Regulation of Expression of Sprouty Isoforms by EGF, FGF7 Or FGF10 in Fetal Mouse Submandibular Glands

Regulation of Expression of Sprouty Isoforms by EGF, FGF7 or FGF10 in Fetal Mouse Submandibular Glands

Kenji Ohno1, Noriko Koyama2, Toru Hayashi2, Yoshiaki Takai1, Edward W Gresik3＊ and Masanori Kashimata2

1Department of Maxillofacial Surgery, Asahi University School of Dentistry 2Department of Pharmacology, Asahi University School of Dentistry 3Department of Cell Biology and Anatomy, The Sophie Davis School of Biomedical Education, The City University of New York

Abstract: Branching morphogenesis of the fetal mouse submandibular gland (SMG) is regulated by signaling through the ErbB and FGF families of tyrosine kinase receptors, whose members activate the ERK-1/2 pathway. The four Sprouty (Spry) proteins are inhibitory modulators of ERK-1/2. There is little information on their expression during pre- and postnatal development of the SMG. Qualitative RT-PCR detected mRNAs for Spry1, 2, and 4 from embryonic day 13 (E13) through postnatal day 7 (P7), but only trace amounts of Spry1 and 2 in adult SMGs. More sensitive quantitative RT-PCR revealed that transcripts for all four Spry isoforms are expressed, and each shows individual patterns of variation across fetal and early postnatal stages, and that there are very low levels of Spry1 and 2, but no Spry3 and 4, in adult glands. EGF, FGF7 and FGF10 upregulate expression of mRNA for Spry1, but only FGF7 upregulates Spry2 mRNA. EGF strongly induces an activating phosphorylation of all four Spry isoforms, but both FGFs do so only minimally. Quanti- tative RT-PCR of samples collected by laser capture microdissection showed that transcripts for Spry1 are conﬁned to the epithelium of E13 SMG rudiments. The isoform-speciﬁc temporal variation in the patterns of expression of Spry1, 2, 3 and 4 suggests a potentially important role for these negative modulators of growth-factor driven ras/ERK-1/2 signaling at stages when the SMG is most actively undergoing branching morphogenesis.

Key words: submandibular gland, Sprouty, EGF, FGF7, FGF10

Introduction Received 1/6/10; revised 8/30/10; accepted 9/9/10. Branching morphogenesis (BrM) is one of the Grant support: National Institutes of Health (NIDCR) ; DE10858 and KAKEN; 22592054. basic processes for the formation of many organs, Requests for reprints: Masanori Kashimata, Department of such as the lungs, the kidneys, and all exocrine Pharmacology, Asahi University School of Dentistry 1851-1 Hozu- glands including the three major salivary glands. mi, Mizuno, Gifu 501-0296, Japan, Phone: ＋81-58-329-1431, Fax: ＋81-58-329-1432 The submandibular gland (SMG) of the fetal ＊ Author to whom correspondence should be addressed: mouse is a widely used model system to study Edward W Gresik, PhD, Department of Cell Biology and Anato- my, The Sophie Davis School of Biomedical Education, The City BrM, since it is capable of undergoing branching University of New York, Harris Hall, Rm 310E, 138th St and Con- in vitro1,2. The mouse SMG initiates development vent Ave, New York, NY, 10031, USA, Phone: 212 650-6857, Fax: 212 650-6812 on day 12 of gestation (E12; the day of discovery 48 Oral Science International Vol. 7, No. 2 of vaginal plug was taken as E0) as a downgrowth factor systems regulating development of this of the oral epithelium into the underlying mesen- gland, we have examined expression of the four chyme, forming an elongated stalk ending in an Spry isoforms across the entire time span of the expanded endbud. A cleft develops on the distal gland's fetal development, and in early postnatal tip of the endbud, and the epithelium bifurcates life and in adults. as two epithelial stalks elongate from this site and again end in expanded endbuds. This process of Materials and Methods clefting and elongation of the epithelium is done repeatedly until the final appropriate extent of 1. Materials branching is reached. Eventually the elongated Recombinant EGF, FGF7, and FGF10 were pur- portions form lumens and become ducts, and the chased from R & D Systems Inc. (Minneapolis, endpieces differentiate into acini. This process is MN, USA). Anti-Spry1, 2, 3 and 4 antibodies were dependent on epitheliomesenchymal interactions from Santa Cruz Biotechnology (Santa Cruz, CA, mediated by integrins and extracellular matrix, USA), and Protein G magnetic beads were from and growth factors and their receptors3,4. New England Biolabs (Ipswich, MA, USA). Several growth factors are known to be required for full development of the fetal SMG, including 2. Organ culture EGF5, FGF6, and PDGF7. The members of all Pregnant mice (ICR strain) were purchased three of these families of ligands activate mem- from Japan SLC Inc. (Hamamatsu, Japan). The brane spanning receptors that are tyrosine day of discovery of the vaginal plug was taken as kinases (RTKs), that trigger intracellular signal- embryonic day 0 (E0). The SMGs were removed ing cascades8. The most commonly used pathways from prenatal (E13 to E18) and postnatal (7 days are those leading to activation of ERK-1/2, PLCγ1, and adult) mice. and PI3K/Akt. All three of these signaling cas- SMG rudiments were cultured as described pre- settes have been shown to be important for BrM of viously17. Briefly, approximately 20 SMGs were the fetal mouse SMG9-11. taken at E13 (one pregnant mouse), and placed on Sprouty (Spry) was first discovered as an antag- a membrane filter (Nuclepore membrane, 0.1 μm onist of fibroblast growth factor (FGF) signaling in pore size; Whatman International, Brantford UK) branching of Drosophila airway, and of epidermal floating on 400 μl of DMEM/F-12 medium (GIBCO growth factor (EGF) signaling in eye and wing Invitrogen Cell Culture, Carlsbad, CA, USA) con- development via inhibition of the Ras/Raf/MEK/ taining 100 units/ml penicillin (GIBCO Invitrogen ERK pathway12. There are four member of the Cell Culture), 100 μg/ml streptomycin (GIBCO Spry family called Spry1, 2, 3, and 4, and they Invitrogen Cell Culture), 150 μg/ml vitamin C have been shown to be general inhibitors of RTK (Sigma, St. Louis, MO, USA), and 50 μg/ml trans- signaling mediated by the Ras/Raf/MEK/ERK ferrin (Sigma). The SMGs were precultured for 3 pathway13,14. It is known that the phosphorylation hrs on the filters in organ-culture dishes (Falcon of tyrosine residues of Spry molecules is required Becton and Dickinson Labware, Franklin Lake, NJ, for their activation. However, depending on the USA) ; then EGF (50 ng/ml), FGF7 (200 ng/ml) phosphorylation site, the functions of the individ- or FGF10 (500 ng/ml) was added to the medium, 15 ual Spry proteins vary . and the rudiments were cultured in 5％ CO2/95％ Expression of Spry genes 1, 2, and 4 has been air and 80％ humidity at 37℃. The experiments detected in the perinatal mouse SMG epithelium were repeated five times (5 pregnant mice were by in situ hybridization16. However, nothing fur- used). ther is known about these proteins in the develop- All procedures of animal handling were con- ment of the SMG. To further analyze possible ducted in accordance with the Guideline for roles for these negative modulators of the ras/ Experimental Animals of Asahi University (Proto- ERK-1/2 pathway activated by many of the growth col No. 07-011). November, 2010 Spry Isoforms in Fetal Mouse SMG 49

3. Total RNA preparation and qualitative mouse Spry1, forward 5'-GAGGTTCTTACTC RT-PCR GGATAAC-3' and reverse 5'-GGTCCAGTCGT Total RNA was extracted from SMGs using TRI AACAGC-3'; mouse Spry2, forward 5'-AAAGC Reagent (Sigma) following the manufacturer's CGACATAGCATTATTAATCA-3' and reverse instructions. The following numbers of SMGs 5'-GCAAACCACAAATCCCCTA-3'; mouse Spry3, were used, with 20-34 SMG rudiments obtained forward 5'-AGCCAGTCTAGTATTTCCAGCT-3' per liter: E13, 6 pregnant mice; E14, 2 pregnant and reverse 5'-AGAGCACCATCGACCTTTG-3'; mice; E15-P7, one pregnant mouse at each age; mouse Spry4, forward 5'-CTCCTAGATCACAT Ad, 3 mice. Total RNA extracts were then dis- GGCTCC-3' and reverse 5'-GTGCTTGTCTAGCTC solved in nuclease-free H2O and digested by TGGC-3'; and mouse β-actin, forward 5'-GTGAC DNase I to remove possible DNA contamination. GTTGACATCCGTAAAGA-3' and reverse 5'- The total RNA (3 μg) was reverse transcribed to CTCAGGAGGAGCAATGATCTT-3'. The PCR cDNA by a Super Script First Strand Synthesis reactions were carried out for 40 cycles at an ini- System for RT-PCR (GIBCO Invitrogen Cell Cul- tial denaturing temperature of 95℃ for 5 min and ture) as described previously5. PCR was carried a final extension at 72℃ for 10 min. Each cycle out in triplicate with the following primers: had a denaturing temperature of 94℃ for 10 sec, mouse Spry1 (Acc#; NM_011896), forward 5'- an annealing temperature specific for each primer GAAAGGACTCATGAAATCATACC-3' and reverse pair for 30 sec, and an extension temperature of 5'-GTTATCCGAGTAAGAACCTC-3'; mouse Spry2 72℃ for 30 sec. The annealing temperatures were (Acc#; NM_011897), forward 5'-AAGCCGACATA 56℃ for Spry1, 2, 3 and 4, 52℃ for β-actin. The GCATTATTAATCA-3' and reverse 5'-TGAATT reactions were carried out in a DNA Engine Opti- GTAGCAAGCGTATCA-3'; mouse Spry3 (Acc#; con System (Bio-Rad) and results were analyzed NM_001030293), forward 5'-CGCTCTACTCACG with Opticon Monitor Analysis Software (Bio-Rad). CTAGCA-3' and reverse 5'-CCTTCAGAGCACCA The amplification amount of each Spry reaction TCGAC-3'; mouse Spry4 (Acc#; NM_011898), for- was compared to the amount of β-actin ampli- ward 5'-GGCCTGTGGAAAGTGTAAGT-3' and fication as an internal standard, and the final reverse 5'-CCGGTTTCTTCCCAATCTAGG'; and results were expressed as the ratio of these two mouse β-actin (Acc#; NM_007393), forward 5'- amounts. GTGACGTTGACATCCGTAAAGA-3' and reverse 5'-CTCAGGAGGAGCAATGATCTT-3'. The PCR 5. Metabolic labeling of Spry proteins with reactions were carried out for 35 cycles at an ini- 32P tial denaturing temperature of 95℃ for 5 min and E13 SMG rudiments were collected from four a final extension at 72℃ for 10 min. Each cycle pregnant mice (128 SMGs) and preloaded with had a denaturing temperature of 94℃ for 1 min, radioisotope by culture under the above condi- an annealing temperature specific for each primer tions for 2.5 h in medium containing 0.5 mCi/ml pair for 1 min, and an extension temperature of 32P-orthophosphoric acid, before addition of EGF, 72℃ for 2 min. The annealing temperatures were FGF7 or FGF10 at the concentrations used above, 50℃ for Spry1 and 2, 58℃ for Spry3 and β-actin, and culture was continued for 30 min. Incubation and 54℃ for Spry4. was terminated by washing the tissues on the fil-

ters with ice-cold PBS containing 1 mM Na3V04, 4. Quantitative RT-PCR and the SMG rudiments were collected and For each PCR tube, 1.0 μl cDNA from the above homogenized in 50 mM Tris-HCl (pH 7.4), 150 mM reverse transcriptase reaction mixture (Super- NaCl, 1 mM EDTA, 1％ Triton X-100, 2.5 mM Na- script) was mixed with 0.5 pmole primers (forward pyrophosphate, 1 mM β-glycerophosphate, pro- and reverse) and 12.5 μl iQ SYBR Green Super- tease inhibitor cocktail (Sigma) and phosphatase mix (Bio-Rad Laboratories, Hercules, CA, USA). inhibitor cocktail 2 (Sigma), and supernatants Primer pairs and temperatures were as follows: were processed for immunoprecipitation, as 50 Oral Science International Vol. 7, No. 2 described previously5. Equal amounts of radioac- Results tivity (1.77 × 105 cpm) were used for immunoprecipitation, and the precipitates were resolved by 1. RT-PCR analyses of Spry isoforms SDS-polyacrylamide gel electrophoresis (SDS- Qualitative RT-PCR of total RNA from intact PAGE) and the gels were dried and exposed to submandibular glands from embryonic day 13 to X-ray film. 18 (E13 to E18), and from newborns (NB), postnatal day 7 (P7), and adults detected mRNAs for 6. Laser capture microdissection Spry1, 2, and 4, but not for Spry3 (Fig. 1). SMG rudiments were collected from E13 mice Quantitative real-time RT-PCR confirmed the (96 SMGs) and embedded into OCT compound, presence of transcripts for Spry1, 2, and 4 and and frozen sections (8 μm) were cut in a cryostat. also detected the mRNAs for Spry3 (Fig. 2). The The SMG sections were stained with crystal violet (LCM Staining Kit, Ambion). From each section, samples were collected from three epithelial sites and one mesenchymal site, each containing about 500 cells, using Laser Capture Microdissection (PALM-MBIII, P.A.L.M. Microlaser Technologies AG, Germany). The three epithelial samples were from (1) ducts, (2) endbud epithelium in contact ( ) with mesenchyme, and 3 inner epithelial cells Fig. 1 RT-PCR for Spry1, 2, and 4 in developing mouse of the endbud (see Fig. 6A). Total RNAs were pre- SMG. Total RNAs from SMGs of E13-E18, new pared from the captured cells using a RNeasy born (NB), postnatal day 7 (P7) and adult (Ad) were transcribed to cDNA, and the target se- Micro Kit (QIAGEN Sciences, Germantown, MD, quences were amplified with each specific primer USA), and then cDNA synthesis and real-time RT- for target sequences. The PCR products were subjected to agarose gel electrophoresis and visual- PCR were performed as described above. ized using ethidium bromide. The experiments were repeated three times.

Fig. 2 Quantitative analyses of Spry mRNAs in developing mouse SMG by real-time RT-PCR. The ampliﬁcation amount of each of the Spry cDNAs (A: Spry1, B: Spry2, C: Spry3, and D: Spry4) which were synthesized from E13-E18, new born (NB), postnatal day 7 (P7) and adult (Ad) mouse SMGs using reverse transcriptase were analyzed. Amount of each Spry gene was expressed as the ratio of the amounts of Spry and β-actin cDNAs. The experiments were repeated three times, with four SMGs per age group. November, 2010 Spry Isoforms in Fetal Mouse SMG 51

Fig. 3 Effect of EGF, FGF7 or FGF10 on morphogenesis of E13 SMG in vitro. E13 SMG rudiments were cultured with none (A), EGF (B, 50 ng/ml), Fig. 4 Upregulation of Spry1 and Spry2 gene expressions FGF7 (C, 200 ng/ml) or FGF10 (D, 500 ng/ml). in cultured E13 SMG by administration of EGF, The photographs were taken after 48 hrs cul- FGF7 or FGF10. E13 SMG rudiments were culture. Original magnification 60 ×. Bar, 0.5 mm. tured for 24 hrs with none (control), EGF (20 ng/ This experiment was repeated five times. ml), FGF7 (200 ng/ml) or FGF10 (500 ng/ml). Amounts of Spry1 and 2 genes were expressed as the ratio of amounts of Sprys and β-actin cDNA. The experiments were repeated three times, with pattern of expression is specific for the mRNA of 5 SMGs per group. ＊ p ＜ 0.05 and ＊＊ p ＜ 0.01 as each isoform across the course of the gland's compared with the control value; Student's t-test. development. Spry1 mRNA is expressed at an unchanging level from E13 through P7, but is dra- matically reduced in adult glands (Fig. 2A). Lev- receptors. As previously reported by us and oth- els of Spry2 transcripts are high at E13 and E14, ers5,18,19, EGF increases branching morphogenesis then decline from E15 to birth, rise again at P7, in E13 SMGs cultured in vitro (Fig. 3A, B), but the and are lowest in adult glands (Fig. 2B). Spry3 two FGF have only a slight or minimal effect on transcripts are highest at E13, but remain low this process in this gland (Fig. 3C, D). from E14 to birth; there is a slight increase at P7, but they are essentially absent in adults (Fig. 2C). 3. Inductions of transcription and phospho- Spry4 mRNA levels are highest at E13 and gradu- rylation of Spry mRNA by growth factors ally decline thereafter, and are apparently absent We found, furthermore, that all three of these in the adult (Fig. 2D). The ranges of CT (CT ＝ ligands of these two RTK families significantly 25-29) values for Spry1 and Spry2 are lower than upregulate expression of Spry1 and Spry2 those of Spry3 and Spry4. mRNAs, with FGF7 having the strongest effect (Fig. 4A, B). Spry proteins must be phosphory- 2. Morphological changes by three major lated in order for them to inhibit Raf-ERK-1/2 sig- growth factors naling20. However, levels of Spry proteins are very Spry proteins are specific inhibitors of the Raf- low, and are very difficult to detect by immunob- ERK-1/2 intracellular signaling cascade, which is lotting studies. To determine if these proteins are activated by many receptor tyrosine kinases, phosphorylated in response to EGF or FGF7 or including both the ErbB and FGFR families of FGF10, we exposed E13 SMGs to these growth 52 Oral Science International Vol. 7, No. 2

Fig. 5 Phosphorylation of Spry proteins in cultured E13 SMG stimulated by EGF, FGF7 or FGF10. E13 SMG rudiments were precultured with 0.5 mCi/ml 32P-orthophosphoric acid, and then culture was continued with no addition (control), EGF (20 ng/ Fig. 6 Preparation of total RNA from the cells of E13 ml), FGF7 (200 ng/ml) or FGF10 (500 ng/ml). After SMG rudiments using laser capture microdissec- 30 min stimulation, SMGs were collected and ho- tion and estimation of the amount of Spry1 mRNA mogenized. Phosphorylated Spry protein in the by real-time RT-PCR. (A) From the stained SMG homogenates was immunoprecipitated using each sections, samples were collected from one mesen- specific Spry antibody and subjected to SDS- chymal site and three epithelial sites: (1) ducts, (2) 32 PAGE. The P-radiolabeled proteins in the dried endbud epithelium in contact with mesenchyme, gels were detected by exposure to X-ray film and (3) inner epithelial cells of the endbud. (B) (Kodak XAR). Phosphorylated Spry1-4 proteins The extracted total RNAs were reverse tran- were estimated at about 35 kD by SDS-PAGE scribed to cDNA and the amount of each Spry analysis. gene was analyzed by real-time RT-PCR. The amount of Spry1 gene was expressed as the ratio of amounts of Spry1 and β-actin cDNAs. The ex- factors in the presence of 32P-orthophosphate. The periments were repeated four times. ＊＊p ＜ 0.01 as compared with the value of ductal epithelium supernatants of these glands were then subjected (lowest value); Student's t-test. to immunoprecipitation with antibodies specific for each of the four Spry isoforms, and the immu- noprecipitates were resolved by electrophoresis tion by immunocytochemical techniques (data not and transferred to membranes for autoradiogra- shown). Thus, we probed the cellular localization phy. The radioactive tag greatly improved the by laser capture microdissection and RT-PCR. sensitivity of these immunoprecipitation tech- Samples were chosen from four sites (Fig. 6A): the niques, and all four Spry isoforms were then epithelium of ducts, the outer epithelium of the detectable in the fetal SMGs (Fig. 5). EGF had endbud in contact with mesenchyme, the inner the greatest effect and strongly phosphorylated all endbud epithelium, and the mesenchyme. All four Spry proteins. FGF7 mildly stimulated three epithelial compartments showed the same Spry2, 3 and 4, and FGF10 had a slightly greater amount of Spry1 transcripts, but the mesenchyme effect on Spry1 and 3. had only negligible message (Fig. 6B).

4. Localization of Spry transcript in E13 SMG Discussion rudiments The present findings demonstrate that the four Since levels of the Spry proteins are so low, we members of the Spry family are expressed in the were unable to demonstrate their tissue localiza- fetal SMG, that mRNAs for Spry1, 2 and 4 are November, 2010 Spry Isoforms in Fetal Mouse SMG 53 present at higher levels than the mRNA for Spry3, that mRNAs for Spry1 and Spry2 were signifi- and that EGF or FGF7 or 10 not only upregulate cantly increased by the administration of EGF, expression of these genes, but also can stimulate FGF7 or FGF10 in E13 SMG rudiments. These phosphorylation and activation of these proteins. data suggest complex regulatory interactions The present study is the first to measure the between these growth factors and these downregu- amounts of mRNAs for Spry1, 2, 3, and 4 that lators of ERK-1/2 signaling during organogenesis are endogenously expressed at all developmental of fetal mouse SMG. stages in the mouse SMG. The results of real- Other investigators using cultured cell systems time RT-PCR showed that mRNAs for Spry1, 2, 3, found that several growth factors were able to and 4 were expressed in the SMG rudiments from stimulate the phosphorylation of tyrosine residues E13 to E18 and newborn, postnatal 7 days and of Spry proteins, and these phosphorylations are adult mice. All Spry mRNAs were decreased at known to be required to activate the inhibitory adult ages. Because mRNAs of Spry1, 2, 3, and 4 functions of Spry proteins. In this study, we are highly expressed in fetal and early postnatal showed that all three growth factors (EGF, FGF7 stages, they most likely are involved mainly in and FGF10) stimulated phosphorylation of Spry1, modulating developmental processes of fetal 2, 3, and 4. EGF was the strongest stimulator for mouse SMGs. the phosphorylation of Sprys, whereas the two Spry transcripts are known to localize to epithe- FGFs were weaker stimulators, and had no effect lium in developing mouse SMG, lung, kidney, and on Spry2. These results imply that the modulat- tooth by in situ hybridization16. The transcripts ing effects of Sprys on ERK signaling would vary were most strongly expressed in the distal leading depending on which growth factor is active in trig- tips of the epithelial endbuds of the SMG and gering this intracellular signaling cascade, and lung, but were not seen in the mesenchyme. We may be relevant in explaining differences in mor- collected total RNA from epithelial and mesenchy- phogenesis events in fetal mouse SMGs induced mal tissues from E13 SMG sections using Laser by different growth factors, such as EGF or the Capture Microdissection, and we showed that FGFs. Spry1 is highly expressed in epithelium by real- There results show that the levels of expression time RT-PCR, but not in the mesenchyme, consis- and/or phosphorylation of Sprys are regulated by tent with Zhang et al.16 the growth factors, and that the Spry isoforms Mailleux et al.21 reported that both Spry2 mRNA may function as modulators of development of the expression and the size of the epithelial endbuds mouse SMG through modulating the ras/ERK-1/2 are increased in E12 lung rudiments cultured in signaling network. Given that Spry proteins act the presence of FGF10. Spry2 most likely regu- as inhibitory modulators of the Raf-ERK-1/2 sig- lates organogenesis of the lung by suppressing naling pathway that is involved in many networks ERK signals stimulated by FGF10. Mai et al.22 triggered by many RTKs responsive to a wide reported agenesis of the SMG and lacrimal glands variety of growth factors, it is at present not possi- in mice with a missense mutation in the Fgfr2 ble to define the precise roles that Spry proteins gene. Patel et al.23 found increased levels of the play in the fetal mouse SMG, but their presence mRNAs for Spry1 and 2, but not Spry4, in mesen- unquestionably has functional importance since chyme-free fetal SMG epithelium in Matrigel cul- the levels of individual Spry isoforms differ from tured in the presence of a synthetic sulfated oligo- each other across development. The observation saccharide, which presumably forms a complex that the growth factors used in this study resulted with FGFs to activate the FGFR1b. both in increased phosphorylation and activation However, there has been no information about of ERK1/2, as well as increased expression of the the actual amounts of the mRNAs for Spry iso- Spry isoforms, suggests that the latter may be act- forms induced by direct stimulation by growth fac- ing as modulators in negative feedback loops in tors in the developing mouse SMG. We showed this signaling cascade. There are no known inhib- 54 Oral Science International Vol. 7, No. 2 itors of the Spry family members, so an approach 1506-1507, 2004. using siRNAs would likely be most effective in 9. Kashimata M., Sayeed S., Ka A., Onetti-Muda A., Sakag- pursuing these investigations17. Indeed, by using ami H., Faraggiana T., and Gresik E.W.: The ERK-1/2 sig- knockout or transgenic mice for Spry1 it has been naling pathway is involved in the stimulation of branching demonstrated that this isoform acts to oppose and morphogenesis of fetal mouse submandibular glands by balance the stimulatory effects of GDNF on EGF. Dev Biol 220：183-196, 2000. 10. Koyama N., Kashimata M., Sakashita H., Sakagami H., branching morphogenesis in the developing kid- and Gresik E.W.: EGF-stimulated signaling by means of ney24. It is reasonable to suggest that the Spry PI3K, PLCγ1, and PKC isozymes regulates branching family of proteins may play similar roles in the morphogenesis of the fetal mouse submandibular gland. developing SMG. Dev Dyn 227：216-226, 2003. 11. Larsen M., Hoffman M.P., Sakai T., Neibaur J.C., Mitchell

Acknowledgment J.M., and Yamada K.M.: Role of PI 3-kinase and PIP3 in This study was supported by KAKEN (22592054) and submandibular gland branching morphogenesis. Dev Biol National Institutes of Health (NIDCR) (DE10858) grants. 255：178-191, 2003. 12. Hacohen N., Kramer S., Sutherland D., Hiromi Y., and

References Krasnow M.A.: sprouty encodes a novel antagonist of FGF 1. Borghese E.: The development in vitro of the submandibu- signaling that patterns apical branching of the Drosophila lar and sublingual glands of Mus musculus. J Anat 84： airways. Cell 92：253-263,1998. 287-302,1950. 13. Reich A., Sapir A., and Shilo B.Z.: Sprouty is a general 2. Spooner B., Thompson-Pletscher H., Stokes B., and Bas- inhibitor of receptor tyrosine kinase signaling. Develop- sett K.: Extracellular matrix involvement in epithelial ment 126：4139-4147, 1999. branching morphogenesis. Developmental Biology. Vol 3. 14. Kim H.J., Taylor L.J., and Bar-Sagi D.: Spatial regulation (Steinberg, M. (ed)) Plenum Press, New York, 1986, pp. of EGFR signaling by Sprouty 2. Curr Biol 17：455-461, 225-260. 2007. 3. Kadoya Y., and Yamashina S.: Salivary gland morphogen- 15. Rubin C., Zwang Y., Vaisman N., Ron D., and Yarden Y.: esis and basement membranes. Anat Sci Int 80：71-79, Phosphorylation of carboxyl-terminal tyrosines modulates 2005. the specificity of Sprouty-2 inhibition of different signaling 4. Patel V.N., Rebustini I.T., and Hoffman M.P.: Salivary pathways. J Biol Chem 280：9735-9744, 2005. gland branching morphogenesis. Differentiation 74： 16. Zhang S., Lin Y., Itaranta P., Yagi A., and Vainio S.: 349-364, 2006. Expression of Sprouty genes 1, 2, and 4 during mouse 5. Kashimata M., and Gresik E.W.: Epidermal growth factor organogenesis. Mech Dev 109：367-370, 2001. system is a physiological regulator of development of the 17. Koyama N., Hayashi T., Ohno K., Gresik E.W., and mouse fetal submandibular gland and regulates expres- Kashimata M.: Signaling pathways activated by EGFR or sion of the alpha-6-integrin subunit. Dev Dyn 208：149- FGFR differentially regulate branching morphogenesis in 161, 1997. fetal mouse submandibular glands. Develop Growth Differ 6. Hoffman, M.P., Kidder B.L., Steinberg Z.L., Lakhani S., 50：565-576, 2008. Ho S., Kleinman H.K., and Larsen M.: Gene expression 18. Nogawa H., and Takahashi Y.: Substitution for mesen- profiles of mouse submandibular gland development: chyme by basement-membrane-like substratum and epi- FGFR1 regulates branching morphogenesis in vitro dermal growth factor in inducing branching morphogen- through BMP- and FGF-dependent mechanisms. Develop- esis of mouse salivary epithelium. Development 112： ment 129：5767-5778, 2002. 855-861, 1991. 7. Yamamoto S., Fukumoto E., Yoshizaki K., Iwamoto T., 19. Jaskoll T., and Melnick M.: Submandibular gland mor- Yamada A., Tanaka K., Suzuki H., Aizawa S., Arakaki M., phogenesis: stage-specific expression of TGF-alpha/EGF, Yuasa K., Oka K., Chai Y., Nonaka K., and Fukumoto S.: IGF, TGF-beta, TNF, and IL-6 signal transduction in nor- Platelet-derived growth factor receptor regulates salivary mal embryonic mice and the phenotypic effects of TGF- gland morphogenesis via fibroblast growth factor expres- beta2, TGF-beta3, and EGF-r null mutations. Anat Rec sion. J Biol Chem 283：23139-23149, 2008. 256：252-268, 1999. 8. Schlessinger J.: Common and distinct elements in cellular 20. Hanafusa H., Torii S., Yasunaga T., and Nishida E.: signaling via EGF and FGF receptors. Science 306： Sprouty1 and Sprouty2 provide a control mechanism for the Ras/MAPK signaling pathway. Nat Cell Biol 4：850- November, 2010 Spry Isoforms in Fetal Mouse SMG 55

858, 2002. Lassiter K.S., Sher I., Yates E.A., Turnbull J.E., Ron D., 21. Mailleux A.A., Tefft D., Ndiaye D., Itoh N., Thiery J.P., and Hoffman M.P.: Spediﬁc heparan sulfate structures Warburton D., and Bellusci S.: Evidence that SPROUTY2 modulate FGF10-mediated submandibular gland epithe- functions as an inhibitor of mouse embryonic lung growth lial morphogenesis and differentiation. J Biol Chem and morphogenesis. Mech Dev 102：81-94, 2001. 283：9308-9317, 2008. 22. Mai S., Wei K., Flenniken A., Adamson S.L., Rossant J., 24. Basson M.A., Watson-Johnson J., Shakya R., Akbulut S., Aubin J.E., and Gong S.G.: The missense mutation Hyink D., Costantini F.D., Wilson P.D., Mason I.J., and W290R in Fgfr2 causes developmental defects from aber- Licht J.D.: Branching morphogenesis of the ureteric epi- rant IIIb and IIIc signaling. Dev Dyn 239：1888-1900, thelium during kidney development is coordinated by the 2010. opposing functions of GDNF and Sprouty1. Dev Biol 23. Patel V.N., Likar K.M., Zisman-Rozen S., Cowherd S.N., 299：466-477, 2006.