Involvement of Functional Groups on the Surface of Carboxyl Group

Feb 21, 2012 - Terminated Polyamidoamine Dendrimers Bearing Arbutin in ... supported that the glucose moiety of arbutin was essential for the high act...
0 downloads 0 Views 490KB Size
Download PDF
Article pubs.acs.org/molecularpharmaceutics

Involvement of Functional Groups on the Surface of Carboxyl GroupTerminated Polyamidoamine Dendrimers Bearing Arbutin in Inhibition of Na+/Glucose Cotransporter 1 (SGLT1)-Mediated DGlucose Uptake Shinji Sakuma,*,† Shun Kanamitsu,† Yumi Teraoka,† Yoshie Masaoka,† Makoto Kataoka,† Shinji Yamashita,† Yoshiyuki Shirasaka,‡ Ikumi Tamai,‡ Masahiro Muraoka,§ Yohji Nakatsuji,§ Toshiyuki Kida,∥ and Mitsuru Akashi*,∥ †

Faculty of Pharmaceutical Sciences, Setsunan University, 45-1 Nagaotoge-cho, Hirakata, Osaka 573-0101, Japan Faculty of Pharmacy, Institute of Medical, Pharmaceutical and Health Sciences, Kanazawa University, Kakuma-machi, Kanazawa, Ishikawa 920-1192, Japan § Faculty of Engineering, Osaka Institute of Technology, 5-16-1 Ohmiya, Asahi-ku, Osaka 535-8585, Japan ∥ Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita, Osaka 565-0871, Japan ‡

ABSTRACT: A carboxyl group-terminated polyamidoamine dendrimer (generation: 3.0) bearing arbutin, which is a substrate of Na+/glucose cotransporter 1 (SGLT1), via a nonbiodegradable ω-amino triethylene glycol linker (PAMAM-ARB), inhibits SGLT1-mediated D-glucose uptake, as does phloridzin, which is a typical SGLT1 inhibitor. Here, since our previous research revealed that the activity of arbutin was dramatically improved through conjugation with the dendrimer, we examined the involvement of functional groups on the dendrimer surface in inhibition of SGLT1-mediated D-glucose uptake. PAMAM-ARB, with a 6.25% arbutin content, inhibited in vitro D-glucose uptake most strongly; the inhibitory effect decreased as the arbutin content increased. In vitro experiments using arbutin-free original dendrimers indicated that dendrimer-derived carboxyl groups actively participated in SGLT1 inhibition. However, the inhibitory effect was much less than that of PAMAM-ARB and was equal to that of glucose moiety-free PAMAM-ARB. Data supported that the glucose moiety of arbutin was essential for the high activity of PAMAM-ARB in SGLT1 inhibition. Analysis of the balance of each domain further suggested that carboxyl groups anchored PAMAM-ARB to SGLT1, and the subsequent binding of arbutin-derived glucose moieties to the target sites on SGLT1 resulted in strong inhibition of SGLT1-mediated Dglucose uptake. KEYWORDS: dendrimer, arbutin, conjugation, Na+/glucose cotransporter, antidiabetic drug

■

INTRODUCTION Diabetes is a metabolic disease caused by either the deficiency of insulin secretion or cellular resistance to insulin, which results in hyperglycemia. The disease is classified as insulindependent type I diabetes mellitus and noninsulin-dependent type II diabetes mellitus; over 90% of diabetic patients suffer from the latter type. Many antidiabetic drugs, including insulin, sulfonylurea derivatives,1 biguanide and thiazolidine derivatives,2,3 and α-glucosidase inhibitors,4 have been used clinically. The aim of most diabetes treatments is to maintain an appropriate blood glucose level.4 It is also important to suppress the rapid increase in the blood glucose level after a meal.5 Carbohydrates in the human diet are hydrolyzed by digestive enzymes in the gastrointestinal tract. The resulting monosaccharides, such as glucose, are absorbed from the small intestine via influx hexose transporters.6,7 There are 2 types of © 2012 American Chemical Society

hexose transporters in human and rat small intestines: sodiumdependent Na+/glucose cotransporters (SGLT) and sodiumindependent glucose transporters (GLUT).8,9 Of these, SGLT1 is predominantly expressed in the apical membranes of the intestinal epithelial cells, and it is understood to contribute mainly to the apical membrane permeation of D-glucose.6 Since the recognition of D-glucose by SGLT1 is the first step of glucose absorption, SGLT1 inhibition can be one of the most effective approaches to suppress the rapid increase in the blood glucose level after a meal.6,10 Antidiabetic drugs that inhibit Dglucose uptake via SGLT families have been investigated, Received: Revised: Accepted: Published: 922

October 1, 2011 February 7, 2012 February 19, 2012 February 21, 2012 dx.doi.org/10.1021/mp300017e | Mol. Pharmaceutics 2012, 9, 922−929

Molecular Pharmaceutics

Article

Table 1. Characterization of the PAMAM-ARB Prepared in This Study PAMAM dendrimer (mmol)a run 1 2 3 4 5 6

conjugate PAMAM-ARB-2/32 PAMAM-ARB-4/32 PAMAM-ARB-10/32 PAMAM-ARB-2/128 PAMAM-ARB-4/128 PAMAM-ARB-18/128

generation: 3.0

EDC

generation: 5.0

(mmol)

−1

1.06 × 10 1.06 × 10−1 1.06 × 10−1

Am-ARB

1.06 1.06 1.06 8.08 4.04 2.02

8.08 × 10−1 4.04 × 10−1 2.02 × 10−1

× × × × × ×

(mmol) −1

10 10−1 10−1 10−1 10−1 10−1

0.53 1.06 2.12 1.01 1.01 1.01

× × × × × ×

−1

10 10−1 10−1 10−1 10−1 10−1

arbutin content in the conjugateb ratio

percentage (%)

2/32 4/32 10/32 2/128 4/128 18/128

6.25 12.5 31.3 1.56 3.13 14.1

a

Molar amount as a carboxyl group-equivalent of PAMAM dendrimers. bThe average ratio and the percentage of the number of terminal carboxyl groups bearing arbutin to the total number of terminal carboxyl groups.

active ingredient in bearberry leaves, which are traditionally used as an herbal medicine. It is a substrate for human SGLT1,23 and orally absorbed arbutin sterilizes the urinary tract while being eliminated. In vitro experiments demonstrated that the inhibitory effect of intact arbutin on SGLT1-mediated Dglucose uptake was about one-thirtieth that of intact phloridzin. This poor activity was dramatically improved when arbutin was immobilized on the dendrimer surface. The inhibitory effect of PAMAM-ARB was comparable to that of phloridzin when oneeighth (12.5%) of the surface carboxyl groups, on an average, were modified with arbutin.15 Here, we report our approaches to assess the involvement of functional groups on the surface of the dendritic conjugate bearing arbutin in inhibition of SGLT1-mediated D-glucose uptake.

including inhibitors of SGLT2, which is specifically expressed in the kidney and plays a major role in renal glucose reabsorption in the proximal tubule; however, these inhibitors have not been used clinically.11,12 We have been investigating polymeric conjugates bearing glucosides with the expectation that orally administered conjugates bind to SGLT1 from the mucosal side, inhibit the glucose absorption via SGLT1, and consequently suppress an increase in the blood glucose level.13−15 This strategy enables us to develop drugs with high safety, because macromolecules remain in the intestinal lumen without systemic exposure.16−19 Poly(γ-glutamic acid) (γ-PGA) and phloridzin were first used as a polymeric platform and a glucoside, respectively, and phloridzin was grafted onto the polymer backbone via a nonbiodegradable ω-amino triethylene glycol linker (PGAPRZ).13 Phloridzin, which is found in the bark and stem of apple trees, is known to inhibit D -glucose transport competitively through the binding of intramolecular glucose moieties to SGLT1.10,13−15,20,21 However, its in vivo activity is rarely observed because of the lack of the glucose moiety through the hydrolysis of a glucoside bond of phloridzin by βglucosidase located on the apical membranes of the intestine. This hydrolysis also results in a serious disadvantage of phloridzin as a drug. Phloretin, which is the aglycon component of phloridzin, is considered to be absorbed from the intestine and inhibit GLUT1, which is responsible for D-glucose uptake in various tissues.22 Steric hindrance of the polymer chain often improves the stability of enzyme-susceptible chemical bonds.16 As expected, PGA-PRZ remained unchanged in the gastrointestinal tract and significantly suppressed an increase in the blood glucose level after oral administration of D-glucose in rats through inhibition of SGLT1-mediated D-glucose transport. The stabilization of the glucoside bond was presumably due to the spontaneous self-assembly of PGA-PRZ that prevented the glucoside bond from being hydrolyzed by β-glucosidase. We have also confirmed that toxic phloretin was stably retained in the polymer backbone.13,14 It appeared that PGA-PRZ had a potential as an oral antidiabetic drug; however, two concerns still remained. The chemical modification of phloridzin resulted in a ca. 90% reduction in its in vitro activity. A gram-order dose is so large that diabetic patients cannot take the drug daily for an extended period of time. The toxicity of phloretin, which is bound to nonabsorbable polymeric chains via the nonbiodegradable linker, is unlikely to appear because it is not released from the conjugate; however, it is preferable that there is no latent toxicity. In order to tackle these concerns, a carboxyl groupterminated poly(amidoamine) (PAMAM) dendrimer (generation: 3.0) bearing arbutin via the ω-amino triethylene glycol linker (PAMAM-ARB) was next designed. Arbutin is the main

■

EXPERIMENTAL SECTION 1. Materials. PAMAM-succinamic acid dendrimer (1,6diaminohexane core; generation, 3.0; terminals, 32; molecular weight (Mw), 10167 Da) and PAMAM dendrimer (ethylenediamine core; generation, 3.0; terminals, 32; Mw, 6908.8 Da), which were defined as a carboxyl group-terminated dendrimer and an amino group-terminated dendrimer, respectively, were obtained from Sigma-Aldrich (St. Louis, MO, USA). The former dendrimer with a different generation (generation, 5.0; terminals, 128; Mw, 41690 Da), arbutin, and β-glucosidase were also purchased from Sigma-Aldrich. D-Glucose and tritiated Dglucose ([2-3H], 370−740 GBq/mmol, 37 MBq/mL, in sterilized water) were obtained from Wako Pure Chemical Industries Co., Ltd. (Osaka, Japan) and Moravek Biochemicals and Radiochemicals (Brea, CA, USA), respectively. All other chemicals were commercial products of analytical or reagent grade and were used without further purification. Centrifugation-type ultrafilters with a nominal molecular weight limit of 100 kDa (Amicon Ultra, Ultracell-100) and membrane filters (HAWP02500; pore size, 0.45 μm) were purchased from Millipore Co. (Billerica, MA, USA). F-kit D-glucose was furnished by J. K. International Co., Ltd. (Tokyo, Japan). 2. Synthesis of PAMAM-ARB. The dendritic conjugate bearing arbutin was prepared in the same manner as described previously.15 Briefly, N-benzyloxycarbonylated ω-amino triethylene glycol was prepared in 3 reaction steps using 2-[2-(2chloroethoxy)ethoxy]ethanol as a starting material. It was then reacted with methanesulfonyl chloride to yield its mesylate. The mesylate was reacted with 4′-phenolic hydroxyl groups of arbutin to yield N-benzyloxycarbonyl-ω-amino triethylene glycol-bound arbutin. The protecting groups were removed by catalytic reduction to yield ω-amino triethylene glycolbound arbutin (Am-ARB), which is a precursor of PAMAMARB. 923

dx.doi.org/10.1021/mp300017e | Mol. Pharmaceutics 2012, 9, 922−929

Molecular Pharmaceutics

Article

Figure 1. Expected D-glucose uptake by rat small intestinal BBMVs in the presence of inhibitors at various concentration.

conditions. The BBMVs were finally dispersed in cold suspension buffer (HEPES, 10 mM; D-mannitol, 100 mM; KCl, 100 mM, pH 7.4) at a concentration of 0.1−0.2 mg of protein/10 μL. BBMVs (10 μL) were incubated at 30 °C for 3 min and then mixed with 90 μL of uptake buffer (NaCl, 100 mM; D-mannitol, 100 mM; HEPES, 10 mM, pH 7.4) containing D-glucose, tritiated D -glucose, and an inhibitor (arbutin and its derivatives). The concentrations of D-glucose (including the tritiated one) and the inhibitor were adjusted to 0.01 mM and 0.01−2 mM as a glucosyl group-equivalent, respectively. The radioactivity of the solution was adjusted to ca. 100 kBq/mL. To create sodium ion-free conditions, NaCl in the uptake buffer was replaced with an equivalent concentration of KCl.25 After incubation of the mixture at 30 °C for 1 min, the uptake reaction was terminated by addition of 1 mL of ice-cold stop buffer (KCl, 100 mM; HEPES, 10 mM, pH 7.4). The solution was immediately filtered through the membrane filter (HAWP02500) under reduced pressure. The filter, on which BBMVs were collected, was subsequently washed with the icecold stop buffer (5 mL × 2) and dried for 30 s under reduced pressure. After the filter was dissolved in 10 mL of scintillation fluid at room temperature, the radioactivity was measured with a liquid scintillation counter (LSC 3500, Aloka Co., Tokyo, Japan). The same test was performed for original dendrimers and glucose moiety-free PAMAM-ARB (the concentration was adjusted as a dendrimer-equivalent). A fundamental principle of data analysis is described in our previous report with a schematic representation.14 To clarify the experimental approach for D-glucose uptake measurements, the schematic representation was also described (Figure 1). When sodium ions are present in the medium, D-glucose uptake under inhibitor-free conditions corresponds to the sum of the uptake of D-glucose via influx hexose transporters and the passive diffusion of D-glucose through the membranes of BBMVs. D-Glucose uptake is expected to be reduced to a great extent when sodium ions are removed from media because SGLT families cannot transport D-glucose in the absence of sodium ions. Since SGLT1 mainly participates in the absorption of D-glucose across the brush border membranes of intestinal epithelial cells, the difference between D-glucose uptake under inhibitor-free conditions with and without sodium ions (difference A) is approximately regarded as the uptake via SGLT1. Furthermore, the difference between D-glucose uptake under inhibitor-free conditions without sodium ions and that in the presence of inhibitors at various concentrations with sodium ions (difference B) is approximately regarded as residual SGLT1-mediated uptake in the presence of inhibitors.

Am-ARB was subsequently immobilized on the surface of carboxyl group-terminated PAMAM dendrimers with a generation of 3.0 and 5.0 through the coupling of the amino groups of Am-ARB with the carboxyl groups of the dendrimers activated by preincubation with 1-ethyl-3-(3dimethylaminopropyl)carbodiimide (EDC). After the unreacted substances were removed by ultrafiltration, the resulting solution was lyophilized to yield PAMAM-ARB. The content of arbutin in PAMAM-ARB was determined by the integral ratio of the phenyl proton signals in arbutin to the methylene proton signals in the 1,6-diaminohexane core of the PAMAM dendrimer on the 1H NMR spectrum of PAMAMARB. The content was expressed as the average ratio and the percentage of the number of terminal carboxyl groups bearing arbutin to the total number of terminal carboxyl groups. On the basis of the generation of the PAMAM dendrimer, dendritic conjugates bearing arbutin were abbreviated as PAMAM-ARBX/32 and PAMAM-ARB-X/128 (Table 1). 3. Hydrolysis of the Glucoside Bond of Arbutin Bound to Dendrimers. The experiment was performed using a modification of the manner described previously for the hydrolysis of a glucoside bond of phloridzin and its derivatives.13 Aqueous solution containing either arbutin or PAMAM-ARB was prepared. The concentration was adjusted to 2.0 mM as a glucosyl group-equivalent. The solution was mixed with an equivalent volume of aqueous solution containing β-glucosidase at a concentration of 1.0 mg/mL. The mixture was incubated at 37 °C for 24 h and then filtrated using the centrifugation-type ultrafilter. The filtrate was condensed to half under reduced pressure. The concentration of D-glucose released from arbutin and its derivatives was measured using the NADP spectrometric method (F-kit Dglucose). Under our experimental conditions, the lower limit of quantification corresponded to 10% of D-glucose released from arbutin. 4. In Vitro D-Glucose Uptake by Rat Small Intestinal Brush-Border Membrane Vesicles. All rat experiments were approved by the Ethical Review Committee of Setsunan University. The experimental conditions of D-glucose uptake by rat small intestinal brush-border membrane vesicles (BBMVs), in which SGLT1 is predominantly expressed,6,24 and data analysis were described previously.14 BBMVs were isolated by the conventional divalent ionprecipitation method. Briefly, the mucosa was scraped off the luminal surface of the rat small intestine. The scrapings were homogenized, and calcium chloride was added to the homogenate. Precipitated BBMVs were collected and purified through centrifugation. All processes occurred under cool 924

dx.doi.org/10.1021/mp300017e | Mol. Pharmaceutics 2012, 9, 922−929

Molecular Pharmaceutics

Article

Table 2. IC50 Values of Arbutin and Its Derivatives against SGLT1-Mediated D-Glucose Uptakea glucoside-based IC50 (mM)

polymeric platform-based IC50 (mM)

inhibitor

arbutin-equivalent

glucosyl group-equivalentc

dendrimer-equivalent

carboxyl group-equivalentd

arbutinb PAMAM-ARB-2/32 PAMAM-ARB-4/32b PAMAM-ARB-10/32 PAMAM-ARB-2/128 PAMAM-ARB-4/128 PAMAM-ARB-18/128

1.0 0.045 0.15 0.36 0.040 0.045 0.11

1.0 0.045 0.15 0.36 0.040 0.045 0.11

0.023 0.038 0.036 0.020 0.011 0.0063

0.68 1.1 0.79 2.5 1.4 0.69

a

Mean of 4−5 experiments. bData cited from our previous article.13 cIdentical to the arbutin-equivalent IC50 value because one molecule of glucose binds to the 1′-phenolic hydroxyl group of hydroquinone, which is the aglycon of arbutin dDendrimer-equivalent IC50 value was multiplied by the number of unreacted carboxyl groups without arbutin.

Figure 2. Glucosyl group-equivalent IC50 value (a) and carboxyl group-equivalent IC50 value (b) as a function of the arbutin content in PAMAMARB. Data, which are shown in Table 2, were plotted regardless of the dendrimer generation. The dotted line indicates the predictive curve.

observed for the preparation of PAMAM-ARB-2/128 and PAMAM-ARB-4/128. However, the reactivity approximately doubled (28.1%) when the molar amount of Am-ARB was set to half the amount of the dendrimer (PAMAM-ARB-18/128). We confirmed that PAMAM-ARB possessed water solubility sufficient to perform the below-mentioned in vitro uptake experiments (more than 4 mM as a glucosyl group-equivalent), irrespective of the arbutin content and the dendrimer generation. 2. D-Glucose Uptake by Rat Small Intestinal BBMVs in the Presence of PAMAM-ARB. Table 2 shows the IC50 values of arbutin and its derivatives against SGLT1-mediated Dglucose uptake. In vitro experiments demonstrated that the glucosyl group-equivalent IC50 value of intact arbutin was 1.0 mM on average. This poor activity was noticeably improved when arbutin was immobilized on the surface of carboxyl group-terminated PAMAM dendrimers, irrespective of the dendrimer generation. The glucosyl group-equivalent IC50 value of PAMAM-ARB decreased with a decrease in the number of arbutin bound to the dendrimer with a generation of 3.0, indicating that the inhibitory effect of PAMAM-ARB became strong as the arbutin number decreased. When the dendrimer with a generation of 5.0 was used, SGLT1-mediated D-glucose uptake was inhibited most strongly by the dendritic conjugate bearing arbutin, whose content was 4/128 or less, as seen in the case of PAMAM-ARB-2/32. A 2.5-fold increase in the glucosyl group-equivalent IC50 value was observed when the arbutin number was set to 18/128. Figure 2a shows the glucosyl group-equivalent IC50 value as a function of the arbutin content in PAMAM-ARB. Data in Table 2 were plotted regardless of the dendrimer generation. The

Here, the percentage of differences B to A is calculated and subtracted from 100%. The resulting value corresponds to the inhibitor-suppressed percentage of SGLT1-mediated D-glucose uptake (0%: no inhibition, 100%: complete inhibition). When the values are plotted as a function of inhibitor concentration, the inhibitor concentration giving half-maximum inhibition (IC50) of the uptake can be calculated by fitting the data into the Michaelis−Menten-patterned curve.

■

RESULTS 1. Characterization of PAMAM-ARB. Characterization of the PAMAM-ARB prepared in this study is summarized in Table 1. Carboxyl group-terminated PAMAM dendrimers with a generation of 3.0 and 5.0 were used as a platform of arbutin. After the ω-amino triethylene glycol linker was introduced to the 4′-phenolic hydroxyl group of arbutin, the amino groups of the resulting Am-ARB were coupled with the carboxyl groups of the dendrimer in the presence of EDC. As shown in Table 1, the number of arbutin bound to the dendrimer was controlled by changing the feed ratio of Am-ARB to the dendrimer. Under a constant amount of the dendrimer, an increase in the amount of Am-ARB resulted in the elevation of the arbutin content in PAMAM-ARB with a generation of 3.0. One-eighth of the added Am-ARB was reacted with the dendrimer when PAMAM-ARB-2/32 and PAMAM-ARB-4/32 were prepared, while a slight increase in the reactivity (from 12.5% to 15.6%) was observed for the preparation of PAMAM-ARB-10/32. When arbutin was immobilized on the surface of the dendrimer with a generation of 5.0, the dendrimer amount was changed under a constant amount of Am-ARB to control the arbutin content in PAMAM-ARB. A similar low reactivity (12.5%) was 925

dx.doi.org/10.1021/mp300017e | Mol. Pharmaceutics 2012, 9, 922−929

Molecular Pharmaceutics

Article

groups. This immobilization often results in an increase in the affinity of ligands for target molecules through their interactions that occur simultaneously and with multiple access.19,27−29 The carboxyl group-terminated PAMAM dendrimer was substituted for γ-PGA, which is a linear macromolecule, with the expectation that the dendrimer-intrinsic unique properties enhance the inhibitory effect of immobilized glucosides on SGLT1-mediated D-glucose uptake. A common size of the dendrimer core (generation: 3.0) was first chosen, and arbutin was bound to the dendrimer via the ω-amino triethylene glycol linker, which was used for γ-PGA−glucoside conjugates. Our previous research certainly revealed that the poor activity of arbutin was improved when it was immobilized on the dendrimer surface (generation: 3.0) at a content of 12.5% (PAMAM-ARB-4/32). On the contrary, a similar phenomenon was not observed when γ-PGA was used as a platform.15 To assess the involvement of functional groups on the surface of the dendritic conjugate bearing arbutin in inhibition of SGLT1-mediated D-glucose uptake, PAMAM-ARB with different contents of arbutin was prepared by changing the feed ratio of Am-ARB, which is a precursor of PAMAM-ARB, to the dendrimer (Table 1). Arbutin was immobilized on the generation-different dendrimer surface by coupling the amino groups of Am-ARB with the EDC-activated carboxyl groups of the dendrimer. On an average, 12.5% of the added Am-ARB was bound to the dendrimer surface when the arbutin content in PAMAM-ARB was low (runs 1−2 and 4−5, Table 1), irrespective of the dendrimer generation. Clear improvement of the low reactivity was observed when PAMAM-ARB-18/128 (run 6) was prepared. Arbutin immobilized on the dendrimer surface might accelerate the subsequent coupling between the amino groups of Am-ARB and the residual carboxyl groups of the dendrimer. The reason for this improvement was unclear; however, we concluded that these dendritic conjugates bearing arbutin would be appropriate for our mechanism studies. The glucosyl group-equivalent IC50 values of PAMAM-ARB against SGLT1-mediated D-glucose uptake are summarized in Table 2. The inhibitory effect of PAMAM-ARB was much stronger than that of intact arbutin, irrespective of the arbutin content and the dendrimer generation. However, the resulting relationship between the arbutin content and the IC50 value (Figure 2a) was contrary to our expectation. If the interactions between the immobilized arbutin and SGLT1 occurred simultaneously and with multiple access, a reduction of the IC50 value of PAMAM-ARB would be observed with an increase in the arbutin amount. Data strongly indicated that the condensation of ligands on the dendrimer surface did not escalate the ligand activity in our case. Here, we hypothesized that not only arbutin-derived glucose moieties recognize the binding site on SGLT1, but dendrimerderived surface activities also cooperate with the glucose moieties in inhibiting SGLT1 through some interaction with the transporter. As is obvious from the IC50 value of the carboxyl group-terminated PAMAM dendrimer (Table 3), this polymeric platform inhibited SGLT1-mediated D-glucose uptake. No inhibitory effect was observed when the same test was performed for plain PAMAM dendrimers with terminal amino groups. We also examined the inhibitory effect of PAMAM-amidoethanol dendrimers with terminal hydroxyl groups (generation: 3.0) and confirmed that D-glucose uptake was hardly ever inhibited when the dendrimer concentration was adjusted to 0.2 mM as a dendrimer equivalent, which

lowest IC50 value was stably observed when the arbutin content in PAMAM-ARB was 6.25% or less. The IC50 value increased linearly when the arbutin content exceeded this threshold value. A similar analysis was performed for the carboxyl groupequivalent IC50 value (Figure 2b). The IC50 value decreased exponentially until the arbutin content reached 6.25%; it was constantly low when the arbutin content was more than this threshold value. No association was observed when the relationship between the arbutin content and the dendrimerequivalent IC50 value was examined (graphical data not shown). The inhibition of SGLT1-mediated D-glucose uptake by arbutin-free original dendrimers was next examined. As shown in Table 3, the dendrimer-equivalent IC50 value of the carboxyl Table 3. IC50 Values of the Original Dendrimers and Glucose Moiety-Free PAMAM-ARB against SGLT1Mediated D-Glucose Uptakea polymeric platform-based IC50 (mM) dendrimerequivalent

inhibitor carboxyl group-terminated PAMAM− succinamic acid dendrimer (generation: 3.0) amino group-terminated PAMAM dendrimer (generation: 3.0) glucose moiety-free PAMAM-ARB-2/32b a

0.20 >1.0 0.19

functional groupequivalent 6.4c >32.0d 5.7c

b

Mean of 3 experiments. Glucose moieties of PAMAM-ARB-2/32 were completely removed through incubation with β-glucosidase. c Carboxyl group-equivalent IC50. dAmino group-equivalent IC50.

group-terminated PAMAM dendrimer with a generation of 3.0, which was used as a platform of PAMAM-ARB, was 0.2 mM on average. Amino group-terminated PAMAM dendrimers hardly ever inhibited SGLT1-mediated D-glucose uptake, even when the concentration was adjusted to 1.0 mM. The IC50 value of PAMAM-ARB-2/32 was finally compared with the glucose moiety-free one. Twenty-four hours after incubation of PAMAM-ARB-2/32 with β-glucosidase, the mixture was filtrated using the centrifugation-type ultrafilter with a nominal molecular weight limit of 100 kDa, which is more than the Mw of the dendrimer and less than the Mw of the enzyme. The filtrate containing PAMAM-ARB and/or its degradation products, such as D-glucose and glucose moietyfree PAMAM-ARB, was consequently collected. D-Glucose in the filtrate was assayed to estimate the stability of the glucoside bond of intact arbutin, and it was confirmed that the glucoside bond was completely degraded during a 24-h incubation. Similar β-glucosidase-susceptibility was observed for arbutin bound to carboxyl group-terminated PAMAM dendrimers (data not shown). The in vitro uptake experiment using the filtrate was performed to measure the IC50 value of glucose moiety-free PAMAM-ARB. As shown in Table 3, the dendrimer-equivalent and carboxyl group-equivalent IC50 values were 0.19 mM and 5.7 mM, respectively, which were almost equal to those of the original PAMAM-succinamic acid dendrimer.

■

DISCUSSION

A dendrimer is a nanometer-sized starburst macromolecule with a generation-dependent number of terminal groups.26 A large number of ligands can be immobilized on the surface of dendrimers through chemical reactions with the terminal 926

dx.doi.org/10.1021/mp300017e | Mol. Pharmaceutics 2012, 9, 922−929

Molecular Pharmaceutics

Article

Figure 3. Chemical structure of PAMAM-ARB-2/32.

Current data indicate that PAMAM-ARB-2/32 is the slimmest arbutin-bound dendritic conjugate with the highest activity for SGLT1 inhibition. In Figure 3, for convenience, a couple of arbutin molecules are symmetrically anchored to the terminal carboxyl groups of PAMAM dendrimers via the linker, although the precise chemical structure of this conjugate, including the distribution of the number of arbutin bound to the dendrimer, has not been confirmed experimentally. The position of the carboxyl group with which Am-ARB is coupled cannot be controlled by current synthetic procedures of PAMAM-ARB. Instrumental analysis, which gives information on the position, has not been established (the NMR analysis does not give the information). However, we consider that Figure 3 is not an irrelevant structure. The cooperation of dendrimer-derived carboxyl groups with arbutin-derived glucose moieties in SGLT1 inhibition may depend on the 3-dimensional distribution, orientation, and/or flexibility of each domain on the dendrimer surface. It is reported that the number of acidic amino acids was more than that of basic amino acids on the extracellular domain of SGLT1.30,31 On the other hand, amino group-terminated PAMAM dendrimers did not affect SGLT1-mediated D-glucose uptake although carboxyl group-terminated PAMAM dendrimers participated in SGLT1 inhibition. These facts imply that the inhibition mechanism of anionic dendrimers bearing arbutin is not simply discussed on the basis of electrostatic interactions between SGLT1 and the dendrimer. It may be critical that the distance between basic amino acids on the extracellular domain of SGLT1 and the sugar binding site matches that between the dendrimer-derived carboxyl groups and the arbutin-derived glucose moiety. The inhibition mechanism of SGLT1 by the carboxyl groups of dendrimers is also unclear although it is predicted that SGLT1 is inhibited competitively by the glucose moieties of arbutin because this SGLT1 substrate is just bound to the surface of the dendrimers.

corresponded to the IC50 value of carboxyl group-terminated PAMAM dendrimers (data not shown). These data indicate that carboxyl groups on the dendrimer surface actively participate in SGLT1 inhibition. Now, as shown in Tables 2 and 3, the polymeric platformbased IC50 values of dendritic conjugates bearing arbutin were much lower than that of the original dendrimer. The IC50 value of glucose moiety-free PAMAM-ARB-2/32 was almost equal to that of the original dendrimer. These data support that the glucose moiety of arbutin is essential for the high activity of PAMAM-ARB in SGLT1 inhibition. We further considered the balance of each domain on the dendrimer surface. As shown in Figure 2, when the arbutin content was 6.25% or less, the inhibitory effect of PAMAM-ARB remained at the highest level; however, there was a large difference in their carboxyl groupequivalent IC50 values. It occurs to us that extra carboxyl groups, which do not participate in SGLT1 inhibition, are possibly located on the surface of PAMAM-ARB-2/128 and 4/ 128. On the other hand, when the arbutin content exceeded 6.25%, a reduction of the PAMAM-ARB activity was observed. In this range, the carboxyl group-equivalent IC50 value reached a plateau. The constant IC50 value suggests that the number of carboxyl groups on the dendrimer surface meets the minimum requirement. It appears that surface carboxyl groups anchor PAMAM-ARB to SGLT1 and the subsequent binding of arbutin-derived glucose moieties to the target sites on SGLT1 results in strong inhibition of SGLT1-mediated D-glucose uptake. However, the comparison between the glucoside-based IC50 value of PAMAM-ARB-2/32 and that of PAMAM-ARB10/32 indicates that the arbutin-derived glucose moieties of the latter conjugate do not always bind to the target sites on SGLT1 effectively. The interference between excess glucose moieties possibly prevents them from binding to the target sites. 927

dx.doi.org/10.1021/mp300017e | Mol. Pharmaceutics 2012, 9, 922−929

Molecular Pharmaceutics

Article

(7) Wood, I. S.; Trayhurn, P. Glucose Transporters (GLUT and SGLT): Expanded Families of Sugar Transport Proteins. Br. J. Nutr. 2003, 89, 3−9. (8) Wright, E. M.; Turk, E. The Sodium/Glucose Cotransport Family SLC5. Pflugers Arch. 2004, 447, 510−518. (9) Uldry, M.; Thorens, B. The SLC2 Family of Facilitated Hexose and Polyol Transporters. Pflugers Arch. 2004, 447, 480−489. (10) Mizuma, T.; Awazu, S. Intestinal Na+/Glucose CotransporterMediated Transport of Glucose Conjugate Formed from Disaccharide Conjugate. Biochim. Biophys. Acta 1998, 1379, 1−6. (11) Isaji, M. Sodium-Glucose Cotransporter Inhibitors for Diabetes. Curr. Opin. Invest. Drugs 2007, 8, 285−292. (12) Washburn, W. N. Evolution of Sodium Glucose Co-Transporter 2 Inhibitors as Anti-diabetic Agents. Expert Opin. Ther. Pat. 2009, 19, 1485−1499. (13) Ikumi, Y.; Kida, T.; Sakuma, S.; Yamashita, S.; Akashi, M. Polymer-Phloridzin Conjugates as an Anti-diabetic Drug That Inhibits Glucose Absorption through the Na +/Glucose Cotransporter (SGLT1) in the Small Intestine. J. Controlled Release 2008, 125, 42− 49. (14) Sakuma, S.; Sagawa, T.; Masaoka, Y.; Kataoka, M.; Yamashita, S.; Shirasaka, Y.; Tamai, I.; Ikumi, Y.; Kida, T.; Akashi, M. Stabilization of Enzyme-Susceptible Glucoside Bonds of Phloridzin through Conjugation with Poly(γ-glutamic acid). J. Controlled Release 2009, 133, 125−131. (15) Sakuma, S.; Teraoka, Y.; Sagawa, T.; Masaoka, Y.; Kataoka, M.; Yamashita, S.; Shirasaka, Y.; Tamai, I.; Ikumi, Y.; Kida, T.; Akashi, M. Carboxyl Group-Terminated Polyamidoamine Dendrimers Bearing Glucosides Inhibit Intestinal Hexose Transporter-Mediated D-Glucose Uptake. Eur. J. Pharm. Biopharm. 2010, 75, 366−374. (16) Sakuma, S.; Lu, Z.-R.; Kopečková, P.; Kopeček, J. Biorecognizable HPMA Copolymer−Drug Conjugates for Colon-Specific Delivery of 9-Aminocamptothecin. J. Controlled Release 2001, 75, 365−379. (17) Sakuma, S.; Matsumoto, T.; Yamashita, S.; Wang, Y.; Lu, Z.-R. Conjugation of Poorly Absorptive Drugs with Mucoadhesive Polymers for the Improvement of Oral Absorption of Drugs. J. Controlled Release 2007, 123, 195−202. (18) Sakuma, S.; Kataoka, M.; Higashino, H.; Yano, T.; Masaoka, Y.; Yamashita, S.; Hiwatari, K.; Tachikawa, H.; Kimura, R.; Nakamura, K.; Kumagai, H.; Gore, J. C.; Pham, W. A Potential of Peanut AgglutininImmobilized Fluorescent Nanospheres as a Safe Candidate of Diagnostic Drugs for Colonoscopy. Eur. J. Pharm. Sci. 2011, 42, 340−347. (19) Sakuma, S.; Yamashita, S.; Hiwatari, K.; Hoffman, R. M.; Pham, W. Lectin-Immobilized Fluorescent Nanospheres for Targeting to Colorectal Cancer from a Physicochemical Perspective. Curr. Drug Discovery Technol. 2011, 8, 367−378. (20) Xia, X.; Lin, J. T.; Kinne, R. K. H. Binding of Phlorizin to the Isolated C-Terminal Extramembranous Loop of the Na+/Glucose Cotransporter Assessed by Intrinsic Tryptophan Fluorescence. Biochemistry 2003, 42, 6115−6120. (21) Tyagi, N. K.; Kumar, A.; Goyal, P.; Pandey, D.; Siess, W.; Kinne, R. K. D-Glucose-Recognition and Phlorizin-Binding Sites in Human Sodium/D-Glucose Cotransporter 1 (hSGLT1): A Tryptophan Scanning Study. Biochemistry 2007, 46, 13616−13628. (22) Ehrenkranz, J. R.; Lewis, N. G.; Kahn, C. R.; Roth, J. Phlorizin: A Review. Diabetes Metab. Res. Rev. 2005, 21, 31−38. (23) Díez-Sampedro, A.; Lostao, M. P.; Wright, E. M.; Hirayama, B. A. Glucoside Binding and Translocation in Na+-Dependent Glucose Cotransporters: Comparison of SGLT1 and SGLT3. J. Membr. Biol. 2000, 176, 111−117. (24) Tomimatsu, T.; Horie, T. Enhanced Glucose Absorption in the Rat Small Intestine Following Repeated Doses of 5-Fluorouracil. Chem.-Biol. Interact. 2005, 155, 129−139. (25) Gopal, E.; Umapathy, N. S.; Martin, P. M.; Ananth, S.; GnanaPrakasam, J. P.; Becker, H.; Wagner, C. A.; Ganapathy, V.; Prasad, P. D. Cloning and Functional Characterization of Human SMCT2 (SLC5A12) and Expression Pattern of the Transporter in Kidney. Biochim. Biophys. Acta 2007, 1768, 2690−2697.

If SGLT1 is inhibited noncompetitively by the dendrimerderived carboxyl groups, the inhibition activity may be influenced considerably by the length of the linker and the size of the monomer unit. Details of the inhibition mechanism will be discussed with experimental data in the future.

■

CONCLUSIONS Since our previous research revealed that the activity of arbutin was dramatically improved when it was conjugated with the carboxyl group-terminated PAMAM dendrimer, the involvement of functional groups on the surface of the conjugate (PAMAM-ARB) in inhibition of SGLT1-mediated D-glucose uptake was examined. In vitro uptake experiments using rat small intestinal BBMVs strongly indicated that the arbutinderived glucose moiety was essential for the high activity of PAMAM-ARB in SGLT1 inhibition, although dendrimerderived carboxyl groups also actively participated in the inhibition. Analysis of the balance of each domain suggested that there were extra carboxyl groups and glucose moieties, which did not participate in SGLT1 inhibition effectively. It appears that PAMAM-ARB (generation: 3.0) with a 6.25% arbutin content is the strongest and slimmest conjugate without extra domains.

■

AUTHOR INFORMATION

Corresponding Author

*S.S.: Faculty of Pharmaceutical Sciences, Setsunan University, 45-1 Nagaotoge-cho, Hirakata, Osaka 573-0101, Japan; telephone, +81-72-866-3124; fax, +81-72-866-3126; e-mail, [email protected]. M.A.: Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita, Osaka 565-0871, Japan; telephone, +81-6-6879-7356; fax, +81-6-68797359; e-mail, [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported in part by a grant-in-aid for scientific research (No. 21590051) from the Ministry of Education, Culture, Sports, Sciences and Technology of Japan (MEXT). The authors thank Meiji Seika Co., Ltd. for gifting γPGA.

■

REFERENCES

(1) Rendell, M. The Role of Sulphonylureas in the Management of Type 2 Diabetes Mellitus. Drugs 2004, 64, 1339−1358. (2) Ohara, S.; Komatsu, R.; Matsuyama, T. Short-Term Effect of Buformin, a Biguanide, on Insulin Sensitivity, Soluble Fraction of Tumor Necrosis Factor Receptor and Serum Lipids in Overweight Patients with Type 2 Diabetes Mellitus. Diabetes Res. Clin. Pract. 2004, 66, 133−138. (3) El-Batran, S. A.; Abdel-Salam, O. M. E.; Nofal, S. M.; Baiuomy, A. R. Effect of Rosiglitazone and Nateglinide on Serum Glucose and Lipid Profile Alone or in Combination with the Biguanide Metformin in Diabetic Rats. Pharmacol. Res. 2006, 53, 69−74. (4) Mertes, G. Safety and Efficacy of Acarbose in the Treatment of Type 2 Diabetes: Data from a 5-Year Surveillance Study. Diabetes Res. Clin. Pract. 2001, 52, 193−204. (5) Chiasson, J. L.; Josse, R. G.; Gomis, R.; Hanefeld, M.; Karasik, A.; Laakso, M. Acarbose for Prevention of Type 2 Diabetes Mellitus: The STOP-NIDDM Randomized Trial. Lancet 2002, 359, 2072−2077. (6) Hediger, M. A.; Coady, M. J.; Ikeda, T. S.; Wright, E. M. Expression Cloning and cDNA Sequencing of the Na+/Glucose Cotransporter. Nature 1987, 330, 379−381. 928

dx.doi.org/10.1021/mp300017e | Mol. Pharmaceutics 2012, 9, 922−929

Molecular Pharmaceutics

Article

(26) Tamolia, D. A.; Baker, H.; Dewald, J.; Hall, M.; Kallos, G.; Martin, S.; Roeck, J.; Ryder, J.; Smith, P. A New Class of Polymers: Starburst-Dendritic Macromolecules. Polym. J. 1985, 17, 117−132. (27) Tsutsumiuchi, K.; Aoi, K.; Okada, M. Globular Carbohydrate Macromolecule “Sugar Balls” IV. Synthesis of Dendritic Nanocapsules with Molecular Recognition Sites of Periphery. Polym. J. 1999, 31, 935−941. (28) Page, D.; Aravind, S.; Roy, R. Synthesis and Biological Properties of Mannosylated Starburst Poly(amidoamine) Dendrimers. Bioconjugate Chem. 1997, 8, 714−723. (29) Turnbull, W. B.; Kalovidouris, S. A.; Stoddart, J. F. Large Oligosaccharide-Based Glycodendrimers. Chem.Eur. J. 2002, 8, 2988−3000. (30) Hediger, M. A.; Rhoads, D. B. Molecular Physiology of SodiumGlucose Cotransporters. Physiol. Rev. 1994, 74, 993−1026. (31) Wright, E. M. Renal Na+-Glucose Cotransporters. Am. J. Physiol. Renal Physiol. 2001, 208, F10−18.

929

dx.doi.org/10.1021/mp300017e | Mol. Pharmaceutics 2012, 9, 922−929