Biochemical Pharmacology, Vol. 41, No. 12, pp. 1879-1886, 1991
`Printed in Great Britain.
`
`0006-2952/91 S3.00 + 0.03
`Pergamon Press plc
`
`STIMULATION OF SECRETION BY THE T84 COLONIC
`EPITHELIAL CELL LINE WITH DIETARY FLAVONOLS
`
`TOAN D. NGUYEN,* ANDREW T. CANADA,1' GREGORY G. HEINTZ, THOMAS W. GETTYS
`and JONATHAN A. COHN
`Departments of Medicine and t Anesthesiology, Duke University School of Medicine and Durham
`V.A. Medical Center, Durham, NC, U.S.A.
`
`(Received 26 July 1990; accepted 3 January 1991)
`
`Abstract—Flavonols are dietary compounds widely distributed in plants and characterized by a 2-
`phenyl-benzo(a)pyrane nucleus possessing hydroxyl and ketone groups at positions 3 and 4, respectively.
`Kaempferol, quercetin, and myricetin are flavonols that are further mono-, di-, or trihydroxylated on
`the phenyl ring, respectively. To test whether these ingested flavonols might exert a direct secretory
`effect on intestinal epithelial cells, monolayers of the T8,, colonocyte cell line were mounted in Ussing
`chambers and examined for ion transport response. Twenty minutes after addition of 100 µM quercetin
`to either the serosal or mucosal side, the short-circuit current change was maximal at 16.6 µA/cm2.
`Kaempferol was less potent than quercetin, while myricetin and glycosylated quercetin (rutin) did not
`induce secretion. The secretion induced by quercetin did not seem to be mediated by the reactive
`oxygen species generated by quercetin through auto-oxidation and/or redox cycling (superoxide,
`hydrogen peroxide, and the hydroxyl radical) because it was neither enhanced by iron, nor inhibited
`by desferroxamine B or catalase (alone or in combination with superoxide dismutase). Like vasoactive
`intestinal peptide, quercetin induced a secretory response that was inhibited by barium chloride and
`bumetanide, and which exhibited synergism with carbachol. Quercetin also stimulated a modest increase
`in intracellular cAMP levels and the phosphorylation of endogenous protein substrates for cAMP-
`dependent protein kinase. Thus, quercetin is a potent stimulus of colonocyte secretion that resembles
`secretagogues which act via a cAMP-mediated signaling pathway.
`
`Flavonoids constitute a class of compounds which
`contain the basic structural feature of a 2-phenyl-
`benzo(a)pyrane nucleus (Fig. 1). Either as free
`aglycones or more commonly glycosylated at carbons
`3,4, or 7, these compounds are universally distributed
`among vascular plants where they may serve as
`natural transport regulators for the plant growth
`substance auxin [1]. Flavonols are a subgroup of the
`flavonoids, characterized by a hydroxyl group at
`position 3 and a ketone at position 4. Flavonols can
`be further hydroxylated at positions 3', 4' or 5' on
`the # phenyl ring to yield the 4'-monohydroxy-
`flavonol kaempferol, the 3 ' ,4' -dihydroxy-flavonoi
`quercetin, and the 3' ,4' ,5'-trihydroxy-flavonol myr-
`icetin (Fig. 1). Both quercetin and myricetin produce
`reactive oxygen species (superoxide, hydrogen
`peroxide, and hydroxyl radical) through auto-
`oxidation and redox cycling [2-4]. Since reactive
`oxygen species may induce intestinal secretion [5, 6],
`we examined the possibility that flavonols might act
`on the intestinal epithelial cell to stimulate ion
`transport.
`In this study, the colonic epithelial cell line TM
`was used as a model to study the effects of flavonols
`on the enterocyte. These cells grow as well-
`differentiated monolayers which exhibit vectorial
`chloride secretion when mounted in Ussing chambers
`and exposed to a variety of neurohormonal stimuli
`[7]. Chloride secretion, monitored by a change in
`
`* Correspondence: Toan D. Nguyen, M.D., Building 2,
`2nd Floor, V.A. Medical Center, 508 Fulton St., Durham,
`NC 27705.
`
`the short-circuit current (Isc) necessary to nullify the
`potential difference across the cell monolayer, is
`stimulated by agents which increase cAMP, such as
`vasoactive intestinal peptide (VIP) or prostaglandin
`El [8, 9], and also by agents which act through Ca2+-
`mediated pathways, such as carbachol, histamine
`and calcium ionophores [10, 11]. Chloride secretion
`occurs through Cl- channels located on the apical
`membrane of confluent monolayers [12], and is the
`result of the Cl- electrochemical gradient generated
`by the concerted action of three transporters: the
`basolateral
`isia+,K+,C1-
`co-transporter,
`the
`Na+,K+-ATPase pump, and K+ channels [9]. The
`Na+ and K+
`imported
`into the cell by the
`co-transporter are recycled to the extracellular
`compartment, respectively, by the Na+ ,K+-ATPase
`pump and by at least two distinct K+ efflux channels
`(activated separately by VIP and by carbachol).
`Because the secretory response in T84 cells is well
`characterized and reflects the direct action of
`secretagogues on the enterocyte, we chose this model
`to characterize the effects of dietary flavonols on
`colonic secretion.
`
`MATERIALS AND METHODS
`
`Chemicals. Quercetin, kaempferol, myricetin,
`rutin, barium chloride, superoxide dismutase (SOD)
`(from bovine erythrocytes), catalase (from bovine
`liver), desferroxamine B, carbachol, and Fe(III)-
`EDTA were obtained from Sigma, St. Louis, MO.
`Bumetanide was provided by Biomol Research
`Laboratories, Plymouth Landing, PA. VIP was from
`Peninsula, Belmont, CA. 32P as inorganic phosphate
`
`1879
`
`MYLAN EXHIBIT - 1036
`Mylan Pharmaceuticals, Inc. v. Bausch Health Ireland, Ltd. - IPR2022-00722
`
`

`

`1880
`
`T. D. NGUYEN et al.
`
`penicillin, and 5000 µg/L streptomycin sulfate) to
`bathe both sides of the cell monolayer. The
`confluent monolayers used for secretory studies were
`maintained for at least 7 days after the filters were
`seeded.
`Secretory studies. The cell monolayers on the
`filter/ring units were mounted in a modified Ussing
`chamber as previously described [7], and both sides
`of the monolayer were bathed with a Ringer's
`solution containing 115 mM NaC1, 1.2 mM CaC12,
`1.2 mM MgC12, 0.4 mM KH2PO4, 2.5 mM K2HPO4,
`25 mM NaHCO3, and 10 mM glucose. Quercetin,
`myricetin, and kaempferol were dissolved in ethanol
`and added to the Ringer's solution at a 1 : 100 dilution
`(final concentrations of ethanol 1%). The medium
`was warmed to 37° with a circulating water jacket
`and gently mixed and oxygenated with a constant
`inflow of 95%O2/5%CO,. During secretory studies,
`spontaneous tissue potential differences were short-
`circuited by an automatic voltage clamp (model
`DVC-1000, World Precision Instruments, New
`Haven, CT) with Ag-AgCl2 electrodes, and the
`current necessary to maintain this short circuit (Ise)
`recorded at 1-min intervals. Instrument calibration
`was performed prior to each experiment using a
`filter/ring unit without cells. All comparative studies
`used matched pairs of monolayers seeded at the
`same time and studied concurrently. Spot checks of
`the T84 monolayers after completion of
`the
`experiments showed that cells exposed to 100 µM
`quercetin for > 1 hr could still exclude trypan blue.
`Cellular protein phosphorylation. T84 cell protein
`phosphorylation
`responses were studied using
`methods previously described [13]. Briefly, cells
`were labeled with 32P, in a phosphate-free buffer,
`exposed to 100 µM quercetin for 5 min, scraped from
`the filters, and homogenized with a glass—Teflon
`homogenizer. Phosphoproteins contained in the
`supernatant fraction after centrifugation at 436,000g
`for 15 min were precipitated with 10% trichloroacetic
`acid and resolved by two-dimensional polyacrylamide
`gel electrophoresis. Phosphoproteins were detected
`by autoradiography using X-OMAT AR5 film
`exposed at room temperature.
`cAMP Assay. T84 cells grown to confluence on
`filters were washed twice with Ringer's solution and
`immersed in 15 mL of Ringer's solution supplemented
`with 0.2 mM 3-isobutyl-1-methylxanthine and 10 mM
`glucose, warmed to 37°, and equilibrated with 95%
`O2, 5% CO2. Quercetin (final concentration 100 µM)
`or VIP (final concentration 1 nM) was added to the
`solution, and after different time periods, the
`filters containing the cells were transferred into
`12 mm x 75 mm plastic tubes and placed in liquid
`nitrogen. The cAMP was then extracted by boiling
`the cells for 7 min in 1 mL of 5 mM KH,PO4, 5 mM
`K2HPO4, 1 mM EDTA, and 0.1 mM 3-isobutyl-1-
`methylxanthine. The supernatant resulting from
`centrifugation at 15,000g for 7.5 min was assayed
`for cAMP according to the method described by
`Gettys et al. [14].
`
`RESULTS
`
`Stimulation of secretion with flavonols. As shown
`in Fig. 2, addition of 100µM quercetin to the mucosal
`
`A FLAVONOIDS
`
`B FLAVONOLS
`
`8
`
`0
`
`H
`
`lel
`
`
`
`0
`
`O -OH
`
`T
`
`OH
`
`fi
`
`0
`
`OH
`
`C KAEMPFEROL
`
`OH
`
`0 H
`
`0
`
`°H
`
`OH
`
`0
`
`I
`
`OH
`
`OUERCET1N
`
`0
`
`O
`
`E MYRICET1N
`
`T
`OH
`
`°H
`
`O
`Fig. 1. Chemical structures of flavonoids (A), flavonols
`(B), kaempferol (C), quercetin (D), and myricetin (E).
`
`was obtained from ICN, Irvine, CA. Culture medium
`was obtained from the Tissue Culture Facility of the
`University of North Carolina, Chapel Hill, NC, or
`from Gibco, Grand Island, NY.
`Growth and maintenance of T84 cells. T84 cells
`were provided by Dr. K. Dharmsathaphorn
`(University of California, San Diego). These cells
`were cultured at 5% CO2 and 37° in a 1 : 1 mixture
`of Dulbecco's modified Eagle's medium and Ham's
`F-12 medium supplemented with 5% (v/v) newborn
`calf serum. Cells were seeded onto collagen-coated
`Nuclepore filters previously glued onto plastic rings
`(surface area: 2.9 cm', approximately 106 cells/
`filter). These filters were then set on glass beads to
`allow medium (supplemented with 5000 units/L
`
`

`

`Stimulation of T8,4 cell secretion with flavonols
`25
`
`(A)
`
`20 -
`
`Ouercetin
`100µM
`
`15
`
`N
`
`E
`< 10
`
`U
`2
`
`5
`
`N
`
`U
`
`% Maximal Ise increase
`
`10
`
`20
`
`40
`30
`Time (min)
`
`50
`
`60
`
`20 Quereelin
`
`N
`
`15
`E v 1n
`
`100µM
`-*-- 300µM
`
`(C)
`
`10
`
`20
`
`40
`30
`Time (min)
`
`50
`
`60
`
`1881
`
`10µM
`100 ptM
`
`(B)
`
`Ouercean
`
`11,
`
`20
`
`15
`
`10
`
`5
`
`0
`
`0
`
`10
`
`20
`30
`Time (min)
`
`40
`
`100
`
`80
`
`60
`
`40
`
`20
`
`Ouereettn
`100µM
`
`1
`
`Mucosal
`Serosal
`
`(D)
`
`20
`40
`Time (min)
`
`60
`
`Fig. 2. Secretory effect of quercetin. TM monolayers were grown to confluence on semipermeable
`membranes, mounted on modified Ussing chambers, and stimulated with quercetin as described in
`Materials and Methods. The resulting Isc's were recorded every minute and the resulting means and
`SEM (alternated for clarity in panels C and D) shown. Panel A (top left): Incubation with 100 µM
`quercetin added to the mucosal compartment (1-36 min: N = 47; 37-66 min: N = 20-46). Panel B (top
`right): Incubation with either 10 or 100µM quercetin added to the mucosal compartment (three matched
`pairs). Panel C (bottom left): Incubation with either 100 or 300 µM quercetin added to the mucosal
`compartment (three matched pairs). Panel D (bottom right): Incubation with 100 µM quercetin added
`to either the serosal or the mucosal side of the T84 cell monolayer. Within each of the seven matched
`pairs, the Isc increases were normalized using the maximal Isc increase; the resulting means and SEM
`are shown [mean maximal Isc response: 53 ± 8.7 µA (18 µA/cm2)].
`
`side of the cell monolayer produced an increase in
`Isc which peaked after 15-20 min to a maximum
`value above baseline of 16.6 µA/cm2(SEM = 0.9 µA/
`cm2, N = 48, 48.1 µA/filter). The Isc was unaffected
`in control filters exposed to ethanol alone at a final
`concentration of 1%.
`A threshold response to quercetin was obtained
`at 10µM (Fig. 2B). The maximal Isc increase
`produced by 300 µM quercetin was the same as that
`produced by 100 µM, but the response was more
`rapid in onset and shorter in duration (Fig. 2C). As
`shown in Fig. 2D, quercetin stimulated a similar
`secretory response whether added to the serosal or
`the mucosal side of the monolayer. However, the
`maximal response elicited from the serosal side was
`only 68% of the maximal response obtained from
`the mucosal side (P = 0.01 for a smaller serosal
`response, paired two-tailed t-test with 6 df, mean
`maximal mucosal Isc response: 11.9 ± 1.9 pA/cm2).
`The effects of flavonols structurally related to
`quercetin were also investigated. The maximal Isc
`increase observed with 100 µM kaempferol was
`26 ± 2%© of the maximal increase obtained with
`100 µM quercetin (three paired experiments). For
`myricetin, secretion was detected only at 300 µM
`and not at 100 µM. A minimal response was obtained
`with a 300µM concentration of the glycosylated
`quercetin-rutinoside (rutin) (data not shown).
`
`Effects of modulators of the metabolism of reactive
`oxygen species on quercetin-stimulated chloride
`secretion. It is possible that the reactive oxygen
`species (superoxide, hydrogen peroxide, or hydroxyl
`radical) produced by quercetin upon auto-oxidation
`and/or redox cycling may mediate its secretory
`effect. In this case, the Isc response should be altered
`by compounds or enzymes which modulate the
`production or degradation of these species. Quer-
`cetin-induced secretion was therefore studied after
`cells were preincubated with superoxide dismutase
`(SOD) (to enhance the conversion of superoxide to
`hydrogen peroxide), catalase (to enhance
`the
`conversion of hydrogen peroxide to water and
`oxygen), iron (to facilitate the Haber—Weiss reaction
`in which hydrogen peroxide is converted to hydroxyl
`radical), or desferroxamine B (to chelate iron and
`prevent the Haber—Weiss reaction). If quercetin-
`induced secretion is mediated by hydrogen peroxide,
`then catalase should inhibit the secretory response
`to quercetin. Similarly, if secretion is mediated by
`hydroxyl radical, then the Isc response should be
`inhibited by desferroxamine and enhanced by Fe3+.
`Finally,
`if extracellular superoxide is the key
`mediator, then secretion should be blocked by the
`combination of SOD and catalase. As shown in
`Table 1, little effect on quercetin-induced secretion
`was observed after preincubation with 50 µM
`
`

`

`1882
`
`T. D. NGUYEN et al.
`
`Table 1. Effects of modulators of the metabolism of reactive oxygen species on quercetin-
`induced secretion in Ts, cells
`
`Agent
`
`SOD (470 units/mL) + catalase (450 units/mL)
`Catalase (450 units/mL)
`Desferroxamine B (50 µM)
`Fe' (50 µM)
`
`Maximal Isc
`(% control)
`
`104 ± 12.6
`117 ± 7.8
`97 ± 5.6
`75 ± 9.2
`
`N
`
`6
`6
`6
`6
`
`P (* 100)
`
`0.78
`0.08
`0.68
`0.04
`
`T84 cells mounted in modified Ussing chambers were incubated with the different modulators
`of reactive oxygen species metabolism for 10-15 min, and quercetin was added to a final
`concentration of 100 µM. In each matched pair, the maximal short-circuit current (Isc) change
`induced by quercetin in the presence of the modulator is expressed as the percentage of the
`maximal Isc change induced by quercetin in the absence of the modulator. Values are means
`± SEM. The mean maximal Isc changes in the control monolayers were, respectively,
`21.9 ± 3.3, 15 ± 2.5, 15.8 ± 1.66, and 18.3 ± 2.7 µA/cm2 for the experiments studying the
`effects of SOD plus catalase, catalase, desferroxamine B, and Fe'. P values were calculated
`using two-tailed t-tests with 5 df.
`
`desferroxamine B or a combination of 470 units/
`mL SOD and 450 units/mL catalase. A slight
`enhancement in secretion, which did not reach
`statistical significance (0.1 > P > 0.05), was noted
`with 450 units/mL catalase, while 50 µM Fe(III)-
`EDTA produced a significant (P < 0.05) inhibition of
`quercetin-induced secretion. However, as discussed
`previously, these last two effects were opposite of
`what was expected should either hydrogen peroxide
`or the hydroxyl radical mediate secretion. In the
`aggregate, these findings do not support a role for
`reactive oxygen species in the secretory response of
`quercetin.
`Intracellular mechanism of secretion. TR4 cell apical
`chloride secretion is dependent on the chloride
`gradient across the mucosal membrane of the cell.
`This gradient
`is generated by the basolateral
`Na+,K+,C1+ co-transporter with the imported Na+
`and K+ recycled outside the cell by the Na+,K+-
`ATPase pump and K+ efflux channels [15]. The role
`of these transport systems in quercetin-induced
`chloride secretion was studied using bumetanide,
`which inhibits the Na+,K+,C1- co-transporter, and
`barium chloride, which inhibits a VIP-responsive K*
`channel [10, 16]. Table 2 demonstrates the inhibitory
`effects of 0.3 mM bumetanide and 6 mM BaCl2 on
`the secretory response elicited by 100 µM quercetin.
`Compared with matched controls, the Isc response
`obtained 20 min after addition of quercetin was only
`19 ± 1% and 31 ± 5% of the expected response for
`bumetanide and barium chloride, respectively. Thus,
`quercetin-induced secretion appears to require active
`Na+,K+ ,C1- co-transport and K+ efflux mechanisms.
`One approach to determining which intracellular
`signalling pathway(s) mediates
`the quercetin
`secretory response is to evaluate whether quercetin
`exhibits synergism when administered with other
`secretagogues. The interactions between carbachol,
`VIP, and quercetin were therefore studied. In Fig.
`3A, cells were exposed either
`to carbachol
`(final concentration 10 µM) or to quercetin (final
`concentration 50µM). After 15 min, quercetin was
`added to the cells previously exposed to carbachol
`and vice versa. In these matched pairs, the effect of
`
`Table 2. Effects of ion transport inhibitors on quercetin-
`induced secretion in TM cells
`
`Agent
`
`Isc
`(% control) N P (* 100)
`
`Bumetanide (0.3 mM)
`Barium chloride (6 mM)
`
`19 ± 1
`31 ± 5
`
`4
`4
`
`0.0001
`0.0007
`
`T84 cells were incubated for 15 min with either bumetanide
`or barium chloride, and quercetin was added to a final
`concentration of 100 µM. In each matched pair, the Isc
`change induced by quercetin after 20 min in the presence
`of the inhibitor is expressed as the percentage of the Isc
`change induced by querecetin alone. Bumetanide was
`dissolved in 0.1 M NaOH and added to the cells at a 1 : 100
`dilution; the corresponding control also contained 1 : 100
`dilution of 0.1 M NaOH. Values are means ± SEM. The
`maximal Isc responses in the control monolayers were
`12.1 ± 0.59 and 13.4 ± 1.2 µA/cm', respectively, for the
`experiments studying the effects of bumetanide and barium
`chloride. P values were calculated using two-tailed t-tests
`with 3 df.
`
`quercetin followed by carbachol (or carbachol
`followed by quercetin) in one monolayer can be
`compared to the initial effect of quercetin (or
`carbachol) alone in the other. Analyzed in this
`fashion, quercetin alone produced an Isc increase of
`9.4 ± 0.4 µA/cm2 (27 ± 1.2 µA, N = 3) after 15 min,
`while carbachol produced an Isc
`increase of
`2.3 ± 1 µA/cm- (7 ± 3.5 µA, N = 3) after 3 min.
`Carbachol added to cells responding maximally to
`quercetin produced an additional Isc increase of
`34.9 ± 5 µA/cm2 (101 ± 14.4 µA) resulting
`in a
`combined total Isc increase of 44.4 ± 5 µA/cm2
`(129 ± 14.4 µA). Quercetin added to cells previously
`exposed to carbachol produced an Isc increase of
`19.1 ± 1.7 µA/cm2 (55 ± 4.8 µA). However, in the
`latter case, quercetin was added after completion of
`the carbachol response and the interaction between
`quercetin and carbachol may be suboptimal. In
`additional experiments,
`the secretory response
`produced by a simultaneous dose of carbachol and
`
`

`

`(a) C
`(o)
`
`(II O
`(0) C
`(%)O8C
`
`E
`
`50
`
`40
`
`30
`
`20
`
`10
`
`Stimulation of T84 cell secretion with
`
`flavonols
`
`1883
`
`(A)
`
`(0) VIP
`1.)
`
`30
`
`20
`
`o
`
`E a
`
`(o) O
`(0) VIP
`
`V
`
`(B)
`
`r
`10
`
`e
`20
`
`•
`
`I
`50
`
`60
`
`40
`20
`30
`Time (min)
`
`50
`
`60
`
`1
`1
`40
`30
`Time (min)
`Fig. 3. Interactions between quercetin, carbachol and VIP. Pairs of confluent T84 monolayers were
`mounted in modified Ussing chambers and their secretory responses (mean Isc and SEM) compared.
`Panel A (left panel): In each pair, one monolayer was exposed to 10 µM carbachol [C] (serosal surface)
`or to 50 µM quercetin [Q] (mucosal surface); after 15 min, 50 µM quercetin was added to the monolayer
`previously exposed to carbachol and vice versa. Additional monolayers were also exposed to a combined
`concentration of 10µM carbachol and 50µM quercetin [Q&C]. The following symbols are used: (II)
`quercetin followed by carbachol (N = 3), (O) carbachol followed by quercetin (N = 3), (A) quercetin
`and carbachol added simultaneously (N = 5). Panel B (right panel): In each pair, one monolayer was
`exposed to 1 nM VIP (serosal surface) and the other to 50 µM quercetin [O] (mucosal surface). After
`15 min, 50 µM quercetin was added to the monolayer previously exposed to VIP and vice versa. The
`following symbols are used: (0) quercetin followed by VIP (N = 3), (O) VIP followed by quercetin
`(N = 3).
`
`quercetin was also evaluated. A peak Isc with an
`intermediate value of 23.6 ± 2.2 µA/cm2 (69
`6.5 µA, N = 5) was obtained 14 min after addition
`of the combined secretagogues. In all the different
`sequences, the Isc increases produced by cells
`exposed to the combination of quercetin and
`carbachol were greater than the sum of the individual
`Isc changes produced by quercetin and carbachol
`(9.4 ± 2.3 = 11.7 µA/cm2). The difference in the
`degrees of synergism probably reflects the different
`timing of the maximal effect of quercetin and
`carbachol. In contrast, as shown in Fig. 3B, when
`the interaction between quercetin and VIP was
`analyzed in the same fashion, no such synergism,
`but a possible inhibitory effect was demonstrated.
`The synergism between carbachol and quercetin,
`but not between VIP and quercetin, suggests that
`quercetin may induce secretion through pathways
`related to the ones activated by VIP, but not by
`carbachol. This possibility was explored further in
`the following phosphorylation studies.
`Phosphorylation and intracellular cAMP studies.
`Previous studies have shown that T84 cells exhibit
`distinct phosphorylation responses to stimuli acting
`via cAMP or via Ca2+ [13, 17]. Phosphoproteins p83,
`p29, and p23 are examples of proteins exhibiting
`increased phosphorylation in cells stimulated by
`agents which act via Ca2+ , such as carbachol,
`histamine, and ionomycin. By contrast, phos-
`phoproteins p37, p18, and p23 exhibit increased
`phosphorylation in cells exposed to agents which act
`via cAMP, such as VIP and forskolin. Each of these
`five phosphoproteins showed increased labeling in
`monolayers stimulated with forskolin plus carbachol
`(Fig. 4, comparing panels A and B). When
`monolayers were stimulated with 100 µM quercetin,
`only three of these five phosphorylation responses
`were observed: quercetin stimulated the phos-
`phorylation of p37, p18, and p23, but not p29 or p83
`(Fig. 4, comparing panels C and D). These results
`
`suggest that quercetin activates intracellular signaling
`mechanisms mediated by cAMP, but not by Ca2+.
`In an attempt to study whether quercetin induces
`the generation of cAMP, this second messenger was
`measured in cells exposed to 100 µM quercetin for
`different time periods ranging from 1 to 20 min.
`Surprisingly, as shown in Table 3, only a modest
`increase in cAMP levels was detected. These
`increases were minimal when compared with the
`mean 85-fold increase in cAMP demonstrated with
`control monolayers exposed to 1 nM VIP for the
`same time periods (data not shown).
`
`DISCUSSION
`
`We have demonstrated that quercetin is a potent
`stimulator of ion transport in T84 cells. When added
`to either the mucosal or serosal side of the cell
`monolayer, quercetin produced a concentration-
`dependent increase in Isc, which peaked 15-20 min
`after the addition of 100 µM quercetin. The observed
`Isc response is consistent with the possibility that
`quercetin can act directly on the enterocyte to
`secretion. Of the related
`stimulate electrogenic
`hydroxylated flavonols, kaempferol was less potent
`than quercetin, while myricetin and the glycosylated
`quercetin-rutinoside (rutin) produced minimal res-
`ponses. Considering the effective mucosal—serosal
`barrier, the observation that quercetin can act from
`either side of the monolayer suggests that its effect
`is not
`initially mediated by cellular surface
`components selectively present on either side of the
`cell (e.g. receptors). It is also possible that quercetin,
`being a small hydrophobic molecule, penetrated the
`cell and produced its effect intracellularly. The
`similar time courses of the secretory responses
`produced by the addition of quercetin to either side
`of the cell do not support the possibility that the
`effect of quercetin is localized to one side of the cell
`
`

`

`A
`
`p83.
`
`•
`
`•
`
`J
`p37
`
`•
`
`tb.
`p29t1
`
`A •
`dpi • *4
`p23
`
`
`•
`
`,p18
`
`•
`
`4401
`
`•
`
`*is
`
`0 w•
`
`711/
`
`1
`
`•
`
`•
`
`•
`
`•
`
`•
`
`•
`
`+0
`•
`- -IND
`•
`eisofillo
`0 , •
`
`•
`
`a
`
`•
`
`"ftvasCr,
`
`•
`
`t. •
`_4'6 a • de*
`„,•,6 • .4116
`
`'-*
`
`Fig. 4. Protein phosphorylation induced by quercetin. As described in Materials and Methods,
`monolayers were labeled with 32P and exposed to different stimuli for 5 min. Soluble phosphoproteins
`were then resolved using two-dimensional gel electrophoresis by isoelectric point from pH 4 to pH 7.5
`(from left to right) and by size (from top to bottom), and detected by radioautography. Panels A and
`B: Compared to a control incubation (panel A), at least five proteins showed increased labeling after
`stimulation with 10µM forskolin and 100 µM carbachol (panel B): p18, p23, p29, p37, and p83. Panels
`C and D: Compared to a control incubation (panel C), exposure to 100 µM quercetin for 5 min (panel
`D) resulted in the increased labeling of p18, p23 and p37. The labeling of p29 and p83 was not increased.
`Similar results were obtained in three additional experiments.
`
`

`

`Stimulation of Tfl4 cell secretion with flavonols
`
`1885
`
`Table 3. Effect of 100 µM quercetin on cAMP production in T84 cells
`
`Incubation time
`(min)
`
`cAMP production
`(% control)
`
`N (df)
`
`P (# 100)
`
`1
`5
`10
`15
`20
`
`94 ± 6
`152 ± 31
`143 ± 13
`116 ± 12
`126 ± 13
`
`6 (5)
`6 (5)
`8 (7)
`8 (7)
`6 (5)
`
`0.19
`0.07
`0.006
`0.11
`0.05
`
`Confluent TR4 cell monolayers were exposed to 100 µM quercetin for the indicated
`time periods, and the levels of cAMP in the corresponding cell homogenates were
`determined as outlined in Materials and Methods. In each paired experiment,
`cAMP production in cells exposed to quercetin is expressed as the percentage of
`the cAMP production in control cells not exposed to quercetin. Values are means
`± SEM. The mean cAMP production values in control cells for the time periods
`1, 5, 10, 15, and 20 min were respectively, 10.5 ± 0.8, 7.4 ± 0.9, 15.9 ± 1.3,
`12.1 ± 0.8, and 11.4 ± 1.2 pmol/filter. P values were calculated using one-tailed t-
`tests with the indicated df.
`
`transported
`to be
`that quercetin needs
`and
`transcellularly when added to the other side.
`We initially postulated that the reactive oxygen
`species produced by flavonol auto-oxidation and/or
`redox cycling may be the ultimate mediators of
`quercetin-induced secretion. However, because the
`secretory response to quercetin was not enhanced
`by Fe3+ and was not inhibited by desferroxamine B
`or catalase (alone or in combination with SOD), we
`were unable to substantiate a role for either
`superoxide, hydrogen peroxide, or the hydroxyl
`radical in the action of quercetin. The possibility
`that quercetin-induced secretion is independent of
`reactive oxygen species is further supported by the
`observation that kaempferol, which does not generate
`reactive oxygen species, stimulates secretion, while
`the converse is true of myricetin, a potent generator
`of reactive oxygen species [4]. When studies were
`performed to identify the intracellular mechanism
`responsible for quercetin-induced secretion, quer-
`cetin resembled other secretagogues known to
`stimulate T84 cells via cAMP: (a) quercetin-induced
`secretion was inhibited by barium chloride and
`bumetanide, (b) quercetin was synergistic with
`carbachol, but not with VIP, and (c) exposure of
`in
`the
`to quercetin resulted
`intact T84 cells
`phosphorylation of endogenous protein substrates
`for cAMP-dependent protein kinase. However,
`quercetin, at concentrations capable of stimulating
`cellular secretion, did not increase cAMP levels to
`the extent demonstrated by VIP. It is possible that,
`similar to the case of the adenosine analogue 5'-(N-
`ethyl)-carboxamido-adenosine [18], there may be a
`shift in the concentration-response curve when
`stimulation of cAMP is studied instead of secretion.
`However, we have not been able to explore this
`possibility because of the poor solubility and potential
`toxicity of high concentrations of quercetin [19]. It
`still remains possible that the modest increase in
`cAMP produced by quercetin was sufficient to
`stimulate the phosphorylation of the substrates for
`cAMP-dependent protein kinase responsible for
`controlling chloride secretion. That the interaction
`between quercetin and the cAMP pathway may be
`
`complex is further suggested by a possible inhibition
`of VIP-induced secretion by quercetin (e.g. less
`potent activation by quercetin of a pathway shared
`with VIP). Should this be the case, quercetin may
`prove to be useful as a probe for further studies of
`the intracellular mechanisms regulating intestinal
`secretion.
`Additional studies will be required to judge
`whether flavonols are physiologically important
`stimuli of intestinal secretion. Flavonols are found
`in the edible portions of many fruits and vegetables
`and the average intake is estimated to be 100 mg/
`day [20, 21], with vegetarians consuming significantly
`larger amounts. Quercetin is the most abundant
`dietary flavonol, and it therefore seems plausible
`that
`luminal concentrations of quercetin may
`reach 50-100 µM (15-30 mg/L) depending on the
`disposition of this flavonol after ingestion. Should
`these concentrations of quercetin be obtained in the
`intestinal lumen, the findings presented in this study
`indicate that this flavonol could stimulate a significant
`secretory response. In
`this regard,
`it seems
`noteworthy that dietary supplementation with fruits
`and vegetables has proved to be useful in the clinical
`management of chronic constipation [22]. Even
`though it has been assumed that the benefit of fruits
`and vegetables results from secretory as well as
`osmotic mechanisms, and from the contained fiber,
`scant information exists concerning the identity of
`the substances that may function as secretory stimuli.
`The present study raises the possibility that the
`secretory action of quercetin and other flavonols
`may account for the beneficial effects of fruits and
`vegetables for patients with constipation.
`
`Acknowledgements—This research was funded in part by
`the Department of Veterans Affairs, by NIH Grants DK
`40506 (T.D.N.), DK 40701 (J.A.C.) and DK 42486
`(T.W.G.), by NIEHS Grant IS 04752 (A.T.C.), and by a
`grant from the Cystic Fibrosis Foundation (J.A.C.). The
`authors are indebted to Margaret Wolfe, Jolanta Kole, and
`Wei Wang for their technical assistance, and to Dr. Helen
`Berschneider for her help in setting up the Ussing chambers.
`
`

`

`1886
`
`T. D. NGUYEN et al.
`
`REFERENCES
`
`1. Jacobs M and Rubery PH, Naturally occurring auxin
`transport regulators. Science 241: 346-349, 1988.
`2. Hodnick WF, Kung FS, Roettger WJ, Bohmont CW
`and Pardini RS, Inhibition of mitochondrial respiration
`and production of toxic oxygen radicals by flavonoids.
`A structure—activity study. Biochem Pharmacol 35:
`2345-2357, 1986.
`3. Hodnick WF, Kalyanaraman B, Pritsos CA and Pardini
`RS, The production of hydroxyl and semiquinone free
`radicals during the autoxidation of redox active
`flavonoids. In: Oxygen Radicals
`in Biology and
`Medicine (Eds. Simic MG, Taylor KA, Ward JF and
`VonSonntag C), pp. 149-153. Plenum Press. New
`York, 1988.
`4. Canada A, Gianella E, Nguyen TD and Mason RP.
`The production of reactive oxygen species by dietary
`flavonols. Free Radic Biol Med 9: 441-449, 1990.
`5. Bern MJ, Sturbaum CW, Karayalcin SS, Berschneider
`HM, Wachsman JT and Powell DW. Immune system
`control of rat and rabbit colonic electrolyte transport:
`Role of prostaglandins and enteric nervous system. J
`Clin Invest 83: 1810-1820, 1989.
`6. Karayalcin SS, Sturbaum CW, Dixon MU and Powell
`DW, Hydrogen peroxide is the reactive oxygen species
`(ROS) which stimulates colonic electrolyte secretion.
`Gastroenterology 94: A216(Abstract), 1988.
`7. Dharmsathaphorn K, McRoberts JA, Mandel KG,
`Tisdale LD and Masui H, A human colonic tumor cell
`line that maintains vectorial electrolyte transport. Am
`J Physiol 246: G204-G208, 1984.
`8. Weymer A, Huott P. Liu W, McRoberts JA and
`Dharmsathaphorn K, Chloride secretory mechanism
`induced by prostaglandin E, in a colonic epithelial cell
`line.