Proc. NatL Acad. Sci. USAVol. 80, pp. 3948-3952, July 1983Biochemistry

Modulation of proline-rich protein biosynthesis in rat parotidglands by sorghums with high tannin levels

(gene expression /cell-free translation /isoproterenol)

HAILE MEHANSHO, ANN HAGERMAN, SCOTT CLEMENTS, LARRY BUTLER, JOHN ROGLER, ANDDON M. CARLSON*Departments of Biochemistry and Animal Sciences, Purdue University, West Lafayette, Indiana 47907

Communicated by John D. Axtell, March 31, 1983

ABSTRACT Feeding of sorghum with a high level of tannin(high-tannin sorghum) to rats caused changes in gene expressionin parotid glands similar to isoproterenol treatment. Within 3 daysthe parotid glands were enlarged about 3-fold and a series of pro-line-rich proteins were increased about 12-fold. Unlike isopro-terenol treatment, no changes were observed in the submandib-ular glands, and a Mr 220,000 glycoprotein in parotid glands wasnot induced. Amino acid analyses, electrophoretic patterns, andcell-free translations of mRNAs all confirmed that the proline-richproteins induced by feeding high-tannin sorghum were identicalto those induced by isoproterenol treatment. Binding curves forproline-rich proteins to tannins showed affinities 10-fold greaterthan bovine serum albumin and tannins.

A series of unusual proteins containing 25-45% proline havebeen isolated from human and rat salivary glands and the se-cretions of salivary glands (1-6). Two families of either acidicor basic proline-rich proteins (PRPs) are commonly found, withvariations in glycosylation and phosphorylation. The phos-phorylated PRPs are postulated to be involved in dental repair(7) because these substances have a very high affinity for hy-droxyapatite. The functions of the basic PRPs, which have iso-electric points of >10 (8), are unknown.The condensed tannins (proanthocyanidins, oligomers of fla-

van-3-ols) present in many plant tissues characteristically bindand precipitate proteins (9). This astringency of the tannins,which are rich in phenolic groups, apparently protects the planttissue against ingestion by some insects (10) and seed-eatingbirds (11). Seeds of bird-resistant cultivars of sorghum, a majorcereal of the semi-arid tropics, contain a high level of tannin(high-tannin sorghum), which diminishes the nutritional valueof the grain by inhibiting protein digestion (12) and possibly byother mechanisms. In studies designed to define the interac-tion of tannins and proteins, Hagerman and Butler (13) wereable to demonstrate that tannins have an unusually high affinityfor proteins rich in proline. Because the gastrointestinal tract-specifically the oral cavity-is a source of PRPs, Hagerman andButler (13) proposed that salivary PRPs could possibly interactwith tannins and serve to protect dietary proteins and digestiveenzymes.

In this study we report on the dramatic effects on rat parotidglands of feeding high-tannin sorghum. High-tannin sorghumingestion mimics some of the phenotypic changes observed aftertreatment of rats with the a-agonist isoproterenol (6, 8). Within3 days of feeding high-tannin sorghum, the parotid glands aregreatly enlarged and synthesis of PRPs is highly induced. Theseproteins are identical to those accumulating after isoproterenoltreatment. Cell-free translation experiments indicate that mRNAs

coding for the PRPs are greatly increased after feeding high-tannin sorghum, similar to results obtained with parotid glandsfrom isoproterenol-treated rats (14).

MATERIALS AND METHODSAll materials were of highest purity available and were pur-chased from commercial sources unless otherwise indicated.The following substances were purchased from the respectivecompanies: L-[3,4-3H(N)]proline (40-60 Ci/mmol; 1 Ci = 3.7X 1010 Bq) and L-['S]methionine (600-1,300 Ci/mmol),Amersham; DL-isoproterenol HCl and I-anilino-8-naphtha-lenesulfonic acid, Sigma; molecular weight standards and Bio-Gel A-1.5m, Bio-Rad; reticulocyte lysate system, BethesdaResearch Laboratories; Sephadex G-25, Pharmacia; andEN3HANCE, New England Nuclear. Sorghum with high (Sa-vanna), medium (IS-8768), and low (RS-610) levels of tanninweregrown at the Purdue UniversityAgronomy Farm. Tannin analysiswas performed as described by using amodified vanillin assay (15,16). The tannin contents expressed in catechin equivalents are0.16%, 1.2%, and 7.7% for RS-610, IS-8768, and Savanna, re-spectively (16).

Feeding Trials. Sprague-Dawley male rats (70 g) (MurphyBreeding Laboratories, Plainfield, IN) were maintained on PurinaLab Chow for 3-6 days before initiating the feeding trials.Sorghum grains were ground and incorporated into diets of thefollowing composition (% of diet): sorghum, 92.6; corn oil, 2.0;minerals (AIN-76), 3.5; vitamins (AIN-76), 1.0; lysine HCI, 0.75;choline chloride, 0.14; and butylated hydroxytoluene, 0.01. Feedand water were provided ad lib. Isoproterenol was adminis-tered as described by Muenzer et al. (6).

Isolation of PRPs. Rats were anesthetized with sodium pen-tobarbital, weighed, and killed by exsanguination. The parotidand submandibular glands were removed, stripped of connec-tive tissues, and weighed. Extraction of trichloroacetic acid-sol-uble components was carried out as described by Mehanshoand Carlson (17). Proteins in the trichloroacetic acid-solublefraction were concentrated by acetone precipitation and sub-jected to Bio-Gel A-1.5m column chromatography (17). The PRPsthat have an apparent M, of 38,000 (PRP38) were isolated bypreparative gel electrophoresis (18). Proteins were eluted frompolyacrylamide gels essentially as described by Walker et al.(19). Proteins were visualized by using 1-anilino-8-naphtha-lenesulfonic acid as the fluorescent probe (20). Fluorescent bandswere cut into 3- to 4-cm squares and proteins were removed byelectrophoresis.

Proteins were isolated from NaDodSO4 and salts by acetoneprecipitation and by dialysis, respectively (17). Glycine con-taminating the preparation was removed by desalting on a

Abbreviation: PRP, proline-rich protein.* To whom reprint requests should be addressed.

3948

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

Dow

nloa

ded

by g

uest

on

Aug

ust 1

5, 2

021

Proc Natl Acad. Sci. USA 80 (1983) 3949

Sephadex G-25 column, which was equilibrated and eluted with0.05 M acetic acid.

Analytical Procedures. Amino acid analyses were performedas described (6). Proteins were estimated by absorbances of 280nm and 230 nm. PRPs do not contain aromatic amino acids andtherefore after purification these were quantitated by 230-nmabsorbance (Ai*m = 25.0) (17). Polyacrylamide gel electropho-resis was carried out as described by Laemmli (18). Unless oth-erwise indicated, 12% polyacrylamide was used. Proteins pre-cipitated with acetone were dissolved in sample buffer (18) andboiled for 5 min. Fixation and staining of PRPs has been de-scribed (17). Radioactivity on gels was detected by fluorogra-phy (21) with EN3HANCE.

bD

be9_

be

:3GO

a)

0E

04

bO

U,

3 6 9 12Time, days

FIG. 1. Effects of feeding high-tannin sorghum. Rats weighing -77g were fed Savanna (o) or RS-610 (.) sorghums. (A) Total body weightgains; (B) parotid gland weight gains; and (C) changes ofPRPs as mea-sured by increases in 230-nm absorption. Data inA are averages fromfour rats and data in B and C are the sums of gland weight and PRPsfrom four rats, respectively.

Binding affinities of proteins and tannin were assayed by acompetitive binding ass-ay (13). Purified sorghum tannin (22)was added to mixtures of l"I-labeled bovine serum albumin (1mg) and proteins to be tested for relative affinities. After 15min, the precipitate was removed by centrifugation (5,000 Xg, 15 min) and 1.0 ml of the supernatant was assayed for ra-dioactivity.

Isolation of RNA and cell-free translations were performedas described (14, 23) with both ["S]methionine and [3H]pro-line. Phosphotungstic acid (1%) was added to 10% trichloro-acetic acid for precipitation of the PR's.

RESULTS AND DISCUSSIONEffects of High-Tannin Sorghum on Rat Parotid and Sub-

mandibular Glands. The parotid glands of rats maintained onhigh-tannin sorghum showed dramatic increases both in weightand in PRPs (Fig. 1). Both of these parameters were at a max-imum after 3 days of feeding Savanna sorghum. No substantialchanges were observed-for either glandular weight or for PRPswith low-tannin sorghum (RS-610) or with rat chow (not shown).A loss of body weight was also observed during the first 3 daysof feeding high-tannin sorghum. Initiation of weight gain ofanimals on high-tannin sorghum is coincident with maximalstimulation of synthesis of PRPs in parotid glands. Changes ingland weight and in PRPs after 3 days, based on whole bodyweight for rats- fed RS-610 and Savanna sorghums, are shownin Table 1. On a comparative basis, PRPs from-rats fed Savannawere about 12-fold higher than proteins from rats fed RS-610.Slight decreases in both weight and PRPs were observed forsubmandibular glands. Parotid glands and PRPs of rats fed Sa-vanna for 3 wk returned to normal levels within 1 wk after shift-ing the diet to rat chow (data not shown).

Electrophoretic analysis of the trichloroacetic acid-solublecomponents of parotid and submandibular glands of rats main-tained for 21 days on RS-610, IS-8768, and Savanna sorghumsis shown Fig. 2. Both the medium-tannin sorghum (IS-8768)and high-tannin sorghum (Savanna) induced the synthesis ofPRPs. The relative amounts of each protein and the apparentmolecular weights were- identical to the proteins induced byisoproterenol treatment. A glycoprotein of an apparent Mr of200,000 is constitutive in parotid glands, whereas a Mr 220,000glycoprotein is induced by isoproterenol treatment (24). How-ever, this latter glycoprotein is not induced by the high-tannindiet. Isoproterenol treatment also causes a dramatic increase inPRPs and a Mr 158,000 glycoprotein in the submandibular glandsof rats. As noted in Table 1, a high-tannin diet did not inducethe synthesis of theserproteins in rat-submandibular glands (Fig.2).

Major changes in phenotypic expression of PRP synthesisoccur by 3 days (Fig. 3). The family of PRPs increase in concert;as noted previously for analogous experiments with isoproter-enol (17). Small amounts of PRPs are present in parotid glandsfrom animals maintained on rat chow or on RS-610.

Isolation and Amino Acid Analysis-of PBPs. The trichloro&acetic acid-soluble fraction from parotid glands of rats fed Sa-

Table 1. Effects of feeding high-tannin sorghum for 3 days ongland weight and PRPs in parotid and submandibular glands

Gland weight, PRPs,mg/g of body weight mg/g of body weight

Diet Parotid Submandibular Parotid SubmandibularRS 610 4.6 3.1 0.04 0.05Savanna 12.5 2.4 0.49 0.04

Values are averages of four rats.

Biochemistry: Mehansho et aL

Dow

nloa

ded

by g

uest

on

Aug

ust 1

5, 2

021

3950 Biochemistry: Mehansho et al

2 3 4 5 6 7

220_ -200

1-.

* t

2.8

a 2.40

N 2.01:~O 1.6- 1.20

s: 0.8

0.4-38

- 32

40 -40-- Vo* --2724 48 72 96

Fraction

FIG. 2. Analysis of trichloroacetic acid-soluble components by 10%NaDodSO4/polyacrylamide gel electrophoresis. Components ex-

tracted by homogenization in 10% trichloroacetic acid were applied toNaDodSO4/polyacrylamide gels. Protein equivalent to 2.0 mg ofparot-id or 4 mg of submandibular glands was added to each lane. Lanes 1,3, 5, and 7 are extracts from parotid glands; lanes 2, 4, and 6 are extractsfrom submandibular glands. Rats were maintained for 21 days on thediets indicated: lanes 1 and 2, RS-610; lanes 3 and 4, Savanna; lanes 5and 6, IS-8768; and lane 7, isoproterenol-treated for 10 days. Molecularweights are shown as Mr x 10-3.

vanna sorghum for 7 days was chromatographed on Bio-Gel A-1.5m (Fig. 4). Similar to results obtained in the glands fromisoproterenol-treated rats (17), three protein peaks were ob-served. The first two did not contain materials absorbing at 280nm. Peak I is the Mr 200,000 glycoprotein and peak II containsthe family of PRPs (Fig. 2). Peak III consists of small peptidesthat migrate faster than the dye front on NaDodSO4/poly-acrylamide gel electrophoresis. The family of PRPs in peak IIwas separated on a preparative NaDodSO4/polyacrylamide gel.The major slow-moving band of Mr 38,000 (PRP38) was iso-lated. The amino acid analysis of PRP38 was essentially iden-tical to PRP38 isolated from parotid glands and from subman-dibular glands of isoproterenol-treated rats (Table 2).

Binding Affinities of PRPs and Tannins. Proteins high inproline and hydroxyproline have high affinities for tannin (13).Binding curves of PRPs, bovine serum albumin, myoglobin,

o'_DAYS 0 1 1 rlc ) p o to) r- ?- 2I

- 92.5-

- 66.2

- 45 0

-31.0

- 21.5

-- 14.4

FIG. 3. Effects of days of feeding on changes in amounts of PRPs.Parotid glands were extracted and protein equivalent to 1.25mg ofglandswas applied to each lane. R, rats fed RS-610; S, rats fed Savanna; andIPR, rats treated with isoproterenol. Molecular weights are shown as

Mr X 10-3.

FIG. 4. Chromatography of the trichloroacetic acid-soluble frac-tion from parotid glands on Bio-Gel A-1.5m. Trichloroacetic acid-sol-uble components from parotid glands of rats fed high-tannin sorghum(Savanna) were fractionated on a Bio-Gel A-1.5m column (2.5 x 102cm). The column was equilibrated and eluted with 25mM Tris-HCl (pH7.4) containing 0.14 M NaCl. Five-milliliter fractions were collected.

and lysozyme to tannin were determined by using the 1251-la-beled bovine serum albumin assay procedure (Fig. 5). PRPshave about a 10-fold greater affinity for tannin than does bovineserum albumin and have binding affinities similar to polyvi-nylpyrrolidone (13).

Cell-Free Translation of PRP mRNAs. Cell-free translationof mRNAs from rat parotid glands has shown that at least sixdifferent PRPs are translated with poly(A)+ mRNA from theglands of animals treated with isoproterenol (14). Similar re-

sults were obtained by cell-free translation studies of parotidglands from rats fed Savanna sorghum (Fig. 6). Each of the PRPstranslated is likely to be a product of a separate, specific mRNAbecause the patterns of labeling with [3S]methionine and [3H]-proline are identical and [35S]methionine is found only as theNH2-terminal amino acid. Methionine is not present in the iso-lated PRPs or in proteins synthesized by tissue slices (14). Thesedata are similar to those reported earlier (14) and suggest thatpost-translational processing of a large polypeptide chain to givethe various PRPs is unlikely.A common characteristic of proteins and polypeptides with

high affinity for tannin is their high proline content (13). Calf-skin gelatin (18 mol % proline and hydroxyproline) and rat par-otid PRPs (43 mol % proline) were the most effective at bindingtannin of all naturally occurring polymers tested (8, 13). Gas-trointestinal mucins that form the protective mucous barrier ofthe gut are also relatively high in proline (12-15 mol %) (25).The affinity of tannins for these glycoproteins could have a del-eterious effect.The decreased growth rate of animals fed high-tannin feeds

Table 2. Amino acid compositions of PRP38 from parotid glandsSavanna, Isoproterenol,

Amino acid mol/100 mol mol/100 mol*

Aspartic acid 2.5 2.9Serine 1.4 1.6Glutamic acid 24.0 23.9Proline 41.3 41.8Glycine 24.2 23.2Leucine 1.3 1.3Histidine 0.2 0Lysine 0.6 1.0Arginine 4.5 4.7

*Ref. 17.

Proc. Natl. Acad, Sci. USA 80 (1983)

Dow

nloa

ded

by g

uest

on

Aug

ust 1

5, 2

021

Proc. Natl. Acad. Sci. USA 80 (1983) 3951

cd

790

- 1 0 10 50100 00

~~~~~~~~~Lysozyme-~60 0

650

FG40 Bovie Myoglobinoieserum

30 albumin

020

"~10

of binding to tannin. Inhibition of binding was calculated as described(13).

may reflect a diminished digestibility of proteins (26-28) anda possible inhibition of digestive enzymes (29). Eggum andChristensen (30) reported a severe negative effect on proteindigestibility when 1.5% tannin was included in the diet of rats.The availability of proline, glycine, and glutamic acid was re-ported to be most severely decreased, as determined by an aminoacid balance study of feed and fecal analyses. However, theinterpretation of these results may be questioned when the ex-perimental animals secrete copious amounts of salivary pro-

l 2 3 4 5 6 7 8

40.x

58

-38

4;j -32

_ _ - 2 27

FIG. 6. Cell-free translations of mRNA. RNA was prepared fromthe parotid glands of rats fed Purina Lab Chow (lanes 1 and 5), RS-610(lanes 2 and 6), and Savanna (lanes 3 and 7) for 3 days. Isoproterenoltreatment was for 10 days (lanes 4 and 8). Cell-free translations wereperformed with rabbit reticulocyte preparations as described (14). Lanes1-4, translation products labeled with [35S]methionine, and lanes 5-8,translation products labeled with [3H]proline. Molecular weights areshown as Mr x 10-3.

teins that contain proline (42%), glycine (24%), and glutamate(24%), as determined by studies reported here. Likely, thesePRPs bind to tannins and are substantially less digestible thanother proteins and therefore are present in large amounts in thefeces. PRPs are present in high concentrations in human saliva(3). The amino acid sequences for both the acidic (31) and basic(32, 33) human PRPs have been partially established.The primary effector regulating the hypertrophic response

and changes in protein synthesis in parotid glands of rats fedhigh-tannin sorghum is not known. The overall responses closelyresemble the effects of isoproterenol treatment with at leasttwo exceptions. The submandibular glands do not respond andthe Mr 220,000 glycoprotein is not induced in the parotid glandsby the high-tannin diet. If dietary tannins induce catechol-amine synthesis and release, these differences in responses mayresult from a relatively constant production but a lower amountof a 3agonist. Preliminary data indicate that the /-antagonistpropranolol will inhibit the hypertrophic and protein re-sponses. If the increased production of PRPs by parotid glandsis beneficial to the rat, this would be a genetic response to atoxic substance.

This research was supported in part by Public Health Service GrantAM19175 (D.M.C.), by U.S. Agency for International DevelopmentIntsormil Project PRF-4 (L.B.), and by National Science FoundationFellowship 7922349 (A.H.). This is journal paper no. 9398 from thePurdue University Agricultural Experiment Station.

1. Oppenheim, F. G., Hay, D. I. & Franzblau, C. (1971) Biochem-istry 10, 4233-4238.

2. Bennick, A. & Connell, G. E. (1971) Biochem. J. 123, 455-464.3. Fernandez-Sorensen, A. & Carlson, D. M. (1974) Biochem. Bio-

phys. Res. Commun. 60, 249-256.4. Kauffman, D. L. & Keller, P. J. (1979) Arch. Oral Biol. 24, 249-

256.5. Mehansho, H., Clements, S. & Carlson, D. M. (1982) Fed. Proc.

Fed. Am. Soc. Exp. Biol 41, 681 (abstr.).6. Muenzer, J., Bildstein, C., Gleason, M. & Carlson, D. M. (1979)

J. Biol Chem. 254, 5623-5628.7. Bennick, A. (1982) Mol. Cell. Biochem. 45, 83-99.8. Muenzer, J., Bildstein, C., Gleason, M. & Carlson, D. M. (1979)

J. Biol Chem. 254, 5629-5634.9. Haslam, E. (1979) Recent Adv. Phytochem. 12, 475-523.

10. Schultz, J. C. & Baldwin, I. T. (1982) Science 217, 149-151.11. McMillan, W. W., Wiseman, B. R., Burns, R. E., Harris, H. B.

& Greene, G. L. (1972) Agron. J. 64, 821-822.12. Jambunatham, R. & Mertz, E. T. (1973)J. Agric. Food Chem. 21,

692-696.13. Hagerman, A. E. & Butler, L. G. (1981)J. Biol. Chem. 256, 4494-

4497.14. Ziemer, M. A., Mason, A. & Carlson, D. M. (1982)J. Biol. Chem.

257, 11176-11180.15. Price, M. L. & Butler, L. G. (1977)J. Agric. Food Chem. 25, 1268-

1273.16. Price, M. L., Van Scoyoc, S. & Butler, L. G. (1978)J. Agric. Food

Chem. 26, 1214-1218.17. Mehansho, H. & Carlson, D. M. (1983)J. Biol. Chem., in press.18. Laemmli, U. K. (1970) Nature (London) 227, 680-685.19. Walker, J. E., Auffret, A. D., Cane, A., Gurnett, A., Hanisch,

P., Hill, D. & Saraste, M. (1982) Eur. J. Biochem. 123, 253-260.20. Hartman, B. K. & Udenfriend, S. (1969) Anal. Biochem. 30, 391-

394.21. Laskey, R. A. & Mills, A. D. (1975) Eur.J. Biochem. 56, 335-341.22. Hagerman, A. E. & Butler, L. G. (1980)J. Agric. Food Chem. 28,

947-952.23. Harding, J. D., Przybyla, A. E., MacDonald, R. J., Pictet, R. L.

& Rutter, W. J. (1978) J. Biol Chem. 253, 7531-7537.24. Humphreys-Beher, M. G., Immel, M. & Carlson, D. M. (1982)

Fed. Proc. Fed. Am. Soc. Exp. Biol. 41, 665 (abstr.).25. Jabbal, I., Kells, D. I. C., Forstner, G. G. & Forstner, J. F. (1976)

Can. J. Biochem. 54, 707-716.26. Rostagno, H. S., Rogler, J. C. & Featherston, W. R. (1973) Poult.

Sci. 52, 772-778.

Biochemistry: Mehansho et aL

Dow

nloa

ded

by g

uest

on

Aug

ust 1

5, 2

021

3952 Biochemistry: Mehansho et al.

27. Featherston, W. R. & Rogler, J. C. (1975) Nutr. Rep. Int. 11, 491-497.

28. Cousins, B. W., Tanksley, T. D., Jr., Knabe, D. A. & Zebrowska,T. (1981) J. Anim. Sci. 53, 1524-1537.

29. Milic, B. L., Stojanovic, S. & Vucarevic, N. (1972)J. Sci. Food Agric.23, 1157-1162.

30. Eggum, B. 0. & Christensen, K. D. (1975) in Proceedings of theSecond Research Co-ordination Meeting of the Seed Protein Im-

Proc. Nat. Acad. Sci. USA 80 (1983)

provement Program (Int. Atomic Energy Agency, Vienna, Aus-tria), pp. 135-143.

31. Wong, R. S. C. & Bennick, A. (1980) J. Biol. Chem. 255, 5943-5948.

32. Isemuza, S., Saitoh, E. & Sanada, K. (1982)J. Biochem. (Tokyo)91, 2067-2075.

33. Kauffman, D., Wong, R., Bennick, A. & Keller, P. (1982) Bio-chemistry 21, 6558-6562.

Dow

nloa

ded

by g

uest

on

Aug

ust 1

5, 2

021