Sialin (SLC17A5) functions as a nitrate transporter in the plasma membrane

Lizheng Qina,1, Xibao Liub,1, Qifei Suna, Zhipeng Fana, Dengsheng Xiaa, Gang Dinga, Hwei Ling Ongb, David Adamsc, William A. Gahlc, Changyu Zhengb, Senrong Qia, Luyuan Jina, Chunmei Zhanga, Liankun Gud, Junqi Hee, Dajun Dengd,2, Indu S. Ambudkarb,2, and Songlin Wanga,e,2 aSalivary Gland Disease Center and Beijing Key Laboratory of Tooth Regeneration and Function Reconstruction, Capital Medical University School of Stomatology, Beijing 100050, China; bMolecular Physiology and Therapeutics Branch, National Institute of Dental and Craniofacial Research, Bethesda, MD 20892; cMedical Genetics Branch, National Human Genome Research Institute, Bethesda, MD 20892; dKey Laboratory of Carcinogenesis and Translational Research, Peking University Cancer Hospital and Institute, Beijing 100142, China; and eDepartment of Biochemistry and Molecular Biology, Capital Medical University School of Basic Medicine, Beijing 100069, China

Edited by Mark Gladwin, University of Pittsburg Medical Center, Pittsburgh, PA and accepted by the Editorial Board June 5, 2012 (received for review October 13, 2011)

In vivo recycling of nitrate (NO −) and nitrite (NO −) is an important Nitrate uptake into salivary glands represents the key initial step 3 2 − alternative pathway for the generation of nitric oxide (NO) and in NO clearance from the serum; however, the mechanism me- 3 − maintenance of systemic nitrate–nitrite–NO balance. More than diating transport of NO3 in salivary gland epithelial cells has not − − fl 25% of the circulating NO3 is actively removed and secreted by yet been established. In this study, we examined NO3 in ux in − + salivary glands. Oral commensal bacteria convert salivary NO3 to salivary gland cells. Here we report that the sialic acid (SA)/H − SLC17A5 NO2 , which enters circulation and leads to NO generation. The cotransporter, sialin ( ), mutations in which result in Salla − transporters for NO3 in salivary glands have not yet been identi- disease and ISSD, is involved in nitrate uptake into salivary glands. fied. Here we report that sialin (SLC17A5), mutations in which cause Our data suggest a similar function for sialin in several other cell fi Salla disease and infantile sialic acid storage disorder (ISSD), func- types as well, including broblasts. We show that sialin is endoge- − + nously localized in the lysosomes as well as in the plasma membrane tions as an electrogenic 2NO3 /H cotransporter in the plasma − fi of salivary gland cells, where it functions as an electrogenic 2NO / membrane of salivary gland acinar cells. We have identi ed an ex- + 3 − H cotransporter mediating influx of nitrate into the cell. We also tracellular pH-dependent anion current that is carried by NO3 or − provide evidence that plasma membrane sialin is a multifunctional sialic acid (SA), but not by Br , and is accompanied by intracellular − + acidification. Both responses were reduced by knockdown of sialin anion transporter that can mediate electrogenic A /H cotransport expression and increased by the plasma membrane-targeted sialin of anions such as SA, glutamate, and aspartate. Importantly, we mutant (L22A-L23A). Fibroblasts from patients with ISSD displayed have assessed the function of sialin in vivo in pig salivary glands and − provide evidence for the physiological relevance of sialin in medi- reduced SA- and NO -induced currents compared with healthy − 3 ating nitrate uptake NO influx into pig salivary glands. In ag- controls. Furthermore, expression of disease-associated sialin mu- 3 + gregate, our findings suggest that sialin is a versatile anion tants in fibroblasts and salivary gland cells suppressed the H -de- − transporter, and that functional defects in the protein may have pendent NO conductance. Importantly, adenovirus-dependent 3 a deleterious impact on several critical physiological functions. expression of the sialinH183R mutant in vivo in pig salivary glands − − decreased NO3 secretion in saliva after intake of a NO3 -rich diet. Results Taken together, these data demonstrate that sialin mediates nitrate − Coupled NO Currents and Intracellular Acidification in Human influx into salivary gland and other cell types. We suggest that the 3 − + Submandibular Gland Cells. Human submandibular gland cell line 2NO /H transport function of sialin in salivary glands can contrib- 3 (HSG) cells did not display constitutive currents in standard ex- ute significantly to clearance of serum nitrate, as well as nitrate − − tracellular solution. Replacement of Cl with 150 mM NO3 recycling and physiological nitrite-NO homeostasis. produced a relatively slow, but significant, spontaneous increase in the outward current that was dramatically enhanced by decreasing pH | proton theexternalpHfrom7.4to4.0(Fig.1A); the current density in the − ± n = − − 150 mM NO3 solution was 42.4 5.4 pA/pF ( 21). Both the he anions NO3 and NO2 were once thought to be inert end constitutive and low pH-induced currents had similar outwardly Tproducts of NO metabolism. However, it is now evident that rectifying characteristics with a reversal potential of −15 ± 2mV nitrate and nitrite can be recycled in vivo to form NO, and thus (n = 19) (Fig. 1B). Low pH also increased the outward current in − these anions complement the nitric oxide synthase (NOS)-de- normal standard extracellular solution (Cl -containing), although − pendent activity (1). The nitrate-nitrite-NO pathway is emerging the amplitude of the current was smaller than that seen with NO3 as a potential therapeutic target in such diseases as myocardial (23.1 ± 4.3 pA/pF; n = 18) (Fig. 1C), and the current reversed at infarction, stroke, gastric ulcers, and pulmonary hypertension (1, 2). There are two major sources of nitrate and nitrite: the L-arginine–NO synthase pathway and diet. Dietary intake of ni- − Author contributions: L.Q., X.L., D.D., I.S.A., and S.W. designed research; L.Q., X.L., Q.S., Z. trate leads to a relatively rapid increase in NO3 concentration F., D.X., G.D., H.L.O., D.A., W.A.G., C. Zheng, S.Q., L.J., L.G., and S.W. performed research; in serum. Although a large part of the anion is excreted via the L.Q., X.L., C. Zheng, and L.J. contributed new reagents/analytic tools; L.Q., X.L., C. Zheng, kidneys, up to 25% of the circulating nitrate is actively taken up L.J., C. Zhang, J.H., D.D., I.S.A., and S.W. analyzed data; and L.Q., X.L., D.D., I.S.A., and S.W. by the salivary glands and concentrated ∼10-fold in the saliva wrote the paper. secreted from the glands (3–5). Conditions that compromise The authors declare no conflict of interest. − salivary gland function have been linked to decreased NO se- This article is a PNAS Direct Submission. M.G. is a guest editor invited by the Editorial − 3 Board. cretion from the salivary glands and increased NO3 levels in the serum and urine (5, 6). Although nitrate can be reduced to nitrite See Commentary on page 13144. by the commensal bacteria in the oral cavity, most of the salivary 1L.Q. and X.L. contributed equally to this work. nitrite escapes gastric conversion to NO and enters the systemic 2To whom correspondence may be addressed. E-mail: [email protected], iambudkar@dir. circulation, where it generates NO. Thus, salivary nitrate is nidcr.nih.gov, or [email protected]. − recycled back to NO and is critical for the maintenance of 2 − This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. physiological levels of NO and NO2 in the serum (1, 2, 7). 1073/pnas.1116633109/-/DCSupplemental.

13434–13439 | PNAS | August 14, 2012 | vol. 109 | no. 33 www.pnas.org/cgi/doi/10.1073/pnas.1116633109 SEE COMMENTARY

− − Fig. 1. Coupled NO3 currents and intracellular acidification in HSG cells. NO3 currents in HSG cells (A–F) and primary huSMG cells (G) measured by the whole-cell patch-clamp technique. NaCl was replaced with NaNO3, NaBr, or Na-gluconate as indicated. Changes in extracellular pH are shown in the traces − (bar). I-V curves are shown in B and D.(H) External pH and NO3 (5 mM)-dependent acidification of HSG cells measured using BCECF fluorescence. (I)pH − dependence of NO3 currents in HSG cells. (J) Intracellular pH changes under the same experimental conditions as shown in I.(K) Data from I and J were used − to determined the relationship of NO3 currents and intracellular acidification (Hill coefficient: 2.0 ± 0.1).

− − 0 mV (Fig. 1D). Decreasing the external pH in Br - or gluconate- conditions demonstrated pH- and NO3 -dependent intracellular containing medium did not generate any detectable current (Fig. 1 acidification (Fig. 1J).Thedatawerefitted (R2 = 0.999) using the E and F). Similar findings were observed in primary human sub- Hill equation, yielding a Hill coefficient of 2.0 ± 0.1 (Fig. 1K), mandibular gland (huSMG) cells (Fig. 1G), human parotid gland consistent with the electrophysiological data. These data strongly − + ductal cells, and freshly dispersed acini cells prepared from mouse suggest that an electrogenic 2NO /H cotransporter is involved in − − 3 salivary glands. Cl channels in salivary gland cells are quite per- mediating NO uptake in salivary gland epithelial cells. − − 3 meable to NO and Br , unlike the activity described here (8–11). 3 − Nonetheless, NO conductance was inhibited by several anion Involvement of Sialin in 2NO −/H+ Cotransport. To examine nitrate 3 − 3 − channel blockers that block various Cl channels (Fig. S1). Fur- transport via salivary gland, [NO ] was measured in the serum − − 3 thermore, unlike known Cl channels in salivary gland cells, NO and saliva of miniature pigs fed with regular or nitrate-rich 3 −

conductance was not regulated by cAMP, intracellular or extra- fodder. In both cases, the [NO ] level in saliva exceeded that in PHYSIOLOGY + 3 cellular Ca2 , or muscarinic or purinergic receptor agonists (Fig. serum within 1 h after feeding (Fig. 2A), suggesting that salivary + + S2). Replacing Na in the medium with Cs also had no affect on glands take up and secrete nitrate. The nitrate transporters the current (Fig. S3 A and B). Relatively lower (and likely more identified to date belong to the highly conserved major facilitator − physiological) NO concentrations (0.05–0.5 mM) also induced superfamily (MFS) (11–15). To determine the molecular com- 3 − + currents in HSG cells, with channel densities of 2.3 ± 0.3 pA/pF ponent mediating 2NO3 /H cotransport in salivary glands, we (n = 5) and 3.4 ± 0.5 pA/pF (n = 6), respectively (Fig. S3C). Taken analyzed the expression of 127 MFS genes (11, 16, 17) in human together, these findings demonstrate the unique properties of the parotid and submandibular glands (n = 3) using Affymetrix − NO conductance detected in salivary gland cells. U133Plus 2.0 microarrays. SLC17A5 (which encodes sialin) was 3 − To examine the modulation of NO conductance by pH, we expressed at high levels in both types of salivary glands (Table S1). 3 + monitored intracellular pH using 2′-7′-bis(carboxyethyl)-5(6)-car- Sialin functions as a lysosomal SA/H transporter involved in SA boxyfluorescein (BCECF). Decreasing the external pH in gluco- efflux, although it is detected in the plasma membrane of neurons − nate-containing medium (ambient [Cl ] in gluconate-containing (18, 19) where it mediates aspartate and glutamate transport. The medium was <10 mM) did not change intracellular pH, whereas protein is widely expressed in such tissues as brain, heart, lung, − including 5 mM NO induced rapid acidification that was reversed liver, kidney, and mouse submandibular glands (20–22). Impor- − 3 by replacing NO with gluconate in the medium (Fig. 1H). When tantly, mutations in sialin are causative factors in the neurode- 3 − external pH was varied, keeping [NO3 ] constant at 5 mM, the generative disorders Salla disease and ISSD (18, 23). We validated outward current was changed as a function of external pH (Fig. 1I). the expression of sialin in various human tissues by qPCR (Fig. Importantly, measurement of intracellular pH under the same 2B) and by Western blot analysis using samples of various tissues

Qin et al. PNAS | August 14, 2012 | vol. 109 | no. 33 | 13435 − + − Fig. 2. Involvement of sialin in 2NO3 /H cotransport. (A) Saliva and serum [NO3 ] in miniature pigs fed on regular fodder or supplemented with 100 mg/kg of nitrate. (B) Validation of gene expression of SLC17A5 based on the MFS gene expression pattern (Table S1). (C) Detection of sialin (i and iv) in human salivary gland. Na/K-ATPase (ii) or LAMP-1 (v) are markers for basolateral membrane and lysosomes, respectively. Colocalization of the two proteins is shown (iii and vi, yellow, indicated by arrows). (D) Nitrate uptake in HSG cells transfected with sh-sialin or scram-sh. The values indicated by * or ** are significantly different from the unmarked values. (P < 0.05 or P < 0.01; n ≥ 3). (E and F) Knockdown of sialin by sh-sialin and overexpression of WT-sialin. (G and H) Effect of − sialin knockdown on constitutive and low-pH–induced NO3 current. Levels of current and intracellular pH in control HSG cells transfected with scram-sh (G)or sh-sialin (H) were as described in Fig. 1. (I) Average data obtained from the experiments shown in G and H. The number of cells tested is indicated. Statistically significant differences are indicated by ** (P < 0.01). from miniature pigs (Fig. S4). Sialin was highly expressed in sali- biopsies and in lysosomes as detected by its colocalization with the + + vary glands, and liver, with lower levels of expression in brain, respective markers Na /K ATPase and LAMP-1 (Fig. 2C). spleen and kidney, and even lower in muscle and pancreas. Sialin Based on the expression and function of known solute carrier was localized in the basolateral regions of human parotid gland (SLC) transporters (Table S1) and the channel activity described in

− + + Fig. 3. Sialin mediates NO3 /H as well as SA/H cotransporter in HSG cells. (A) Current measurement in cells perfused with medium at pH 4.0 containing − − 5 mM SA, NO3 , or both. (B) I-V curves of the current (Erev= 0mVfor5mMSAorNO3 ) obtained under the different conditions shown in A.(C) Intracellular − − pH in medium of pH 4.0 containing either SA or NO3 .(D) Currents generated by substituting NO3 with glutamate (Glu) or aspartate (Asp), with other anions − transported by sialin. (E and F) SA and NO3 currents induced at pH 4.0 in control HSG cells (E) and cells treated with sh-sialin (F). The y-axis scale is the same in E and F.(G) Average of data from the experiments shown in E and F.(H) Effect of expression of plasma membrane-targeted mutant of sialin L22A-L23A on − NO3 and SA currents. (I) Average data and statistical evaluation for H.(J) Surface expression of sialin in WT cells and in cells transfected with scram-sh, sh- − sialin, or L22A-L23A mutant (AA). (K)NO3 and SA uptake measured at 10 min after loading in HSG cells transfected with sh-sialin or scram-sh. *Values significantly different from the unmarked values (P < 0.05; n = 6).

13436 | www.pnas.org/cgi/doi/10.1073/pnas.1116633109 Qin et al. − + Fig. 1, we assessed the involvement of sialin in 2NO /H co- the action of commensal bacteria in the oral cavity. This is not a − 3 transport. HSG cells accumulated NO , and this function was completely unexpected finding, given that the physiological SEE COMMENTARY 3 − significantly decreased in cells expressing sh-sialin compared to function of salivary gland cells is to mediate transepithelial NO3 cells expressing scrambled shRNA (scram-sh) (Fig. 2D; protein transport and concentrate the anion in saliva. A similar conclu- expression and mRNA levels are shown in Fig. 2 E and F). In- sion was reached in an earlier study, which showed that salivary + + − − tracellular levels of K ,Na ,andCl were not changed by sh-sialin glands deliver NO from the serum into the oral cavity with − 3 expression. Importantly, NO3 current in HSG cells maintained at minimal metabolism (2). normal or low pH (Fig. 2 G–I) was significantly reduced in sh- − + + sialin–treated cells. Sialin Mediates NO3 /H and SA/H Cotransport in HSG Cells. In- In addition, incubation of HSG and other cell types, including clusion of SA, a major substrate for sialin, in the external me- human colon carcinoma (RKO), human gastric mucosal epi- dium also induced currents when ambient pH was decreased. − thelial cells (GES-1), and human umbilical vein endothelial cells Although NO3 and SA individually (5 mM each) generated − (HE-CV-304), in medium containing NO resulted in increased outward currents at low external pH (Fig. 3A), the amplitude 3 − levels of intracellular NO as detected by diaminofluorescein with NO3 (11.2 ± 2.2 pA/pF; n = 15) was greater than that with (DAF) (Fig. S5 A and B); nitrite was not detected in the nitrate SA (5.4 ± 0.7 pA/pF; n = 7). When both anions were present loading solution. Athough HSG cells displayed nitrate conduc- together (10.6 ± 0.7 pA/pF; n = 9), the current amplitude was tance when exposed to physiological levels of nitrate (50–1,000 slightly more than that seen with SA alone, although it was not μM), indicating that these cells have an influx pathway that can a sum of the individual currents. The characteristics of currents − transport nitrate at relatively low [NO ], generation of NO and were similar in all of the conditions (Fig. 3B). As seen with 3 − cGMP was detected only at relatively high, nonphysiological levels NO3 , decreasing extracellular pH in the presence of 5 mM SA − fi C of the anions (≥15 mM nitrate and 3 mM NO2 , respectively) also led to intracellular acidi cation (Fig. 3 ). In addition, (Fig. S5 C and D). Note that other tissues, such as liver, can proton-dependent aspartate and glutamate currents, similar to − effectively metabolize both anions at lower concentrations (Fig. NO3 current, were detected in these cells (Fig. 3D), consistent S5E). This indicates that direct nitrate uptake via sialin is not a with a previous study indicating that sialin transports both of physiological pathway for NO generation in these cell types; these anions (20). Sialin knockdown decreased low-pH stimu- rather, once secreted, salivary nitrate is metabolized to nitrite by lated currents in HSG cells perfused with media containing SA + PHYSIOLOGY

− Fig. 4. Assessment of sialin-mediated NO3 transport in fibroblasts from patients with ISSD, showing the effect of the Salla disease and ISSD sialin mutations − on anion transport. (A–D)NO3 and SA (5 mM each) currents in HSG cells transfected with sialinR39C (B) or sialinH183R (C). (C, Insert) Average data and − − statistical evaluation. (D) I-V curve of the NO3 current shown in A and B.(E–H) Sialin-mediated current (150 mM SA or NO3 )infibroblasts from healthy volunteers (HV) (E and F) and patients with ISSD (G and H). (I and J) Effect of sialinH183R expression on anion transport in fibroblasts from patients with ISSD. (K) Average data from the experiments with cells from patients.

Qin et al. PNAS | August 14, 2012 | vol. 109 | no. 33 | 13437 − − − NO3 , SA, or NO3 (Fig. 3 E–G). Importantly, expression of the levels of NO3 in the saliva (152.5 ± 17.3 μM vs. 221.6 ± 8.5 μMat sialin mutant L22A-L23A, which reportedly is targeted to the 30 min and 204.1 ± 31.5 vs. 305.3 ± 27.8 μM after 60 min of feeding; plasma membrane (24, 25), increased the currents by approxi- P < 0.05; data obtained from fours pigs and eight parotid glands in − − mately twofold with either NO3 or SA in the external solution each group) (Fig. 5A). [NO3 ] in the saliva from animals over- (Fig. 3 H and I). This increase was also detected when the anions expressing WT-sialin was not significantly different than that from − were used at low concentrations (5 mM). Surface biotinylation control animals, and although serum and urine [NO3 ] levels were confirmed the presence of endogenous sialin in the plasma higher after feeding, there was no difference between the two membrane, which was decreased by the expression of sh-sialin groups. Western blot analysis (Fig. 5B) revealed greater sialin ex- J − and increased by expression of sialin L22A-L23A (Fig. 3 ). NO3 pression in parotid glands in pigs receiving the sialinH183R plasmid − > and SA were transported into HSG cells (NO3 uptake SA compared with controls (receiving AdCMV-EGFP-PEI alone). uptake), and uptake of both anions was reduced when sialin Furthermore, immunohistochemistry demonstrated a higher sialin expression was suppressed by sh-sialin (Fig. 3K). − signal in glands receiving sialinH183R compared with those in the Sialin also transports NO , as detected by activation of low- C D 2 − − control group (Fig. 5 and ). Taken together with the data pH currents when either NO3 or NO2 was included in the bath shown in Fig. 4, these data strongly suggest that nitrate transport D − ± n = solution (Fig. S3 ). NO2 current (7.5 1.0 pA/pF; 8) was in salivary glands could be potentially altered in patients with − ± n = smaller than NO3 current (11.2 2.2 pA/pF; 15), but was ISSD or Salla disease. not additive in solutions containing both anions (12.8 ± 2.4 pA/ pF; n = 5). Nitrite uptake was confirmed in cells for which in- Discussion + cluding nitrite or nitrate in the medium led to intracellular Herein we report a novel function for the lysosomal SA/H increases in the accumulation of the respective anion, and in cells SLC17A5 fi cotransporter, sialin ( ), which has been linked to Salla for which including SA with the anions did not signi cantly alter disease and ISSD. We show that sialin can also function as an the uptake of either anion (Fig. S3E). − + electrogenic 2NO3 /H cotransporter in the plasma membrane. − Other findings include that (i) sialin mutants associated with Assessment of Sialin-Mediated NO Transport in Fibroblasts from − 3 Salla disease and ISSD induce dominant suppression of 2NO / + 3 Patients with ISSD and Effect of the Salla Disease and ISSD Sialin H cotransporter activity, (ii) cotransporter activity is relatively Mutations on Anion Transport. To establish the link between sia- low in fibroblasts obtained from patients with ISSD compared lin and nitrate transport, we examined the effect of two non- with healthy volunteers, and (iii) expression of the ISSD-asso- functional sialin mutants that have been associated with Salla ciated sialin mutant (H183R) suppresses cotransporter function disease (R39C) and ISSD (H183R) (23–25). Expression of each in cells from healthy volunteers. Taken together, these data of these mutants in HSG cells induced dominant suppression of + − − provide strong evidence that sialin mediates H -dependent both NO and SA currents without altering the current-voltage − 3 NO uptake into cells. Although it is well established that the (I-V) characteristics (Fig. 4 A–D). A major finding of the present 3 + − study was that fibroblasts from patients with ISSD displayed lysosomal H /SA transport function mediated by sialin con- − a substantial reduction in pH-dependent NO or SA currents tributes to lysosomal recycling of SA, further studies are needed 3 to determine the exact physiological function of nitrate transport compared with the current amplitudes in cells from healthy fi volunteers (Fig. 4 E and G) with no change in the I-V re- mediated by sialin in broblasts and other cell types. However, it − lationship (Fig. 4 F and H). Note that a relatively higher [NO ] has been well established that salivary glands serve as a major − 3 level was required to detect NO3 conductance in fibroblasts. Although we could not induce recovery of function in cells obtained from patients with ISSD by expressing the WT-sialin (likely due to a dominant negative effect of the endogenously expressed mutant), expression of sialinH183R in cells from healthy volunteers strongly suppressed transporter function to levels seen in the cells from patients with ISSD (Fig. 4 I–K). Although we appreciate that including data demonstrating sali- vary deficiencies in patients would have strengthened our find- ings, several major issues preclude conducting such a study at the present time. There are very few patients with Salla disease or ISSD worldwide. In Salla disease, found primarily in Finland, newborns develop intellectual impairment gradually. ISSD, al- though not geographically restricted, is more severe, and patients generally die early in childhood or even in utero. Both disorders cause developmental delays. The majority of patients are chil- dren with severe developmental defects and poor overall health. Thus, although assessment of saliva in patients might yield sig- nificant data, this is beyond the scope of the present study. De- spite lack of such data, the findings presented herein establish a strong link between nitrate transport and sialin. We further demonstrate that disease-causing mutants of sialin also decreases − NO3 transport.

Effect of Adenovirus-Dependent Expression of Sialinh183r Mutant in Vivo in Pig Salivary Glands on Salivary Nitrate Secretion. To provide Fig. 5. Effect of in vivo suppression of sialin function in salivary glands on further evidence that sialin mediates salivary gland nitrate trans- salivary nitrate secretion. (A) Salivary nitrate concentrations at 30 min and 60 port, we used adenoviral vector containing cytomegalovirus pro- min after feeding a nitrate-rich diet to miniature pigs after in vivo delivery moter (AdCMV)-EGFP-polyethylenimine (PEI) complex as of plasmids encoding WT-sialin or sialinH183R (Methods). *Values signifi- a carrier to deliver sialinH183R or WT-sialin vector into pig salivary < − cantly different from the control values (P 0.05). (B) Sialin expression in glands. At 3 d after infection, the animals were fed a NO3 -rich diet control parotid glands and glands receiving the plasmid-PEI-Ad complex as for 30 min and then stimulated with pilocarpine to induce saliva determined by Western blot analysis. (C and D) Detection of sialin in salivary secretion. Compared with control animals or those that received glands at 3 d after transduction with AdCMV-EGFP-PEI plus sialinH183R WT-sialin, the pigs receiving sialinH183R displayed relatively lower vector (C) or in control parotid gland (D). (Scale bar: 50 μM.)

13438 | www.pnas.org/cgi/doi/10.1073/pnas.1116633109 Qin et al. − route for NO clearance from the serum. Approximately 25% large amount of ingested nitrite survives hydrolysis in the stomach 3 − of the circulating NO3 is taken up by the salivary glands, where and enters the systemic circulation, where it is reduced to NO and SEE COMMENTARY it is concentrated and secreted (at millimolar levels) in the saliva. other bioactive nitrogen oxides (1, 2). Thus, salivary nitrate fl − Thus, it is likely that transepithelial ux of NO3 occurs across transport provides an alternative, noncanonical pathway for the the salivary gland cells as the anion is taken up from the serum generation of nitrite and NO. This pathway appears to be par- and secreted into saliva. fi − ticularly signi cant under conditions of hypoxia and acidosis. We propose that sialin can function as the NO uptake system − 3 Taken together, these findings suggest that NO3 secretion via the in salivary gland cells. Consistent with this proposal, we found that gland can significantly impact the nitrate–nitrite–NO balance in the protein is localized in the basal and lateral regions. Further − fl the serum, which is of major importance in such conditions as studies are needed to identify the NO3 ef ux pathway likely to be high blood pressure, platelet aggregation, and vascular damage localized in the apical membrane of the cells. The physiological − (26–30). We suggest that disruption of sialin function, as seen in relevance of sialin in NO3 transport via the salivary gland is fi fi fi patients with Salla disease or ISSD, can have signi cant effects con rmed by our nding that in vivo expression of the dominant – – negative sialin mutant (sialinH183R) in pig parotid glands re- on the nitrate nitrite NO balance, in addition to the previously − − recognized impairment in SA storage and aspartergic neuro- duced NO3 secretion in the saliva in response to a NO3 -rich diet compared with the function in control animals and those transmission. Furthermore, loss of salivary gland secretory ac- expressing WT-sialin in the glands. Together with our findings tivity as a result of radiation treatment for head and neck cancers − or autoimmune disease (e.g., Sjogren’s syndrome) can have po- demonstrating that sialin mediates NO3 uptake into salivary gland cells, these data provide evidence for a physiological role tential systemic consequences owing to an imbalance of nitrite– of sialin in nitrate uptake into these glands. In addition, sialin NO homeostasis. Together, the findings presented herein dem- appears to be a versatile anion transporter that also has the ability onstrate that sialin, as a nitrate cotransporter in the salivary + − to mediate H -dependent transport of SA, NO2 , aspartate, or glands, can play an important role in the physiological regulation glutamate. Given the protein’s relatively wide distribution in dif- of systemic nitrate–nitrite–NO balance. ferent tissues, its function in the plasma membrane as well as ly- sosomes can have a significant impact on cell function, including Methods SA recycling as well as nitrate–NO balance. In neuronal tissues, All reagents and detailed methods are described in SI Methods. These include there might be additional consequences related to sialin’s ability cell culture, electrophysiology, confocal imaging, and biochemical techni- to transport glutamate and aspartate. ques such as surface biotinylation and western blotting. Descriptions of In summary, our data provide insight into the function of sialin experiments with miniature pigs are also provided in SI Methods. and demonstrate its unique function as a nitrate transporter. In − mammals, the diet is a major source of NO ; absorption of di- − 3 − ACKNOWLEDGMENTS. We thank Dr. Jim Turner, Dr. Shmuel Muallem, Dr. etary NO3 results in an increase in serum NO3 . Approximately Jing Jiang, and Professor Liangbiao Chen for their invaluable help during − the course of this work. This study was supported by the National Nature 25% of the circulating NO3 is taken up into salivary gland and secreted via saliva, where it is reduced to nitrite by the action of Science Foundation of China (Grants 30430690, 30125042, and 81170975), the National Basic Research Program of China (Grants 2007CB947304 and commensal bacteria in the oral cavity and then converted to NO 2010CB944801), and the Divisions of Intramural Research of the National in the stomach. NO is suggested to have an important role in the Institute of Dental and Craniofacial Research and the National Human protection of gastric tissues from stress-induced injury; however, a Genome Research Institute.

1. Lundberg JO, Weitzberg E, Gladwin MT (2008) The nitrate-nitrite-nitric oxide path- 16. Brown S, Chang JL, Sadée W, Babbitt PC (2003) A semiautomated approach to gene way in physiology and therapeutics. Nat Rev Drug Discov 7:156–167. discovery through expressed sequence tag data mining: Discovery of new human 2. Jansson EA, et al. (2008) A mammalian functional nitrate reductase that regulates transporter genes. AAPS PharmSci 5:E1. nitrite and nitric oxide homeostasis. Nat Chem Biol 4:411–417. 17. Nishimura M, Naito S (2008) Tissue-specific mRNA expression profiles of human solute 3. Djekoun-Bensoltane S, Kammerer M, Larhantec M, Pilet N, Thorin C (2007) Nitrate and carrier transporter superfamilies. Drug Metab Pharmacokinet 23:22–44. nitrite concentrations in rabbit saliva: Comparison with rat saliva. Environ Toxicol 18. Aula N, Jalanko A, Aula P, Peltonen L (2002) Unraveling the molecular pathogenesis Pharmacol 23:132–134. of free sialic acid storage disorders: Altered targeting of mutant sialin. Mol Genet 4. Pannala AS, et al. (2003) The effect of dietary nitrate on salivary, plasma, and urinary Metab 77:99–107. nitrate metabolism in humans. Free Radic Biol Med 34:576–584. 19. Aula N, Kopra O, Jalanko A, Peltonen L (2004) Sialin expression in the CNS implicates 5. Xia DS, Deng DJ, Wang SL (2003) Destruction of parotid glands affects nitrate and extralysosomal function in neurons. Neurobiol Dis 15:251–261. nitrite metabolism. J Dent Res 82:101–105. 20. Miyaji T, et al. (2008) Identification of a vesicular aspartate transporter. Proc Natl 6. Xia D, Deng D, Wang S (2003) Alterations of nitrate and nitrite content in saliva, Acad Sci USA 105:11720–11724. serum, and urine in patients with salivary dysfunction. J Oral Pathol Med 32:95–99. 21. Yarovaya N, et al. (2005) Sialin, an anion transporter defective in sialic acid storage 7. Benjamin N, et al. (1994) Stomach NO synthesis. Nature 368:502. diseases, shows highly variable expression in adult mouse brain, and is de- 8. Arreola J, Melvin JE (2003) A novel chloride conductance activated by extracellular velopmentally regulated. Neurobiol Dis 19:351–365. ATP in mouse parotid acinar cells. J Physiol 547:197–208. 22. He M, et al. (2011) Postnatal expression of sialin in the mouse submandibular gland. 9. Barbier-Brygoo H, et al. (2000) Anion channels in higher plants: Functional charac- Arch Oral Biol 56:1333–1338.

terization, molecular structure and physiological role. Biochim Biophys Acta 1465: 23. Verheijen FW, et al. (1999) A new gene, encoding an anion transporter, is mutated in PHYSIOLOGY 199–218. sialic acid storage diseases. Nat Genet 23:462–465. 10. Gong X, Linsdell P (2003) Mutation-induced blocker permeability and multiion block 24. Morin P, Sagné C, Gasnier B (2004) Functional characterization of wild-type and of the CFTR chloride channel pore. J Gen Physiol 122:673–687. mutant human sialin. EMBO J 23:4560–4570. 11. Pao SS, Paulsen IT, Saier MH, Jr. (1998) Major facilitator superfamily. Microbiol Mol 25. Wreden CC, Wlizla M, Reimer RJ (2005) Varied mechanisms underlie the free sialic Biol Rev 62:1–34. acid storage disorders. J Biol Chem 280:1408–1416. 12. Guo FQ, Wang R, Crawford NM (2002) The Arabidopsis dual-affinity nitrate trans- 26. Lundberg JO, Carlström M, Larsen FJ, Weitzberg E (2011) Roles of dietary inorganic porter gene AtNRT1.1 (CHL1) is regulated by auxin in both shoots and roots. J Exp Bot nitrate in cardiovascular health and disease. Cardiovasc Res 89:525–532. 53:835–844. 27. Cosby K, et al. (2003) Nitrite reduction to nitric oxide by deoxyhemoglobin vasodilates 13. Jia W, Tovell N, Clegg S, Trimmer M, Cole J (2009) A single channel for nitrate uptake, the human circulation. Nat Med 9:1498–1505. nitrite export and nitrite uptake by Escherichia coli NarU and a role for NirC in nitrite 28. Gladwin MT, et al. (2000) Relative role of heme nitrosylation and beta-cysteine 93 export and uptake. Biochem J 417:297–304. nitrosation in the transport and metabolism of nitric oxide by hemoglobin in the 14. Martín Y, Navarro FJ, Siverio JM (2008) Functional characterization of the Arabidopsis human circulation. Proc Natl Acad Sci USA 97:9943–9948. thaliana nitrate transporter CHL1 in the yeast Hansenula polymorpha. Plant Mol Biol 29. Modin A, et al. (2001) Nitrite-derived nitric oxide: A possible mediator of ‘acidic- 68:215–224. metabolic’ vasodilation. Acta Physiol Scand 171:9–16. 15. Wang R, Liu D, Crawford NM (1998) The Arabidopsis CHL1 protein plays a major role 30. Dejam A, Hunter CJ, Gladwin MT (2007) Effects of dietary nitrate on blood pressure in high-affinity nitrate uptake. Proc Natl Acad Sci USA 95:15134–15139. [letter]. N Engl J Med 356:1590–author reply 1590.

Qin et al. PNAS | August 14, 2012 | vol. 109 | no. 33 | 13439 Supporting Information

Qin et al. 10.1073/pnas.1116633109 SI Methods (Thermo Scientific). Reaction was stopped with PBS + 100 mM Cell Culture. HSG and other cells were cultured as described glycine, cells were lysed, and biotinylated fractions were iso- previously (1). ShRNA against sialin (sh-sialin) plasmid DNA lated using NeutrAvidin beads (Thermo Scientific). Samples (TR309394) and scram-sh (TR30003, OriGene) were transfected were eluted and separated by PAGE. Sialin was detected by into HSG cells for 48 h. Various mutants of sialin and fibroblast Western blot analysis using 1:1000 anti-sialin antibody and HRP- cells from patients diagnosed with ISSD and healthy volunteers conjugated secondary antibody (KPL), followed by enhanced were used as reported previously (2, 3). chemiluminescence (ECL) method. Parotid gland, submandi- bular gland, liver, and spleen tissues from normal miniature pigs Electrophysiological Measurements. Standard whole-cell patch-clamp were harvested, and samples were used to assess sialin expres- methodology was used to assess membrane currents as described sion by a similar Western blot analysis technique using the anti- previously (1), NaCl was replaced with NaNO3,NaNO2,freeSA, sialin antibody. NaBr, or Na-gluconate as indicated. huSMG cells were dispersed as described previously (4), and acinar cells were identified under Uptake of Nitrate or Sialic Acid into Cells. Cells were incubated in a microscope for electrophysiological measurements. PBS containing 15 mM nitrate or sialic acid (SA) and then washed, lysed, and centrifuged. The supernatant was used for analysis of Immunofluorescence Staining and Confocal Microscopy Examination. nitrate or SA by HPLC using a C18 reversed-phase column (125 mm Salivary gland sections were treated with the required primary × 4.6 mm i.d., 5 μm particle size). The mobile phase included 83% antibodies [anti-sialin (1:100 dilution), anti-LAMP-1 (1:50 dilution), 3.0 mM tetrabutylammonium hydroxide titrant and 2.0 mM so- and anti-Na+/K+ ATPase (1:50 dilution); all from Santa Cruz Bio- dium phosphate buffer (pH 3.9) and 17% of acetonitrile organic technology], followed by treatment with the Dylight 649-conjugated solvent (Sigma-Aldrich), with a flow rate of 0.4 mL/min. Nitrate rabbit anti-goat IgG (KPL) or FITC-conjugated donkey anti-mouse was measured by reading the UV absorbance at 205 nm (11). For IgG (Proteintech). Goat anti-human IgG isotype was used for SA measurement, SA in samples was first labeled with O-phe- negative controls. Fluorescence was measured using a Leica nylenediamine (OPD) as described previously (12). OPD-SA confocal microscope. derivatives were detected by the HPLC. The mobile phase was μ 92% of 0.2% H3PO4, including 1.0% tetrahydrofuran, and 8% Measurement of Intracellular pH. HSG cells were loaded with 1 M acetonitrile organic solvent; the flow rate was 1.0 mL/min. OPD-SA BCECF-AM (Invitrogen) and exposed to either SA or NO3− in an derivatives were measured as UV absorbance at 227 nm. [Nitrate] external solution (pH indicated) in which 145 mM external Cl− or [SA] was normalized based on [protein]. was replaced with gluconate. BCECF fluorescence was recorded with excitation at 440 nm and 490 nm and emission at 510 nm. pH Dietary Nitrate Loading in Miniature Pigs and Effect of SA on Nitrate was calibrated using a 490/440 nm ratio as described previously or Nitrite Uptake in HSG Cells. Miniature pigs (male, age 6 mo, fl (5, 6). The BCECF uorescence ratio (490/440 nm) was linear weight 35–45 kg) were fed regular food or food containing po- – within a pH range of 6.2 7.6. tassium nitrate (100 mg/kg of body weight) after a 12-h fast. Parotid saliva and blood samples were collected at different time In Vivo Knockdown of Sialin in Parotid Glands of Miniature Pigs. points (10), and nitrate concentration was measured as described Adenovirus–PEI–plasmid complex was used to deliver sialin-en- above. After incubation in PBS containing 5 mM nitrate or ni- coding plasmids into miniature pig salivary glands in vivo (7). A trite without or with same concentration of SA for 25 min, HSG mixture of 50 μg of plasmid DNA/gland, 0.1 mM PEI, and 2 × 10 11 cells were washed and lysed. Nitrate and nitrite concentrations in particles of AdCMV-EGFP in a volume of 3 mL was delivered supernatant were measured with nitrate and nitrite kit (R&D). into a single parotid gland by duct cannulation (8, 9). Twelve – animals (male miniature pigs, 5 mo old, weighing 25 30 kg) were Intracellular NO Detection by Flow Cytometry Analysis Using the divided into three groups (control, overexpression, and knock- Fluorescent Probe Diaminofluorescein. Suspensions of HSG, RKO, down groups). At 3 d after virus delivery, the animals were fed GES-1, and HECV304 cells (∼0.5 × 106 cells/mL) were incu- potassium nitrate (40 mg/kg body weight), then anesthetized and batedwith10μMDAF-2/DA(dissolvedinKrebs–Henseleit buffer treated with 0.5 mg/kg of pilocarpine i.m. Serum, urine, and pa- containing 2% BSA) for 2 h. Experimental conditions were (i) rotid saliva (8, 9) samples were collected at 30 and 60 min. Samples control cells incubated with anion-free medium, (ii) cells incu- were prepared and analyzed as reported previously (10). Parotid bated with different concentrations of nitrate (0–30 mM) with or glands were excised and used to detect sialin expression. without 100 μM allopurinol, (iii) cells incubated with different concentrations of nitrate (0–3 mM) with or without 100 μM Statistical Analyses. All statistical analyses were done using Origin iv μ 8 (OriginLab) or SPSS 12.0 software. Data for normal distributions allopurinol, and ( ) cells incubated with 100 M NO donor are presented as mean ± SEM. The independent t test was used to (SNP; Sigma-Aldrich) as the positive control. After incubation fl determine the statistical analysis of normally distributed means. A and washing, intracellular uorescence of DAF-triazol (DAF- fl P value < 0.05 was considered to indicate statistical significance. 2T,anoxidizedformofDAF-2/DA)wasanalyzedby ow cytometry. Fluorescence-activated cell sorting analysis data Analysis of Sialin Expression. The following primers were used for were expressed as mean fluorescence intensity (percentage of qPCR analysis for sialin (encoded by SLC17A5; 231 bp): forward, control). 5′-AAC TTC CAA GGC ATG TCC AG-3′; reverse, 5′-ATG For control experiments using liver, liver tissue samples from GGA GTG AAC AGG GTG AG-3′. β-actin or GAPDH was male minipigs were homogenized in 10 mM Tris-HCl plus 250 mM sucrose medium, and supernatant fractions were collected used as a reference control. – (13). Aliquots (8 mg mL 1 protein) were incubated with 0.5 mM Biotinylation Assays. Surface biotinylation of sialin was measured NaNO3 or 0.1 mM NaNO2, with or without 100 μM allopurinol in human salivary gland (HSG) cells using sulfo-NHS-LC-biotin in medium containing 1 mM NADPH, 2 mM UDP glucuronic

Qin et al. www.pnas.org/cgi/content/short/1116633109 1of6 acid, 0.5 mM glutathione, 0.5 mM NAD+, and NADH (all from treatment of cells. Cells were pretreated with a phosphodiesterase Sigma-Aldrich). NO was measured by DAF fluorescence with a 5 inhibitor, sildenafil(5μM; Sigma-Aldrich), for 15 min, followed Promega Glomax multidetector system, with excitation at 490 by treatment with sodium nitrate (30 mM), sodium nitrite nm and emission at 530 nm. (3 mM), or SNP (100 μM) in the presence of sildenafil for 3 h, μ Measurement of cGMP. Intracellular cGMP concentration was with or without 100 M allopurinol. cGMP levels were measured measured in the cell lysates using cGMP kit (BioVersion) after in the supernatants. The data are expressed as pmol/g protein.

1. Liu X, Liao D, Ambudkar IS (2001) Distinct mechanisms of [Ca2+]i oscillations in HSY 8. Li J, et al. (2004) Developing a convenient large animal model for gene transfer to and HSG cells: Role of Ca2+ influx and internal Ca2+ store recycling. J Membr Biol 181: salivary glands in vivo. J Gene Med 6:55–63. 185–193. 9. Shan Z, et al. (2005) Increased fluid secretion after adenoviral-mediated transfer of 2. Kleta R, et al. (2003) Biochemical and molecular analyses of infantile free sialic acid the human aquaporin-1 cDNA to irradiated miniature pig parotid glands. Mol Ther storage disease in North American children. Am J Med Genet A 120A:28–33. 11:444–451. 3. Renlund M, Tietze F, Gahl WA (1986) Defective sialic acid egress from isolated 10. Xia DS, Deng DJ, Wang SL (2003) Destruction of parotid glands affects nitrate and fibroblast lysosomes of patients with Salla disease. Science 232:759–762. nitrite metabolism. J Dent Res 82:101–105. 4. Arreola J, Melvin JE (2003) A novel chloride conductance activated by extracellular 11. Zuo Y, Wang C, Van T (2006) Simultaneous determination of nitrite and nitrate ATP in mouse parotid acinar cells. J Physiol 547:197–208. in dew, rain, snow and lake water samples by ion-pair high-performance liquid 5. Muallem S, Zhang BX, Loessberg PA, Star RA (1992) Simultaneous recording of cell chromatography. Talanta 70:281–285. volume changes and intracellular pH or Ca2+ concentration in single osteosarcoma 12. Anumula KR (1995) Rapid quantitative determination of sialic acids in glycoproteins cells UMR-106-01. J Biol Chem 267:17658–17664. by high-performance liquid chromatography with a sensitive fluorescence detection. 6. Thomas JA, Buchsbaum RN, Zimniak A, Racker E (1979) Intracellular pH measurements Anal Biochem 230:24–30. in Ehrlich ascites tumor cells utilizing spectroscopic probes generated in situ. 13. Jansson EA, et al. (2008) A mammalian functional nitrate reductase that regulates Biochemistry 18:2210–2218. nitrite and nitric oxide homeostasis. Nat Chem Biol 4:411–417. 7. Zheng C, Baum BJ (2005) Evaluation of viral and mammalian promoters for use in gene delivery to salivary glands. Mol Ther 12:528–536.

− Fig. S1. Effect of anion channel inhibitors on the NO3 currents in HSG cells. Effects of four anion channel blockers—DIDS (A), CHC (B), ECA (C), and NPPB (D)— − on NO3 currents in HSG cells. Each trace is representative of between four and six cells for each experiment. Here 100 μM ECA, 5 mM DIDS, 100 μM NPPB, and 1 mM CHC were added to the external solution. The apparent order of inhibition of the outward current is ECA > DIDS > NPPB > CHC.

Qin et al. www.pnas.org/cgi/content/short/1116633109 2of6 2+ − Fig. S2. Effects of CCh, ATP, cAMP, and Ca on NO3 currents in HSG cells. Effects of the muscarinic agonist CCh, purinergic agonist ATP, intracellular cAMP, 2+ − and external and internal Ca were investigated. In the presence of high external NO3 , a spontaneous current was seen that was not changed by perfusion with 100 μM CCh (A) or 5 mM ATP (B). The low pH-induced current was activated when exposed to low pH, as indicated by corresponding bars. (C) Here 20 μM − − cAMP was included in the patch pipette solution. (D) The cell was perfused first with Cl and then with NO3 solution, as indicated by corresponding bars. The respective current-voltage (I-V) curves represent the time points indicated in C and D.(E and F) External Ca2+-containing solution was replaced with Ca2+-free − medium, and 10 mM EGTA was included in the patch pipette solution as indicated. Each trace is a representative of three or four cells for each group. NO3 conductance was not regulated by cAMP or by muscarinic or purinergic receptor stimulation and showed no dependence on intracellular or extracellularCa2+.

Qin et al. www.pnas.org/cgi/content/short/1116633109 3of6 − + + + Fig. S3. Characteristics of NO3 current in HSG cells. (A and B) Effects of replacing Na with Cs on low-pH-induced nitrate current. 145 mM Na was replaced + − with equal amount of Cs and currents were recorded at different carriers indicated by the bars in A, and the corresponding I-V curve is shown in B.(C)NO3 currents recorded at relatively low nitrate concentration, ranging from 50 μM to 5 mM. (D) Currents carried by nitrite and/or nitrate. (E) Intracellular nitrate or nitrite accumulation after the cells were loaded in 5 mM anion-containing medium (nitrate or nitrite) for 25 min (Left), and the intracellular nitrate or nitrite level after the addition of 5 mM SA together with 5 mM nitrate or nitrite loading (Right).

Fig. S4. Sialin expression in different tissues from miniature pigs as demonstrated by Western blot analysis. Sialin was highly expressed in salivary glands and liver, with lower levels of expression in brain, spleen, and kidney and even lower levels in muscle and pancreas.

Qin et al. www.pnas.org/cgi/content/short/1116633109 4of6 Fig. S5. Effect of nitrate loading on intracellular NO levels. (A) HSG, RKO, GES-1, and HECV-304 cells were incubated in DAF-2/DA Krebs–Henseleit buffer with 30 mM nitrate (pH 7.4) for 2 h at 37 °C. Intracellular NO production of cells was recorded as mean fluorescence intensity increases of 154%, 151%, 112%, and 117% compared with respective controls, which were incubated in nitrate-free medium. (B) The increase in intracellular NO was dependent on extracellular nitrate concentration. (C) Increases in DAF fluorescence in HSG cells after incubation with 30 mM nitrate, 3 mM nitrite, and 100 μM SNP. Allopurinol (100 μM) was added where indicated (+A). **Statistically significant difference compared with unmarked value (control) as well as those marked ** (allopurinol- treated); P < 0.01. (D) Intracellular cGMP level after incubation of the cells in SNP, nitrate, or nitrite in the presence or absence of allopurinol (A). *Values significantly different from unmarked values and those marked **P < 0.01. (E) Increase in DAF fluorescence in liver homogenates incubated with 500 μM nitrate, 100 μM nitrite, or 100 μM SNP. Allopurinol (100 μM) was added where indicated (+A). *Values significantly different from unmarked values and those marked **P < 0.01.

Qin et al. www.pnas.org/cgi/content/short/1116633109 5of6 Table S1. Expression of MFS genes in human parotid and submandibular glands (n =3) Expressed gene Description

SLC17A5 H+/anion cotransporter inhibitted by DIDS SLC15A2 H+/peptide cotransporter SLC15A4 H+/peptide cotransporter SLC16A1 Monocarboxylate transporter inhibited by DIDS and CHC SLC16A2 Monocarboxylate transporter inhibitted by DIDS and CHC SLC16A3 Monocarboxylate transporter inhibitted by DIDS and CHC SLC16A4 Monocarboxylate transporter inhibited by DIDS and CHC SLC16A7 Monocarboxylate transporter inhibited by DIDS and CHC SLC16A11 Monocarboxylate transporter inhibited by DIDS and CHC SLC16A14 Monocarboxylate transporter inhibited by DIDS and CHC SLC18A2 Amine/ proton anti-porter, not inhibited by DIDS SLC2A8 Class III glucose transporter SLC2A9 Class II glucose transporter SLC2A10 Class III glucose transporter SLC20A1 Type III sodium-dependent phosphate cotransporter SLC21A2 Solute carrier organic anion transporter SLC22A3 Organic cation transporter SLC22A5 Organic cation transporter SLC22A18 Organic cation transporter SLC33A1 Acetyl-CoA transporter SLC37A2 Sugar-phosphate exchanger SLC37A3 Sugar-phosphate exchanger SLC43A3 pH-independent L amino acid transporter 4 MFSD1 Major facilitator superfamily domain containing 1 NR3C1 Glucocorticoid receptor DIRC2 Disrupted in renal carcinoma 2 C6orf192 Chromosome 6 ORF 192

Gene expression was analyzed in normal human tissues from three donors (male, age 18–30 y) using micro- array chips (1).

1. Sun QF, Sun QH, Du J, Wang S (2008) Differential gene expression profiles of normal human parotid and submandibular glands. Oral Dis 14:500–509.

Qin et al. www.pnas.org/cgi/content/short/1116633109 6of6