6 male mice

6 male mice

Mechanisms o f Ageing and Development, 34 ( 1986) 175-189 175 Elsevier Scientific Publishers Ireland Ltd. EFFECT OF ADVANCED AGE ON THE INDUCTION B...

2MB Sizes 0 Downloads 23 Views

Mechanisms o f Ageing and Development, 34 ( 1986) 175-189

175

Elsevier Scientific Publishers Ireland Ltd.

EFFECT OF ADVANCED AGE ON THE INDUCTION BY ANDROGEN OR THYROID HORMONE OF EPIDERMAL GROWTH FACTOR AND EPIDERMAL GROWTH FACTOR mRNA IN THE SUBMANDIBULAR GLANDS OF C57BL/6 MALE MICE

EDWARD W. GRESIK, KAREN WENK-SALAMONE, ANDREA ONETTI-MUDA, RUTH M. GUBITS and PHYLLIS A. SHAW Department o f Anatomy, Mount Sinai School o f Medicine o f the City University o f New York. New York, N Y 10029 (U.S.A.]

iReceived September 17th, 1985) (Revision received December 23rd, 1985) SUMMARY We have compared the responsiveness of the submandibular glands of mature (12 month old) and senescent ( 2 6 - 2 8 month old) male C57BL/6 mice to dihydrotestosterone (DHT) or triiodothyronine (T3) in terms of steady state levels of epidermal growth factor (EGF) protein and EGF mRNA. Northern blot analyses did not disclose any differences with age in the apparent sizes of EGF mRNA species. In untreated animals, submandibular glands of 26-2g-month-old mice contained approximately 50% less EGF, and 75% less EGF mRNA than those of 12-month-old males. With advanced age, there was a 20% reduction in the absolute volume of the granular convoluted tubule (GCT) compartment, which is the exclusive site of EGF and EGF mRNA in the gland. In general, GCTs of old mice were composed of smaller cells with fewer secretion granules, but there was considerable cell-to-cell variation. In addition, there was greater variation in the intensity of immunocytochemical staining for EGF in senescent GCT cells, which also gave a lower and more variable in situ hybridization signal for EGF mRNA. After hormonal stimulation for l week with either tri-iodothyronine (T3) or dihydrotestosterone (DHT), EGF protein concentration in the glands was induced to the same level at both ages. However, EGF mRNA was 50% less abundant in old hormonally stimulated glands, compared to similarly treated young ones. Although many GCT cells in treated glands of senescent males respond to hormonal stimulation by increases in size and in content of secretion granules, there was cell-to-cell variation in responsiveness, especially after treatment with T3. These findings indicate that the decreases seen in the entire gland in EGF and EGF mRNA are caused by a wide-spread deterioration of the

Address all correspondence to: E.W. Gresik, PhD, Department of Anatomy, Mt Sinai School of Medicine, 5th Ave and 100th St., New York, NY 10029, U.S.A.

0047-6374/86/$03.50 Printed and Published in Ireland

© 1986 Elsevier Scientific Publishers Ireland Ltd.

176 GCT cells themselves, which apparently can be reversed in many but not all GCT cells by stimulation with supraphysiologic doses of either T3 or DHT.

Key words: Epidermal growth factor; Epidermal growth factor mRNA; Aging; Submandibular gland; Dihydrotestosterone ; Triiodothyronine INTRODUCTION Epidermal growth factor (EGF) is a biologically potent polypeptide, whose richest known source is the granular convoluted tubule (GCT) cell in the submandibular gland of the mouse [1]. These cells, and the content of EGF [2-5] and EGF mRNA [6-8] within the gland, are under the multihormonal control of testicular androgens and thyroid hormones. In senescent male C57BL/6 mice, the relative volume of the GCT compartment is reduced to about one-third of its value at 12 months of age [9]. Moreover, individual GCT cells in old mice are smaller and have fewer secretion granules, and the concentration of submandibular EGF falls to 17% of its peak values seen at 1 year [10]. Since serum levels of gonadal and thyroid hormones have not consistently been tound to be decreased in senescent C57 male mice [11-13], it is not likely that the above changes are due to inadequate circulating hormone levels. Moreover, the cytologic changes seen in senescent GCT cells are different from those seen in castrated males or hypothyroid mice [ 14,15]. We postulate, therefore, that the defect lies in these senescent GCT cells themselves, and in the present studies have tested the ability of these cells to respond to dihydrotestosterone or to triiodothyronine. MATERIALS AND METHODS Mature (12 months of age) and senescent ( 2 6 - 2 8 months of age) C57BL/6 male mice were received from Charles River Breeding Laboratories, Inc. (Stone River, NY) from a colony maintained under contract to the National Institute on Aging. Mice of each age were divided into three groups: control mice were untreated; androgen-treated mice were given two injections of 500/ag dihydrotestosterone (DHT) in sesame oil 4 days apart, and killed on the 8th day after the first injection; thyroid hormone*treated mice were injected with 0.25 tig/g body wt triiodothyronine (T3) in 0.005 N NaOH in saline for 7 consecutive days, and killed on the 8th day. All injections were subcutaneous. All mice were weighed, then killed by cervical dislocation, and exsanguinated by thoracic transection. Submandibular glands were freed of surrounding tissue, including the sublingual gland, and weighed. Pieces of the glands were taken from each mouse for the following procedures: (1) fixation in Bouin's solution, embedding in paraffin, for staining by either hematoxylin and eosin (H & E) or by the immunocytochemical pro-

177 cedure for EGF [16];(2) fixation in 4% paraformaldehyde/2% glutaraldehyde (0.1 M phosphate buffer, pH 7,4); postfixation in 1% OSO4, and embedding in Epon, for 1 /am sections stained with toluidine blue; (3) quenching in isopentane chilled with liquid N2, for in situ hybridization of EGF mRNA [17] ; (4) quenching directly in liquid N2 for isolation of RNA [18] ; and (5) freezing at -20°C for preparation of 5% homogenates for determination of protein [19] and EGF content by radioimmunoassay [20]. Tissues for RNA extraction were pooled for a given experimental group. Total RNA extracts were analyzed by Northern blotting [21] and dot-blotting [22]. Differences in intensity of signal in the autoradiograms were estimated visually. For Northern blots, RNA was electrophoresed in a 1.5% agarose, 6% formaldehyde gel and transferred to nitrocellulose paper (Schleicher and SchueU) [23]. Dot-blots were prepared on nitrocellulose paper in a Schleicher and Schuell dot-blotting apparatus. Both blots were baked for 2 h at 80°C under vacuum. Blots were hybridized either with (1) a a2P-labelled, nick-translated cDNA of the 754 bp Sma I/Pvu II restriction fragment of the recombinant plasmid pmegfl0 [6], containing the coding region for EGF, or (2) a 32P-labelled riboprobe representing the same fragment. In situ hybridizations were performed with the same 754 bp fragment labelled with aH by nick-translation, and more extensively nicked with pancreatic DNase I to yield a probe with an average size of 50-100 bp [17]. Morphometry Sections (5 /am) of paraffin-embedded, Bouin's-fixed submandibular glands, stained with H & E, were analyzed by point counting to determine the relative volume of the GCT compartment (VvGCT) in each gland. Using a IOX ocular lens fitted with a square reticle of 100 regularly spaced points, 10 fields were scored for points over GCTs at 400X magnification. The VvGET was multiplied by the weight of the gland (mg), which was taken as a close estimate of the glandular volume (mma), to compute the absolute volume of the GCT compartment (VGCT). All quantitative data were analyzed by a two factor analysis of variance (age X hormone), and individual pairs of data were further tested by Student's t-test for significance of difference of means [24].

RESULTS All quantitative data are presented in Table I. In general, body weight did not vary with age, or after hormonal treatment. The slight decline in the absolute weight of the submandibular gland with age was not significant. At both ages, Ta or DHT treatment resulted in an increase in absolute submandibular weight (compared to same age controis), that was significant only for the young DHT-injected group. Furthermore, after hormonal treatment, glands of 12-month-old males weighed more than those of 26-28 month olds; the difference was significant after DHT treatment. However, the weight of

49.6 56.1 74.2 655.3

_+ 34.3 + 6.3 _+ 17.5 _+432.6

1.50

_+ 36.1

2.38 _+

44.4

1.68

_+ 7.5 _+ 26.5 _+105.6

15.86_+

32.2 132.6 425.6

97.9 68.3 104.9 904.8

+ 69.3 d _+ 3.3 a _+ 24.5 a -+521.8

~

1.94 h

_+ 38.8

3.75 _+

68.8

1.60 d

+ 6.2 _+ 35.6 _+ 68.3

17.93 +_

34.8 154.1 436.3

*Means _+ standard deviations. ap < 0.001, compared to 12-month-old + DHT. b p < 0.02, compared to 12-month-old controls. < 0.03, compared to 2 6 - 2 8 - m o n t h - o l d controls. P < 0.005, compared to 12-month-old controls. ep < 0.01, compared to 12-month-old controls.

protein) pgEGF/gland VvGCT(%) V G C T ( m m 3) ngEGF/mm 3 of GCT

(ug/mg

Body w t ( g ) Glandwt(mg) Relative gland wt (mg/100 g BW) Protein (mg/100 mg tissue) EGF (~8/100 mg tissue) EGF

+ 2.5 +_ 11.5 b + 34.9

_+ 60.2 d -+ 4.5 d'e _+ 7.1 a +619.9 e

20.2 49.9 59.2 338.2

_+ 16.3 e _+ 7.5 b -+ 10.8 e +-246.7 b

0.76 e

_+ 23.0 h

1.15 _+

21.6

1.82 e

_+ 2.0 _+ 16.7 _+ 52.2

14.13_+

30.8 119.0 386.2

91.4 60.0 78.1 1143.1

_+ 32.3 g _+ 6.0 j'k -+ 21.2 n ' ° _+459.2 g

1.44 g

_+ 28.0 g

4.15 _+

71.2

1.46 f

_+ 6.6 + 25.2 _+ 87.5

16.83 _+

32.3 128.3 395.6

T3 (n = 9)

l p < 0.002, compared to 2 6 - 2 8 - m o n t h - o l d controls. gP < 0.001, compared to 2 6 - 2 8 - m o n t h - o l d controls. h p < 0.05, compared to 12-month-old controls. ip. < 0.025, compared to 12-month-old controls. P < 0.001, compared to 12-month-old + T 3. P < 0.005, compared to 2 6 - 2 8 - m o n t h - o l d controls.

116.4 63.8 98.5 1189.1

2.02 i

_+ 39.4 h

4.23 _+

76.2

17.66 _+ 1.15 d

34.6 154.9 450.8

Con (n = 15)

DHT (n = 11)

Con (n = 18)

T3 (n = 13)

26-28 month

12 m o n t h

84.5 _+ 40.3 g 51.8 _+ 6.7 I'm 66.9 + 11.81 1232.6 -+486.3 g

1.34 g

_+ 26.4 g

3.75 _+

64.0

1.56 g

+ 3.1 a _+ 15.7 a _+ 68.3 c

t6.77 +

29.4 129.3 442.1

DHT (n = 11)

EFFECTS OF T R E A T M E N T WITH T 3 OR D H T ON S U B M A N D I B U L A R G L A N D S OF Y O U N G (12 MONTH OLD) A N D SENESCENT ( 2 6 - 2 8 MONTH OLD) C57BL/6 MALE MICE*

TABLE I

O0

179 the submandibular gland, relative to body weight, was not different among the various groups, except for a significant increase following DHT treatment of the old mice. The concentration of total protein was slightly, but significantly, less in senescent submandibular glands, compared to young control mice. Moreover, although treatment with either hormone caused a significant increase in protein concentration in the glands of both young and old mice, no interactive effects of hormone and age were disclosed by analysis of variance. Both the concentration of EGF (per 100 mg tissue, or per mg protein) and the total amount of EGF in the glands of old untreated males were 50% less than in the glands of young mice. In 12-month-old animals the EGF concentration was increased by 60% after T 3 treatment, and by 80% after DHT treatment. By contrast, in senescent mice EGF levels were increased by more than 200% after treatment with either hormone. Analysis of variance showed a significant effect of either hormone on EGF levels, but no interactive effects of hormone and age. Northern blot analysis revealed that the size of the major and minor species of EGF mRNA did not vary either with age or hormonal treatment (Fig. 1A). Furthermore, the abundance of EGF mRNA declined with age in the untreated mice, but after treatment with either hormone the concentration of EGF mRNA was increased, more so after DHT than after Ta. Dot blot analyses of total RNA extracted from the glands of untreated mice revealed a 75% reduction in the abundance of EGF mRNA with age (Fig. 1B). However, there was approximately 4-fold induction of message by T3 and approximately 8-fold induction by DHT, compared to same age controls. Thus, relative induction ofEGF mRNA by either hormone did not vary with age. In 12-month-old males GCTs were formed of tall columnar cells filled with uniform apical secretion granules (Fig. 2). These ceils were intensely stained immunocytochemically for EGF (Fig. 3) and gave a strong signal for EGF mRNA by in situ hybridization (Fig. 4). By contrast, GCTs of senescent males were composed of smaller cells with variable amounts of secretion granules of diverse size (Fig. 5). Many ceils had appreciable amounts of lipofuscin granules. There was cell-to-ceil variation in the intensity of immunocytochemical staining for EGF (Fig. 6), and of the in situ hybridization signal for EGF mRNA, which was much weaker than that seen in the 12-month-old mice (Fig. 7). GCT cells of 12*m6nth-old males treated with Ta were composed of uniformly large cells filled with secretion granules of highly variable size (Fig. 8). The presence of some exceptionally large granules was characteristic of this group. Immunocytochemically, GCT cells of young T3-treated males were uniformly strongly stained (Fig. 9). GCT ceils of senescent males, by contrast, did not demonstrate a consistent response to Ta. Some tubules were formed of large ceils packed with secretion granules, whereas other GCTs contained smaller cells with few or no secretion granules (Fig. 10). Cell-to-cell variation in immunocytochemical staining for EGF was also seen in these GCTs (Fig. 11). After androgen treatment of 12-month-old males, large GCT cells were filled with secretion granules of homogeneous size (Fig. 12), and stained strongly for EGF (Fig. 13).

180

a

b

c

d

e

f

g

A S

S

a

B

b

c

d

e

f

Q

i

g

181 In senescent males given DHT, GCT cells were larger and contained more secretion granules than old control mice. However, some GCT ceils contained only small numbers of minute secretion granules (Fig. 14). Immunocytochemical staining for EGF was strong in the GCTs of androgen-treated old mice (Fig. 15). Thus, although most GCT cells of senescent males were strongly induced by T3 or DHT, some cells seemed to show varying degrees of blunted responsiveness to these two hormones. The relative volume of the gland occupied by the GCT compartment (VvGCT) was 10% less in the senescent mice, compared to young ones (Table I). Treatment with T3 resulted in a 20% increase in the VvGCT in both young and old mice. After DHT treatment, there was a 15% increase in the VvGCT in 12-month-old males, but no change over control values in 2 6 - 2 8 m o n t h old mice. By contrast, the absolute volume of the GCTs (VGCT) was 20% less in senescent males, compared to 12-month-old mice. At 12 months of age either hormone caused a 3 5 - 4 0 % increase in the amount of GCTs, whereas at 2 6 - 2 8 months of age the increase was slightly, but significantly, less after T3 (30%), and considerably less after DHT (12%) (Table I). There was 50% less EGF per unit volume of GCT at 2 6 - 2 8 months, compared to 12 months of age, but either hormone raised the concentration of EGF in the GCT compartment to the same level at both ages (Table I). At 12 months of age the increase over control levels was 38% after Ts, and 82% after DHT. In senescent males either hormone raised the tubular concentration of EGF about 250%.

Fig. 1. Comparison of EGF mRNA levels in total RNA extracts from the submandibular glands of young and senescent C57BL/6 male mice with and without treatment with Ta or DHT. A. Northern blot hybridization: Electrophoresis of total RNA extracts was carried out in 1.5% agarose, 6% formaldehyde gels [21]. RNA samples were heated to 65°C for 5 min and then cooled on ice before being loaded onto the gel. The RNA was then blotted onto nitrocellulose paper [23], and hybridized overnight at 65°C with a s2P-labelledriboprobe corresponding to the EGF-coding region of pmegfl0 [6], according to the conditions of Wahl et al. [31 ], with the addition of 0.4 mg/ml yeast total RNA. The blot was stringently washed (0.1 × SSC, 0.1% SDS, 65°C, 15 min), and exposed overnight at -80°C to Kodak XAR-5 ~ m with an intensifying screen. 1 × SSC = 0.15 M NaCI, 0.015 M Na-citrate. Lanes a to leach contained 5 ~g of RNA: a, 12 month control; b, 12 month treated with Ts;c, 12 month treated with DHT; d, 28 month control; e, 28 month treated with T~;.f, 28 month treated with DHT. Lane g contained 1.25 t~g of total RNA extracted from the submandibular glands of 3-month-old male mice, used as a standard. B. Dot blot hybridization: Samples of total submandibular RNA were serially diluted with water, and yeast total RNA was added to make the final amount of RNA the same in each tube. Formaldehyde and SSC were added to a t-real concentration of 14% and 7.5×, respectively. RNA samples were heated at 65°C for 5 min, cooled on ice, and applied to nitrocellulose paper in a Sehleieher and Schuell dot-blotting apparatus. The dried dot blot was hybridized overnight at 42°C to a nick-translated s2P-cDNA probe corresponding to the EGF-eoding region of pmegfl0 [6], according to Gubits et al., [22]. The dot-blot was washed and exposed to X-ray f'am as above. The first dot in each dilution series contained 5 tag of submandibular RNA, except for the first dots in columns d and g, which contained 10 t~g RNA. a, 12 month control; b, 12 month treated with Ts; c, 12 month treated with DHT; d, 28 month control; e, 28 month treated with T~; f, 28 month treated with DHT; g, total liver RNA from an untreated 3-month-old female mouse, showing a lack of non-specific hybridization.

6

L

!

Fig. 2. Twelve month control. GCTs (G) axe composed of tall columnar cells filled with dense apical secretion granules. Aeini (A). Epon section, toluidine blue, 630X. Fig. 3. Twelve month controL Immtmoeytoehemical straining for EGF shows strong and uniform reactivity eontrmed to cells of the GCTs (G), while aeini (A) and other glandular components are umeactive. Paxaffm section, 250×. Fig. 4. Twelve month control In situ hybridization signal for EGF mRNA is intense over GCT cells (G), and absent over acirfi (A). Cryostat section, autoradiograph, toluidine blue eounterstain, 630X. Fig. 5. Twenty-eight month control. GCT cells (G) axe smaller and contain fewer secretion granules than in young mice. Some cells have few or no granules (arrows). Acinar cells (A) frequently have large vacuoles. Epon section, toluidine blue, 630×. Fig. 6. Twenty-eight month control Smaller GCT cells show variable intensity of immunoeytoehernical staining for EGF, with some cells essentially unreactive (arrows). Pamff'm section, 250X. Fig. 7. Twenty-eight month control. A weak to absent in situ hybridization signal is seen over the smaller GCT cells (G), with acini (A) free of signal Cryostat section, automdiograph, toluidine blue eounterstain, 630×.

Figs. 2-15. All photomicrographs of submandibular glands of C57BL/6 male mice.

toluidine blue, 630×. Fig. 9. Twelve month old, treated with T 3. GCT cells are strongly and uniformly immunoreactive for EGF. Paraff'm section, 250X. Fig. 10. Twenty-eight month old, treated with T 3. Some GCTs are formed of large columnar cells with abundant granules (upper left), whereas others contain smallei ceils with fewer oI apparently no (arrows) g~anules. Epon section, toluidine blue, 630X. Fig. 11. "Iwenty-eight month old, treated with T~. GCTs vary in size and in staining intensity for EGF. Parafffm section, 250X.

Fig. 8. Twelve month old treated with T~. Uniformly large GCT cells are/tiled with secretion granules of varying sizes, some being quite large. Epon section,

I1

]o

Fig. 12. Twelve m o n t h old, treated with DHT. Large cells are uniformly packed with secretion granules. Epon section, toluldine blue, 630×. Fig. 13. Twelve m o n t h old, treated with DHT. GCT cells give a strong and uniform immunoreaction for EGF. Paraft-m section, 250X. Fig. 14. Twenty-eight m o n t h old, treated with DHT. Most GCT cells are enlarged and filled with granules, but a few (lower central tubule) contain only a !ew small granules. Epon section, toluidine blue, 630×. Fig. 15. Twenty-eight m o n t h old, treated with DHT. In general GCT cells are intensely immunoreactive for EGF. Paraffin section, 250×.

186 DISCUSSION These studies confirm our earlier findings [9,10] that both submandibular EGF levels and the relative volume of the GCT compartment are reduced in senescent C57 male mice. Furthermore, we have now shown that this decline in EGF content is accompanied by a substantial decrease in EGF mRNA, and is thus likely to be due primarily to disturbances at pretranslational levels in GCT cells. The quantitative differences in EGF concentration and relative GCT volume seen in the present study and those previously reported are probably due to the fact that the two sets ofC57 mice were obtained from colonies maintained by two different breeders. Moreover, the fall with senescence in the content of EGF per unit volume of the GCT compartment firmly establishes that the decrease in the glandular concentration of EGF is due to a defect in the GCT cells rather than merely to a relative increase in the acinar compartment in the glands of old animals. This is further supported by the cytologic findings of smaller GCT cells with fewer secretion granules, by their less intense and more variable immunocytochemical staining for EGF, and by the weaker in situ hybridization signal for EGF mRNA. Although these findings conclusively demonstrate an age-related defect in GCT cells themselves, they also show that there is considerable cell-to-cell variation in the extent of this impairment. Thus, the decline seen at the level of the whole gland represents an average measure of the wide variations seen among individual GCT cells. Hence, while it is tempting to conclude that there is a general decline in senescent GCT ceils, we observed that some GCT cells of otd mice are apparently not different from those generally seen in the glands of young mice~ whereas others show signs of considerable impairment. This ceU-to-cell variation must be kept in mind in suggesting a possible general cellular mechanism to explain the age-related decline seen at the level of the whole organ. Furthermore, the decline in function seen in the whole senescent gland represents an average of the varying functional capabilities of individual GCT cells. This marked intercellular variation renders any elementary stereological analysis meaningless (e.g., average volume of a senescent GCT cell). In 12-month-old males both T3 and DHT exerted a clear inductive response on the concentration of total protein, EGF, and EGF mRNA in the gland. In GCT ceils there was a uniform increase in the content of secretion granules and uniform intensity of immunocytochemical staining for EGF. Either hormone caused a significant increase in the total glandular content of EGF, as well as in the concentration of EGF in the GCT compartment. These findings differ from those of Walker et al. [25], who reported no inductive effects of Ta on EGF levels in the submandibular glands of Swiss-Webster male mice. Moreover, the magnitude of the inductive effect of these hormones is less than that seen in the glands of similarly treated female mice [4,26]. At 28 months of age either hormone increased total protein and EGF content in the gland to the same level as seen in hormonally treated 12-month-old males. By contrast, EGF mRNA levels were induced to only about half their concentrations in young treated

187 males, although the relative induction over the age-matched controls was the same at 12 and 2 6 - 2 8 months of age. Although the relative volume of the GCT compartment increased to the same level in old mice as in young ones after hormone treatment, the absolute volume of GCTs in old treated mice was slightly but significantly less than in young mice. However, the concentration of EGF within the GCTs was raised to the same levels as in young treated mice. In contrast to young glands, there was some intercellular variation in the responsiveness of GCT cells in senescent glands, especially to T 3 administration. The reason for this refractoriness to hormonal induction by some GCT cells is not known, but conceivably could be due to variation among cells in their content of hormone receptors. These findings indicate that, although the majority of GCT ceils - including those showing some signs of impairment - is quite responsive to T3 of DHT, some cells have passed a critical threshold of deterioration, and are no longer inducible by these hormones° Although not entirely consistent, the literature generally describes a decline in total RNA content with advanced age in a variety of rodent tissues [see reviews 2 7 - 2 9 ] . Moreover, decreases in the rates of transcription have been reported for total RNA, hnRNA, poly(A)÷RNA, and tRNA [same reviews]. The present report is among the few to describe an age-related decline in a specific mRNA species in both untreated and hormonally stimulated mice. One factor leading to reduced abundance of EGF mRNA in senescent glands may well be a decrease in transcription in this message. The reduced in situ hybridization signal over senescent GCT cells supports this interpretation. Moreover, in an independent study, we have found that the total number of GCT cells per gland decreased with age (20.5 +- 2.5 X 106 ceils at 12 months, n = 8, vs. 14.3 +- 2.9 × 106 at 28 months, n = 9, P < 0.001) (Gresik, unpublished). Thus, the smaller populations of GCT cells in the glands of old males may also contribute to the reduction in EGF mRNA content at advanced ages. Whether the proposed transcriptional defect is due to a decrease in the rate or amount of EGF mRNA synthesis remains to be investigated. The lack of close correspondence in the extent of the decline in the steady state levels of EGF mRNA and EGF protein seen in this study was unexpected. Thus, in all three senescent groups the levels of EGF mRNA were lower than in the corresponding young groups. However, EGF protein levels were reduced in untreated old animals, but did not differ with age in hormonally stimulated ones. Related findings on senescent animals have been reported for mouse liver and other tissues, where there were no age-related changes in content of specific proteins, in spite of decreases in the rates of synthesis of protein and or poly(A)÷RNA [see reviews 2 8 - 3 0 ] . Maintenance of adequate levels of many specific proteins in senescent tissues has been associated with decreased turnover times of both proteins and of poly(A)÷RNA, the latter apparently due to increased stabilization [29]. w.~ su,iamary, these studies have shed some light on the cause of the senescent decline in EGF levels in the whole organ, namely a wide-spread deterioration in the component GCT cells themselves, which may be primarily pretranslational, since steady state levels

188 of E G F m R N A are reduced. F u r t h e r m o r e , they indicate that although glandular E G F levels are induced by T 3 o f DHT to the same extent in the submandibular glands of y o u n g and senescent mice, there is variable responsiveness by individual GCT cells with senescence. ACKNOWLEDGEMENTS We thank Dr Tibor Barka for his helpful discussion t h r o u g h o u t the course o f this work, and for his critical reading o f the manuscript. We thank Dr. William R u t t e r for the initial supply o f p m e g f l 0 . This w o r k was supported by grants AG-3508 f r o m N I A and AM-19753 f r o m NIH. REFERENCES 1 G. Carpenter and S. Cohen, Epidermal growth factor. Annu. Rev. Biochem., 48 (1979) 193-216. 2 P. Barthe, L. Bullock, I. Mowszowicz, C . Bardin and D. Orth, Submaxillary gland epidermal growth factor: a sensitive index of biologic androgen activity. Endocrinology, 95 (1974) 10191025. 3 K. Hosoi, S. Kobayashi, T. Ueha, S. Sato, T. Takuma and M. Kumegawa, Induction of androgendependent protease and serous-like granules by tri-iodothyronine in the submandibular gland of mice with testicular feminization. J. Endocrinol., 83 (1979) 429-434. 4 E. Gresik, I. Schenkein, H. van der Noen and T. Barka, Hormonal regulation of epidermal growth factor and protease in the submandibular gland of the adult mouse. Endocrinology, 109 (1981) 924 -929. 5 C.M. Wilson, J. Griffin, R. Reynolds and J.D. Wilson, The interaction of androgen and thyroid hormones in submandibular gland of the genetically hypothyroid (hyt/hyt) mouse. Endocrinology, 116 (1985) 2568-2577. 6 J. Scott, M. Urdea, M. Quiroga, R. Sanchez-Pescador, N. Fong, M. Selby, W. Rutter and G. Bell, Structure of a mouse submaxillary messenger RNA encoding epidermal growth factor and seven related proteins. Science, 221 (1983) 236-240. 7 L. Rail, J. Scott, G. Bell, R. Crawford, J. Penschow, H. Niall and J. Coghlan, Mouse preproepidermal growth factor synthesis by the kidney and other tissues. Nature, 313 (1985) 228231. 8 R.M. Gubits, P.A. Shaw, E. Gresik, A. Onetti-Muda and T. Barka, Epidermal growth factor gene expression is regulated differently in mouse kidney and submandibular gland. (1986) (in press, Endocrinology). 9 E. Gresik and E. Azmitia, Age related changes in NGF, EGF, and protease in the granular convoluted tubules of the mouse submandibular gland. A morphological and immunocytochemical study. J. Gerontol., 35 (1980) 520-524. 10 E. Gresik, M. Brennan and E. Azmitia, Age-related changes in EGF and protease in submandibular glands of C57BL/6J mice. Exp. Aging Res., 8 (1982) 87-90. I1 J, Nelson, K. Latham and C. Finch, Plasma testosterone levels in C57BL/6 male mice: effects of age and disease. Acta Endocrinol., 80 (1975) 744-752. 12 B. Eleftheriou and L. Lucas, Age-related changes in testes, seminal vesicles, and plasma testosterone in male mice. Gerontologia, 20 (1975) 231-238. 13 C. Finch, Studies on hormonal regulation and target cell response in the aging C57BL/6J mouse. In A. Cherkin et al. (eds.), Physiology and Cell Biology o f Aging, Raven Press, NY, 1979, pp. 71-85. 14 J. Raynaud, Control hormonal de la glande sousmaxilhke de la Souris. Bull. Biol., 94 (1960) 4 0 0 525. 15 M. Chertien, Action of testosterone on the defferentiation and secretory activity of a target organ: the submaxillary gland of the mouse. Int. Rev. Cytol., 50 (1977) 333-396.

189 16 E. Gresik and T. Barka, Immunocytochemical localization of epidermal growth factor. J. Histochem. Cytochem., 25 (1977) 1027-1035. 17 E. Gresik, R.M. Gublts and T. Barka, In situ localization of mRNA for epidermal growth factor in the submandibular gland of the mouse. J. Histochem. Cytochem., 33 (1985) 1235-1240. 18 J. Chirgwin, A, Przybla, R. Macdonald and W. Rutter, Isolation of biologically active ribonucleic acid from sources rich in ribonuclease. Biochemistry, 18 (1979) 5294-5299. 19 O. Lowry, N. Rosenbrough, A, Farr and R. Randall, Protein measurement with the Folin phenol reagent. J. Biol. Chem., 193 (1951) 265-275. 20 T. Barka, H. van der Noen, E, Gresik and T. Kerenyi, Immunoreactive epidermal growth factor in human amniotic fluid.Mount Sinai& Med., 45 (1978) 679-684. 21 H. Lehrach, D. Diamond, J. Wozney and H. Boedtker, RNA molecular weight determinations by gel electrophoresis under denaturing conditions, a critical reexamination. Biochemistry, 16 (1977) 4743 -4751 .. 22 R.M. Gubits, K. Lynch, A. Kulkarni, K. Dolan, E. Gresik, P. Hollander, and P. Feigelson, Differential gene regulation of ~x-2u-globulin gene expression in liver, lachrymal gland, and salivary gland. J. Biol. Chem., 259 (1984) 12803-12809. 23 P. Thomas, Hybridization of denatured RNA and small DNA fragments transferred to nitrocellulose. Prec. Nat. Acad. Sci. (USA), 77 (1980) 5201-5205. 24 B. Weiner, Statistical Principles in Experimental Design, McGraw-Hill Book Co., NY, 1962. 25 P. Walker, M. Weichsel, S. Heath, R. Poland and D. Fisher, Effect of thyroxine, testosterone, and corticosterone on nerve growth factor (NGF) and epidermal growth factor (EGF) concentrations in adult female mouse submaxillary gland: dissociation of NGF and EGF responses. Endocrinology, 109 (1981) 582-587. 26 M. Tom-Moy and T. Barka, Epidermal growth factor in the submand~uhr glands of inbred mice. Am. J. Anat., 160 (1981) 267-276. 27 A. Richardson and H.T. Cheung, The relationship between age-related changes in gene expression, protein turnover, and the responsiveness of an organism to stirnufi. Life ScL, 31 (1982) 605-613. 28 M. Rothstein, BiochemicalApproaches to Aging, Academic Press, NY, 1982, pp. 132-197. 29 A. Richardson, M. Birchenall-Sparks, and J. Staecker, Aging and transcription. Rev. Biol. Res. Aging, 1 (1983) 275-294. 30 A. Richardson and M. Birchanall-Sparks, Age-related changes in protein synthesis. Rev. Biol. Res. Aging, 1 (1983) 255-273. 31 G. Wahl, M. Stern and G. Stark, Efficient transfer of large DNA fragments from agarose gels to diazobenzyloxymethyl-paper and rapid hybridization using dextran sulfate. Prec. Natl, Acad. Sci. (USA), 76 (1979) 3683-3687.