Stimulation of colonic secretion by lipoxygenase metabolites of arachidonic acid.

Both 5-hydroperoxyeicosatetraenoic acid (5-HPETE) and 5-hydroxyeicosatetraenoic acid (5-HETE) increased the short-circuit current (Isc) in rabbit colonic mucosa mounted in vitro in Ussing chambers. Measurements of chlorine-36 fluxes indicated that the Isc response to 5-HPETE is due to stimulation of active chlorine secretion. 9-, 11-, and 12-HPETE's and leukotrienes C4 and B4 produced either very small increases in Isc or no increase. In contrast to results in rabbit colon, no HPETE, HETE, or leukotriene was effective in rabbit ileal mucosa. The effects of 5-HPETE in the rabbit colon were unaffected by mepacrine, but could be partially blocked by indomethacin. These results suggest that drugs which block both cyclooxygenase and lipoxygenase may be effective antidiarrheals in patients with colitis.

Correction to: Role of intestinal microbiome in American ginseng-mediated colon cancer prevention in high fat diet-fed AOM/DSS mice.

This article was originally published with the wrong title.

Role of intestinal microbiome in American ginseng-mediated colon cancer prevention in high fat diet-fed AOM/DSS mice [corrected].

OBJECTIVE: Chronic intestinal inflammation is a risk factor for colorectal cancer (CRC) initiation and development. Diets that are rich in Western style fats have been shown to promote CRC. This study was conducted to investigate the role of intestinal microbiome in American ginseng-mediated CRC chemoprevention in a mouse model. The population and diversity of enteric microbiome were evaluated after the ginseng treatment. METHODS: Using an azoxymethane (AOM)/dextran sulfate sodium (DSS)-induced gut inflammation and tumorigenesis mouse model, the effects of oral American ginseng on high fat diet-associated enteric pathology were determined. After establishment of a 16S rRNA illumina library from fecal samples, MiSeq sequencing was carried out to reveal the microbial population. The alpha and beta diversities of microbiome were analyzed. RESULTS: American ginseng significantly attenuated AOM/DSS-induced colon inflammation and tumorigenesis by reducing the colitis score and colon tumor multiplicity. The MiSeq results showed that the majority of sequences fell into three phyla: Firmicutes, Bacteroidetes and Verrucomicrobia. Further, two significant abundance shifts at the family level, Bacteroidaceae and Porphyromonadaceae, were identified to support ginseng's anti-colitis and anti-tumor effects. In addition, alpha and beta diversity data demonstrated that ginseng led to a profound recovery from the AOM/DSS-induced dysbiosis in the microbial community. CONCLUSION: Our results suggest that the CRC chemopreventive effects of American ginseng are mediated through enteric microbiome population-shift recovery and dysbiosis restoration. Ginseng's regulation of the microbiome balance contributes to the maintenance of enteric homeostasis.

Bradykinin-stimulated electrolyte secretion in rabbit and guinea pig intestine. Involvement of arachidonic acid metabolites.

Bradykinin (BK) increases short-circuit current (Isc) when added to the serosal side of rabbit or guinea pig ileum or rabbit colon. Significant effects on Isc are seen at concentrations as low as 10(-10) M. Anion substitution experiments and unidirectional 36Cl flux measurements indicate that this effect of BK on Isc is due to Cl secretion. The effect of BK on Isc can be partially blocked (60-70% inhibition) by cyclooxygenase inhibitors (indomethacin and/or naproxen) and completely blocked by the phospholipase inhibitor, mepacrine. The combined cyclooxygenase/lipoxygenase inhibitors BW 755 and eicosa-5,8,11,14-tetraynoic acid (ETYA) also completely block the effect of BK on Isc but the slow-reacting substance of anaphylaxis (SRS-A) antagonist FPL 55712 has no effect. None of the above inhibitors diminish the effect on Isc of other exogenously added secretory stimuli such as vasoactive intestinal peptide (VIP), theophylline, or prostaglandin E2 (PGE2). Prior desensitization of rabbit ileum to PGE2 blocks the effect on Isc of BK but not those of VIP or theophylline. Conversely, prior desensitization of rabbit ileum to BK greatly reduces the effect of PGE2 on Isc. BK also stimulates the synthesis of PGE2 in rabbit ileal and colonic mucosa and this effect can be blocked by prior addition of either indomethacin or mepacrine. These effects of BK are similar to those of exogenously added arachidonic acid (AA). AA also stimulates Cl secretion and increases PGE2 synthesis and its effect on Isc can be inhibited by prior desensitization to PGE2 or by prior addition of indomethacin. The above results indicate that BK stimulates active Cl secretion in both small and large intestine and suggest that this effect is due to the intracellular release of AA. Although the prostaglandins appear to be the major products of AA metabolism contributing to the secretory response, lipoxygenase products may also play a role.

Na+/H+ exchangers, NHE-1 and NHE-3, of rat intestine. Expression and localization.

Na-H exchange (NHE) is one of the major non-nutritive Na absorptive pathways of the intestine and kidney. Of the four NHE isoforms that have been cloned, only one, NHE-3, appears to be epithelial specific. We have examined the regional and cellular expression of NHE-3 in the rat intestine. NHE-3 message in the small intestine was more abundant in the villus fractions of the small intestine than in the crypts. Analysis of NHE-3 mRNA distribution in the gut by in situ hybridization demonstrated epithelial cell specificity, as well as expression preferential to villus cells. NHE-1 message, in contrast, was ubiquitous, with slightly greater expression exhibited in the differentiating crypt and lower villus cells of the small intestine. Isoform-specific NHE-3 fusion protein antibody identified a 97-kD membrane protein in the upper villus cells of the small intestine, which was exclusively localized in the apical membrane. In contrast, antibody previously developed against the COOH-terminal region of human NHE-1 (McSwine, R. L., G. Babnigg, M. W. Musch, E. B. Chang, and M. L. Villereal, manuscript submitted for publication) identified a 110-kD basolateral membrane protein. These data suggest that unlike NHE-1, which probably serves a "housekeeping" function, NHE-3 may be involved in vectorial Na transport by the intestine.

Dexamethasone protection of rat intestinal epithelial cells against oxidant injury is mediated by induction of heat shock protein 72.

Although the therapeutic actions of glucocorticoids are largely attributed to their anti-inflammatory and immunosuppressive effects, they have been implicated in enhancing tissue and cellular protection. In this study, we demonstrate that dexamethasone significantly enhances viability of IEC-18 rat small intestinal cells against oxidant-induced stress in a dose-dependent fashion. This protective action is mediated by induction of hsp72, the major inducible heat shock protein in intestinal epithelial cells. Dexamethasone stimulates a time- and dose-dependent response in hsp72 protein expression that parallels its effects on cell viability. Furthermore, the induction of hsp72 is tissue dependent, as nonintestinal epithelioid HeLa cells show differential induction of hsp72 expression in response to the same dexamethasone treatment. Antisense hsp72 cDNA transfection of IEC-18 cells abolishes the dexamethasone-induced hsp72 response, without significantly affecting constitutive expression of its homologue, hsc73. Dexamethasone treatment also significantly induces hsp72 protein expression in rat intestinal mucosal cells in vivo. These data demonstrate that glucocorticoids protect intestinal epithelial cells against oxidant-induced stress by inducing hsp72.

Ca-mediated stimulation of Cl secretion by reactive oxygen metabolites in human colonic T84 cells.

Monochloramine (NH2Cl), a granulocyte-derived reactive oxygen metabolite (ROM), increases short-circuit current (Isc) in cultured T84 monolayers in a concentration-dependent manner up to nonlethal concentrations of 75 microM. Isc increases slowly after NH2Cl, reaching a peak value of 18 +/- 2 microA/cm2 20 min after addition. The Isc changes are persistent (lasting over 20-30 min), depend on medium Cl, and are inhibitable with bumetanide. 36Cl flux studies demonstrated that NH2Cl increases serosa-to-mucosa flux of Cl without changing mucosa-to-serosa flux, consistent with stimulation of electrogenic Cl secretion. Isc responses to NH2Cl, but not PGE2, are dependent on medium calcium. As demonstrated in fura-2-loaded T84 cells, NH2Cl increases free cytosolic calcium by influx of extracellular Ca2+ and by release of Ca2+ from endogenous stores. However, NH2Cl had no effect on phosphatidylinositol metabolism or cyclic nucleotide levels. We conclude that ROM directly stimulate electrolyte secretion, an effect in part mediated by increases in cytosolic Ca2+, possibly through increasing Ca2+ permeability of cellular membranes.

Localization of the Na+/H+ exchanger isoform NHE-3 in rabbit and canine kidney.

The distribution and subcellular localization of Na+/H+ exchanger isoform NHE-3 was studied in rabbit and canine kidney using polyclonal antibodies to an NHE-3 fusion protein. Western blot analyses were performed against microsomal, brush-border, and basolateral membranes isolated from rabbit kidney cortex, outer medulla, and inner medulla. Immunoblots indicated that NHE-3 antibody recognized a strong band with 95-100 kDa molecular mass in cortical microsomes. Subcellular localization studies showed that NHE-3 was expressed in brush-border membranes of kidney cortex. Expression of NHE-3 in the medullary regions was studied by immunoblot analysis of NHE-3 antibody against the microsomal membranes from the outer and inner medulla. NHE-3 antibody specifically labelled a 95-100 kDa protein in outer but not inner medulla. Subcellular localization studies demonstrated that NHE-3 is localized to the brush-border membranes of the outer medulla. Immunoblot analysis against brush-border membranes from canine kidney cortex and outer medulla demonstrated the presence of an 83-90 kDa protein. The above experiments suggest that NHE-3 in rabbit kidney is a 95-100 kDa protein and is expressed in brush-border membranes of the cortex and outer medulla. The canine kidney NHE-3 is a 83-90 kDa protein and is expressed in brush-border membranes of the cortex and outer medulla. Based on its subcellular localization, we conclude that NHE-3 may be involved in vectorial Na+ and HCO3- transport and pHo regulation.

Surgical stress shifts the intestinal Escherichia coli population to that of a more adherent phenotype: role in barrier regulation.

BACKGROUND: We have shown that the combination of surgical stress and starvation in mice is associated with a defect in epithelial permeability and increased numbers of mucosa-associated Escherichia coli in the cecum. The aim of this study was to determine the specific role of mucosa-associated E coli on epithelial barrier dysfunction in this model. METHODS: Cecal E coli were harvested from mice 48 hours after a sham operation (control mice) or after a 30% surgical hepatectomy with only water provided ad libitum (short-term starvation) after the surgical procedure. Strains were tested for their ability to adhere to and alter the transepithelial electrical resistance (TEER) of cultured young adult mouse colon epithelial cells. TEER changes were further characterized by mannitol fluxes to confirm a defect in paracellular permeability. RESULTS: Strains of cecal E coli harvested from hepatectomy-starved mice adhered to and altered the permeability of young adult mouse colon cells, whereas E coli from the cecum of control mice were less adherent and had no effect on epithelial permeability. The effect of the strains harvested from mice after hepatectomy on the TEER of young adult mouse colon cells was inhibited by mannose and reversed by ciprofloxacin. CONCLUSION: The combination of surgical stress and short-term starvation is associated with a greater abundance of adherent and barrier-altering strains of E coli in the mouse cecum.

Calcium mediated neurohumoral inhibition of chicken enterocyte Na influx: role of phosphatidylinositol metabolites.

Carbachol (CCH), serotonin (5HT), divalent ionophore A23187, cAMP, and certain neuropeptides, i.e. substance P (SP), inhibit the initial rate of uptake (influx) of 22Na into isolated chicken villus enterocytes. All these agents also increase cytosolic Ca. However, the increases stimulated by CCH, 5HT, and cAMP are not blocked by chelation of extracellular Ca, whereas those of A23187 and SP are. Only CCH and 5HT stimulate hydrolysis of membrane phosphoinositides to form inositol phosphates. CCH and 5HT also stimulate incorporation of [32P]-PO4 into membrane polyphosphoinositides. These studies suggest that at least three mechanisms exist to increase cytosolic Ca in chicken enterocytes and thereby inhibit Na influx. Certain neurohumoral agents such as SP open a plasma membrane permeability for Ca, permitting extracellular Ca to enter the cell down its electrochemical gradient. These agents do not stimulate phosphatidylinositol breakdown. CCH and 5HT stimulate phosphatidylinositol breakdown and via the formation of inositol trisphosphate release Ca from intracellular stores. A third mechanism exists for cAMP which mobilizes Ca from intracellular stores, but does not involve the metabolism of membrane phosphatidylinositols.

Calcium ionophore and chemotactic peptide stimulation of peptidoleukotriene synthesis in DMSO-differentiated HL60 cells.

The human promyelocytic leukemia cell line HL60 can be differentiated to mature granulocytes upon exposure to DMSO (1.3%, 6 days). The ability of these cells to metabolize arachidonic acid via the 5-lipoxygenase pathway to form 5-HETE, LTB4, and 5,12-diHETEs, has been previously documented. However, the production of peptidoleukotrienes by DMSO-differentiated HL60 cells has not been previously reported. Arachidonic acid metabolites produced via 5-lipoxygenase were identified by reverse-phase, high-performance liquid chromatography, immunoreactivity specific for peptidoleukotriene, glutamyl transpeptidase transformation, characteristic UV spectra, and GC mass spectra. Leukotriene synthesis in the DMSO-differentiated HL60 cell is maximal at 5 min when stimulated with the calcium ioniphore, A23187 (1 microM), in the presence of calcium. These cells produce 12.94 +/- 1.8 ng/10(6) cells of LTC4 and 3.8 +/- 0.4 ng/10(6) cells of LTB4. LTC4 and LTB4 are also synthesized in the undifferentiated cell when stimulated with 1 microM A23187 and 1 mM Ca2+, but in much smaller quantities, i.e., 1.91 +/- 0.42 ng/10(6) cells of LTC4 and 0.41 ng +/- 0.06/10(6) cells of LTB4. The synthetic chemotactic peptide, f-Met-Leu-Phe, also elicits formation of LTC4 and LTB4 in a dose-dependent manner in the presence of exogenously added calcium. Maximal stimulation of DMSO-differentiated cells with f-Met-Leu-Phe produces 2.5 +/- 0.2 ng of LTC4 and 1.45 +/- 0.2 ng of LTB4 per 10(6) cells. The observation that DMSO-differentiated HL60 cells produce LTC4, as well as other 5-lipoxygenase products, increases the utility of this cell line for unraveling the regulation of leukotriene biosynthesis by granulocytes.

Cell membrane surface expression and tyrosine kinase regulate the osmolyte channel (skAE1) in skate erythrocytes.

A regulatory volume decrease response occurs when skate (Raja erinacea) erythrocytes are exposed to hyposmotic medium; they swell and then reduce their volume by releasing organic osmolytes (primarily taurine) and associated water. The response involves the red blood cell anion-exchanger skate anion-exchanger 1 (skAE1), which has been previously shown to be affected by tyrosine phosphorylation, to form tetramers and to change binding affinities to the cytoskeletal proteins, ankyrin and band 4.1. Our recent studies are focused on determining the sequence and mechanism of these events to better understand the activation of skAE1 upon hyposmotic stimulation. Under isosmotic conditions a large portion of skAE1 is found not only on the plasma membranes but also associated intracellularly in detergent-resistant membranes or lipid rafts. We hypothesize that an important step in the hyposmotic-induced increase in taurine transport involves the movement of skAE1 from an intracellular association with lipid rafts into the cell membrane. Inhibition of tyrosine phosphorylation of skAE1 with piceatannol reduces the hyposmotic-induced increase in taurine transport and also inhibits both the decreased binding of skAE1 to band 4.1 and the increased affinity to ankyrin. However, the phosphorylation inhibitor does not block the movement of the transporter into the plasma membrane or the formation of tetramers. This suggests that tyrosine phosphorylation is important in the hyposmotic-induced taurine transport but other steps that do not require phosphorylation play an important role.

Antigenic stimulated release of arachidonic acid, lipoxygenase activity and histamine release in a cloned murine mast cell MC9.

A cloned murine mast cell MC9 expresses phospholipase and lipoxygenase activity when stimulated with IgE and hapten. Addition of DNP-BSA to sensitized MC9 cells causes release of 58% of the cell histamine and 127 pmoles LTC4/10(6) cells. Prelabelling studies with [1-14C]-arachidonic acid showed that LTC4 production was proceeded by the release of arachidonic acid from membrane phospholipids. Approximately 8.7% of the cell arachidonic acid was released and half of this was converted to LTC4. The remaining radioactivity was converted to diHETES including LTB4 (15%), 5-HETE (10%), free arachidonic acid (10%), reesterified 5-HETE and arachidonic acid (8%) and prostaglandins (7%). This stimulation was dependent on hapten (DNP-BSA) and extracellular Ca++. Under identical conditions the calcium ionophore A23187 stimulated the release of 10.3% of the total cell arachidonic acid, and 51% of this was metabolized to LTC4. In addition the ionophore stimulated the release of 61% of the total cellular histamine.

Antigen-stimulated metabolism of inositol phospholipids in the cloned murine mast-cell line MC9.

Cells of the murine mast-cell clone MC9 grown in suspension culture were sensitized with an anti-DNP (dinitrophenol) IgE and subsequently prelabelled by incubating with [32P]Pi. Stimulation of these cells with DNP-BSA (bovine serum albumin) caused marked decreases in [32P]polyphosphoinositides (but not [32P]phosphatidylinositol) with concomitant appearance of [32P]phosphatidic acid. Whereas phosphatidylinositol monophosphate levels returned to baseline values after prolonged stimulation, phosphatidylinositol bisphosphate levels remained depressed. Stimulation of sensitized MC9 cells with DNP-BSA increased rates of incorporation of [32P]Pi into other phospholipids in the order: phosphatidylcholine greater than phosphatidylinositol greater than phosphatidylethanolamine. In sensitized cells prelabelled with [3H]inositol, release of inositol monophosphate, inositol bisphosphate and inositol trisphosphate, was observed after stimulation with DNP-BSA. When Li+ was added to inhibit the phosphatase activity that hydrolysed the phosphomonoester bonds in the sugar phosphates, greater increases were observed in all three inositol phosphates, particularly in inositol trisphosphate. The IgE-stimulated release of inositol trisphosphate was independent of the presence of extracellular Ca2+. In addition, the Ca2+ ionophore A23187 caused neither the decrease in [32P]polyphosphoinositides nor the stimulation of the release of inositol phosphates. These results demonstrate that stimulation of the MC9 cell via its receptor for IgE causes increased phospholipid turnover, with effects on polyphosphoinositides predominating. These data support the hypothesis that hapten cross-bridging of IgE receptors stimulates phospholipase C activity, which may be an early event in stimulus-secretion coupling of mast cells. The results with the Ca2+ ionophore A23187 indicate that an increase in intracellular Ca2+ alone is not sufficient for activation of this enzyme.

Hypotonicity stimulates phosphatidylcholine hydrolysis and generates diacylglycerol in erythrocytes.

Exposure of skate erythrocytes to hypotonic medium stimulates a rapid increase in levels of 1,2-diacylglycerol. Other treatments which produce cell swelling such as replacement of a portion of medium NaCl with the permeant solutes ethylene glycol or ammonium chloride also stimulate increases in diacylglycerol. Whereas the reduction of medium osmolarity to 460 mosm (from 940) stimulated a persistent diacylglycerol increase, the increase after reduction to 660 mosm was transient, peaking at 2.5 min and then slowly declining. This decline could be prevented by preincubation with the diacylglycerol kinase inhibitor R59022. To investigate the source of the increased diacylglycerol, the rate of incorporation of [32P]PO4 into each major phospholipid was measured. Reduction of osmolarity to 660 mosm stimulated the incorporation of phosphate into phosphatidylcholine markedly, with a smaller increase observed into phosphatidylinositol. To demonstrate phosphatidylcholine hydrolysis, erythrocytes were prelabeled with [32P]PO4. Subsequent exposure to hypotonic (660 mosm) medium stimulated a decrease in radioactivity in phosphatidylcholine and a large increase in radioactivity in phosphatidic acid. When stimulated in the presence of ethanol, 32PO4-labeled phosphatidylethanol was formed, suggesting activation of phospholipase D. In addition, the initial formation of 32PO4-labeled phosphatidic acid was not sensitive to inhibition of diacylglycerol kinase, supporting the role of direct activation of phospholipase D. These results indicate that hypotonicity and the accompanying cell swelling induce cell membrane phospholipid turnover, predominantly phosphatidylcholine, and production of the protein kinase C activator, diacylglycerol, which appears to occur via activation of phospholipase D.

Modulation of erythrocyte band 4.1 binding by volume expansion.

Erythrocyte band 4.1 is an important protein in the control and maintenance of the cytoskeleton. Skate erythrocyte band 3, the anion exchanger, appears to play a pivotal role in the regulation of volume-stimulated solute efflux during volume expansion. Because band 4.1 interacts with band 3, we tested whether their interaction might change during volume expansion. Skate red blood cells were volume-expanded in either hypotonic media (one-half osmolarity) or were swollen under isoosmotic conditions by inclusion of ethylene glycol or ammonium chloride in the medium. Microsomal membranes isolated from red cells under volume expanded conditions demonstrated a significant decrease in the amount of band 4.1 bound to band 3. In unstimulated cells, approximately one third of the binding of band 4.1 occurred to band 3. This binding was characterized as being sensitive to competition by the peptide IRRRY. The majority of band 4.1 is bound to glycophorin (as demonstrated in other species), and this binding does not change during volume expansion. The alteration in band 4.1:band 3 interaction occurs within 5 min after volume expansion and is transient, returning to near normal interaction within 60 min. Two drugs that promote band 3 oligomerization, pyridoxal-5'-phosphate and DIDS, also decreased band 4.1 interaction with band 3. Band 4.1 and ankyrin binding to band 3 may be reciprocally related as high-affinity ankyrin binding sites to band 3 observed under volume-expanded conditions are decreased by inclusion of band 4.1 in the binding reactions. J. Exp. Zool. 289:177-183, 2001.

K-independent Na-Cl cotransport in bovine tracheal epithelial cells.

Uptakes of 22Na, 36Cl, and 86Rb into isolated bovine tracheal epithelial cells were measured in the presence and absence of 10 microM bumetanide. Preincubation with ouabain (0.5 mM) did not alter initial rates of Na and Cl uptakes but prolonged from 0.5 or less to 2 min the period in which Na uptake is linear with time. Initial rates of bumetanide-inhibitable Na and Cl uptakes (influxes), measured for 2 min in identical solutions, were similar in magnitude (bumetanide-sensitive Cl influx/bumetanide-sensitive Na influx = 1.2). Omission of K did not affect bumetanide-sensitive Na or Cl influx. Cl influx was not affected by 4-acetamido-4'-isothiocyanostilbene-2,2'-disulfonic acid. Amiloride (0.1 mM) partially inhibited the bumetanide-insensitive Na influx but had no effect on the bumetanide-sensitive Na influx. Rb influx was not affected by bumetanide but was markedly reduced by ouabain and slightly reduced by Ba, the combination being additive. Half-maximally inhibitory concentration values for inhibition of Cl influx by 4 sulfamoylbenzoic acid derivatives were as follows (in mumol/l): 0.125 benzmetanide; 0.64 bumetanide; 16 piretanide; and 26 furosemide. Affinities of Na and Cl for the bumetanide-inhibitable cotransport process were determined by measuring bumetanide-sensitive Cl influx at varying [Na] and bumetanide-sensitive Na influx at varying [Cl]. Both plots were hyperbolic, and the K0.5 values for Na and Cl were 4.1 and 53.9 mM, respectively. Bumetanide-inhibitable Cl influx was not altered by secretory stimuli (epinephrine, A23187) but was more than doubled by osmotic shrinkage (200 mM mannitol or sucrose).(ABSTRACT TRUNCATED AT 250 WORDS)

High affinity binding of ankyrin induced by volume expansion in skate erythrocytes.

Volume expansion of little skate (Raja erinacea) erythrocytes increases the affinity of ankyrin binding without altering in the number of binding sites. Potassium iodide-stripped inside-out vesicles (KI-IOV) were used to assess ankyrin binding under volume-expanded conditions. Under isoosmotic conditions, ankyrin binds nearly exclusively to a single class of sites (Bmax, 52 +/- 12 microg/mg; Kd, 150 +/- 28 nM). KI-IOV from volume-expanded cells (either with one-half osmolarity medium or with inclusion of the permeant solute ethylene glycol) demonstrate two ankyrin-binding populations. A high affinity population occurs transiently under volume-expanded conditions. This population has a Bmax of 18 +/- 7 microg/mg and a Kd of 25 +/- 9 nM. Total binding of high and low affinity sites is 57 +/- 17 microg/mg. This change in ankyrin affinity is reversible on volume regulatory decrease. A major target protein in the KI-IOV was identified as the skate homolog of the mammalian red cell anion exchanger band 3. Inclusion of the purified cytoplasmic domain of band 3 competes away more than 80% of the ankyrin binding. To determine whether increased ankyrin affinity is due to band 3 tetramer formation that occurs in volume expansion, cells were treated with pyridoxal 5-phosphate or 4,4'-dinitrostilbene-2,2'-disulfonic acid, two agents that increase tetramer formation under isoosmotic conditions. Both treatments altered the binding affinity with a shift toward higher affinity binding without significant alteration in the number of binding sites.