- Email: [email protected]
Sites of glandular kallikrein gene expression in fetal mice
Molecular and Cellular Endocrinology, 88 (1992) 23-30 0 1992 Elsevier Scientific Publishers Ireland, Ltd. 0303-7207/92/$05.00
MOLCEL
23
02822
Sites of glandular kallikrein gene expression in fetal mice Jennifer D. Penschow and John P. Coghlan The Howard Florey Institute of Experimental Physiology and Medicine, Royal Parade, Parkuille, vie. 3052, Australia (Received
Key words: Kallikrein,
fetal; Salivary
gland;
Kidney,
3 April 1992; accepted
8 June 1992)
fetal; Nasal gland
Summary As part of an ongoing study of the cell-specific expression of glandular kallikrein genes in mice, we have investigated cellular sites of expression of the renal/pancreatic kallikrein gene, mGK-6, during fetal life. Expression of a1 and P-subunit genes of Na+K+ATPase and bradykinin binding were used as an indication of the functional maturity of the fetal epithelial tubules in which mGK-6 expression was identified. mGK-6 mRNA was first observed at embryonic day 16 (E16) in the submandibular main duct, then at El8 in the sub-lingual main duct, at El9 in renal tubules and at El9 in ducts of the nasal glands. All of these ducts contained detectable epithelial Na+K+ATPase mRNAs from an earlier gestational age than mGK-6 mRNA, suggesting their capacity for electrolyte transport. Bradykinin binding was evident in renal tubules at E18. This study established that renal/pancreatic kallikrein is synthesized in fetal epithelial tubules which are mature functionally.
Introduction Glandular or tissue kallikreins are a group of serine protease enzymes encoded in several species by a family of highly homologous genes (Evans et al., 1987; Riegman et al., 1991; Wines et al., 1991). The glandular kallikrein most highly conserved between species encodes a kininogenase, often referred to as ‘true’ kallikrein, which is expressed in kidney, pancreas, salivary glands and a variety of other tissues from several mammalian species (for review see Clements, 1989). This renal/pancreatic kallikrein has an important
Correspondence to: Jennifer D. Penschow, The Howard Florey Institute of Experimental Physiology and Medicine, The University of Melbourne, Parkville, Vie. 3052, Australia. Tel. (03) 344-5654; Fax (03) 347-0479..
indirect role in the regulation of the local circulation through its production of bradykinin by enzymic cleavage of the substrate kininogen (Carretero and Scicli, 1989). There is increasing evidence that renal/pancreatic kallikrein is also involved in electrolyte transport within the epithelial duct systems in which it is synthesized (Cuthbert et al., 1985; Croxatto et al., 1986; Cuthbert and MacVinish, 1986; Lewis and Alles, 1986; Obika, 1989). Developmental studies of kallikrein gene expression in rodents show that most glandular kallikrein gene family members are expressed from puberty, whereas expression of the gene encoding renal/pancreatic kallikrein is detectable in some rat and mouse tissues from neonatal life (Clements et al., 1990; Penschow et al., 1991). The question of endogenous production of renal/pancreatic kallikrein during fetal
24
life remains largely unanswered. Expression of this gene was not detectable in Northern blots from late gestation fetal rat kidney and salivary glands (Clements et al., 1990) but there is evidence of urinary kallikrein excretion in fetal sheep (Robillard et al., 1982). The level of urinary kallikrein excreted by the ovine fetus begins to rise during gestation but its source has not been determined. The fetal kidney is not necessarily the site of synthesis of the excreted kallikrein but may be the main route for clearance from plasma of glandular kallikreins produced elsewhere. Such is the case in adult rats, in which salivary glands produce more of the kallikrein cleared by the kidneys than the kidneys synthesize themselves (Lawton et al., 1981). Although there is little data available which points to a fetal glandular kallikrein-kinin system in rodents, high molecular weight (HMW) kininogen, which may be cleaved by renal/pancreatic kallikrein (Kato et al., 19851, is present in fetal rats and sheep during the last third of gestation (Heymann et al., 1969; Gelly et al., 1991). However, as HMW kininogen is a multi-purpose protein (Miiller-Ester1 et al., 19861, its presence during fetal life is not necessarily as a substrate for renal/pancreatic kallikrein. Due to our interest in cell-specific expression of mouse glandular kallikrein genes, we were keen to identify sites of renal/pancreatic kallikrein gene expression (encoded by the gene mGK-6) in fetal mice at specific gestational ages and to place the findings in a physiological context. This includes assessment of the functional maturity of fetal tissue in which the kallikrein is synthesized and identification of sites of bradykinin binding. These studies may help to determine whether endogenous kallikreins have a potential role as part of a fetal glandular kallikrein-kinin system in mice. Materials
and methods
Animals
Outbred Swiss mice were mated overnight and pregnancy ascertained by the appearance of a vaginal plug on the following morning, designated El (embryonic day 1). Adult mice were killed by cervical dislocation.
Tissue preparation for hybridization histochemistry
Tissues were freeze-embedded in OCT compound (Lab-Tek, Naperville, IL, USA) by immersion in hexane-dry ice, as described previously (Penschow et al., 1989). Fetal mice were freezeembedded whole. Sections were cut at 6 pm on a cryo-microtome and fixed in 3% glutaraldehyde (Merck, Darmstadt, Germany) in 0.1 M phosphate buffer with 20% ethylene glycol. Probes and controls
Oligodeoxyribonucleotide probes of 30-34 nucleotides were synthesized using an Applied Biosystems Model 380A DNA Synthesizer and labelled with [Y-~~P]ATI? or [ 3H]dNTP as described previously (Penschow et al., 1989). Probe sequences corresponded to mRNAs encoded by the following genes: mouse renal/pancreatic (true) kallikrein (mGK-6) (nucleotides 2157-2186; van Leeuwen et al., 19861, rat Na+K+ATPase a-subunit (nucleotide 387-420; Shull et al., 19861, rat Na+K+ATPase P-subunit (nucleotides 506535; Young et al., 1987). A common kallikrein oligonucleotide probe, which detects mRNAs encoded by all 12 mouse glandular kallikreins (van Leeuwen et al., 19861, was used as a screening probe for transcripts of these 12 genes and a probe specific for transcripts of the mouse photoreceptor protein rhodopsin was used as a negative control. Additional controls were adjacent sections hybridized after treatment with ribonuclease A (Sigma, St. Louis, MO, USA) or with an excess of unlabelled probe added to the probe mixture. Hybridization
Pre-hybridization, hybridization, washing and autoradiography procedures were as described previously (Penschow et al., 1989). Briefly, probes were diluted to 400 ng/ml in hybridization buffer with 40% formamide and hybridization effected at 40°C for all except the mGK-6 probe, which was used at 50°C to avoid cross-hybridization with other glandular kallikrein mRNAs (Penschow et al., 1991). Sections were washed at 50°C in 1 x SSC (0.015 M trisodium citrate and 0.15 M NaCl, pH 7.0) and autoradiographed using X-ray film followed by liquid emulsion.
25
Bradykinin binding
Tissues were freeze-embedded and sectioned as for hybridization histochemistry. Sections were collected on chilled slides coated with 0.03% poly-r_-lysine (Sigma), sealed with silica gel and
left overnight at -20°C. Sections were removed from freezer and brought to room temperature. (The following incubation mixtures and labelling procedure were based upon studies by Lewis et al., 1985.) One of each pair of serial sections was
Fig. 1. Liquid emulsion autoradiographs, showing 6 pm frozen sections of fetal mice at various ages, after hybridization with the mGK-6 probe (a, b, d, e, f) or the NaC K+ATPase a1 probe (cl. (a-d) show an extraglandular region of the main submandibular (SM) duct (arrowhead) and sub-lingual (SL) duct in serial sections at El6 (a-c) or at El8 (d). Labelling of the SM duct with 3ZP-labelled mGK-6 is evident in (a) and is negative in ribonuclease-treated control (b). In (c) both the SM and SL ducts have labelled with the 32P-labelled Na+ K+ATPase (~1 probe and in (d) with the 3H-mGK-6 probe. Bar = 200 pm. (e and f) are serial sections showing areas of the SM (top of field) and SL glands (lower half of field) at E19. Ducts in the SL (arrowheads) in (e) are lightly labelled with the 32P-mGK-6 probe compared with the strong signal in SM ducts at top of field. Secretory material is evident in some SL acini in centre field. Negative control (f) was hybridized with 32P-labelled mGK-6 probe containing 5-fold unlabelled mGK-6 probe. Photomicrographs in combined bright and dark field. Bar = 800 pm.
26
pre-incubated for 5 min at room temperature CRT) in 0.1% bovine serum albumin (Sigma) in 0.05 M potassium phosphate pH 7.4 with 3 PM 1,10-o-phenanthroline (B.D.H., Poole, UK) and 3 PM SQ 14225 (Squibb, Princeton, NJ, USA). The alternate section of the pair, used as a blocking control, was pre-incubated in the above mixture containing 25 PM unlabelled Tyr-bradykinin (Sigma). Slides were then incubated at RT for 50 min in a mixture of the above ingredients and 3 p M [ I25IlTyr-bradykinin (Sigma). For blocking controls, the incubation mixture also contained 50 PM of unlabelled ligand. After incubation with the iodinated ligand, all slides were washed in three changes of 0.05 M potassium phosphate at 4°C (30 ml per slide), then fixed in 3% glutaraldehyde (Merck) in the same buffer, rinsed twice in buffer at 4°C then dried in a stream of air at RT. Autoradiography was performed as described previously for hybridization histochemistry studies (Penschow et al., 1989). Results
In order to evaluate expression of all murine glandular kallikrein genes in fetal mice, screening of adjacent sections was undertaken with a probe common to all of the glandular kallikrein genes.
This located only the same sites of expression revealed by the mGK-6 probe, which specifically detects transcripts of renal/pancreatic kallikrein (encoded by the gene, mGK-6). In all ducts in which mGK-6 expression was found, Na+K+ATPase aI and /?-subunit genes were expressed concomitantly (Figs. 1 and 4) and at earlier gestational ages (data not shown). Expression of the al subunit gene always coincided with /?-subunit gene expression and vice versa, therefore we have illustrated expression of one or other of these genes, but not of both. The gestational ages at which renal/pancreatic kallikrein gene expression was first detectable in murine fetal tissues by hybridization histochemistry are summarized in Table 1.
Gene expression in salivary ducts
The luminal region of the main submandibular (SM) excretory duct en route from the submandibular gland to the mouth at El6 was the first site at which mGK-6 expression was detected (Fig. la and b). Na+K+ATPase (YI- and P-subunit mRNAs co-localized in the SM duct with mGK-6 mRNA (Fig. 1~). Na+K+ATPase subunit genes were also expressed at El6 in the main sub-lingual (SL) duct (Fig. lc), whereas mGK-6 expression in this duct was first detectable at El8 (Fig. Id). By E19, mGK-6 expression was at a
Fig. 2. Liquid emulsion autoradiographs of 6 pm frozen serial sections of fetal mouse kidney cortex at El9 after hybridization with the 32P-labelled mGK-6 probe. (a) is the positive and (b) is the control, which has been pre-treated with ribonuclease A. The blood vessel (x) and glomerulus (G) are landmarks which appear in both sections. (a) shows tubules (arrowheads) which have hybridized and adjacent tissue stuctures which show no hybridization signal. The control (b) is negative. Photomicrographs in combined bright and dark field. Bar = 150 Wm.
27
high level in numerous smaller ducts within the SM gland and at a lower level in occasional ducts in the SL gland (Fig. le and f).
Expression of mGK-6 in fetal kidney
mGK-6 expression was not detected in the kidney until E19, in occasional cortical distal
Fig. 3. Liquid emulsion autoradiographs of 6 km frozen sections of kidney from a fetal mouse at which show show sites of lzsI-Tyr-bradykinin binding. (a and b) show a field corresponding to control serial section (c and d). Both pairs of micrographs show the same field photographed by bright-field (a and cl or dark-field (b and d) microscopy. The tubule (x) and adjacent structures clearly correspond in (a) and (cl. Labelling of tubules (some marked with arrowheads) is evident in (b) and weak or absent in corresponding tubules in control Cd), which was incubated with excess unlabelled ligand. Bar = 150 pm. (e) shows the field (boxed) from (a and b) and surrounding renal tissue at low magnification. Bar = 700 pm. (f and g) show a corresponding field in serial sections of adult kidney, which were incubated concurrently with the sections in (a-d). There is a strong signal in the outer medulla (arrowheads) in (f) which is absent in control (g). The papilla (P) and inner stripe are clearly negative in both sections. Bar = 500 pm. Note: The morphological detail of these tissues is not as clear as in the other figures, as the method precludes fixation prior to incubation with the ligand.
28
tubules (Fig. 2a and b) and co-localized with Na+K+ATPase subunit mRNAs, which were already present at high levels in renal tubules at El6 (data not shown).
within a group (Fig. 4b) and co-localized with Na+K+ATPase subunit mRNAs (Fig. 4~). Controls were negative (Fig. 4d). Discussion
Bradykinin binding in fetal kidney
At El& binding of lz51-bradykinin (Fig. 3a and b), blockable by incubation with excess bradykinin (Fig. 3c and d), was located in renal tubules (Fig. 3e). Control sections of adult kidney prepared identically to the fetal tissue show strong binding to the outer medulla (Fig. 3f 1, which was blocked by excess unlabelled bradykinin (Fig. 3g) as in fetal kidney. Bradykinin binding was not investigated at earlier gestational ages. Gene expression in fetal nasal glands
At E19, mGK-6 expression was detected in the lateral nasal glands (Fig. 4a), in the largest duct
Our previous studies have shown that of the 12 functional glandular kallikrein genes in mice, expression of 11 genes is induced by the hormonal changes which occur in the SM gland at puberty. The renal/pancreatic kallikrein gene, mGK-6, is the only gene expressed in salivary glands between 1 and 21 days post-natal and the only gene expressed in kidney (van Leeuwen et al., 1986; Penschow et al., 1991). Screening of paired fetal tissue sections with the common kallikrein probe and the mGK-6-specific probe confirmed that all of the sites of expression found with the screening probe corresponded to sites of mGK-6 ex-
Fig. 4. Liquid emulsion autoradiographs of 6 pm frozen horizontal sections from a fetal mouse at El9 after hybridization 3ZP-labelled probes for mGK-6 (a, b, d) or Na’K+ATPase P-subunit (c) mRNAs. (a) shows the location of the lateral glands (arrowhead) in relation to the septum (S) and nasal cavities (N). Bar = 1 mm. (b-d) are serial sections in which the marked in (a) with an arrowhead appears in each section, marked similarly. (b) shows labelling of a few cells in the marked (tip of arrowhead) with the mGK-6 probe. (c) shows labelling of the same duct with Na+K+ATPase and control (d) shows RNAase pretreatment abolishes the mGK-6 signal. Bar = 125 pm.
with nasal duct duct that
29 TABLE 1 SITES OF GLANDULAR SION IN FETAL MICE Fetal age
KALLIKREIN GENE EXPRES-
Site Submandibular gfand
El6 El7 El9
Extraglandular excretory ducts (Fig. 1) Hilar excretory ducts (not shown) Interlobular ducts (Fig. 1) Subf~ngual gfand
El8 El9
Extraglandular excretory ducts (Fig. 1) Hilar excretory and interlobular ducts (Fig. 1) Parotid glands
3-day-old
Excretory ducts (not shown) Kidney
El9
Cortical distal tubules (Fig. 2) Lateral nasal glands
El9
Excretory ducts (Fig. 4)
pression. Thus all detectable glandular kallikrein transcripts in these fetal tissues were attributed to the mGIS-6 gene. We have shown in the present study that renai/pan~reatic kallikrein is synthesized in SM excretory ducts as early as El6 in mouse (Fig. 1). There may also be a potential substrate for renal/pancreatic kallikrein in the circulation of fetal mice, as HMW kininogen has been shown in fetal rat liver as early as El3 (Gelly et al,, 1991) and in ovine fetal arterial blood during the first half of gestation (Heymann et al., 1969). As these components of the glandular kallikrein-kinin system may be present and the fetal circulation is well-developed in mice by El6 (Theiler, 19721, there could be a system of local regulation of blood flow in SM glands at El6 which corresponds to that shotin in adult rats by Berg et al. 0989). Their data and studies by Lawton et al. (1980, which showed that the SM and SL glands are the main source of glandular kallikrein in plasma of adult rats, suggest that the renal kaIlikrei~ produced by salivary duct cells enters the circulation. The route by which this occurs is unclear; however, Carretero and Scicli (1989) suggest that renal kallikrein secreted from the basolateral surface of duct cells in the SM gland or kidney produces kinins in the interstitial space
adjacent to underlying blood vessels which are the target tissues for the kinins produced. Thus, the renal/pancreatic kallikrein produced by mouse SM duct cells at El6 may be part of a glandular kallikrein-kinin system similar to that involved in the regulation of the microcirculation of the adult SM gland. The onset of renal/pancreatic kallikrein gene expression in SL ducts occurred at E18, a later gestational age than in SM ducts (Fig. 11, which is in accordance with the comparatively later development of the SL duct tree. Although SL ducts develop later than SM ducts, the reverse is the case with secretory endpieces, which develop during fetal life in the SL (Leeson and Booth, 1961) but not until post-natal life in the SM and parotid glands (Jacoby and Leeson, 195’9; Redman and Sreebny, 1970). Despite the relative immaturity of the salivary secretory apparatus during late gestation, the secretion of a kallikrein-containing fetal salivary fluid is nevertheless a distinct possibili~. SL endpieces contain secretory material by El9 (Fig. 1) and SM duct cells can themselves produce a salivary secretion in the absence of secretory endpieces. The capacity of SM duct cells to secrete water and electrolytes at flow rates approaching those of intact glands, was demonstrated using ligated adult rat SM glands, which had functional secretory endpieces (Schneyer and Schneyer, 1961). Furthermore, by El6 in mice SM duct cells can produce both cYI-and P-subunits of Na+K+ATPase (Fig. 11, which suggests that they already have the capability to generate the ion gradient which drives ductal salivary secretion (Martinez, 1987; Young et al., 1987). Expression of mGK-6 in fetal kidneys was first detectable at El9 (Fig. 21, much later in gestation than in the submandibular ducts. Considering that bradykinin binding was detectable in renal tubules at El8 (Fig. 3), there may be also transcription, translation and secretion of the products of mGK-6 expression occurring at this time. Although mGK-6 expression was undetectable at El8 by hybridization histochemist~, low levels of mRNA may have been present. However, in a comparable study (Clements et al., 1990), Northern analysis was not sufficiently sensitive to detect renal/pancreatic kallikrein mRNA in rat kidney at a comparable stage of gestation.
30
We have shown previously high level mGK-6 expression in excretory ducts of the nasal glands (Penschow et al., 1991) of neonatal mice. In the present study, we have demonstrated expression of mGK-6 at E19, in the same duct of the lateral nasal glands as mRNAs encoding epithelial Na+K+ATPase (Fig. 41. This shows that these kallikrein-synthetic ducts can also transport electrolytes. This is another example of renal/pancreatic kallikrein and epithelial Na+K+ATPase mRNAs in the same fetal duct cells, seen also in SM ducts at El6 (Fig. 1) and in renal tubules at El9 (data not shown). This may not be a mere co-incidence, as there is evidence that renal/pancreatic kallikrein may have a role in the regulation of electrolyte transport in renal tubules (Cuthbert et al., 1985; Croxatto et al., 1986; Cuthbert and MacVinish, 1986) and such a system may also operate in other epithelial tubules, such as salivary and nasal ducts. In conclusion, we have shown that renal kallikrein is synthesized in mature transporting epithelia of a number of tissues during fetal life in mice. Furthermore, the binding of bradykinin at El8 in renal tubules and synthesis of renal kallikrein by the fetal kidney at much the same fetal age suggests that the renal kallikrein-kinin system is operating during late gestation. Acknowledgements
We thank Dr. Jim Haralambidis, Ms. Suzie Khoury, Dr. Ross Fernley and Ms. Elizabeth Cooper for preparation and labelling of probes and ligands. This work was supported by grantsin-aid from the National Health and Medical Research Council of Australia, the Myer Family Trusts, the Ian Potter Foundation and the Howard Florey Biomedical Foundation. References Berg, T., Carretero, O.A., Scicli, A.G., Tilley, B. and Stewart, J.M. (1989) Hypertension 14, 73-80. Carretero, O.A. and Scicli, G.A. (1989) in Endocrine Mechanisms in Hypertension (Laragh, J.H., Brenner, B.M. and Kaplan, N.M., eds.), pp. 219-239, Raven, New York. Clements, J.A. (1989) Endocr. Rev. 10, 393-419.
Clements, J.A., Matheson, B.A. and Funder, J.W. (1990) J. Biol. Chem. 265, 1077-1081. Croxatto, H.R., Corthorn, J., Roblero, J., Villalo, P. and Perez, F. (1986) Am. J. Physiol. 250, F400-F406. Cuthbert, A.W. and MacVinish, L.J. (1986) Adv. Exp. Med. Biol. 198A, 203-21 I. Cuthbert, A.W., George, A.M. and MacVinish, L. (1985) Am. J. Physiol. 249, F439-F447. Evans, B.A., Drinkwater, C.C. and Richards, R.I. (1987) J. Biol. Chem. 262, 8027-8034. Gelly, J.L., Richoux, J.P., Grignon, G., Bouhnik, J., Baussant, T., Alhenc-Gelas, F. and Corvol, P. (1991) Histochemistry 96, 7-12. Heymann, M.A., Rudolph, A.M., Nies, A.S. and Melmon, K.L. (1969) Circ. Res. 25, 521-534. Jacoby, F. and Leeson, C.R. (1959) J. Anat. (London) 93, 201-216. Kato, H., Kei-ichi, E., Miyata, T., Hayashi, I., Oh-ishi, S. and Iwanaga, S. (1985) Biochem. Biophys. Res. Commun. 127, 289-295. Lawton, W.J., Proud, D., Frech, M.E., Pierce, J.V., Keiser, H.R. and Pisano, J.J. (1981) Biochem. Pharmacol. 30, 1731-1737. Leeson, C.R. and Booth, W.G. (1961) J. Dent. Res. 40, 838-845. Lewis, R.E., Childers, S.R. and Phillips, MI. (1985) Brain Res. 346, 263-272. Lewis, S.A. and Alles, W.P. (1986) Proc. Natl. Acad. Sci. USA 83, 5345-5348. Martinez, J.R. (1987) J. Dent. Res. 66 (Special Issue), 638-647. Miiller-Esterl, W., Iwanaga, S. and Nakanishi, S (19861 Trends Biochem. Sci. 11, 336-339. Obika, L.F.O. (1989) Clin. Sci. 77, 21-27. Penschow, J.D., Haralambidis, J., Pownall, S. and Coghlan, J.P. (1989) Methods Neurosci. 1, 222-238. Penschow, J.D., Drinkwater, C.C., Haralambidis, J. and CoghIan, J.P. (1991) Mol. Cell Endocrinol. 81, 135-146. Redman, R.S. and Sreebny, L.M. (1970) Anat. Rec. 168, 127-138. Riegman, P.H.J., Vlietstra, R.J., van der Korput, H.A.G.M., Romijn, J.C. and Trapman, J. (1991) Mol. Cell Endocrinol. 76, 181-190. Robillard, J.E., Lawton, W.J., Weisman, D.N. and Sessions, C. (1982) Kidney Int. 22, 594-601. Schneyer, C.A. and Schneyer, L.H. (1961) Am. J. Physiol. 201, 939-942. Schull, G.E., Greeb, J. and Lingrel, J.B. (1986) Biochemistry 25, 8125-8132. Theiler, K. (1972) The House Mouse, Springer, Berlin. van Leeuwen, B.H., Evans, B.A., Tregear, G.W. and Richards, RI. (1986) J. Biol. Chem. 261, 5529-5535. Wines, D.R., Brady, J.M., Southard, E.M. and MacDonald, R.J. (1991) J. Mol. Evol. 32, 476-492. Young, J.A. and van Lcnnep, E.W. (1979) in Membrane Transport in Biology (Giebisch, G., Tosteson, D.C. and Ussing, H.H., eds.), pp. 563-613, Springer, Berlin. Young, R.M., Schull, G.E. and Lingrel, J.B. (1987) J. Biol. Chem. 262, 4905-4910.
MOLCEL
23
02822
Sites of glandular kallikrein gene expression in fetal mice Jennifer D. Penschow and John P. Coghlan The Howard Florey Institute of Experimental Physiology and Medicine, Royal Parade, Parkuille, vie. 3052, Australia (Received
Key words: Kallikrein,
fetal; Salivary
gland;
Kidney,
3 April 1992; accepted
8 June 1992)
fetal; Nasal gland
Summary As part of an ongoing study of the cell-specific expression of glandular kallikrein genes in mice, we have investigated cellular sites of expression of the renal/pancreatic kallikrein gene, mGK-6, during fetal life. Expression of a1 and P-subunit genes of Na+K+ATPase and bradykinin binding were used as an indication of the functional maturity of the fetal epithelial tubules in which mGK-6 expression was identified. mGK-6 mRNA was first observed at embryonic day 16 (E16) in the submandibular main duct, then at El8 in the sub-lingual main duct, at El9 in renal tubules and at El9 in ducts of the nasal glands. All of these ducts contained detectable epithelial Na+K+ATPase mRNAs from an earlier gestational age than mGK-6 mRNA, suggesting their capacity for electrolyte transport. Bradykinin binding was evident in renal tubules at E18. This study established that renal/pancreatic kallikrein is synthesized in fetal epithelial tubules which are mature functionally.
Introduction Glandular or tissue kallikreins are a group of serine protease enzymes encoded in several species by a family of highly homologous genes (Evans et al., 1987; Riegman et al., 1991; Wines et al., 1991). The glandular kallikrein most highly conserved between species encodes a kininogenase, often referred to as ‘true’ kallikrein, which is expressed in kidney, pancreas, salivary glands and a variety of other tissues from several mammalian species (for review see Clements, 1989). This renal/pancreatic kallikrein has an important
Correspondence to: Jennifer D. Penschow, The Howard Florey Institute of Experimental Physiology and Medicine, The University of Melbourne, Parkville, Vie. 3052, Australia. Tel. (03) 344-5654; Fax (03) 347-0479..
indirect role in the regulation of the local circulation through its production of bradykinin by enzymic cleavage of the substrate kininogen (Carretero and Scicli, 1989). There is increasing evidence that renal/pancreatic kallikrein is also involved in electrolyte transport within the epithelial duct systems in which it is synthesized (Cuthbert et al., 1985; Croxatto et al., 1986; Cuthbert and MacVinish, 1986; Lewis and Alles, 1986; Obika, 1989). Developmental studies of kallikrein gene expression in rodents show that most glandular kallikrein gene family members are expressed from puberty, whereas expression of the gene encoding renal/pancreatic kallikrein is detectable in some rat and mouse tissues from neonatal life (Clements et al., 1990; Penschow et al., 1991). The question of endogenous production of renal/pancreatic kallikrein during fetal
24
life remains largely unanswered. Expression of this gene was not detectable in Northern blots from late gestation fetal rat kidney and salivary glands (Clements et al., 1990) but there is evidence of urinary kallikrein excretion in fetal sheep (Robillard et al., 1982). The level of urinary kallikrein excreted by the ovine fetus begins to rise during gestation but its source has not been determined. The fetal kidney is not necessarily the site of synthesis of the excreted kallikrein but may be the main route for clearance from plasma of glandular kallikreins produced elsewhere. Such is the case in adult rats, in which salivary glands produce more of the kallikrein cleared by the kidneys than the kidneys synthesize themselves (Lawton et al., 1981). Although there is little data available which points to a fetal glandular kallikrein-kinin system in rodents, high molecular weight (HMW) kininogen, which may be cleaved by renal/pancreatic kallikrein (Kato et al., 19851, is present in fetal rats and sheep during the last third of gestation (Heymann et al., 1969; Gelly et al., 1991). However, as HMW kininogen is a multi-purpose protein (Miiller-Ester1 et al., 19861, its presence during fetal life is not necessarily as a substrate for renal/pancreatic kallikrein. Due to our interest in cell-specific expression of mouse glandular kallikrein genes, we were keen to identify sites of renal/pancreatic kallikrein gene expression (encoded by the gene mGK-6) in fetal mice at specific gestational ages and to place the findings in a physiological context. This includes assessment of the functional maturity of fetal tissue in which the kallikrein is synthesized and identification of sites of bradykinin binding. These studies may help to determine whether endogenous kallikreins have a potential role as part of a fetal glandular kallikrein-kinin system in mice. Materials
and methods
Animals
Outbred Swiss mice were mated overnight and pregnancy ascertained by the appearance of a vaginal plug on the following morning, designated El (embryonic day 1). Adult mice were killed by cervical dislocation.
Tissue preparation for hybridization histochemistry
Tissues were freeze-embedded in OCT compound (Lab-Tek, Naperville, IL, USA) by immersion in hexane-dry ice, as described previously (Penschow et al., 1989). Fetal mice were freezeembedded whole. Sections were cut at 6 pm on a cryo-microtome and fixed in 3% glutaraldehyde (Merck, Darmstadt, Germany) in 0.1 M phosphate buffer with 20% ethylene glycol. Probes and controls
Oligodeoxyribonucleotide probes of 30-34 nucleotides were synthesized using an Applied Biosystems Model 380A DNA Synthesizer and labelled with [Y-~~P]ATI? or [ 3H]dNTP as described previously (Penschow et al., 1989). Probe sequences corresponded to mRNAs encoded by the following genes: mouse renal/pancreatic (true) kallikrein (mGK-6) (nucleotides 2157-2186; van Leeuwen et al., 19861, rat Na+K+ATPase a-subunit (nucleotide 387-420; Shull et al., 19861, rat Na+K+ATPase P-subunit (nucleotides 506535; Young et al., 1987). A common kallikrein oligonucleotide probe, which detects mRNAs encoded by all 12 mouse glandular kallikreins (van Leeuwen et al., 19861, was used as a screening probe for transcripts of these 12 genes and a probe specific for transcripts of the mouse photoreceptor protein rhodopsin was used as a negative control. Additional controls were adjacent sections hybridized after treatment with ribonuclease A (Sigma, St. Louis, MO, USA) or with an excess of unlabelled probe added to the probe mixture. Hybridization
Pre-hybridization, hybridization, washing and autoradiography procedures were as described previously (Penschow et al., 1989). Briefly, probes were diluted to 400 ng/ml in hybridization buffer with 40% formamide and hybridization effected at 40°C for all except the mGK-6 probe, which was used at 50°C to avoid cross-hybridization with other glandular kallikrein mRNAs (Penschow et al., 1991). Sections were washed at 50°C in 1 x SSC (0.015 M trisodium citrate and 0.15 M NaCl, pH 7.0) and autoradiographed using X-ray film followed by liquid emulsion.
25
Bradykinin binding
Tissues were freeze-embedded and sectioned as for hybridization histochemistry. Sections were collected on chilled slides coated with 0.03% poly-r_-lysine (Sigma), sealed with silica gel and
left overnight at -20°C. Sections were removed from freezer and brought to room temperature. (The following incubation mixtures and labelling procedure were based upon studies by Lewis et al., 1985.) One of each pair of serial sections was
Fig. 1. Liquid emulsion autoradiographs, showing 6 pm frozen sections of fetal mice at various ages, after hybridization with the mGK-6 probe (a, b, d, e, f) or the NaC K+ATPase a1 probe (cl. (a-d) show an extraglandular region of the main submandibular (SM) duct (arrowhead) and sub-lingual (SL) duct in serial sections at El6 (a-c) or at El8 (d). Labelling of the SM duct with 3ZP-labelled mGK-6 is evident in (a) and is negative in ribonuclease-treated control (b). In (c) both the SM and SL ducts have labelled with the 32P-labelled Na+ K+ATPase (~1 probe and in (d) with the 3H-mGK-6 probe. Bar = 200 pm. (e and f) are serial sections showing areas of the SM (top of field) and SL glands (lower half of field) at E19. Ducts in the SL (arrowheads) in (e) are lightly labelled with the 32P-mGK-6 probe compared with the strong signal in SM ducts at top of field. Secretory material is evident in some SL acini in centre field. Negative control (f) was hybridized with 32P-labelled mGK-6 probe containing 5-fold unlabelled mGK-6 probe. Photomicrographs in combined bright and dark field. Bar = 800 pm.
26
pre-incubated for 5 min at room temperature CRT) in 0.1% bovine serum albumin (Sigma) in 0.05 M potassium phosphate pH 7.4 with 3 PM 1,10-o-phenanthroline (B.D.H., Poole, UK) and 3 PM SQ 14225 (Squibb, Princeton, NJ, USA). The alternate section of the pair, used as a blocking control, was pre-incubated in the above mixture containing 25 PM unlabelled Tyr-bradykinin (Sigma). Slides were then incubated at RT for 50 min in a mixture of the above ingredients and 3 p M [ I25IlTyr-bradykinin (Sigma). For blocking controls, the incubation mixture also contained 50 PM of unlabelled ligand. After incubation with the iodinated ligand, all slides were washed in three changes of 0.05 M potassium phosphate at 4°C (30 ml per slide), then fixed in 3% glutaraldehyde (Merck) in the same buffer, rinsed twice in buffer at 4°C then dried in a stream of air at RT. Autoradiography was performed as described previously for hybridization histochemistry studies (Penschow et al., 1989). Results
In order to evaluate expression of all murine glandular kallikrein genes in fetal mice, screening of adjacent sections was undertaken with a probe common to all of the glandular kallikrein genes.
This located only the same sites of expression revealed by the mGK-6 probe, which specifically detects transcripts of renal/pancreatic kallikrein (encoded by the gene, mGK-6). In all ducts in which mGK-6 expression was found, Na+K+ATPase aI and /?-subunit genes were expressed concomitantly (Figs. 1 and 4) and at earlier gestational ages (data not shown). Expression of the al subunit gene always coincided with /?-subunit gene expression and vice versa, therefore we have illustrated expression of one or other of these genes, but not of both. The gestational ages at which renal/pancreatic kallikrein gene expression was first detectable in murine fetal tissues by hybridization histochemistry are summarized in Table 1.
Gene expression in salivary ducts
The luminal region of the main submandibular (SM) excretory duct en route from the submandibular gland to the mouth at El6 was the first site at which mGK-6 expression was detected (Fig. la and b). Na+K+ATPase (YI- and P-subunit mRNAs co-localized in the SM duct with mGK-6 mRNA (Fig. 1~). Na+K+ATPase subunit genes were also expressed at El6 in the main sub-lingual (SL) duct (Fig. lc), whereas mGK-6 expression in this duct was first detectable at El8 (Fig. Id). By E19, mGK-6 expression was at a
Fig. 2. Liquid emulsion autoradiographs of 6 pm frozen serial sections of fetal mouse kidney cortex at El9 after hybridization with the 32P-labelled mGK-6 probe. (a) is the positive and (b) is the control, which has been pre-treated with ribonuclease A. The blood vessel (x) and glomerulus (G) are landmarks which appear in both sections. (a) shows tubules (arrowheads) which have hybridized and adjacent tissue stuctures which show no hybridization signal. The control (b) is negative. Photomicrographs in combined bright and dark field. Bar = 150 Wm.
27
high level in numerous smaller ducts within the SM gland and at a lower level in occasional ducts in the SL gland (Fig. le and f).
Expression of mGK-6 in fetal kidney
mGK-6 expression was not detected in the kidney until E19, in occasional cortical distal
Fig. 3. Liquid emulsion autoradiographs of 6 km frozen sections of kidney from a fetal mouse at which show show sites of lzsI-Tyr-bradykinin binding. (a and b) show a field corresponding to control serial section (c and d). Both pairs of micrographs show the same field photographed by bright-field (a and cl or dark-field (b and d) microscopy. The tubule (x) and adjacent structures clearly correspond in (a) and (cl. Labelling of tubules (some marked with arrowheads) is evident in (b) and weak or absent in corresponding tubules in control Cd), which was incubated with excess unlabelled ligand. Bar = 150 pm. (e) shows the field (boxed) from (a and b) and surrounding renal tissue at low magnification. Bar = 700 pm. (f and g) show a corresponding field in serial sections of adult kidney, which were incubated concurrently with the sections in (a-d). There is a strong signal in the outer medulla (arrowheads) in (f) which is absent in control (g). The papilla (P) and inner stripe are clearly negative in both sections. Bar = 500 pm. Note: The morphological detail of these tissues is not as clear as in the other figures, as the method precludes fixation prior to incubation with the ligand.
28
tubules (Fig. 2a and b) and co-localized with Na+K+ATPase subunit mRNAs, which were already present at high levels in renal tubules at El6 (data not shown).
within a group (Fig. 4b) and co-localized with Na+K+ATPase subunit mRNAs (Fig. 4~). Controls were negative (Fig. 4d). Discussion
Bradykinin binding in fetal kidney
At El& binding of lz51-bradykinin (Fig. 3a and b), blockable by incubation with excess bradykinin (Fig. 3c and d), was located in renal tubules (Fig. 3e). Control sections of adult kidney prepared identically to the fetal tissue show strong binding to the outer medulla (Fig. 3f 1, which was blocked by excess unlabelled bradykinin (Fig. 3g) as in fetal kidney. Bradykinin binding was not investigated at earlier gestational ages. Gene expression in fetal nasal glands
At E19, mGK-6 expression was detected in the lateral nasal glands (Fig. 4a), in the largest duct
Our previous studies have shown that of the 12 functional glandular kallikrein genes in mice, expression of 11 genes is induced by the hormonal changes which occur in the SM gland at puberty. The renal/pancreatic kallikrein gene, mGK-6, is the only gene expressed in salivary glands between 1 and 21 days post-natal and the only gene expressed in kidney (van Leeuwen et al., 1986; Penschow et al., 1991). Screening of paired fetal tissue sections with the common kallikrein probe and the mGK-6-specific probe confirmed that all of the sites of expression found with the screening probe corresponded to sites of mGK-6 ex-
Fig. 4. Liquid emulsion autoradiographs of 6 pm frozen horizontal sections from a fetal mouse at El9 after hybridization 3ZP-labelled probes for mGK-6 (a, b, d) or Na’K+ATPase P-subunit (c) mRNAs. (a) shows the location of the lateral glands (arrowhead) in relation to the septum (S) and nasal cavities (N). Bar = 1 mm. (b-d) are serial sections in which the marked in (a) with an arrowhead appears in each section, marked similarly. (b) shows labelling of a few cells in the marked (tip of arrowhead) with the mGK-6 probe. (c) shows labelling of the same duct with Na+K+ATPase and control (d) shows RNAase pretreatment abolishes the mGK-6 signal. Bar = 125 pm.
with nasal duct duct that
29 TABLE 1 SITES OF GLANDULAR SION IN FETAL MICE Fetal age
KALLIKREIN GENE EXPRES-
Site Submandibular gfand
El6 El7 El9
Extraglandular excretory ducts (Fig. 1) Hilar excretory ducts (not shown) Interlobular ducts (Fig. 1) Subf~ngual gfand
El8 El9
Extraglandular excretory ducts (Fig. 1) Hilar excretory and interlobular ducts (Fig. 1) Parotid glands
3-day-old
Excretory ducts (not shown) Kidney
El9
Cortical distal tubules (Fig. 2) Lateral nasal glands
El9
Excretory ducts (Fig. 4)
pression. Thus all detectable glandular kallikrein transcripts in these fetal tissues were attributed to the mGIS-6 gene. We have shown in the present study that renai/pan~reatic kallikrein is synthesized in SM excretory ducts as early as El6 in mouse (Fig. 1). There may also be a potential substrate for renal/pancreatic kallikrein in the circulation of fetal mice, as HMW kininogen has been shown in fetal rat liver as early as El3 (Gelly et al,, 1991) and in ovine fetal arterial blood during the first half of gestation (Heymann et al., 1969). As these components of the glandular kallikrein-kinin system may be present and the fetal circulation is well-developed in mice by El6 (Theiler, 19721, there could be a system of local regulation of blood flow in SM glands at El6 which corresponds to that shotin in adult rats by Berg et al. 0989). Their data and studies by Lawton et al. (1980, which showed that the SM and SL glands are the main source of glandular kallikrein in plasma of adult rats, suggest that the renal kaIlikrei~ produced by salivary duct cells enters the circulation. The route by which this occurs is unclear; however, Carretero and Scicli (1989) suggest that renal kallikrein secreted from the basolateral surface of duct cells in the SM gland or kidney produces kinins in the interstitial space
adjacent to underlying blood vessels which are the target tissues for the kinins produced. Thus, the renal/pancreatic kallikrein produced by mouse SM duct cells at El6 may be part of a glandular kallikrein-kinin system similar to that involved in the regulation of the microcirculation of the adult SM gland. The onset of renal/pancreatic kallikrein gene expression in SL ducts occurred at E18, a later gestational age than in SM ducts (Fig. 11, which is in accordance with the comparatively later development of the SL duct tree. Although SL ducts develop later than SM ducts, the reverse is the case with secretory endpieces, which develop during fetal life in the SL (Leeson and Booth, 1961) but not until post-natal life in the SM and parotid glands (Jacoby and Leeson, 195’9; Redman and Sreebny, 1970). Despite the relative immaturity of the salivary secretory apparatus during late gestation, the secretion of a kallikrein-containing fetal salivary fluid is nevertheless a distinct possibili~. SL endpieces contain secretory material by El9 (Fig. 1) and SM duct cells can themselves produce a salivary secretion in the absence of secretory endpieces. The capacity of SM duct cells to secrete water and electrolytes at flow rates approaching those of intact glands, was demonstrated using ligated adult rat SM glands, which had functional secretory endpieces (Schneyer and Schneyer, 1961). Furthermore, by El6 in mice SM duct cells can produce both cYI-and P-subunits of Na+K+ATPase (Fig. 11, which suggests that they already have the capability to generate the ion gradient which drives ductal salivary secretion (Martinez, 1987; Young et al., 1987). Expression of mGK-6 in fetal kidneys was first detectable at El9 (Fig. 21, much later in gestation than in the submandibular ducts. Considering that bradykinin binding was detectable in renal tubules at El8 (Fig. 3), there may be also transcription, translation and secretion of the products of mGK-6 expression occurring at this time. Although mGK-6 expression was undetectable at El8 by hybridization histochemist~, low levels of mRNA may have been present. However, in a comparable study (Clements et al., 1990), Northern analysis was not sufficiently sensitive to detect renal/pancreatic kallikrein mRNA in rat kidney at a comparable stage of gestation.
30
We have shown previously high level mGK-6 expression in excretory ducts of the nasal glands (Penschow et al., 1991) of neonatal mice. In the present study, we have demonstrated expression of mGK-6 at E19, in the same duct of the lateral nasal glands as mRNAs encoding epithelial Na+K+ATPase (Fig. 41. This shows that these kallikrein-synthetic ducts can also transport electrolytes. This is another example of renal/pancreatic kallikrein and epithelial Na+K+ATPase mRNAs in the same fetal duct cells, seen also in SM ducts at El6 (Fig. 1) and in renal tubules at El9 (data not shown). This may not be a mere co-incidence, as there is evidence that renal/pancreatic kallikrein may have a role in the regulation of electrolyte transport in renal tubules (Cuthbert et al., 1985; Croxatto et al., 1986; Cuthbert and MacVinish, 1986) and such a system may also operate in other epithelial tubules, such as salivary and nasal ducts. In conclusion, we have shown that renal kallikrein is synthesized in mature transporting epithelia of a number of tissues during fetal life in mice. Furthermore, the binding of bradykinin at El8 in renal tubules and synthesis of renal kallikrein by the fetal kidney at much the same fetal age suggests that the renal kallikrein-kinin system is operating during late gestation. Acknowledgements
We thank Dr. Jim Haralambidis, Ms. Suzie Khoury, Dr. Ross Fernley and Ms. Elizabeth Cooper for preparation and labelling of probes and ligands. This work was supported by grantsin-aid from the National Health and Medical Research Council of Australia, the Myer Family Trusts, the Ian Potter Foundation and the Howard Florey Biomedical Foundation. References Berg, T., Carretero, O.A., Scicli, A.G., Tilley, B. and Stewart, J.M. (1989) Hypertension 14, 73-80. Carretero, O.A. and Scicli, G.A. (1989) in Endocrine Mechanisms in Hypertension (Laragh, J.H., Brenner, B.M. and Kaplan, N.M., eds.), pp. 219-239, Raven, New York. Clements, J.A. (1989) Endocr. Rev. 10, 393-419.
Clements, J.A., Matheson, B.A. and Funder, J.W. (1990) J. Biol. Chem. 265, 1077-1081. Croxatto, H.R., Corthorn, J., Roblero, J., Villalo, P. and Perez, F. (1986) Am. J. Physiol. 250, F400-F406. Cuthbert, A.W. and MacVinish, L.J. (1986) Adv. Exp. Med. Biol. 198A, 203-21 I. Cuthbert, A.W., George, A.M. and MacVinish, L. (1985) Am. J. Physiol. 249, F439-F447. Evans, B.A., Drinkwater, C.C. and Richards, R.I. (1987) J. Biol. Chem. 262, 8027-8034. Gelly, J.L., Richoux, J.P., Grignon, G., Bouhnik, J., Baussant, T., Alhenc-Gelas, F. and Corvol, P. (1991) Histochemistry 96, 7-12. Heymann, M.A., Rudolph, A.M., Nies, A.S. and Melmon, K.L. (1969) Circ. Res. 25, 521-534. Jacoby, F. and Leeson, C.R. (1959) J. Anat. (London) 93, 201-216. Kato, H., Kei-ichi, E., Miyata, T., Hayashi, I., Oh-ishi, S. and Iwanaga, S. (1985) Biochem. Biophys. Res. Commun. 127, 289-295. Lawton, W.J., Proud, D., Frech, M.E., Pierce, J.V., Keiser, H.R. and Pisano, J.J. (1981) Biochem. Pharmacol. 30, 1731-1737. Leeson, C.R. and Booth, W.G. (1961) J. Dent. Res. 40, 838-845. Lewis, R.E., Childers, S.R. and Phillips, MI. (1985) Brain Res. 346, 263-272. Lewis, S.A. and Alles, W.P. (1986) Proc. Natl. Acad. Sci. USA 83, 5345-5348. Martinez, J.R. (1987) J. Dent. Res. 66 (Special Issue), 638-647. Miiller-Esterl, W., Iwanaga, S. and Nakanishi, S (19861 Trends Biochem. Sci. 11, 336-339. Obika, L.F.O. (1989) Clin. Sci. 77, 21-27. Penschow, J.D., Haralambidis, J., Pownall, S. and Coghlan, J.P. (1989) Methods Neurosci. 1, 222-238. Penschow, J.D., Drinkwater, C.C., Haralambidis, J. and CoghIan, J.P. (1991) Mol. Cell Endocrinol. 81, 135-146. Redman, R.S. and Sreebny, L.M. (1970) Anat. Rec. 168, 127-138. Riegman, P.H.J., Vlietstra, R.J., van der Korput, H.A.G.M., Romijn, J.C. and Trapman, J. (1991) Mol. Cell Endocrinol. 76, 181-190. Robillard, J.E., Lawton, W.J., Weisman, D.N. and Sessions, C. (1982) Kidney Int. 22, 594-601. Schneyer, C.A. and Schneyer, L.H. (1961) Am. J. Physiol. 201, 939-942. Schull, G.E., Greeb, J. and Lingrel, J.B. (1986) Biochemistry 25, 8125-8132. Theiler, K. (1972) The House Mouse, Springer, Berlin. van Leeuwen, B.H., Evans, B.A., Tregear, G.W. and Richards, RI. (1986) J. Biol. Chem. 261, 5529-5535. Wines, D.R., Brady, J.M., Southard, E.M. and MacDonald, R.J. (1991) J. Mol. Evol. 32, 476-492. Young, J.A. and van Lcnnep, E.W. (1979) in Membrane Transport in Biology (Giebisch, G., Tosteson, D.C. and Ussing, H.H., eds.), pp. 563-613, Springer, Berlin. Young, R.M., Schull, G.E. and Lingrel, J.B. (1987) J. Biol. Chem. 262, 4905-4910.