WO1997039766A1 - Regulation of cellular volume via chloride conductive polypeptide channels - Google Patents

Abstract

Therapeutic compositions, as well as methods for treatment of a disease that is characterized by defective chloride conduction across a cellular membrane, are disclosed. These compositions contain a polypeptide that has one or more of the following characteristics: 1) the ability to spontaneously form a channel in a lipid bilayer, 2) the ability to selectively transport chloride ions across that bilayer, and 3) the ability to dissolve in an aqueous physiological solution. These polypeptides include, but are not limited to, IC1n and porin molecules, such as aerolysin, cryptdin 2 and cryptdin 3.

Description

REGULATION OF CELLULAR VOLUME VIA CHLORIDE CONDUCTIVE POLYPEPTIDE CHANNELS The work described herein was supported in part by grants from the National Institutes of Health (POl DK33506-11, DK48106, DK35932, and DK47662). Therefore, the government may have certain rights in the invention.

Field of the Invention The field of the invention is volume-activated chloride channels.

Background of the Invention Cystic fibrosis is the most common genetic defect seen in the Caucasian population, and is associated with very high rates of morbidity and mortality. The disease is caused by a defect in chloride secretion that results from mutation of the CFTR (cystic fibrosis transmembrane regulator) gene. The defective gene product prevents epithelial cells from hydrating their lumenal surface. The epithelial cells most affected include those lining the respiratory tract, the gastrointestinal tract, the biliary tract, pancreatic ductules, and other mucosa. The inability to hydrate the lumenal surfaces of these cells leads to the clinical manifestations of the disease, which include pancreatic failure and respiratory insufficiency.

Apart from proposals to correct the genetic defect in the CFTR, treatment regimes are concerned primarily with symptomatic relief, which can be achieved to some degree by pancreatic enzyme replacement and mobilization of inspissated secretions. Summary of the Invention As described herein, two channel-forming entities have been found to spontaneously form channels in cell membranes and to transport Cl" across the membrane without utilizing ATP. Accordingly, the invention features therapeutic compositions containing a polypeptide that mediates the flux of ions across a cellular membrane. Methods of treatment in which these compositions are administered are also described. The polypeptides include, but are not limited to, I_cln and cryptdins 2 and 3.

I_cln is a channel-forming polypeptide that can spontaneously form channels in lipid-based membranes, such as cellular plasma membranes, that do not require ATP to transport monovalent anions, particularly Cl~, across the plasma membrane. Without wishing to be limited to a specific theory, it is believed that I_cln is the volume-sensitive organic osmolyte-anion channel (VSOAC) , or a functional subunit of VSOAC, and is therefore critical for regulating cellular volume. Ic_ln and other related molecules provide the structural basis for therapeutic compositions and methods, as described below. Similarly, cryptdins 2 and 3 have been found to spontaneously form anion selective channels in the lipid bilayer of a cellular membrane, and to elicit a chloride secretory response in intestinal and respiratory epithelial cells poised for chloride secretion. Accordingly, these polypeptides are useful in some or all of the therapeutic compositions and methods for which I_cln can be used.

A therapeutic composition of the invention therefore contains a polypeptide that has one or more of the following characteristics: (1) the ability to spontaneously form a channel in a lipid bilayer, (2) the ability to selectively transport chloride ions across that bilayer, and (3) the ability to dissolve in an aqueous physiological solution. The polypeptide can have an amino acid sequence that is at least 70% identical, more preferably at least 85% identical, and most preferably at least 95% identical to one or more of the polypeptides that naturally form a β barrel. These polypeptides include, but are not limited to, I_cln, a porin molecule, annexion, and aerolysin (Wilmsen et al., J . Mem . Biol . 115:71-81, 1990; Benz, Ann. Rev . Microbiol . 4.2:359-393, 1988) . Alternatively, the polypeptide can have an amino acid sequence that conforms to the concensus sequence of a cryptdin (SEQ ID NO:l; see below) . For example, the polypeptide can have the amino acid sequence of cryptdin 2 or cryptdin 3, or a fragment thereof.

Also included within the invention are "functional polypeptides," which possess one or more of the biological functions or activities of I_Cι_n, cryptdin 2, or cryptdin 3. These functions or activities are described in detail below and concern, primarily, regulation of cellular volume via ion flux, primarily Cl^" ion flux, across the plasma membrane of the cell. It is well within the abilities of skilled artisans to determine, without resort to undue experimentation, whether a polypeptide, regardless of size, retains the functional activity of a polypeptide of the invention.

Functional polypeptides can contain a primary amino acid sequence that has been modified from the known sequences of I_cln (or β barrel pore-forming polypeptides) , cryptdin 2, or cryptdin 3 (for the sequence of I_cln in various species, those of skill in the art can consult, for example, Krapivinsky et al., Cell 7^:439-448, 1994, or Paulmichl, Nature 356:238-241, 1992; and for the sequence of cryptdin polypeptides, can see, for example, Selsted et al., U.S. Patent No. 5,422,424, in which particular amino acid sequences and consensus amino acid sequences of cryptdins are disclosed) . The modifications can consist of an addition, deletion, or substitution of amino acid residues. In the event amino acid substitutions are introduced, the substituted amino acids can be conservative amino acid substitutions, such as those within the groups listed below.

Given that the therapeutic composition can be used to treat a patient who has a disease that is characterized by defective conductance of chloride ions across the cell membrane, and that this defect is the underlying cause of cystic fibrosis, the therapeutic composition can be formulated for inhalation or aspiration therapy. According to the method of the invention, a patient who has a disease that is characterized by defective conductance of chloride ions across the cell membrane would be administered an amount of a polypeptide of the invention that is sufficient to normalize chloride conduction. The polypeptide may be administered via routes that are well known in the pharmaceutical art. These routes include intravenously, intramuscularly, orally, intraventricularly, subcutaneously, intraperitoneaUy, transmucosally, and topically. In the event that chloride conductance is defective in the respiratory system, the expected route of administration is by inhalation.

A preferred dosage for administration by inhalation is 0.01 mg to 100 mg/ml, and dosages may be repeated as necessary. Determination of an optimum dosage for any given application is well within the abilities of one of ordinary skill in the art of pharmacology. As is well known in the medical arts, dosages for any one patient depend on many factors, which are reviewed below. As used herein "protein" or "polypeptide" refer to any chain of more than two amino acids linked by peptide bonds, regardless of length or post-translational modifications, such as phosphorylation or glycosylation. A "therapeutic composition" is a mixture that contains an active ingredient formulated with a physiologically acceptable carrier, such as physiological saline. The therapeutic composition can be administered to a mammal in a dosage that is sufficient to effectively restore the conductance of ions, particularly chloride ions, across a cell membrane and thereby restore the volume of the cell to within a normal range. Volume is considered within a normal range if it is about 50%, more preferably about 70%, even more preferably about 85%, and most preferably about 95% or more of the volume of an average cell that is of the same phenotype but that does not exhibit defective chloride conductance. Similarly, chloride conduction is said to be normalized if it is about 50%, more preferably about 70%, even more preferably about 85%, and most preferably about 95% or more of the chloride conductance of an average cell that is of the same phenotype but that does not exhibit defective chloride conductance. Determining whether or not chloride conduction is defective is well within the abilities of skilled artisans.

By "spontaneously forms a channel" is meant that the substance in question has the ability to: (1) integrate into a lipid bilayer, such as those that constitute the cellular membranes (also called plasma membranes) of living mammals, when applied to the surface of the lipid bilayer, and (2) form a structure that allows the passage of ions, particularly monovalent anions, such as Cl^~ across the membrane. Any given polypeptide may be tested for the ability to spontaneously form a channel by, for example, the methods described herein where recombinant polypeptides are produced and applied to a planar lipid bilayer.

"Dissolve in aqueous physiological solution" means that the polypeptide is soluble in water and aqueous buffers at physiological pH and ionic strength.

"Identical," as used herein in reference to polypeptide sequences, refers to the subunit sequence identity between two molecules. When a subunit position in both of the molecules is occupied by the same monomeric subunit, i.e., the same amino acid, then the molecules are identical at that position. The identity between two amino acid sequences is a direct function of the number of matching positions. For example, if half of the positions in two amino acid sequences are the same, then the sequences are 50% identical. For polypeptides, the length of comparison sequences will generally be at least 16 amino acids, preferably at least 20 amino acids, more preferably at least 25 amino acids, and most preferably 35 amino acids. Sequence identity can be measured using sequence analysis software such as the Sequence Analysis Software Package of the Genetics Computer Group, University of Wisconsin Biotechnology Center (1710 University Avenue, Madison, WI, 53705) and the default parameters thereof. Where a particular polypeptide is said to have a specific percent identity to a reference polypeptide of a defined length, the percent identity is relative to the reference polypeptide. Thus, a polypeptide that is 50% identical to a reference polypeptide that is 100 amino acids long can be a 50 amino acid polypeptide that is completely identical to a 50 amino acid long portion of the reference polypeptide. It might also be a 100 amino acid long polypeptide that is 50% identical to the reference polypeptide over its entire length. Of course, many other polypeptides will meet the same criteria. In the case of amino acid sequences that are less than 100% identical to a reference sequence, the non- identical positions are preferably, but not necessarily, conservative substitutions for the reference sequence. Conservative substitutions typically include substitutions within the following groups: glycine, alanine; valine, isoleucine, leucine; aspartic acid, glutamic acid; asparagine, glutamine; serine, threonine; lysine, arginine; and phenylalanine and tyrosine. A particular advantage of the polypeptides used in the compositions and methods of the invention is that they can spontaneously insert into a lipid-based membrane. This property obviates the need for complex delivery methods, such as those required when a polypeptide is administered via gene therapy. Other features and advantages of the invention will be apparent from the detailed description and from the claims.

Brief Description of the Drawings Fig. 1 is a bar graph depicting the ³H-taurine efflux (%/10 min) measured during 3 of 5 successive cycles of cellular swelling, which was induced by exposure to 200 mOsm medium, and shrinkage. Rat C6 glioma cells were transfected with DNA from the pCDNAIII vector that either contained the I_Cι_n insert (C6-I_cln, solid bars) or that lacked the insert (C6-Neo, open bars) . Stably transfected clones were selected by neomycin resistance. The inset is a photograph of a Western blot. Protein was extracted from transfected cells, separated by SDS-PAGE, transferred to a solid support, and stained with an affinity purified polyclonal antibody generated against a C6 cell recombinant I_cln protein. One of four representative experiments is shown. Fig. 2 is a schematic diagram of the putative structure of the I_cln channel (Paulmichl et al., Nature 356:238-241, 1992). The top panel is a diagram of the proposed membrane topology of an I_cln monomer. The putative membrane spanning region consists of four antiparallel β strands. Large open circles denote amino acid residues that face the pore lumen. The bottom panel is a diagram of the proposed structure of the I_cln channel. The channel is postulated to consist of a dimer of the protein with the pore formed by an eight-stranded antiparallel β barrel (Paulmichl et al., supra) . Open and shaded regions of the channel denote individual I_cln monomers.

Fig. 3 is a photograph of a silver-stained polyacrylamide gel. Protein was purified from bacteria that were transformed with a pGEX-GST vector encoding only the GST protein or with PGEX-GST-I_cln, encoding the GST-I_cln fusion protein. Gels were loaded with 2 μl of I_cln (600 ng) or 2 μl of "control" (<20 ng) protein. Arrows without labels are unknown contaminates.

Fig. 4 is a pair of tracings obtained from single channel recordings of I_cln reconstituted into a planar lipid bilayer.

Fig. 5 is current-voltage (l-V) curve depicting the current-to-voltage relationship and anion versus cation selectivity of the I_cln channel. A 10-fold trans to cis KCl gradient shifts the reversal potential by -52 mV, demonstrating that the channel is highly selective for anions over cations. Fig. 6 is a pair of tracings obtained from single channel recordings with ATP (1 mM ATP; bottom tracing) and without ATP (control; top tracing) . ATP was added to the presumed extracellular side of the channel.

Figs. 7A-7E are a series of graphs illustrating that a Cl^" secretory response is elicited from human intestinal T84 cell monolayers that have been exposed to cryptdin-enriched peptide fractions of mouse intestine. Fig. 7A illustrates the time course of Cl^" secretion elicited by the peptide fraction (50 μg/ml in HBSS, filled squares), or vehicle alone (0.01% acetic acid, open squares) when applied to apical membranes of T84 cell monolayers. The cAMP agonist vasoactive intestinal peptide (VIP) was added at 52 minutes. In Figs. 7B and 7C, cAMP and cGMP responses to peptide fraction are shown, respectively. Monolayers exposed to vehicle alone (Basal) or to a cGMP-agonist (guanylin) or to a cAMP- agonist (forskolin) provide two point calibration. Fig. 7D illustrates the time course of Cl^" secretion elicited by vehicle alone (buffer) or peptide fractions (further purified by HPLC) in the presence or absence of the adenosine inhibitor 8-phenyltheophyline (1 μM) . Fig. 7E illustrates the time course of Cl- secretion elicited by adenosine (10 μM) in the presence or absence of 8- phenyltheophyline. Figs. 8A-8C are a series of graphs illustrating that cryptdins 2 and 3 elicit a fully reversible Cl^" secretory response from human intestinal T84 cell monolayers. Fig. 8A illustrates the time course of Cl^" secretion elicited by purified murine cryptdins 1-6. Cryptdin 1 (20 μg/ml) or cryptdins 2-6 (40 μg/ml) in HBSS were applied to apical membranes of T84 cell monolayers at 37°C. At 120 minutes, cryptdins were removed from the apical reservoir by "washout" in > 199 volumes of HBSS containing 0.1% bovine serum albumin (BSA). Fig. 8B illustrates the dose dependency of Cl^" secretion for cryptdin 3. Data represent peak currents 30 minutes after apical administration of cryptdin 3 (circles and squares represent two independent experiments) . Data were fit to Michaelis Menten model (R²=0.98). Fig. 8C illustrates the effect of removing Cl^~ ions from transport buffers on the secretory response elicited by cryptdin 3 or vehicle alone (0.01% acetic acid). T84 cells were preincubated in Cl^" free buffer or HBSS and exposed to cryptdin 3 (100 μg/ml) or vehicle alone. At 42 minutes, Cl^" was added back (110 mM final) to Cl^" free buffers. At 55 minutes (asterisk) , the cAMP-agonist VIP (5 nM) was applied to control monolayers not treated with cryptdin 3 but incubated sequentially with Cl^" free and then Cl replete buffers as described above. Cl^" free buffer, consisted of 140 mM Na⁺ gluconate, 5 mM K⁺ gluconate, 1.25 mM Ca⁺⁺ acetate, 1 mM Mg⁺⁺ gluconate, 5 mM KH₂P0₄/Na₂HP0₄ 20 mM HEPES, 5 mM glucose (pH 7.4) .

Fig. 9 is a panel of photographs illustrating the formation of channels in T84 cell membranes following treatment with cryptdin 3. The photographs on the left- hand side show that cryptdin 3 permeabilizes T84 cells to the membrane impermeant fluorophore BCECF-acid (2'7-bis- (2-carboxyethyl)-5(and-6)-carboxyfluorescein; Is = 1 second exposure; 10s = 10 second exposure; Bar = 11 μm) . Fig. 10 is a graph illustrating a synergistic effect between cryptdin 3 and the muscarinic agonist carbacol. T84 cell monolayers were treated with cryptdin 3 (100 μg/ml) or vehicle alone (closed squares) , and, 25 minutes later, were exposed basolaterally to the muscarinic agonist carbacol (100 μM) .

Fig. 11 is a graph illustrating the chloride secretory response from JM15 airway cells treated with cryptdin 3. Detailed Description Examples of the preferred methods and materials will now be described. These examples illustrate the claimed invention and are not to be considered limiting.

Example I; Cryptdins

Recently, it was found that Paneth cells may contribute to the defense of the epithelial surface in a host's intestine by releasing 3-4 kDa antimicrobial peptides into the crypt lumen (Selsted et al., J . Cell Biol . 118:929-936, 1992) . These peptides have been termed intestinal defensins or cryptdins because they are homologous in structure and function to neutrophil defensins (Selsted et al., supra ; Ouellette et al., FEBS Letters 304 :146-148, 1992; and Ouellette et al., Infect . Immun . .6J2:5040-5047, 1994). The I_cln protein is highly conserved across species (see Krapivinsky et al., Cell 26_.:439-448, 1994, for comparison of amino acid sequences of I_cln clones from Xenopus oocytes, rat atria, and MDCK cells) . Paulmichl et al. (Nature 156:238-241, 1992) originally proposed that I_cln is a novel channel-forming protein, consisting of a dimer of the protein with the pore being formed by an eight-stranded antiparallel β- barrel. However, upon further experimentation, it was concluded that I_cln is probably not an ion channel but rather, a regulatory protein required for channel activation (Krapivinsky et al., Cell 7j5:439-448, 1994).

The ability of the endogenous antimicrobial peptides to act as soluble inducers of channel-like activity when applied to apical membranes of Cl^" secreting epithelial cells in culture was tested as described below (for additional information regarding materials and methods, persons of skill in the art may consult Dharmsathaphorn et al., Methods in Enzymol . 192:345-359, 1990) . Certain intestinal defensins, cryptdins 2 and 3, selectively permeablized apical cell membranes of the human intestinal cell line T84 to elicit a physiologic secretory response. These data define cryptdins 2 and 3 as novel intestinal secretagogues, and identify a previously undescribed mechanism of paracrine signaling, which may involve the transfer of ion conductive channels from one cell type to another in vivo.

Culture of Human Intestinal Cells Human intestinal T84 cells were obtained from the American Type Culture Collection (A.T.C.C. Accession No. CCL-248) , and cultured and passaged as previously described in Dharmsathaphorn et al. (Methods in Enzymol . 192 :345-359 , 1990). When grown on permeable supports, T84 cells form confluent monolayers of columnar epithelia that display polarized apical and basolateral membranes, high transepithelial resistances, and regulated Cl^" secretory pathway analogous similar to that found in native crypt epithelium (Dharmsathaphorn et al. supra ) .

Electrophvsiology and Cyclic Nucleotide Assay Cl^" secretion was assessed as a short circuit current (Isc) using standard electrophysiologic techniques (Lencer et al., J . Cell Biol . 117: 1197-1209 _r 1992) . cAMP and cGMP were assessed in ethanol extracts of T84 cell monolayers by radioimmune assay kit (NEM, New Bedford, MA) . Hank's Balanced Soft Solution (HBSS) , containing 0.185 g/L CaCl₂, 0.098 g/L MgSθ₄, 0.4 g/L KCl, 0.06 g/L KH₂P0₄, 8 g/L NaCl, 0.048 g/L Na₂HP0₄, and 1 g/L glucose, to which was added 10 mM HEPES (pH 7.4), was used for all assays unless otherwise stated.

Cryptdin purification and synthesis Cryptdin enriched peptide fractions were prepared as previously described (Selsted et al. , J. Cell Biol . 118:929-936. 1992). Murine cryptdins 1-6 were obtained from these fractions after subsequent fractionation by HPLC (Selsted et al., supra ; and Ouellette et al., FEBS Letters 304:146-148, 1992). Cryptdin 3 was synthesized on solid supports as previously described (Selsted et al. , supra ) .

Crvptdin-induced pore formation Cryptdin-induced pore formation was assessed using T84 cells grown on glass coverslips incubated at 37°C. After 10 minutes, the coverslips were washed in fresh HBSS and examined by epifluorescense (490 nM excitation, 520 emission) and bright field microscopy using Nomarsky optics.

Application of a Cryptdin-Enriched Peptide Fraction Elicits an Increase in Transepithelial Short Circuit Current

When applied to apical membranes of T84 cells, a cryptdin enriched peptide fraction elicited an increase in transepithelial short circuit current (Isc, Fig. 7A) without a concomitant rise in intracellular cGMP or cAMP (Figs. 7B and 7C) . These data indicate that the secretory response was not due to the known apically- acting secretory agonists guanylin (Currie et al., Proc . Natl . Acad . Sci . USA 89_.:947:951, 1992) or E . coli heat stable toxin (Field et al., Proc . Natl . Acad . Sci . USA 7_5:2800-2804, 1978). These agonists elicit a secretory response via cGMP-mediated signal transduction. In addition, the adenosine-receptor inhibitor

8-phenyltheophyline had no detectable effect on the Isc response elicited by cryptdin-enriched fractions (Figs. 7D and 7E) , indicating that the Isc observed was not due to the apically acting agonist adenosine (Barrett et al., Am . J . Physiol . 265:C197-C203. 1989). As guanylin, the heat stable toxin derived from E . coli , and adenosine are the only compounds known to elicit a secretory response by interacting with receptors on the apical membrane, these results suggested the presence of a novel secretory agonist in the cryptdin-enriched peptide fraction.

Cryptdins were Found to be Directly Responsible for the Secretory Response to the Crvptdin- Enriched Peptide Fraction Described Above

To determine whether cryptdins were directly responsible for the observed secretory response, murine cryptdins 1-6 were purified to homogeneity and utilized as described below (for additional guidance, skilled artisans may consult Selsted et al., J. Cell Biol . 118_:929-936, 1992; Ouellette et al. , FEBS Letters 304:146-148, 1992; and Ouellette et al., Infect . Immun . _Y__\:5040-5047, 1994). At a concentration of 40 μg/ml, cryptdins 2 and 3 elicited a Isc response that was completely reversed by removing cryptdins from the apical reservoir (Fig. 8A) The effect of cryptdin 2 on Isc is shown in Fig. 8A by open squares, while that of cryptdin 3 is shown by closed circles. The removal of cryptdins from the apical reservoir is indicated by the "Wash" arrow.

Identification of Arginine at Amino Acid Postion 15 as Important for Bioactivity The secretory activity was highly specific for individual peptides, as cryptdin 1 and cryptdins 4-6 were inactive under these conditions. Since cryptdins 1-3 and cryptdin 6 differ only at positions 10, 15, 29, and 31 (Ouellette et al., Infect . Immun . §_2 :5040-5047, 1994), these data identify arginine (at amino acid position 15) as essential for bioactivity on human T84 cells. As cryptdins 2 and 3, but not cryptdins 1 and 6, display antimicrobial activity against Giardia lamblia (Aley et al., Infect . Immun . _2 :5397-5403, 1994) , it may be that the arginine residue at position 15 accounts for the ability of enteric defensins to interact with eukaryotic membranes.

Although the primary structures of cryptdins 2 and 3 are identical except at position 10, where threonine replaces arginine, respectively, the potencies of cryptdins 2 and 3 in eliciting a secretory response were not equal. These data suggest that the molecular locus at position 10 may also be important in cryptdin function on T84 cells. By analogy with the known crystal structures of human neutrophil defensins HNP-l and HNP-3 and rabbit defensins NP-1 and NP-5, these amino acids at position 10 are predicted to be involved in a conserved turn on the peptide surface, and are well positioned to influence the interaction between cryptdin and cell membranes.

Dose-Dependence

The effect of cryptdin 3 on T84 cells Isc was dose-dependent. These studies were carried out using synthetic, folded and oxidized cryptdin 3, prepared as described previously (Selsted et al., J . Cell Biol . 118:929-936, 1992) . Synthetic and natural cryptdin 3 displayed identical physiochemical and antimicrobial characteristics (Selsted, In Investigational approaches for studying the structures and biological functions... , Setlow, Ed., Plenum Press, New York, Vol. 15, pp. 137- 147, 1993).

When applied to apical membranes, cryptdin 3 elicited an increase in Isc with maximal currents of 65 μA/cm² at 600 μg/ml, and an apparent ED50 of 250 μ/ml (Fig. 8B) . Given the small volume of crypt lumen, and the high density of cryptdins on Paneth cell secretory granules, such concentrations are likely to be achieved within the crypt lumen in vivo . At low doses (<300 μg/ml for 30 minutes at 37°C) , the Isc responses to cryptdin 3 were fully reversible. However, at high doses, 30 minute incubations with cryptdin 3 resulted in a gradual reduction of Isc which paralleled a progressive and ultimately complete loss of monolayer resistance, presumably due to cytotoxicity (Lichtenstein, J . Clin . Invest . 8jL.93-100, 1991) .

The source of the Isc induced by cryptdin 3 was identified as a Cl^" current, the primary transport event responsible for the secretory response across mucosal surfaces. Substitution of membrane impermanent gluconate for Cl^" in transport buffers attenuated completely the Isc induced by cryptdin 3, presumably due to depletion of intracellular Cl^~ (Fig. 8C, open diamonds) . This interpretation was confirmed by adding Cl^~ back to the basolateral reservoir at 43 minutes, which restored the Isc. Replenishing Cl^" in buffers on control monolayers that were not exposed to cryptdin 3 had no effect on Isc's (Fig. 8C, open squares), indicating that this maneuver was specific for cryptdin 3 treated monolayers. The viability of control monolayers was demonstrated by the brisk secretory response to vasoactive intestinal peptide (VIP, 5 nM) at 57 minutes. These data show that the Isc induced by cryptdin 3 displayed dependency on Cl^". Further evidence that the cryptdin induced Isc's represent Cl^" transport was provided by the use of bumetanide (10 μM) , a specific inhibitor for the Na/K/2C1 uptake pathway. Na/K/2C1 co- transport is primarily responsible for Cl^" uptake in secretory epithelia, and inhibition of this cotransporter will also deplete intracellular Cl~. In agreement with data obtained by ion substitution (Fig. 8C) , bumetanide reduced cryptdin-induced Isc's by 75%.

Taken together, these data indicate that cryptdin 3 elicits a Cl^" secretory response from intestinal T84 cells. Such a response could occur by activation of pre-existing apically located Cl^" channels, or by formation of anion-conductive pores resulting from the insertion of cryptdins into the apical membrane.

Cryptdin 3 Elicits a Chloride Secretory Response from Cystic Fibrosis Affected Airway Cells

In the work described below, cryptdin 3 was shown to elicit a chloride secretory response from cystic fibrosis affected airway cells, i.e., JM15 cells (which are a passaged relative of the cells described by Jefferson et al., in Am. J. Physiol. 259:L496-505, 1990) . JM15 cells are affected with the Δ508 mutation in CFTR, which is the most common mutation seen in cystic fibrosis patients.

JM15 cells containing CFTR with the Δ508 mutation were cultured as described by Jefferson et al. (supra ) and either treated (+ cryptdin, as shown in Fig. 11) or not treated (control, and + forskolin, Fig. 11) . Fig. 11 illustrates the relationship between chloride current (I_cl in μA/cm²) and applied voltage (in millivolts, mV) in treated and untreated cells. The slope of the curves represents a measure of membrane conductance for chloride ion transport (i.e., a measure of chloride channel activity) . A steeper slope indicates higher chloride conductance and, thus, greater channel activity. As shown in Fig. 11, cells treated with cryptdin 3 display a significantly higher conductance to chloride ions than do control cells that were not treated with cryptdin 3. This is reflected by a steeper slope in the current-voltage relationship, and indicates that cryptdin 3 elicits an increase in apical membrane conductance to chloride, most likely due to pore formation by cryptdin 3 as it inserts into the membrane. In contrast, a separate group of cells treated with the cAMP agonist forskolin (but not treated with cryptdin 3) did not secrete chloride, as evidenced by little or no change in the slope of the curve as compared to that of control cells. Normally (i.e., in cells where chloride conductance is not compromised) , forskolin causes a strong cAMP response that activates CFTR (wild type) , and this results in a chloride secretory response. As shown in Fig. 11, forskolin has little or no effect on chloride conductance in JM15 cells; the current-voltage relationship for "+ forskolin" and "control" cells nearly overlap one another.

These data support the conclusion that JM15 cells are excellent models of cystic fibrosis in that they do not respond to forskolin by activation of CFTR and that cryptdin 3 elicits a chloride secretory response from cystic fibrosis affected cells and can correct the fundamental defect in chloride conductance that is seen in this disease.

EXAMPLE II; I_cln

Overexpression of l_cln in Rat C6 Glioma Cells Increases Swelling-Induced Taurine Efflux

Overexpression of I_cln in C6 cells increases swelling-induced organic osmolyte efflux (Fig. l) . Rat C6 glioma cells, which are available from the American Type Culture Collection (A.T.C.C. #CCL-107) , were cultured in Eagle's minimum essential medium (MEM; GIBCO, Gaithersburg, MD) with 10% fetal bovine serum and penicillin-streptomycin. These cells were transfected with PCDNAIII containing an I_cln insert or, as a control, PCDNAIII lacking the insert. In addition, the vector contained the neo^r gene which allowed stably transfected clones to be selected by standard techniques with G418.

Cultured C6 glioma cells were swollen by exposure to 200 mOsm medium and the efflux of tritiated taurine was measured by liquid scintillation counting during the first, second, and fifth successive cycles of swelling and shrinking. The rate of ³H-taurine efflux was quantified as the percentage of total cell counts per minute released during a 10 minute efflux period.

At each cycle, the efflux of ³H-taurine was greater from cells that overexpressed I_cln. "Rundown" of VSOAC current and organic osmolyte efflux occur with repetitive swelling (see e.g., Ackerman et al., J. Gen . Physiol . 103:153-179, 1994; Hand et al., J . Gen . Physiol . 105:37a, 1995). This phenomenon is apparent in Fig. 1.

In order to confirm that I_cln was overexpressed by C6 glioma cells that were transfected with PCDNAIII containing the I_Cι_n insert, Western blot analysis was performed. Sibling cultures of rat C6 glioma cells were transfected with vectors containing, as well as lacking, the I_cln insert, as described above. The cells were harvested and the proteins were extracted, separated by SDS-PAGE, and transferred to a solid support. The membrane-bound proteins were then incubated with an affinity purified polyclonal antibody generated against a C6 cell recombinant I_cln protein. The staining was greater in the glioma cells that were transfected with the I_Cι_n insert, indicating the overexpression of I_cln (see inset, Fig. 1) . This, in turn, indicates that I_cln is responsible for the observed increase in organic osmolyte efflux. I_cln May Be a "Porin-like" Channel that Spontaneously Inserts into the Plasma Membrane; The Anchor-Insertion Model

Beta barrels are a defining structural characteristic of porins, which are an evolutionarily ancient class of channel that transports organic solutes in a broadly selective fashion across bacterial outer membranes (Benz, Ann. Rev. Microbiol . 4^:359-393, 1988; Benz et al., Eur. J. Biochem . 176:1-19. 1988; Rosenbusch, Experientia £6:167-173, 1990). Porins have also been called "peptidoglycan-associated proteins," "peptidoglycan-associated general diffusion pore proteins," and "matrix proteins." Structurally, porins are soluble, highly acidic proteins that lack long stretches of hydrophobic amino acids (Benz, supra ; Benz et al., supra ; Eisele et al. , J . Biol . Chem . 265: 10217- 10220, 1990; Pfaller et al., J. Biol . Chem . 260:8188- 8193, 1985; Rosenbusch, Experientia 16:167-173, 1990) . Thus, they can insert spontaneously into lipid bilayers, and once inserted, they have a P₀ near unity and exhibit brief spontaneous channel closures (Benz, supra ; Benz et al., supra ; Berrier et al., FEBS Letters 306:251-256, 1992; Eisele et al., J. Biol . Chem . 265: 10217-10220, 1990; Trias et al., Science 258:1479-1481, 1992). As demonstrated herein, the functional and electrophysiological characteristics of I_cln are similar to those of porins. Thus, it is likely that I_cln forms the VSOAC channel, which is "porin-like" in nature. A model for the activation of I_cln/VSOAC in response to cell swelling is also presented below. This model is termed the "anchor-insertion model," and is based on the proposal that I_cι_n is normally anchored in the cytoplasm by binding to proteins such as those described by Krapivinsky et al. (Cell 7_6:439-448, 1994) , and is released and inserted spontaneously into the cell membrane in response to cell swelling. The OFF state of the channel would represent the cytoplasmically anchored form of I_cln/ and abrupt switching to the ON state would reflect its insertion into the cell membrane. Our proposed model is supported strongly by the data described herein, which were gathered using a variety of cellular, biochemical and molecular approaches.

I_cln Functions as a Rectifying, Anion- Selective Channel in Planar Lipid Membranes

Preparation of I_cln protein

To test the channel hypothesis and the anchor- insertion model, pure I_cln protein was reconstituted, and the reconstituted channel activity was compared to that of VSOAC. Toward this end, I_cln was cloned from a rat C6 glioma cell CDNA library by homology screening. The open reading frame of the C6 cell I_cln was subcloned by PCR and ligated into a PGEX-4T-1 glutathione-S-transferase (GST) gene fusion vector (Pharmacia) . After induction of the protein in transformed E . coli , the GST—Iς;in fusion protein was purified from bacterial lysates by affinity chromatography using glutathione Sepharose 4B~ (Pharmacia) . Icin ^was cleaved from the Sepharose- glutathione-GST complex with thrombin. As a control, bacteria were transformed with a PGEX-4T-1 vector coding only for the GST protein.

After induction of GST or the GST-I_cln fusion protein, equal quantities of the bacteria were lysed, and the lysates were subjected to identical affinity chromatography protocols. The purified I_cln protein and protein from the control preparation were separated by SDS-PAGE and silver-stained as shown in Fig. 3. The control preparation lacks I_cln but contains the same potential contaminants (e.g., trace quantities of thrombin, GST, and unknown proteins) as the I_cln preparation. There was no detectable (<0.010 mg/ml) protein in the "control" preparation as measured by BCA protein assay (Pierce) . The I_cln preparation contained 0.3 mg I_cln/ml. Therefore, the purity of the I_cln preparation is >97%.

Construction of a Planar Bilaver Containing I_cln Bilayer reconstitution experiments were carried out as previously described by Oh et al. (Am. J. Physiol . 2__: 1489-1499, 1993) and Ismailov et al. (J. Biol . Chem . 2___ι 10235-10241, 1994) . Briefly, proteoliposom.es were prepared from phosphatidylcholine by passing detergent solubilized I_cln or control protein through an Extracti- Gel™ D column (Pierce) . Bilayer membranes were constructed from a mixture of diphytanoyl- phosphatidylethanolamine, diphytanoly-phosphatidylserine, and oxidized cholesterol in n-octane (2:1:2 w/w/w; final lipid concentration equaled 25 mg/ml) . The lipid solution was spread over a 200 μm aperture drilled in a piece of polystyrene. Bilayer formation was indicated by an increase in membrane capacitance to final values of 300-400 pF. Membranes were formed in salt solutions containing 100 mM KCl and buffered to pH 7.4 with 10 mM MOPS. Proteoliposomes containing I_cln were spread over bilayers using a fire polished glass capillary. The trans side of the bilayer chamber served as the virtual ground. Current digitization and data storage were performed as described (Ismailov et al. , supra ) , and data analysis was carried out using pCLAMP software.

Channel activity is seen reproducibly in bilayers reconstituted with I_cln proteoliposomes. Incorporation of a single channel into the bilayer occurs approximately once in every 70 attempts. Multiple channel incorporations are seen more frequently. In contrast, channel activity has never been seen in bilayers reconstituted with control protein proteoliposomes that lack I_cι_n- Over 2,000 reconstitution attempts have been carried out using proteoliposomes prepared from bacterial lysates lacking I_cln. If the channels observed were due to contaminants, -30 single channels should have been seen.

In addition to spreading I_cι_n proteoliposomes over the bilayer, channel activity was also seen when a fire polished glass capillary was dipped into an I_cln solution and spread onto the bilayer, and when the I_cln protein was added directly to the solution bathing the bilayer. However, the proteoliposome approach provides a much more controlled method for reconstituting single channels.

Electrophysiological Characteristics of I_cln An example of typical single channel currents measured at +80 and -80 Mv is shown in Fig. 4. As can be clearly seen, the channel is rectified. Single channel conductance in 100 mM KCl is 62 ± 0.7 pS (mean ± S.E.M.; n=37) at +80 mV and 28 ± 0.5 pS (mean ± S.E.M.; n=37) at

-80 mV. Mean channel P_Q is 0.87 ± 0.01 (n=32) . With symmetrical 100 mM KCl solutions, the reversal potential is close to 0 mV (Fig. 5) . A 10:1 trans-to-cis KCl gradient shifts the reversal potential by -52 ± 1.5 mV

(n=4) , demonstrating that the channel is highly selective for anions over cations. The anion selectivity of the I_cln channel has not been fully assessed but it is clear that the channel is more permeable to I^" and Br^" than to Cl^" (I^" > Br^" > Cl^") .

The I_cln channel orients in the bilayer in a random fashion. Rectification of single channel current occurred in 18 experiments when the cis bath was positive. In another 22 experiments, rectification was observed when the cis bath was held at negative potentials. Because VSOAC is outwardly rectifying, the side of the bath chamber in which rectification was observed was operationally defined as the intracellular side of the channel. In other words, if the single channel conductance was -60 pS when the cis bath was +80 mV and -30 pS when the cis bath was -80 mV, the intracellular side of the channel was assumed to be facing the cis bath.

Extracellular ATP Decreases Single Channel Conductance

Exposure of the I_cln channel to 1 mM ATP reduced the single channel current -40%, increased open channel noise, and reduced channel P₀ to 0.32 ± 0.02 (Fig. 6; mean ± S.E.M.; n=6) . ATP was effective only when added to the presumed extracellular side of the I_cln channel.

Table 1 summarizes the shared characteristics of the VSOAC and I_cln channels based on the studies presented herein and others (Strange et al., Am. J. Physiol . , 1995) .

Table 1 Shared Characteristics of VSOAC and the Reconstituted I_cln Channel

From the data presented herein, it can be concluded that I_cι_n is the VSOAC channel, or a functional subunit of the channel. Other Embodiments Identification of Channel-Forming Polypeptides In light of the demonstration that I_Cχn and cryptdins 2 and 3 spontaneously form channels for selective transport of monovalent anions, those skilled in the field will recognize that they can: (1) obtain other channel-forming polypeptides that have relevant characteristics similar to those of I_cln or the cryptdins, as described herein, and (2) use these channel-forming polypeptides, or mutant forms thereof, to treat disorders that are caused by a disturbance in the conductance of anions across the cell membrane in vivo .

Homologues of the polypeptides of the invention, including human homologues, can readily be obtained by those of skill in the art. For example, homologues of I_cln or of the cryptdins could be obtained by screening a genomic or cDNA library generated from a given tissue or cell line with an appropriate I_cln or cryptdin cDNA probe under conditions that allow the probe to hybridize with the I_cln-like gene of that species.

The probe can be designed based on the sequence of I_cln (as disclosed, for example, by Krapivinsky et al., Cell 26:439-448, 1994, or Paulmichl, Nature 356:238-241. 1992) or the sequence of cryptdin 2 or cryptdin 3 or of the concensus sequence of a cryptdin (as disclosed, for example by Selsted et al., U.S. Patent No. 5,422,424) . Human cryptdins are discussed below.

The libraries can be prepared from tissues that are known to express I_cln or cryptdins, such as the intestine. Methods for generating and screening libraries are well known to persons skilled in the art of molecular biology (see e.g., Sambrook et al., In Molecular Cloning: A Laboratory Manual, 2nd Ed., Cold Spring Harbor Press, Cold Spring Harbor, New York, 1989) . In addition, genomic and cDNA libraries from many species (including human) are commercially available.

Typically, libraries are screened using hybridization conditions that are of low to moderate stringency. These conditions favor specific interactions between completely complementary sequences, but allow some non-specific interaction between less than perfectly matched sequences to occur as well. After hybridization, the nucleic acids can be "washed" under moderate or high conditions of stringency to dissociate duplexes that are bound together by some non-specific interaction (the nucleic acids that form these duplexes are thus not completely complementary) .

As is known in the art, the optimal conditions for washing are determined empirically, often by gradually increasing the stringency. The parameters that can be changed to affect stringency include, primarily, temperature and salt concentration. In general, the lower the salt concentration and the higher the temperature, the higher the stringency. Washing can be initiated at a low temperature (for example, room temperature) using a solution that contains salt at a concentration that is equivalent or lower than the concentration of salt in the hybridization solution. Subsequent washing can be carried out using progressively warmer solutions having the same salt concentration. As alternatives, the salt concentration can be lowered and the temperature maintained in the washing step, or the salt concentration can be lowered and the temperature increased. Additional parameters can also be altered. For example, use of a destabilizing agent, such as formamide, alters the stringency conditions.

In reactions where nucleic acids are hybridized, the conditions used to achieve a given level of stringency will vary. There is not one set of conditions, for example, that will allow duplexes to form between all nucleic acids that are 85% identical to one another; hybridization also depends on unique features of each nucleic acid. The length of the sequence, the composition of the sequence (for example, the content of purine-like nucleotides versus the content of pyrimidine- like nucleotides) and the type of nucleic acid (for example, DNA or RNA) affect hybridization. An additional consideration is whether one of the nucleic acids is immobilized (for example, on a filter) .

An example of a progression from lower to higher stringency conditions is the following, where the salt content is given as the relative abundance of SSC (a salt solution containing sodium chloride and sodium citrate; 2X SSC is 10-fold more concentration than 0.2X SSC). Nucleic acids are hybridized at 42°C in 2X SSC/0.1% SDS (sodium dodecylsulfate; a detergent) and then washed in 0.2X SSC/0.1% SDS at room temperature (for conditions of low stringency); 0.2X SSC/0.1% SDS at 42°C (for conditions of moderate stringency); and O.IX SSC at 68°C (for conditions of high stringency) . Washing can be carried out using only one of the conditions given, or each of the conditions can be used (for example, washing for 10-15 minutes each in the order listed above. Any or all of the washes can be repeated. As mentioned above, optimal conditions will vary and can be determined empirically.

Once detected, the nucleic acid molecules can be isolated by any of a number of standard techniques (see, for example, Sambrook et al., Molecular Cloning, A

Laboratory Manual, 2nd Ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1989).

A second standard technique that can be used to identify homologues of the polypeptides of the invention is PCR-based cloning. This technique is enabled by the disclosure of nucleic acid sequences encoding I_cln and cryptdins 2 and 3, which can be employed to select suitable oligonucleotide primers for the reaction, e.g. the primers can consist of almost any sequence over 8 bases from the sequences of I_cln and cryptdins 2 and 3, which are in the public domain.

Candidate cryptdin polypeptides, to be evaluated by the methods given herein for use in the invention, include HD-5 and HD-6 as cited in Jones et al. (J. Biol . CheJii. 267:23216-23225. 1992; FEBS Lett . 315:187-192, 1993) , hereby incorporated by reference.

It is also possible to use the sequence of DNA encoding β barrel channel-forming proteins to identify and isolate a channel-forming peptide with the desired characteristics, for example, porin, aerolysin, and other molecules described above.

Obtaining Channel-Forming Polypeptides Once isolated, nucleic acids that are homologous to I_cln or to cryptdins 2 or 3 (for example, human homologues of these nucleic acids) can be used to produce substantial amounts of the encoded polypeptides, which may in turn, be examined to determine whether it has characteristics that are substantially similar to those of the I_cln or cryptdin polypeptides disclosed herein. Polypeptides thus obtained are recognizable as I_cι_n or cryptdin 2 or cryptdin 3, and are therefore useful as therapeutics based on their water-solubility, as well as their ability to spontaneously insert into a lipid bilayer and to selectively conduct the transport of chloride ions across the bilayer.

Optimizing Channel-Forming Polypeptides The I_cln and cryptdin polypeptides disclosed herein, or any other form of these polypeptides (for example, those obtained by homology screening as described above) , as well as suitable 0-barrel forming pores, can be used to generate useful mutants. Standard methods for generating mutant polypeptides are well known to skilled artisans. These methods can be used to generate site-specific point mutations within the coding sequence of a gene encoding a polypeptide of the invention. Alternatively, sites could be mutated by deletion. The amino acid residues that are the strongest candidates for mutation are residues 35-94 of I_cln. Without limiting the scope of the invention, these residues are selected because they face the lumen of the pore. Candidate mutations are substitutions or insertions of bulky amino acids that would alter the channel to exclude larger organic anions, thus rendering the channel especially suitable for therapeutic application. Skilled artisans may be further guided in selecting desired mutations by consulting publications regarding the structure and function of porin molecules, such as the review by Benz (Ann. Rev. Microbiol . 42:359- 393, 1988).

Administering the Polypeptides of the Invention The polypeptides of the invention can be administered to a patient in order to restore chloride flux across a cell membrane by a variety of routes that are well known to persons skilled in the art of pharmacology. In cystic fibrosis and other respiratory diseases, it is expected that the preferred route will be by inhalation (aspiration) . The polypeptide could be administered alone or in conjunction with a lipid.

Furthermore, the polypeptide could be administered with a physiologically acceptable carrier, possibly with surfactants. For administration by inhalation, the polypeptides for use according to the present invention are conveniently delivered in the form of an aerosol spray presentation from pressurized packs or a nebulizer, with the use of a suitable propellant, for example, dichlorodifluoromethane, trichlorofluoromethane, dichlorotetrafluoroethane, carbon dioxide or other suitable gas. In the case of a pressurized aerosol the dosage unit may be determined by providing a valve to deliver a metered amount. Capsules and cartridges of, for example, gelatin for use in an inhaler or insufflator may be formulated containing a powder mix of the compound and a suitable powder base such as lactose or starch.

The dosage and length of any treatment are known to depend on the nature of the disease and to vary from patient to patient as a function of age, weight, sex, and general health, as well as the particular compound to be administered, the time and route of administration, and other drugs being administered concurrently. The dosage of the polypeptides of the invention can be 0.01 mg to 100 mg. For example, 0.1 to 100 mg, or l to 10 mg, or 2 to 10 mg, or 5 to 10 mg can be administered one or more times per day. Furthermore, administration need not be limited to a single type of polypeptide. For example, both cryptdin 2 and cryptdin 3 can be administered.

Skilled artisans will be guided in their determination of the dosage at which to administer I_cln and cryptdins 2 and 3 by, for example, Gregoriadis (1979, Drug Carriers in Biology and Medicine , Academic press) and Goodman and Gilman (The Pharmacological Basis of Therapeutics , 6th edition) .

What is claimed is:

Claims

CLAIMS - 32 -

1. A therapeutic composition comprising a polypeptide formulated for inhalation or aspiration therapy, the polypeptide being characterized by the ability to: (a) spontaneously form a channel in a lipid bilayer,

(b) selectively transport chloride ions across said lipid bilayer, and

(c) dissolve in an aqueous physiological solution.

2. The therapeutic composition of claim 1, wherein said polypeptide is at least 70% identical to a polypeptide that forms a naturally occurring β barrel pore.

3. The therapeutic composition of claim 2, wherein said β barrel pore-forming polypeptide is selected from the group consisting of I_cln, a porin, annexion, and aerolysin.

4. The therapeutic composition of claim 3, wherein said polypeptide is I_cln.

5. The therapeutic composition of claim 4, wherein said polypeptide is human I_cln-

6. The therapeutic composition of claim 1, wherein said polypeptide is a cryptdin.

7. The therapeutic composition of claim 6, wherein said cryptdin is a human cryptdin.

8. The therapeutic composition of claim 6, wherein said cryptdin is cryptdin 2.

9. The therapeutic composition of claim 6, wherein said cryptdin is cryptdin 3.

10. A therapeutic composition comprising a cryptdin polypeptide formulated for inhalation or aspiration therapy.

11. A method of treating a patient who has a disease that is characterized by defective chloride conduction across a cellular membrane, said method comprising administering to said patient an amount of a polypeptide sufficient to spontaneously insert into said cellular membrane and to form channels that normalize chloride conduction.

12. The method of claim 11, wherein said polypeptide is soluble in an aqueous physiological solution.

13. The method of claim 12, wherein said polypeptide is at least 70% identical to a polypeptide that forms a naturally occurring β barrel pore.

14. The method of claim 13, wherein said β barrel pore is selected from the group consisting of I_cln, a porin, annexion, and aerolysin.

15. The method of claim 14, wherein said polypeptide is I_cln.

16. The method of claim 15, wherein said polypeptide is human I_Cι_n-

17. The method of claim 13, wherein said administration is by aspiration.

18. The method of claim 13, wherein said patient has cystic fibrosis.

19. The method of claim 11, wherein said polypeptide is a cryptdin.

20. The method of claim 19, wherein said cryptdin is a human cryptdin.

21. The method of claim 19, wherein said cryptdin is cryptdin 2.

22. The method of claim 19, wherein said cryptdin is cryptdin 3.

23. The method of claim 19, wherein said administration is by aspiration.

24. The method of claim 19, wherein said patient has cystic fibrosis.

25. A method of treating a patient who has a disease that is characterized by defective chloride conduction across a cellular membrane, said method comprising administering to said patient an amount of a cryptdin polypeptide sufficient to spontaneously insert into said cellular membrane and to form channels that normalize chloride conduction.

232847.Bll