The identification of urogastrone in serum, saliva, and gastric juice

GASTROENTEROLOGY

77:313-318,1979

The Identification of Urogastrone in Serum, Saliva, and Gastric Juice H. GREGORY, S. WALSH, and C. R. HOPKINS ICI Pharmaceuticals Limited, Mereside, Alderley Park, Macclesfield, Cheshire, and University Liverpool, Department of Histology and Cell Biology, Liverpool, England

ffrogastrone, a peptide isolated from human urine, is known to cause inhibition of gastric acid secretion and proliferation of fibroblasts in culture; furthermore immunoj’luorescent localization techniques show it to be present in submandibular and Brunner’s glands. Serum, saliva, and gastric juice samupon Sephadex Gples have now been fractionated 200 and G-50 and the immunoreactive urogastrone located using a specific radioimmunoassay. Biologic activity was shown by mitogenic studies with 3T6 fibroblasts. In serum, the major immunoreactive component was ca. 2-2 x 105 daltons, but trypsin treatment then gave a smaller biologically active species in the same position as pure urogastrone on Sephadex G-50. Both saliva and gastric juice showed major components at the position defined by urogastrone, and these also stimulated the uptake of rH)thymidine into the fibroblasts. It is concluded that a urogastrone-like molecule can be released enzymically from a high molecular weight serum precursor and that the small biologically active peptide is also a normal component of saliva and gastric juice.

Urogastrone has been shown to have the two distinctive properties of causing inhibition of gastric acid secretion’ and of functioning as a mitogen on fibroblasts in culture.’ Although it has been known for many years that extracts of human urine would inhibit gastric secretion3 it was only after urogastrone was characterized that the mitogenic effects became apparent as a consequence of the structural correlation with mouse epidermal growth factor (mEGF).’ This compound was isolated from mouse submaxillary glands, and of the 53 amino acid resiReceived August 7,1978.Accepted March 14,1979. Address reprints requests to: Dr. H. Gregory, ICI Pharmaceuticals Limited, Mereside, Alderley Park, Maclesfield, Cheshire SK10 4TC, England. 0 1979 by the American Gastroenterological Association 9915-5935/79/939313-96$92.99

of

dues comprising the single chain of both urogastrone and mEGF, 37 were commonly located.5 Mouse epidermal growth factor is known to have mitogenic effects upon many types of cells in culture: and in the comparative studies thus far conducted, urogastrone and mEGF have similar actions upon epithelial growth and upon gastric acid secretion. Urogastrone will inhibit acid secretion in the human at doses of 0.25 pg kg-’ hr-‘,’ and in cultured human fibroblasts, the ED, for thymidine incorporation occurs at 0.3 ng ml-‘.’ It is not known whether the action as a human epithelial growth factor or as a gastrointestinal hormone is of primary importance or whether both effects are a manifestation of a single mechanism. Large amounts of mEGF occur in male mouse submaxillary glands, but only recently has urogastronelike immunoreactivity been identified in duct cells of human submandibular glands and in Brunner’s glands of the duodenum.’ Further detailed studies have confirmed the presence of immunoreactivity in the acinar cells and in the secretory ducts of these glands.” Although urogastrone is isolated from human urine as a relatively small peptide, it is probable that it does not occur in blood in this form alone.” We have now carried out further studies on the nature of urogastrone in blood, saliva, and gastric juice and sought to correlate immunoreactivity with one biologic property-the mitogenic action upon fibroblasts in culture, which is essential to draw any conclusions as to the physiologic relevance of the observations.

Methods Samples of human serum and unstimulated saliva were obtained as occasional samples, and gastric juice aspirates were provided by Mr. J. Elder, Manchester Royal Infirmary. Saliva samples were used immediately, but sera and gastric juices were stored for short periods at -20’32. Radioimmunoassay measurements indicated up to 1.5 ng ml-’ urogastrone in intact samples, and the first stage fractionations were upon Z-ml aliquots.

GREGORY ET AL.

GASTROENTEROLOGY Vol. 77,No.2

i

Fraction ~l.lmll Figure 1. Gel chromatography on Sephadex G-ZOOof Z-ml samples of serum (A-A), saliva (O----O), and gastric juice (U) with the observed immunoreactivity.

Gel chromatography was carried out at 4°C on columns of Sephadex G-200 (105 x 1.1 cm) and Sephadex G-50 superfine (100 x 0.9 cm) in a buffer at pH 7.2 composed of 0.04 M phosphate, 0.15 M sodium chloride, 0.01 M EDTA, and 0.1% sodium azide. Combined fractions were desalted between stages and before looking at mitogenic effects, by chromatography upon Sephadex G-25 superfine (35 X 1 cm) in 0.05 M acetic acid solution. Radioimmunoassay measurements were carried out as previously described,” using aliquots of 250 ~1 from col-

umn fractions per assay tube. Smaller aliquots from the final desalted samples were made to 250 ~1 in the standard albumin containing assay buffer. The assay system measured down to 5 pg per tube. Tryptic digestion of the high molecular weight blood proteins containing the immunoreactivity was carried out typically as follows: 94 mg salt-free protein that were derived from 4 ml serum were taken into 600 ~10.1 M ammonium bicarbonate; 50 ~1 0.2 M calcium chloride were added, followed by 100 ~1 of a freshly prepared 10 mg/ml solution of TPCK trypsin (Worthington Biochemical Corp. Freehold, N.J.) in 0.1 M ammonium bicarbonate. The solution was kept at 37°C for 3 hr, cooled to 4’C, and immediately applied to the gel column. The fractions for biologic study were derived from 2-5 ml of the original secretions to provide approximately similar amounts for bioassy. The desalted samples were taken into water containing 2 mg ml-’ human albumin (Kabi, Stockholm) and the quantities determined by RIA (see Table 3). For studies on fibroblasts, samples were reconstituted in distilled water and then diluted 1: 1 with doublestrength Dulbecco modified minimum essential medium (DMEM). 3T6 fibroblasts were grown in DMEM and 10% fetal calf serum (FCS). In preparation for assay, they were subcultured at an initial density of 2 x lo4cells/5 cm diameter Petri dish cultured for 3 days in DMEM/lO% FCS and then maintained 3 days in DMEM/O.l% FCS. For assay, a minimum of five replicates were used for each test fraction. The dishes were washed three times in DMEM minus serum and then incubated 24 hr in either control or experimental media. For the last 8 hr of the incubated [3H]thymidine (Radiochemical Centre, Amsterdam) was added to the medium (to give a final concentration 3 pCi/ml). After the 24hr incubation, the dishes were rinsed three times in DMEM minus serum, fixed in 1% formaldehyde, and prepared for autoradiography. The labeling index (percentage of labeled nuclei) was estimated in the Geisma-stained

Figure 2. Gel chromatography on Sephadex G-50 of the high molecular weight serum protein from G-200 trypsin treatment (U) compared with 25% of the sample as control (0- - - -0) with the observed immunoreactivity. t

Fraction (1.7mll

UROGASTRONE IN SERUM, SALIVA, AND GASTRIC

August 1979

JUICE

315

_,

w i? L c

Figure 3. Gel chromatography on Sephadex G-50 of the low molecular weight components

::

80;

‘\

;

\

/

:

gz $” 60;

from G-200 of saliva (e - - -0) and gastric juice (C--O) and their observed immunoreactivity.

i

40-

; & 20-

VO

“I

Frxtion

autoradiographs. A stock sample mEGF was used routinely to check the response of successive preparations over a period of months.

Results Fractionation of human serum upon Sephadex G-200 gave a major immunoreactive component at a much higher molecular weight region than the position defined by pure isolated urogastrone, and it seemed to be ca. 1-2 x lo5 daltons (Figure 1). It was only possible to obtain RIA measurements at the calibrated urogastrone position by combining fractions and then desalting these so that the whole sample could be measured. This gave values of lo-20 pg ml-’ serum, whereas the high molecular weight fractions gave values up to 1 ng ml-‘. Treatment of the combined high molecular weight protein fractions with trypsin caused breakdown, to give a peak of immunoreactivity in the urogastrone position on Sephadex G-50, whereas the control incubation with a smaller amount of blood protein retained high molecular weight character (Figure 2). Whole saliva showed some apparent immunoreactivity at the very high molecular weight region where large mucopolysaccharides would occur, but

clearly, a low molecular weight component also existed (Figure 1). On the second column, some of this emerged again at the urogastrone region and some at a larger molecular position (Figure 3). A very similar pattern of behavior occurred with gastric juice samples. Fractions were combined as indicated (see Table 2), desalted, and quantitated using the RlA system and the mitogenic effects observed. Five percent FCS was clearly mitogenic upon the 3T6 fibroblasts, whereas 0.1% FCS produced little effect (Table 1). Of the test samples, only saliva gave unambiguous results at 1 ng ml-‘, because at this concentration, the other samples caused the cells to round up and detach from the dish. The addition of 2 mg ml-’ soya bean trypsin inhibitor (SBTI) prevented cell detachment in all samples except the untreated high molecular weight fraction from serum (Figure 2,13-20), indicating that detachment was probably due to contaminating protease. The addition of SBTI had little effect on the labeling index (Table 2) in control and FCS-stimulated preparations and was therefore included in assays. Results given in Tables 2 and 3 were derived from the same assay. In the presence of SBTI and at concentrations of

Table 2. Table

A. The effect of Fetal Calf Serum (FCS), mEGF,

Urogastrone, and Fractionated Saliva upon 3T6 Fibroblasts Sample 1. 0.1% FCS 2. 3. 4. 5.

5% FCS mEGF, 1 ng ml-’ Urogastrone, 1 ng ml-l Saliv,a (24-32), 1 ng ml-’

Labeling

index

8.14 83.40 33.33 65.83 74.83

(SE, P-value

(1.32) (1.89, (2.47, (3.89, (4.56,

P < P< PC P<

0.001) 0.001) 0.001) 0.001)

cf 1)

11.7mli

The Effect of FCS and mECF Upon 3T6 Fibroblasts With and Without Soya Bean Trypsin Inhibitor (SBTI)

Sample 1. 2. 3. 4. 5. 6.

0.1% FCS 0.1% FCS + SBTI 2 mg ml-’ 5% FCS 5% FCS + SBTI 2 mg ml-’ mEGF 1 ng ml-’ mEGF 1 ng ml-’ + SBTI 2 mg ml-’

Labeling

index

6.88 5.68 66.33 63.63 26.75 28.67

(SE, P-value

(0.82, (1.06) (3.99, (6.10, (3.43, (3.29,

NS) P< PC P c P <

0.001) 0.001) 0.001) 0.001)

cf 2)

316

Table

GREGORY ET AL.

3.

The Effect of Various Upon 3T6 Fibroblasts

GASTROENTEROLOGY

Fractions fromSephadex Determined in the Same

Total immunoreactive urogastrone (ng)

Sample 1. Serum trypsin (13-23) 2. Serum trypsin (25-33) 3. Serum control (13-20) 4. Saliva (14-23) 5. Saliva (24-32) 6. Gastric juice (15-22) 7. Gastric juice (23-32)

Vol. 77, No. 2

G-50 of Serum, Saliva, and Gastric Juice with 2 mg ml-’ SBTi Assay as in Table 2 Labeling index (SE, P-value cf. 2, Table 2) __~______-.___ ____ .___ 0.1ng ml-’ 0.01 ng ml-’

4.3 4.9 4.2 5.3 4.4 1.8 2.0

0.1 ng ml-‘, significant mitogenic affects were observed with the saliva, gastric juice, and trypsintreated serum fractions from the defined urogastrone chromatographic region. In a separate assay, urogastrone at 0.1 ng/ml gave a labeling index four times that of the 0.1% FCS. None of the other results was significantly different from the 0.1% FCS control values. In the presence of SBTI, the high molecular weight fraction of the serum control at 0.1 ng ml-’ still caused some detachment of the cells. However, at 0.01 ng ml-‘, it gave similar values to the trypsintreated sample; these were different from the control but with low significance. Direct comparison with the serum control was not possible at the higher concentration, but it is worth noting that the residual high molecular weight sample from the tryptic digest did not give mitogenic effects at 0.1 ng ml-’ (Table 3, No. 1).

Discussion Sensitive radioimmunoassay systems can be influenced nonspecifically by large amounts of serum or other biologic fluids, and it is necessary to carry out some fractionation to increase the significance of measurements upon them. In addition to immunoreactivity as a means of identification, it is highly desirable to have some alternative confirmation to specificity by other physical or biologic characteristic. Previously we have reported measurements of urine levels of urogastrone by RIA upon volumes of l-2 ~1” but have regarded measurements upon larger amounts of intact serum or saliva with reservation. We have now fractioned these upon two gel chromatography systems which were selected to allow for the separation of immunoreactive components from the bulk of higher and lower molecular weight contaminants. The radioimmunoassay system used to identify urogastrone-like reactivity” has good specificity for the peptide; e.g., mEGF with 70% homology with urogastrone is re-

5.66(1.79, NS) 26.20(3.01, PC 0.001) 6.64(1.75, NS) 14.17(2.24, P < 0.01) 7.60(1.91, NS) 29.75(3.14, P < 0.001)

11.33(2.17P< 0.05) 12.17(2.36, P c 0.05)

19.56(2.46, P-C 0.001)

quired at over thousandfold greater concentrations to displace the radiolabel in a comparable manner. Inhibition of acid secretion by urogastrone occurs in vivo at doses of 0.1-l pg kg-‘, whereas mitogenic effects take place in vitro at 0.1-1.0 ng ml-‘.’ Because RIA measurements upon serum, for example, show levels up to 1 ng ml-‘, it is possible to seek effects upon isolated cell systems using quite small volumes of serum. The influence of fractionated serum or se cretions was therefore studied using 3T6 fibroblasts which had been shown to respond clearly to pure mEGF or urogastrone (Table 1). In urine, urogastrone occurs almost entirely as the characterized low molecular weight peptide, but preliminary data indicated that more than one immunoreactive species occurs in blood.‘” This was clearly shown by fractionation of serum upon Sephadex G-200, where the major detectable component was only just retarded and appeared to have a molecular weight of l200,000 daltons. Only by combining fractions at the position defined by the pure peptide could a low molecular weight entity be shown by RIA. There is considerable precedent for the occurrence of biologically active peptides in multiple forms in blood. Gastrin occurs as several species of increasing molecular size” which are contained within the largest molecule. Intracellularly, hormones can also be associated with a higher molecular weight carrier such as neurophysin with oxytocin13 or indeed mEGF, which is associated with a higher molecular weight protein in mouse submaxillary glands.” Many of the known prohormones have sequences of two or more basic amino acid residues at the points where cleavage can occur to give the smaller units. Although the precise enzymes involved are not known, trypsin is also capable of effecting cleavage of Such bonds. This enzyme will also break additional bonds involving basic amino acids, but we have previously established that pure urogastrone is relative stable toward this enzyme.‘5 Trypsin was

August 1979

UROGASTRONE

therefore used to treat the mixture of high molecular weight proteins under conditions whereby a control H-Ser-Pro-Ser-Arg ( Leu-Pro-Gly-Tyr-NH,, peptide, was split into two fragments. Gel chromatography of the protein digest showed that a new peak of immunoreactivity appeared in the urogastrone position (Figure 2). Material from this position showed mitogenie effects upon cultured fibroblasts, whereas the residual high molecular immunoreactive material did not. The existence of urogastrone in urine implies that it is transported in blood from the glands of origin, entering either as an endocrine secretion or by absorption from the gastrointestinal tract. The action of trypsin on the blood is presumptive evidence that biologically active peptide can be released in the bloodstream from a larger protein of lower, if any, activity. When pure urogastrone is infused into humans at 0.25 pg kg-’ hr-‘, profound inhibition of stimulated secretion occurs,’ and if the half life in blood is similar to that in dogs,” then plateau levels would be no more than 100-200 pg ml-‘. Total immunoreactivity in blood samples frequently is of the order of 0.5-l ng ml-’ before, during, and after meals (J.B. Elder, I.E. Gillespie, and H. Gregory, unpublished observations), and presumably, therefore, much of this is in a less potent form. Immunoreactivity can be demonstrated in whole saliva, but because of the viscous nature of the secretion, nonspecific effects cannot be discounted. After chromatography on Sephadex G-200 (Figure l), a major portion of the measurable material was found in the same low molecular weight region as urogastrone, although some reaction was observed in the large mucopolysaccharide region. The lower molecular weight sample gave two regions of immunoreactivity on the second column, one of which again coincided with urogastrone, and furthermore, this sample gave positive mitogenic effects (Table 3, No. 4). This was not seen in the higher molecular fractions (Table 3, No. 4) and the significance of this region is questionable-although clearly detectable in the particular sample illustrated, it was rather less in other samples. Similar results were obtained with gastric juice and thus confirm that mitogenic activity related to urogastrone immunoreactive characteristics can be detected in salivary glands exocrine secretion and survives in gastric juice. It is not known whether the latter derives from saliva, from gastric secretion, or somehow from the duodenal glands. Although hormones function via the blood stream, it has been suggested that local direct actions may be important,” and hormones such as gastrin occur in gastric juice according to RIA evidence.” Even though not stable in gastric juice, it has been suggested that gastrin does have a physiologic role in

IN SERUM,

SALIVA,

AND GASTRIC

JUICE

317

it.18 Urogastrone does have good in vitro stability to acid and pepsin in terms of its gastric acid inhibitory properties,” and it also appears to be functional in the exocrine secretions as a mitogen. Immunohistochemical studies’~’ showed that positive staining was found in submandibular and Brunner’s glands; in both cases, granules were located in duct cells, and in the latter glands in particular, the secretory ducts themselves gave positive fluorescence. Our present results indicate that exocrine secretion of urogastrone is a normal event. Because of the small amounts of secretions utilized in the present studies, direct effects upon gastric acid secretion could not be sought. However, over many years, reports have described the existence of inhibitors of acid secretion in saliva” and gastric juice.” Although not characterized, these entities have been reported as being of higher molecular weight than urogastrone, but with the present indication that urogastrone occurs in the secretions concerned, it remains possible that it could account for the observed acid inhibitory properties. The actions upon gastric secretion and upon cell proliferation may both play a role in the healing of experimental gastric erosions” when urogastrone is given parenterally. There is no evidence of such action by oral administration as yet, and it remains a matter of speculation whether urogastrone may have a normal functional role in maintaining the integrity of the gastrointestinal tract. Further indications of a physiologic role for this putative hormone may now be obtained by studies upon its release into a variety of exocrine secretions under differing circumstances.

References 1. Gregory

2.

3.

4. 5.

6. 7.

6.

H, Willshire IR: The isolation of the urogastronesinhibitors of gastric acid secretion from human urine. Hoppe Seylers Z Physiol Chem 356:1765-1774,1975 Hollenberg MD, Gregory H: Human urogastrone and mouse epidermal growth factor share a common receptor in cultured human fibroblasts. Life Sci Z&267-274.1977 Culmer CU, Atkinson AJ, Ivy AC: Depression of gastric secretion by the anterior pituitary-like fraction of pregnancy urine. Endocrinology 24:631-637,1939 Savage CR, Jr., Inagami T. Cohen S: The primary structure of epidermal growth factor. J Biol Chem 247:7612-7621,1972 Gregory H: The isolation and structure of urogastrone and its relationship to epidermal growth factor. Nature 257:325-327, 1975 Gospodarowicz D, Moran JS: Growth factors in mammalian cell culture. Ann Rev Biochem 45:531-556,1976 Gillespie IE, Elder JB. Ganguli PC, et al: Effect of urogastrone on gastric secretion and plasma gastrin levels in normal subjects. Gut 16667~893,1975. Elder JB, Williams G, Lacey E, et al: Cellular localisation of human urogastrone/epidermal growth factor. Nature 271:466467.1976

318

GREGORY ET AL.

GASTROENTEROLOGY

9. Heitz Ph II, Kaspar M, van Noorden S, et al: Immunohistochemical localisation of urogastrone to human duodenal and submandibular glands. Gut 19:498-413,1978 10. Gregory H, Bower JM, Willshire IR: Urogastrone and epiderma1 growth factor. In: Growth Factors, 11th FEBS Meeting. Edited by KW Kastrup, JH Nielsen. Oxford, Pergamon Press, 1978 p 75-84 11. Gregory H, Holmes JE, Willshire IR: Urogastrone levels in the urine of normal adult humans. J Clin Endocrinol Metab 45:668-872,1977 12. Rehfeld JF, Stadil F, Vikelsoe J: Immunoreactive ponents in human serum. Gut 15:102-111,1974

gastrin com-

13. Hollenberg MD, Hope DB: The isolation of the native hormone-binding proteins from bovine pituitary posterior lobes. Biochem J 196:557-564,1968

Vol. 77, No. 2

14. Cohen S, Taylor JM: Part I. Epidermal growth factor: chemical and biological characterisation. Ret Prog Horm Res 30:551-572,1974 15. Gregory H, Preston BM: The primary structure of human urogastrone. Int J Pept Protein Res 9:107-118,1977 16. Wingate D: The eupeptide system: A general theory of gastrointestinal hormones. Lancet 1:529-532.1976 17. Uunas-Wallenstein K, Rehfeld JF: Type of gastrin released by vagal stimulation in anaesthetised cats. Gastroenterology 72:825, 1977 18. Fiddian-Green RG: Is peptic ulceration a hormonal disease. Lancet 1:74-78,1977 19. Code CF. Ratke HV, Livermore GR, et al: Occurrence of gastric secretory inhibitory activity in fresh gastric and salivary mucin, Fed Proc 8:26,1949 20. Glass GBJ: Gastrone. Gastroenterology 67:746-742.1974