Identification of Cytosolic and Microsomal Bile Acid-binding Proteins in Rat Ileal Enterocytes*

Studies were performed to determine the subcellular fractions and proteins involved in the intracellular transport of bile acids in rat ileal cells. The photolabile derivative 7,7-azo-taurocholate inhibited the Na+-de- pendent uptake of taurocholate into rat ileal entero- cytes reversibly in the dark and irreversibly following photolysis. When photolabeled cells were submitted to subcellular fractionation, greatest radioactivity was found in the soluble protein (SP) fraction with decreas- ing radioactivity in the brush-border-(BBM), basolateral- (BLM), mitochondria-(MT), microsome-(MC), and Golgi-(GO) enriched fractions. Following trichlo- roacetic acid precipitation, delipidation, and correction for loss of marker enzyme activity, radioactivity photolabeled cells

Studies were performed to determine the subcellular fractions and proteins involved in the intracellular transport of bile acids in rat ileal cells. The photolabile derivative 7,7-azo-taurocholate inhibited the Na+-dependent uptake of taurocholate into rat ileal enterocytes reversibly in the dark and irreversibly following photolysis.
When photolabeled cells were submitted to subcellular fractionation, greatest radioactivity was found in the soluble protein (SP) fraction with decreasing radioactivity in the brush-border-(BBM), basolateral-(BLM), mitochondria-(MT), microsome-(MC), and Golgi-(GO) enriched fractions. Following trichloroacetic acid precipitation, delipidation, and correction for loss of marker enzyme activity, protein bound radioactivity was in SP > BBM > MC > BLM > GO > MT. When photolabeled cells were first fractionated and then submitted to sodium dodecyl sulfate-polyacrylamide gel electrophoresis, a 99-kDa polypeptide was associated with BBM, 54-and 59-kDa polypeptides with BLM, and 6%kDa polypeptides with SP and a 20-kDa polypeptide with MC fractions.
Immunoprecipitation with known antisera identified the 68-kDa polypeptide as albumin and the 43-kDa polypeptide as actin. No precipitation on the 14-kDa polypeptide was noted with anti-hepatic and anti-intestinal fatty acid-binding proteins.
No precipitation of the 35-kDa polypeptide occurred with antibody to the hepatic cytosolic bile acid-binding protein. These studies reveal a previously unrecognized 20-kDa microsomal, and 14-and 35-kDa cytosolic bile acid-binding polypeptides which may be involved in the trancellular movement of bile acids.
Bile acids are acidic sterols that undergo an enterohepatic circulation which involves their synthesis from cholesterol in the liver, secretion in bile, and storage in the gallbladder. Upon food ingestion the gallbladder contracts, emptying its contents into the lumen of the small intestine where bile acids are absorbed across the brush-border (microvillus) membrane by passive diffusion and, in addition, by a sodium-dependent active transport process in the ileum (1). Bile acids may exit the enterocyte via anion-exchange across the basolateral membrane (2) prior to their entry into the portal circulation and return to the liver. Affinity labeling with photolabile bile acid derivatives has been used successfully to identify bile acid-binding polypeptides in blood (3,4), hepatocytes (5)(6)(7), liver plasma membranes (3, E&10), and brush-border membranes (11) and basolateral membranes (12) of the intestine. These intestinal studies demonstrated that the photolabile derivatives shared transporters with natural bile acid and identified a 99-kDa polypeptide and 54-and 59-kDa polypeptides which may be involved in bile acid transport across brush-border and basolatereal membranes of the ileal epithelial cell, respectively.
In contrast to our knowledge regarding bile acid transport across the intestinal plasma membranes, relatively little is known regarding those events involved in the intracellular movement across the enterocyte. A distinct advantage of photoaffinity labeling is that a covalent linkage between the photoprobe and binding protein remains intact during cell processing. This advantage has been exploited in the present studies which involve the photoaffinity labeling of enterocytes followed by cell fractionation to localize the subcellular components and putative proteins involved in bile acid transport. without 0.5 mM 7,7-azo-TC for 10 min under red lighting. After washing the cells free of 7,7-azo-TC, the uptake of 0.1 mM taurocholate was measured in the presence or absence of Na' in the incubation media. As shown in Fig. 3 (bars 1 and 2), no significant difference in taurocholate uptake was seen between cells preincubated without or with the photolabile derivative. This experiment demonstrated that the inhibition by 7,7-azo-TC under subdued light was completely reversible. The next experiments were performed to determine whether photolysis of cells in the presence of the photoaffinity probe would result in an irreversible inhibition of taurocholate uptake. Irradiation of the cells for 10 min in the absence of 7,7azo-TC had no effect on subsequent uptake of taurocholate ( Fig. 3, bar 3) when compared with non-irradiated cells (Fig.  3, bar I). However, if ileal cells were exposed to UV light for 10 min in the presence of 0.5 mM 7,7-azo-TC, the subsequent uptake of Na+-dependent taurocholate was significantly reduced (Fig. 3, bar 4). Photoaffinity labeling under all conditions had no effect on taurocholate uptake in the absence of Na'. These studies suggest that photoreacted 7,7-azo-TC irreversibly inhibited the Na+-dependent bile acid uptake system in ileal cells.

Interaction of 7,7-Azo-TC with Taurocholate Uptake by Zleal
Cells under Red Lighting-In order to identify putative transport proteins by photoaffinity labeling, it first must be demonstrated that the photolabile derivative can share the transport system with natural substrates (23). The addition of Na+ to the incubation media was shown previously to stimulate bile acid uptake by isolated ileal cells (24). Therefore, the uptake of 7,7-azo-TC into ileal cells was studied in the presence and absence of Na'. Fig. 2A shows the effect of Na' and taurocholate on [3H]7,7-azo-TC uptake by ileal cells under red lighting. The presence of sodium instead of choline in the incubation media resulted in the stimulation of 0.1 mM 7,7azo-TC uptake. In addition, sodium-dependent 7,7-azo-TC uptake was inhibited by 0.25 mM taurocholate, whereas uptake of 7,7-azo-TC in the absence of sodium was unaffected by the presence of taurocholate in the incubation media. Conversely, Fig. 2B shows the interaction of 7,7-azo-TC with the uptake of taurocholate by ileal cells. The uptake of 0.1 mM taurocholate was stimulated by the presence of Na' in the incubation media. The Na+-dependent uptake of taurocholate was completely abolished by the presence of 0.5 mM 7,7-azo-TC, whereas 7,7-azo-TC had no effect on taurocholate uptake in the absence of Na'. Inhibition of Taurocholate Uptake in Zleal Cells by Photoaffinity Labeling with 7,7-Azo-TC-To test whether the inhibition of taurocholate uptake by 7,7-azo-TC was reversible without photolysis, the ileal cells were preincubated with or

Distribution of Marker Enzyme Activities-After ileal cells were photolyzed
in the presence of 7,7-azo-TC and disrupted in the Parr bomb, the resulting organelles, membrane fragments, and cytosol were fractionated by differential centrifugation and sorbitol density gradient separation (Fig. 1). Marker enzyme activities were used to measure the relative enrichment of the subcellular fractions. Galactosyl transferase, K+-stimulatedp-nitrophenyl phosphatase, alkaline phosphatase and NADPH-cytochrome c reductase were used as marker enzymes for Golgi (GO), basolateral membranes (BLM), brush-border membranes (BBM), and microsomes (MC), respectively. Specific activities of these enzymes were measured in the differential centrifugation and density gradient fractions.
The marker enzyme distribution of these fractions are summarized in Table I  tosyl transferase in the 20/30% sorbitol fraction identified this fraction as being relatively enriched in Golgi membranes. The specific activities of Kc-stimulated p-nitrophenyl phosphatase and NADPH-cytochome c reductase were negatively enriched in this fraction. The relative enrichment of 9.4-fold over starting cell homogenate for K+-stimulated p-nitrophenyl phosphatase in the 30/40% sorbitol fraction indicated that this fraction was relatively enriched in basolateral membranes. The specific activities of Golgi, brush-border membrane, and microsomal enzymes were enriched 3.5, 0.4-, and l.O-fold, respectively. The 40/50% and 50% sorbitol fractions contained relatively enriched specific activities for alkaline phosphatase and NADPH-cytochrome c reductase. Magnesium precipitation was used to separate the brush-border membrane-and microsomal membrane-enriched fractions. As shown in Table I, the addition of 10 mM magnesium resulted in a 7-fold enrichment in the alkaline phosphatase and 5-fold enrichment in the NADPH cytochrome c reductase activities in the supernatant and pelleted fractions, respectively. With the exception of galactosyl transferase in the supernatant fraction, the specific activities of the marker enzymes were negatively enriched. The specific activities of the four marker enzymes also were not enriched with respect to starting cell homogenate in the mitochondrial (MT) and soluble protein (SP) fractions. Finally, the activities of the marker enzymes were not enriched in the 50/60%, 60%, and pelleted sorbitol fractions (data not shown).
Distribution of Radioactivity-The intracellular distribution of [3H]7,7-azo-TC was assessed initially by measuring the total amount of radioactivity associated with each enriched subcellular fraction. Eighty-two % of the radioactivity in the cell homogenate was accounted for in the subcellular fractions; the remaining radioactivity presumably was in the uncollected sorbitol gradient fractions. As shown in Table II, column 2, the majority of the radioactivity (70.7%) was found in the 200,000 x g supernatant (cytosolic or soluble protein fraction). To determine the amount of radioactivity that was protein-bound, each fraction was incubated with 10% trichloroacetic acid at 4 "C overnight.
The reaction mixtures were filtered through 0.45~pm pore size cellulose nitrate filters and washed with 5 ml of 5% trichloroacetic acid, and the amount of radioactivity remaining on the filters was counted. Table  II (compare columns 2 and 3) shows that 37% of the total radioactivity in the cell homogenate fraction was found in the trichloroacetic acid-precipitable cell homogenate fraction. The failure to account for all cell homogenate total radioactivity in the trichloroacetic acid-precipitable cell homogenate may be explained by incomplete photolysis of 7,7-azo-TC. To test for this possibility, lo-p1 aliquots of cell homogenate were submitted to thin layer chromatography using the solvent system n-butanollacetic acid/water (9:2:1, by volume) and cochromatographed with unphotolyzed and photolyzed 7,7-azo-TC as standards. After the plates were developed in the dark, 53.8 f 1.84% (n = 4) of the radioactivity had an RF value identical to that of unphotolyzed 7,7-azo-TC. No activity was found with an RF value identical to that of photolyzed 7,7-azo-TC, and the remainder of radioactivity was found at the origin of the plates. Inasmuch as cell homogenate is not expected to run in the solvent system, these observations suggested that photolyzed 7,7-azo-TC was bound to membrane protein which remained at the origin.
To reduce the likelihood of nonspecific partitioning into membrane lipid, filters containing the trichloroacetic acidprecipitated fraction were further washed with 5 ml of chloroform/ethanol (2:1, by vol), and the radioactivity remaining on the filters was counted. As shown in Table II, column 4, this extraction step resulted in little reduction of radioactivity for each fraction. The majority of radioactivity found in the TCA-precipitated, lipid-extracted cell homogenate was accounted for in the subcellular fractions with decreasing percentage of radioactivity in these fractions as follows: SP, 32.3%; BBM, 12.2%; BLM, 4.6%; MC, 3.9%; GO, 0.7%. As shown in Table II (3.9) 27.3 60 a % equals radioactivity in trichloroacetic acid-precipitable, delipidated fraction divided by total radioactivity in cell homogenate and the percent yield (Y%) of corresponding marker enzyme activity. Additional studies addressed the question of specific covalent labeling of subcellular fractions by 7,7-azo-TC. Ileal cells were exposed to [3H]7,7-azo-TC in the presence of UV light with 0 or 0.5 mM taurocholate or absence of both UV light and taurocholate.
Following these incubations the ileal cells were fractionated, and the subcellular fractions were precipitated with 10% trichloroacetic acid and extracted with chloroform/methanol. The amount of radioactivity remaining was divided by the amount of protein in each fraction, i.e. the specific radioactivity.
As shown in Fig. 4, highest specific radioactivity was found in the brush-border membrane-enriched fraction with progressively decreasing specific radioactivities in BLM, MC, homogenate, SP, GO, and MT fractions. Moreover, photolysis in the presence of 0.5 mM TC reduced the specific radioactivities by upward of 50%, suggesting the natural bile acid competed with [3H]7,7-azo-TC for the protein-binding site or a proximal transport step. Finally, incubation of ileal cells with 7,7-azo-TC under subdued lighting resulted in specific radioactivities that were <lo% of those values encountered when each fraction was obtained from cells that were photolyzed with the photoprobe. cell proteins were solubilized in SDS buffer and subjected to gel electrophoresis.
As shown in Fig. 5, photoaffinity labeling of ileal cells resulted in the incorporation of radioactivity into several polypeptides.
The apparent molecular weights of the clearly labeled polypeptides were 99,000, 68,000, 59,000, 54,000, 43,000, 35,000, 20,000, and 14,000. Fig. 5 also demonstrates that the presence of 0.5 and 1.0 mM taurocholate during photoaffinity labeling progressively inhibited the incorporation of radioactivity, suggesting the photolabile derivative is bound to the same polypeptides as natural bile acid. The next series of studies were performed to determine the intracellular location of bile acid-binding polypeptides. Ileal cells that were photolyzed in the presence of [3H]7,7-azo-TC, were submitted to fractionation as outlined in Fig. 1 (12). Similar results were again obtained in the presence studies (data not shown), leaving unaccounted the 66, 43-, 35-, 20-, and 14-kDa polypeptides. Fig. 6A shows the photoaffinity labeling pattern of the soluble protein fraction. Greatest incorporation of radioactivity was seen in polypeptides of 68,000, 59,000, 43,000, 35,000, and 14,000 molecular weight (Fig. 6A, solid line, closed circles). In order to gain further evidence that the photolabile derivative was bound to the same polypeptide as natural bile acid, photoaffinity labeling of ileal enterocytes was performed in the presence of 0.5 mM taurocholate. As shown in Fig. 6A (solid line, open circles), taurocholate selectively inhibited photoaffinity labeling of these polypeptides. Fig. 6B shows the pattern of incorporation of radioactivity into polypeptides of the microsomal fraction that was obtained from ileal cells irradiated with [3H]7,7-azo-TC. As shown in Fig. 6B (solid line, closed circles), greatest incorporation of radioactivity was seen in a PO-kDa polypeptide. Fig. 6B (solid line, open circles) also shows that taurocholate inhibited the labeling of this polypeptide.
To further confirm which proteins have specific binding, direct photoaffinity labeling of the cytosolic fraction was performed. The soluble protein fraction was first obtained from non-irradiated ileal cells as outlined in Fig. 1, then photolyzed in the presence of ["H]7,7-azo-TC, followed by SDS-PAGE. The greatest incorporation of radioactivity was seen again in polypeptides of 68,000, 59,000, 43,000, 35,000 and 14,000 molecular weight (Fig. 7, solid line, closed circles). When photoaffinity labeling of the soluble protein fraction was carried out in the presence of 0.5 mM taurocholate, the bile acid selectively inhibited the labeling of the 68,000, 43,000,35,000, and 14,000 molecular weight polypeptides ( Polypeptides of the soluble protein fraction, labeled by [3H]7,7-azo-TC, were transferred from SDS gels to nitrocellulose membrane and cut into strips. The strips were incubated with rabbit antiserum against rat L-BABP (A), L-FABP (B), and I-FABP (C). Strip D was cut into 2-mm slices, and the radioactivity was counted in each slice. Strip D was aligned with strips A-C, and the positions corresponding to L-BABP, L-FABP, and I-FABP are indicated by the arrows. 7,7-azo-TC prior to flash photolysis. After the photolyzed cells were fractionated, each subcellular fraction was submitted to SDS-PAGE, and the amount of radioactivity incorpo-rated into the separated polypeptides was determined. Fig. 3.4 shows the incorporation of radioactivity following 0-, 7-, 60-, and 120-s pulses. The incorporation of radioactivity at 90 s and 5 min (data not shown) was identical to that obtained at 120 s. In addition, cells preincubated for 120 s with [3H]7,7azo-TC were chased by incubation with 0.5 mM taurocholate for O-120 s prior to flash photolysis. Fig. 8B indicates that the taurocholate chase resulted in a progressive decline in photolabeling of polypeptides suggesting the natural bile acid competes with photoprobe for protein-binding sites.
The similarity of molecular masses suggested that the 35and 1CkDa polypeptides were the rat 33-kDa liver bile acidbinding protein (L-BABP) and the 14-kDa rat intestinal fatty acid-binding protein (I-FABP), or rat liver fatty acid-binding protein (L-FABP), respectively (25, 26). Therefore, these cytosolic bile acid-binding proteins were characterized by immunoblotting. As shown in Fig. 9, polypeptides that were labeled with [3H]7,7-azo-TC, were transferred from SDS gels to nitrocellulose membrane and cut into strips. The strips were incubated with rabbit antiserum against the L-BABP (Fig. 9, strip A), the L-FABP (Fig. 9, strip B), and the I-FABP (Fig. 9, strip C). In Fig. 9, strip D was cut into 2-mm slices, and the radioactivity was counted in each slice. When strip D was carefully aligned with strips A-C, the migration distance of the radioactive peaks corresponded to that of the blot for L-FABP but not to those for L-BABP or I-FABP.
The identity of the bile acid-binding polypeptides was further explored using radioimmunoprecipitation. Fig. 10 shows that immunoprecipitation experiments were performed with [3H]7,7-azo-TC-labeled cytosolic polypeptides using rabbit When the gels were sliced and counted for radioactivity, radioimmunoprecipitation was noted with antiserum against albumin and actin but not with anti-L-BABP, anti-L-FABP, or anti-I-FABP. The failure to note immunoprecipitation against rabbit antisera may be explained by the masking of antigenic sites due to covalent linkage between photoprobe and binding protein.
To test this possibility cytosolic protein was first incubated with rabbit antiserum adsorbed to protein A-Sepharose. Following centrifugation the supernatant was photolabeled with [3H]7,7-azo-TC and subjected to SDS-PAGE. The amount of radioactivity incorporated into bile acid-binding protein in the supernatant was compared with the amount of radioactivity incorporated into bile acid-binding protein that was not incubated with antibody (control). The incubation of cytosolic protein with rabbit anti-rat L-FABP did not result in loss of radioactivity when compared with control (CL Fig.  11, D with E).

DISCUSSION
A criteria for the identification of transport proteins using photoaffinity labeling techniques is that upon photolysis the photoprobe irreversibly inhibits the transport system (23 Equal amounts of cytosolic protein were incubated in the absence and presence of anti-L-FABP conjugated protein A-Sepharose. The protein incubated with anti-L-FABP was centrifuged. The supernatant fraction and cytosolic protein not incubated with anti-L-FABP were labeled by photolysis with ["H]7,7-azo-TC. Protein standards (A), unincubated cytosolic protein (B), and supernatant (C) were subjected to SDS-PAGE. The solid lines D and E indicate the distribution of radioactivity in the sliced gels B and C, respectively. tography, 50% of the radioactivity had an RF value identical to that of unreacted 7,7-azo-TC. This finding suggested that the loss of radioactivity was due, at least in part, to photolysis products of 7,7-azo-TC or possibly incomplete photolysis of 7,7-azo-TC with failure to form a covalent linkage between the photoprobe and amino acid residues. However, the amounts of radioactivity which remained were sufficient to be assessed quantitatively. Moreover, experiments, performed under conditions identical for photolysis but in the absence of UV light, suggested that this radioactivity from the photoprobe was covalently linked, i.e. the labeling by the unreacted photoprobe represented only a small fraction (<lo%) of the amount of labeling obtained under UV light. Thus, photoaffinity labeling of intact enterocytes and their subsequent fractionation appears to offer the opportunity to identify those organelles, cytosolic components, and/or binding proteins that are involved in the transcellular movement of bile acids.
Although the intracellular location of bile acids has not been studied previously in the enterocyte, numerous studies have examined the distribution of the bile acid pool between subcellular fractions of rat liver. Okishio and Nair (27) reported that 70% of the bile acid pool was recovered in the cell supernatant with the rest of the pool distributed among the membranous subfractions. Strange et al. (28) made a similar observation, however, from partition coefficients they constructed an in uiuo compartmental model showing larger precentages of bile acids in the nuclear, microsomal, and mitochondrial fractions (29). Simion et al. (30) recovered most of the bile acid pool with the liver cell supernatant although a small proportion was recovered form the total microsomal fraction. All of these studies suffered from potential artifacts induced by the loss and/or redistribution of bile acids between subcellular fractions during their isolation. To circumvent this difficulty, Simion et al. (30,31) measured the ability of purified subcellular fractions to take up and bind bile acids, reasoning that subcellular fractions involved in intrahepatic bile acid transport should contain specific transport processes or high affinity binding sites for these compounds. Based on their observations, the authors proposed that taurocholate uptake into the cytoplasm occurs across the sinusoidal membrane by a facilitated transport system; taurocholate then moves from the sinusoidal membrane via a vesicular pathway that includes the smooth endoplasmic reticulum and the Golgi apparatus, and is secreted into bile by exocytosis as the vesicles fuse with the canalicular membrane (30, 31). Electronmicroscopic studies have supported this hypothesis. Autoradiographic studies of the localization of the bile acid analogues '251-cholylglycyltyrosine (32) and '251-cholylglycylhistamine (33) have shown a preferential distribution of electron-dense grains over the Golgi apparatus; grains were also associated with the smooth endoplasmic reticulum. More recent electronmicroscopic studies employed an indirect immunoperoxidase technique with an antibody against conjugated cholic acid and ursodeoxycholic acid. Electron-dense deposits were observed mostly on vesicles of the Golgi apparatus and sometimes in the smooth endoplasmic reticulum (34).
The information obtained in the present study is not entirely consistent with the above hypothesis proposed for the intracellular movement of bile acids in the liver. When the intracellular distribution of radioactivity from [3H]7,7-azo-TC was examined following trichloroacetic acid precipitation, lipid extraction and correction for incomplete enrichment of the subcellular fractions, the greatest percentage of radioactivity was found in the soluble protein fraction (43.2%) with lesser amounts in the microsomal (27.3%) and Golgi (2.2%) fractions. Although these data suggest the participation of these organelle fractions, they also suggest that larger amounts of bile acid may be transported in the enterocyte cytoplasm bound to cytosolic proteins. Three distinct groups of hepatic cytosolic bile acid binding proteins with different molecular masses have been identified: (a) glutathione Stransferase (45-50 kDa), (b) Y' binders (33 kDa), and (c) fatty acid binding proteins (FABP) (14 kDa). These cytosolic proteins may facilitate vectorial bile acid transport by minimizing back diffusion from the cell and retaining bile acids within the cytosolic compartment, thereby preventing distribution into the membranous compartments (35). It is also conceivable that directionality of transport is maintained should the binding proteins have specific membrane affinity sites at the pole of the cell opposite from where bile acids are initially taken up. The association of the 59-kDa polypeptide with both the soluble protein and basolateral membrane fractions in the present studies suggests this latter possibility.
The similarity of molecular masses suggested that the 35 and 14-kDa bile acid-binding polypeptides, identified by photoaffinity labeling, were Y binder and FABP, respectively. The presence of intestinal proteases has prevented an accurate measurement of Y binders in the small intestine, although a 33-kDa protein band was identified on immunoblotting of small intestinal cytosoi (35). In contrast, the rat small intestinal epithelium has been shown to contain two abundant FABPs. The L-FABP, present in both liver and intestine, is a polypeptide of 14,184 which is identical to Z protein; the I-FABP, present in the intestine, is a X,063-Da protein and represents l-2% of the soluble intestinal proteins (36). Studies have supported a role for FABP in intestinal fatty acid uptake, intracellular transport and utilization, but the question remains whether or not FABP serves a broader function in the intracellular handling of amphipathic small molecules such as bile acids. Bile acids are capable of binding to the L-FABP (37). However the lower binding affinity of bile acids to L-FABP than the affinity of long chain fatty acids has argued against an important role for L-FABP in bile acid intracellular transport (35). In the present studies, the specific antibodies to the 33-kDa cytosolic liver bile acid binding protein (L-BABP) and the two fatty acid-binding proteins (I-FABP and L-FABP) recognized on immunoblotting 33-and 14-kDa polypeptides, respectively. Except for L-FABP, the blotted protein bands had a different migration distance than the photoaffinity radiolabeled polypeptides. Moreover, the specific antibodies to L-BABP, I-FABP, and L-FABP failed to immunoprecipitate the radiolabeled polypeptides.
It is unlikely that immunoprecipitation was prevented by the covalent linkage between the photoprobe and the binding protein because immunoprecipitation of cytosolic protein did not reduce the subsequent photoaffinity labeling of bile acidbinding protein. Whatever the reason, specific identification of these newly described cytosolic bile acid-binding proteins (as well as the 20-kDa microsomal binding protein) in the intestine awaits their purification and amino acid sequencing.