

Transforming growth factor-alpha and epidermal growth factor expression in the exocrine pancreas of azaserine- treated rats: modulation by cholecystokinin or a low fat, high fiber (calorie restricted) diet. Carcinogenesis 1995; 16: 2075-2082.

18 Smith LM, Sanders JZ, Kaiser RJ, Hughes P, Dodd C, Connell CR, Heiner C, Kent SB, Hood LE. Fluorescence detection in automated DNA sequence analysis. Nature 1986; 321: 674-679.

19 Vaughan TJ, Pascall J C, James P S, Brown KD. Expression of epidermal growth factor and its mRNA in pig kidney, pancreas and other tissues. Biochem J 1991; 279:315-318.

20 Jaeger LA, Lamar CH. Immunolocalization of epidermal growth factor (EGF) and EGF receptors in the porcine upper gastrointestinal tract. Am J Vet Res 1992; 53: 1685-1692.

21 Logsdon CD, Keyes L, Beauchamp RD. Transforming growth factor-13 (TGF-[31) inhibits pancreatic acinar cell growth. Am J Physiol 1992; 262: G364-G368.

22 Pour P.M. Islet cells as a component of pancreatic ductal neoplasms. I. Experimental study. Ductular cells, including islet cell precursors, and primary progenitor cells of tumors. Am J Pathol 1978; 90: 295-316.

23 Pour PM. Histogenesis of exocrine pancreatic cancer in the hamster model. Envir Health Persp 1984; 56: 229-243.

24 Ishikawa O, Ohigashi H, Imaoka S, Nakai I, Mitsuo M, Weide L, Pour PM. The role of pancreatic islets in experimental pancreatic carcinogenicity. Am JPathol 1995; 147: 1456-1464.

25 Yamada K, Iwai K, Okada Y, Mori M. Immunohisto- chemical expression of epidermal growth factor receptors in salivary gland tumors. Virchow' s Archiv A Patho! Anat Histopathol 1989; 415: 523-531.


Documents

Species differences in the expression of transforming growth factor-alpha (TGF-α) in the submandibular gland and pancreas