WO1997004311A2 - Cell sorting with fluorescent peptides - Google Patents

Abstract

A method for sorting cells included in a cell population is described. The method includes the step of first exposing the cell population to a biologically active fluorescent peptide containing peptide and light-emitting moieties. A first group of cells in the cell population are then labelled when the peptide of the biologically active fluorescent peptide binds to a corresponding receptor contained on (or in) each cell in the first group of cells. The first group of cells or a group of cells excluding the first group of cells are then sorted from the cell population.

Description

CELL SORTING WITH FLUORESCENT PEPTIDES Background

This invention relates to methods for cell sorting and purification.

Replacing or transplanting cells within a tissue or organ is promising as a permanent cure for diseases such as Parkinson's and diabetes. In general, the goal of cell replacement or transplant therapy is to replenish the body's stores of neurotransmitters, peptides, or hormones by permanently substituting or augmenting damaged cells with new, healthy replacements. These therapies are complicated by the fact that tissues or organs used as cell sources are usually a mixture of different cells. For example, the human pituitary gland includes at least six different cell types secreting luteinizing hormone (LH) , follicle-stimulating hormone (FSH) , TSH, ACTH, growth hormone, and prolactin. Consequently, it is difficult, if not impossible, to obtain well-characterized, homogeneous cell populations from this source tissue.

Cell-sorting techniques isolate specific types of cells which, in turn, improve the efficacy of the cell transplant or replacement therapy. Cell sorting requires that cells in a population to be sorted include a clear, distinguishing feature (e.g., size, granularity) which distinguishes them from other, undesired cells. In one sorting method, antigens on the cells are first ^■■tagged" with primary antibodies. The primary antibody is detected using a secondary antibody labeled with a fluorophore. The cell population is then sorted using fluorescence-associated cell sorting (FACS) . Primary antibodies used in FACS are directed toward unique cell surface antigens on cells of the hematopoietic and other lineages. Labelled antibodies have also been used to detect unique intracellular antigens, such as the enzymes necessary for the synthesis of individual hormones. In this case, the cells are permeabilized with alcohol or aldehydes so that the primary and secondary antibodies can reach intracellular antigens; this process typically results in death of the cell.

In many cases, the antibodies necessary to selectively sort cells of interest are either unavailable or unknown. For example, antibodies have not been raised for surface antigens included on many cells of the pituitary gland, brain, pancreas, and other tissues. Cell cultures of the hypothalamic tissue contain a multiplicity of cell types (including ependymal cells, cells containing dopamine, beta-endorphin, neuropeptide Y, somatostatin, and other hormone-containing neurons) . The isolation of any one of these cell types is currently impossible. Thus, most primary cell cultures derived from these tissues include mixtures of cells.

Summary

Applicants' invention addresses these and other limitations by providing an improved method for cell sorting which provides cell samples having improved purity. The method of the invention, which sorts cells included in a cell population containing a first group of cells, involves first contacting the cell population with a biologically active fluorescent peptide (described below) which contains both peptide and light-emitting moieties. The first group of cells are labelled when the peptide moiety of the biologically active fluorescent peptide binds to a corresponding receptor contained on (or in) each cell. The first group of cells, or a group of cells excluding the first group of cells, are then sorted from the cell population. The sorting step is preferably carried out using flow cytometry in which the entire cell population is irradiated to induce fluorescence in the light-emitting moieties attached to the first group of cells. A time period of between 5 and 120 minutes preferably elapses between the labelling and irradiating steps. The cells attached to the fluorescing, light-emitting moieties are then sorted from the cell population. Alternatively, all but the cells attached to fluorescing, light-emitting moieties are then sorted from the cell population.

In another embodiment, the biologically active fluorescent peptide is attached to an electrically or magnetically active material. In this case, the sorting step includes exposing the electrically or magnetically active material to an electric or magnetic field, and then sorting the cells attached to materials which exhibit an electric or magnetic response.

The biologically active, fluorescent peptides used in the cell-sorting method have the following structure:

X

R-_j_ C R

where R_x is a light-emitting moiety; R₂ is a peptide of between 2 and 200 amino acids and is not neurotensin; and X is selected from the group consisting of =0, =S, -OH, =C=0, =NH, -H, -OR, -NR, -R, and -R₃R₄, where each R, R₃, and R₄, independently, is H or a C1-C6 branched or unbranched, substituted or unsubstituted, alkyl. Preferably C is bonded to R₂ through an amino residue of an alpha carbon atom. In other aspects, R₂ is a peptide of between 2 and 200 amino acids bound to C by a binding moiety selected from the group consisting of the residues Ala, Arg, Asn, Asp, Cys, Gin, Gly, His, Ile, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, and Val. In this case X is as described above. In still other aspects, R₂ is a peptide of between 2 and 200 amino acids, and X is selected from the group consisting of =0, -OH, =C=0, =NH, -H, -OR, -NR, -R, -R₃R₄, wherein each R, R₄ and R₃ are as described above.

In other aspects, a C1-C6 branched or unbranched, substituted or unsubstituted, alkyl serving as a "linker group" is disposed between the R₂ and C-X moieties. In this case, R₂ is preferably an opioid peptide and R₇ is preferably -NH(CH₂)₅NH-. In still other aspects, R₂ is neurotensin or an analogue thereof.

Preferably R₂ is selected from the group consisting of endothelin, galanin, and somatostatin (i.e., somatostatin-14 or D-Trp-somatostatin-14) .

Preferably, R_χ is bound, through C, to a region of the R₂ peptide which is not associated with the peptide's biological activity. When R₂ is endothelin, R_χ is bonded to the ninth amino acid residue of R₂ (i.e., the e-amine group of Lys) . When R₂ is galanin, R i*³ bonded to the fifth amino acid residue of R₂ (i.e., the e-amine group of Lys) . When R₂ is neurotensin or an analogue thereof, R_λ is bound, through C, to an Arg-Pro-Trp or Arg-Pro-Tyr moiety of R₂. When R₂ is somatostatin or an analogue thereof, is bound, through an amino or thiol moiety of an -Ala- residue to R₂.

R is preferably selected from the group consisting of fluorescein, FTC, Texas Red, phycoerythrin, rhodamine, carboxytetramethylrhodamine, DAPI, indopyra dyes. Cascade blue cou arin, NBD, Lucifer Yellow, propidium iodide, a porphyrin, Bodipy, CY³, CY⁵, and derivatives and analogues thereof. The R moiety is can be attached to C-X through a linking group selected from the group consisting of indoacetamide, maleimide, isothyocyanate, succinimidyl ester, sulfonyl halide, aldehyde, glyoxal, hydrazine, and derivatives thereof.

The invention provides a substantially pure population of cells (preferably dopamine cells) sorted by the method described above. Once sorted, the cells function as normal, primary cultured cells.

In all cases, "biologically active fluorescent peptide" means a labelled peptide which binds to a receptor having an affinity for the labelled peptide which is at least 0.25% of that of the corresponding unlabelled peptide. The receptor affinity for the labelled peptide is preferably at least 1.0% of that of the corresponding unlabelled peptide; most preferably, the receptor affinity for the labelled peptide is greater than 95% of that of the corresponding unlabelled peptide. Receptor affinity, in this case, is determined either using known methods involving competition binding with radioactively labeled peptides or other fluorescence methods for measuring Kd for the receptor-peptide interaction.

By "peptide" is meant a chain of amino acids of any length. Included in this term are proteins and polypeptides.

By "substantially pure" or "substantially purified" is meant that 60% or more of the isolated cell population is homogeneous; more preferably greater than 90% of the cell population is homogeneous. By "homogeneous populations of cells" is meant a cell population wherein the cells share a common peptide receptor.

The invention has a number of advantages. The biologically active fluorescent peptides described herein are highly selective markers for a wide range of cell types, and are therefore effective when used for cell sorting. As an example, neurotensin receptors are selectively associated with dopamine cells: 90% of dopamine neurons of the substantia nigra and ventral tegmental area possess neurotensin receptors, while fewer than 10% of adjacent cell populations bear these receptors. Biologically active fluorescent neurotensin (i.e., a neurotensin peptide labelled with a fluorophore) can therefore be used to sort dopamine cells.

Biologically active fluorescent peptides typically bind to peptide receptors located of the cell's surface. This labeling mechanism advantageously does not inhibit the activity of the receptor or other cellular processes, and thus preserves the cell's biological activity. The fluorescent peptides do not adversely affect cellular function, and thus can be used to label and sort live cells (something that is not possible with antibodies directed toward intracellular antigens) . Furthermore, the fluorescent peptides are non-toxic and may be rapidly degraded within tissue after cell sorting.

Fluorescent peptides can also be combined with other techniques for cell sorting, such as techniques which sort cells according to their size. In addition, rare subpopulations of cells, such as transfected cells expressing a cell surface receptor, may be selected and sorted using the method and compounds of the invention. Substantially pure cell populations obtained by sorting cells with the fluorescent peptides of the invention also have many advantages, particularly when used in cell transplant and replacement therapies. Compared to cell mixtures, substantially purified cells are relatively homogeneous and are therefore easily characterized (both structurally and functionally) prior to being transplanted. In this way it is possible to gauge how particular cell types survive and integrate into the host tissue. Because of their homogeneity, the number of cells taken from substantially purified cell populations can be accurately controlled.

When used in transplant therapies, embryonic cells isolated with the fluorescent peptides of the invention advantageously respond "normally" (i.e., similar to the cells being replaced) to physiological signals. For example, purified embryonic cells secrete a desired chemical only when the appropriate physiological signals are received; these cells, therefore, accurately mimic the normal function of the lost or damaged cells.

In addition to their use in cell transplant or replacement therapies, substantially purified cell populations can be used to answer several basic research questions which can not be addressed using heterogeneous cell mixtures. For example, in vitro studies on substantially purified cell populations permit accurate investigation of phenomena such as receptor binding, gene expression, gene isolation, release of signalling molecules, response to drugs, apoptosis, and the selective application of the polymerase chain reaction. These studies are used to investigate a wide range of biological processes. Cells isolated with fluorescent peptides can, for example, be used to clone specific gene sequences suspected to be present in only specific cell types. In general, co-culturing experiments with substantially purified cell types, when combined with various manipulation of the cells' external environment, is an invaluable tool for the study of cell regulation. Purified cell preparations are also advantageous when used in drug toxicity studies. Testing the actions of drugs on a system of homogeneous cells, as compared with a mixture of cells, can identify direct drug actions more accurately. This type of study also offers the possibility of facile identification of the drug's action mechanisms or toxicities. This reduces both the total number of experiments required for drug characterization and the need for certain in vivo testing procedures. For example, if a drug-stimulating dopamine release is desired, the drug may be tested in a substantially purified culture of dopamine neurons. Such experiments answer the question of whether or not the drug stimulates dopamine release directly without requiring in vivo testing. The toxicity of compounds can also be examined with substantially purified cell types to identify the mechanisms of action of certain drugs on particular cell types. Substantially purified cell populations can also serve as live vectors for gene therapy or drug delivery. The fluorescent peptides described herein can be used to sort a wide range of cell types, including neurons and pancreatic, ovarian, liver, testicular, kidney, lung, adrenal, and intestinal cells.

These and other advantages will be apparent from the following detailed description, and from the claims. Brief Description of the Drawings Fig. 1 is a flow chart outlining the preferred steps used for purification of homogeneous cell populations;

Fig. 2 is a chemical structure of a fluorescent peptide according to the invention; Fig. 3 is a flow chart showing the general synthetic procedure used to synthesize fluorescent peptides of the invention;

Figs. 4A and 4B are chemical structures of, respectively, flourescent endothelin and fluorescent galanin;

Fig. 5 is a graph showing the effect of temperature on binding of fluorescent neurotensin to the surface of dissociated rat mesencephalic neurons;

Fig. 6 is a graph illustrating the time-dependent binding of fluorescent neurotensin to the surface of dissociated rat mesecenphalic neurons and the time-dependent survival as a percentage of the total number of cells;

Figs. 7A-7F show flow cytometric profiles of specific and non-specific binding of fluorescent neurotensin to, respectively, SN17 control cells (Fig. 7A and 7D) , a Chinese Hamster Ovary (CHO) cell line transfected with the neurotensin receptor (Figs. 7B-7E) , and a CHO cell line without with the neurotensin receptor (Figs. 7C and 7F) ;

Figs. 8A-8B show, respectively, flow cytometric profiles of non-specific and specific binding of fluorescent neurotensin to a cell culture comprised mainly of glial cells in the absence of levocabastine; Figs. 8C-8D show, respectively, flow cytometric profiles of non-specific and specific binding of fluorescent neurotensin to a cell culture comprised mainly of glial cells in the presence of levocabastine; Figs. 9A and 9D show, respectively, flow cytometric profiles of non-specific and specific binding of fluorescent neurotensin to control COS-7 cells;

Figs. 9B and 9E show, respectively, flow cytometric profiles of non-specific and specific binding of fluorescent neurotensin to COS-7 cells transfected to express the somatostatin receptor;

Figs. 9C and 9F show, respectively, flow cytometric profiles of non-specific and specific binding of fluorescent neurotensin to COS-7 cells transfected to express the neurotensin receptor; Figs. IOA and 10D show flow cytometric profiles of, respectively, non-specific and specific binding of fluorescent neurotensin to mesecenphalic neurons after 0 days;

Figs. 10B and 10E show, respectively, flow cytometric profiles of non-specific and specific binding of fluorescent neurotensin to mesecenphalic neurons after 7 days in the absence of levocabastine;

Figs. IOC and 10F show, respectively, flow cytometric profiles of non-specific and specific binding of fluorescent neurotensin to mesecenphalic neurons after 7 days in the presence of levocabastine; and,

Figs. IIA and 11B show, respectively, flow cytometric profiles of non-specific and specific binding of fluorescent somatostatin to cortical cell cultures.

Detailed Description

Cell Sorting

The invention provides a method for cell sorting using biologically active fluorescent peptides containing peptides and light-emitting moieties. During the sorting process, cells are labelled when the peptide moiety binds to peptide receptors on the surface of each cell to be sorted. The light-emitting moiety of the labelled cell is then optically excited to induce fluorescence. The fluorescence is detected and analyzed to sort the cells. Fig. 1 shows a flow chart listing a series of preferred steps for cell sorting. Desired cells in the tissue of interest are first dissociated from a tissue (step 10) . A cell suspension is then formed and plated in a cell culture (step 14) . Cultured cells are dissociated from the cell culture (step 16) , incubated for a short time period (step 18) , and then condensed using a centrifuge (step 20) .

The cells are then incubated with the biologically active fluorescent peptides (step 22) described in detail below. Selective populations of cells endowed with peptide receptors are labeled during the incubation step. The fluorescent peptide retains a high affinity for its peptide receptors and thus ensures that peptide receptors - li ¬ on the surface of the cells are labelled with a high binding affinity.

Labeled cells are isolated from unlabeled cells using a conventional flow cytometer and cell sorter (step 24) . Flow cytometry is typically performed at low temperatures (e.g. 4°C) to prevent dissociation of the fluorescent peptide from the receptor. In this method, single cells contained in liquid droplets are rapidly passed as a stream of liquid through a series of laser beams. The laser beams induce fluorescence from fluorescent peptides attached to individual cells; this process electrostatically changes the droplets. Optical detectors then measure the presence and intensity of a fluorescent signal. Based on the presence or absence of a signal, the electrostatically charged droplets are deviated from the stream by electrically biased deflection plates. The droplets are then collected in different receptacles to sort the cells.

Once the cell population has been sorted, the fluorescent peptides can be removed by displacement with an unlabeled peptide (step 26) . Alternatively, the fluorescent peptide can be left on the cell (step 28) , as it typically has no adverse effect on the cell's biological activity. Since the fluorescent peptides maintain the biological properties of endogenous peptides, they are entirely non-toxic and may be rapidly degraded within the cells after cell sorting. Fluorescent Peptides

Fig. 2 shows a generic structure of a biologically active fluorescent peptide 60 for use in the method described above. The fluorescent peptide 60 includes a light-emitting moiety 62, such as a fluorescent dye, attached by a -(C-X)- bond to a peptide moiety 64. The peptide moiety contains both a receptor binding region 66 and one or more "inactive" regions 68. The receptor binding region 66 binds to the peptide's associated receptor located on the cell surface; the inactive regions 68 do not significantly participate in the peptide/receptor binding process. The light-emitting moiety 62 is chemically attached to the peptide's biologically inactive region 68 so that the peptide retains a biological activity for a fluorescent peptide that is comparable to that of the native peptide. In this way, the light-emitting moiety does not sterically hinder or otherwise significantly affect the region involved in receptor binding; the biological activity of the compound is thus maintained at a high level.

The compounds of the invention preferably have a 1:1 molar ratio between the light-emitting 62 and peptide 64 moieties. The attachment of more than one light-emitting moiety per peptide may, in some cases, reduce biological activity. The 1:1 molar ratio is also important for the quantification of cells and receptors during labelling applications, as the measured fluorescence intensity can be easily analyzed to determine the number of binding peptide moieties. In addition, multiple fluorophores attached to a single peptide quench the fluorescence, as a loss of emission intensity sometimes occurs when multiple fluorophores are present in very close proximity to each other.

Once bound, the light-emitting moieties preferably retain optical properties similar to those of the unbound fluorophore. In this way, the compound, even when bound to its receptor, emits light following absorption of an incident optical field.

The general synthetic method for generating light- emitting biologically active compounds of the invention is shown in Fig. 3. This method begins with the step 80 of incubating the peptide and fluorophore of choice to form a mixture of compounds. Incubation is performed under conditions which permit optimal peptide labelling (see below) . The mixture of compounds includes biologically active and inactive whole peptides, cleaved fragments of peptides, and singly and multiply labelled peptides.

The light-emitting and peptide moiety form covalent bonds in the solution. Unbound fluorophore is then removed (step 82) . This selection process is performed using techniques such a column chromatography or other analytical techniques known in the art. The resultant eluent contains a mixture of labelled biologically active and inactive peptides.

This solution is then collected and subjected to a high-stringency pharmacological binding assay (step 84) . In this assay, only biologically active compounds are bound to tissue receptors; inactive compounds are washed away. The assay is typically performed on tissue sections, receptor- coated columns, or membrane homogenates. An aliquot of the fluorescent peptide mixture is typically first dissolved in an aqueous solution. The solution is then incubated with an immobilized tissue sample containing high numbers of the peptide's receptor. The selection process is designed to separate compounds exhibiting substantial biological activity from relatively inactive compounds. If necessary, during the assay, binding of the biologically active compounds may be rapidly observed visually (from the sections) , in a fluorometer (from precipitated membrane homogenates) , or by using more sensitive techniques such as fluorescence polarization spectroscopy.

The receptor-bound, biologically active fluorescent peptides are then removed from the tissue surface (step 86) . Inactive compounds are removed either by centrifugation of membrane homogenates or, in the case of sections, by rapidly rinsing the sections in incubation buffer. The membranes are then resuspended in binding buffer. The biologically active compounds are removed from the cell surface by incubation in a high salt/acid wash solution.

Once isolated, biologically active compounds are analyzed (step 88) using known techniques, such as laser- induced capillary zone electrophoresis, capillary electrophoresis and mass spectroscopy, and HPLC following carboxypeptidase digestion or other means of amino acid sequencing. Such analysis is used to identify the inactive site at which the light-emitting moiety is attached to the peptide. Once the attachment site is determined, the appropriate amino acid is attached directly to the light-emitting moiety prior to synthesis of the peptide (step 90) . In this way the compound is synthesized in a highly purified form using standard techniques, such as solid-phase peptide synthesis. If desired, the resulting complex can be further purified (step 92) , preferably using a column-based method such as HPLC, and then eluted. In this way, large quantities of biologically active, labelled peptide compounds are easily generated in an automated fashion.

Peptides and light-emitting moieties are coupled together by modifying amino acid functional groups on the peptide. Reactive groups on the light-emitting moiety, such as unsaturated alkyl groups, react with the modified amino acid to form the fluorescent peptides. Typically, a thiol or amine group of the peptide is modified as described in the "Handbook of Fluorescent Probes and Research Chemicals - 5th Edition" by Richard P. Haugland (1992) , the contents of which are incorporated herein by reference. Thiols react with alkylating groups (R'-Z) to yield relatively stable thiol ethers (R-S-R'), with the leaving group Z preferably being a halogen (e.g., Cl, Br, or I) or a similar moiety. The most common reagents for derivatization of thiols are haloacetyl derivatives. Reaction of these reagents with thiols proceeds rapidly at or below room temperature in the physiological pH range.

Light-emitting moieties can also be attached to amine groups on an amino acid of the peptide. The conditions used to modify these groups depend on the class of amine (e.g., aromatic or aliphatic) and its basicity. Aliphatic amines, such as the α-amino group of lysine, are moderately basic and reactive with acylating reagents. The concentration of the free-base form of aliphatic amines below pH 8 is very low; thus, the kinetics of acylation reactions of amines by isothiocyanates, succinimidyl esters, and other reagents is strongly pH-dependent. Although amine acylation reactions should usually be carried out above pH 8.5, the acylation reagents degrade in the presence of water, with the degradation rate increasing as the pH increases. The α-amino function of the amino terminus usually has a pKa of -7, so that it is selectively modified by reaction at neutral pH. The peptide moiety is preferably attached to the light-emitting moiety by a -(CX)- bond so that the biological activity of the fluorescent peptide is maintained. This bond includes groups such as C=0, C=S, CH(OH), C=C=0, C=NH, CH2, CHOH, CHOR, CNR, CH-R, and C-R3R4, wherein each R, R3, and R4, independently, is H or a C1-C6 branched or unbranched, substituted or unsubstituted, alkyl.

In certain peptides, the carboxy terminus is the only part of the molecule which can be attached to a fluorophore without disrupting the peptide's biological activity. In these cases, it is therfore necessary to add a separate "linker" group to the peptide. Since the N-hydroxysuccinimide esters (NHS) or isothiocyanate forms of fluorophores do not readily react with carboxylic groups or carboxyl amine groups, these groups must first be modified to a provide a functional site (e.g., a primary amino group) for conjugation with fluorophores. For example, fluorescent opioid peptides include linker groups to maintain their biological activity. In this case, an aminopentyl group is grafted onto the C-terminal amino acid by aminolysis of the opioid peptide with 1,5 diaminopentane as described in Example IC, below. Aminopentyl linker groups can also be added to a peptide when the peptide is incubated with carbodimides. Water soluble carbodimides are widely used for carboxyl-amine conjugation and may also serve to link fluorophores to the carboxy terminus of peptides.

Whether or not to include a linker group is usually determined empirically by testing a fluorescent peptide labeled at various amino acid sites and finding that it has lost biological activity. For some peptides, structure-activity studies show that the entire amino terminus and central portion of the peptide are involved in receptor binding. This suggests that only the carboxy terminus of the peptide can be modified without disrupting biological activity.

The chemical structure of the light-emitting moiety can affect the synthetic route used to synthesize the fluorescent peptide. It may be necessary, for example, to modify the light-emitting moiety so that it includes a reactive group prior to contact with the desired peptide.

Any peptide having an affinity for its corresponding receptor on a cell surface can be used to make a biologically active fluorescent peptide for cell sorting. Peptides can be synthesized using techniques known in the art, extracted from natural systems, or obtained from commercial sources (e.g., Peninsula, Neosystems, Sigma, and BASF) . Typically, the peptide is either purchased or synthesized using conventional solid-phase synthetic techniques.

Preferred peptides are included in the group consisting of adrenocorticotrophic hormone, amylin, an amyloid beta-fragment, an atrial natriuretic peptide, bombesin, bradykinin, cadherin, calcitonin, a calcitonin-gene-related peptide, a casopmorphin, a morphiceptin, cholecystokinin, corticotropin-releasing factor, deltorphin, a dermorphin, dynorphin, an endorphin, endothelin, enkephalin, fibronectin, galanin, a gonadotropin- associated peptide, a gonadotropin-releasing peptide, a growth factors or growth factor-related peptide, gastrin, glucagon, growth hormone-releasing factor, somatostatin, GTP-binding protein fragments, inhibin, insulin, interieukin, luteinizing hormone-releasing hormone, magainin, melanocyte-stimulating hormone, a morphiceptin, a neurokinin, a neuromedin, neuropeptide-Y, an opioid peptide, oxytocin, a pancreatic polypeptide, parathyroid hormones, vasoactive intestinal polypeptide, Peptide YY, substance P, thyroid-releasing hormone, a toxin, vasopressin, and fragments, derivatives, and analogues thereof.

Peptides useful in the invention include those whose sequences differ from the wild-type peptide sequence by only conservative amino acid substitutions. For example, one amino acid may be substituted for another with similar characteristics (e.g. valine for glycine, arginine for lysine) or by one or more non-conservative amino acid substitutions, deletions, or insertions which do not abolish the peptide's biological activity. Other useful modifications include those which increase the peptide's stability. For example, the peptide may contain one or more non-peptide bonds (which replace a corresponding peptide bond) or D-amino acids in the peptide sequence.

Other peptides which may be used include those described in the Peninsula Laboratories Inc. catalogue, 1992-1993; SIGMA-Peptides and Amino Acids catalogue, 1993-1994; and PIERCE, Catalog & Handbook, Life Science & Analytical Research Products, 1994.

The light-emitting moiety can be any moiety which emits an optical field (e.g., fluorescence or phosphorescence) following excitation. Preferably, the moiety is a standard fluorophore selected from the group consisting of fluorescein, FTC, Texas Red, phycoerythrin, rhodamine, carboxytetramethylrhodamine, DAPI, indopyra dyes, Cascade blue coumarins, NBD, Lucifer Yellow, propidium iodide, Bodipy, CY3, CY5 or derivatives thereof. For example, a standard fluorophore which is derivatized to include additional chemical bonds is a "light emitting moiety". Other light-emitting moieties commonly used in labelling or other applications may be attached to the compound in place of the above. Suitable light-emitting moieties are described, for example, in Molecular Probes, Handbook of Fluorescent Probes and

Research Chemicals, 1992-1994; and Richard P. Haugland et al., "Design and Application of Indicator Dyes", Noninvasive Techniques in Cell Biology: 1-20, Wiley-Liss Inc. , (1990) . Preferred light-emitting moieties possess at least one side group capable of reacting with amino acids to form chemical bonds. Such side groups include indoacetamide, maleimide, isothyocyanate, succinimidyl ester, sulfonyl halide, aldehyde, glyoxal and hydrazine derivatives. Amino acids including alanine, arginine, asparagine, aspartic acid, cysteine, glutamine, glutamic acid, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, proline, serine, threonine, tryptophan, tyrosine, and valine may be labeled in this fashion.

Figs. 4A and 4B show, respectively, fluorescent endothelin and fluorescent galanin made by the general method described above. Each fluorescent peptide features fluorescein (a light-emitting moiety) bound to an individual peptide at an amino acid position which preserves the peptide's biological activity. In each of these compounds, the C-X bond is an acyl moiety.

The following Examples are used to more particularly point out the synthesis of fluorescent peptides having substantial biological activity, and the use of these peptides in cell-sorting methods.

Examples Example 1 - Synthesis of Fluorescent Peptides IA - Fluorescent Galanin, and Endothelin Synthesis of fluorescent galanin and fluorescent endothelin is described in U.S.S.N. 08/504,856 entitled "Fluorescent Peptides", filed July 20, 1995, the contents of which are incorporated herein by reference. Briefly, galanin (25 μg; Peninsula Laboratories, Inc., San Carlos, CA) was dissolved (1:100) in a solution of 40 μl of 50mM bicarbonate buffer (NaH₂C0₃) , pH 9.3, to a final dilution of 0.8 mM. Endothelin (20 μg) from the same supplier was dissolved in 10 μl of the bicarbonate buffer to achieve the same final dilution. NHS-fluorescein (N-hydroxy-succinimidyl ester; Pierce Chemical Company, Rockford, Illinois) was then dissolved in 100 μl of DMSO. 50-100 μl of stock NHS-fluorescein (2.1 μmol) was then mixed with the peptide solution. The resulting solution was then placed on ice, incubated for one hour at pH 9.3, and then brought to pH 8 by the addition of 500mM Tris HCl. Incubation proceeded for the next 18 hours at 4°C.

Following incubation, unreacted fluorescein was removed using G-50 column chromatography (Pharmacia Biotech, Upsala, Sweden) . Biologically active and inactive compounds were eluted with O.l M phosphate buffered saline (PBS) at pH 7.4 by spinning the column in a table-top centrifuge at 3000 rpm for 10 minutes. The eluent was dissolved and incubated with known quantities of rat renal membrane homogenates for endothelin, and rat brain homogenates for galanin.

After allowing binding to occur over a 60-minute period, membranes were precipitated from solution by centrifugation at 3000 rpm for 5 minutes at 4°C. Membranes were then resuspended in PBS and incubated with a solution equivalent to 0.5 M NaCl and 0.2 M acetic acid at pH 3.1 to strip surface-bound fluorescent peptides from their corresponding receptorε. Using this method, the biologically active compounds were collected for amino acid analysis.

As shown in Figs. 4A and 4B, the sites of attachment of fluorescein to endothelin and galanin were confirmed to be respectively, the ninth amino acid (i.e., lysine) on the epsilon amino group and the fifth amino acid residue (lysine) on the epsilon amino group. The molar ratio of fluorescein to peptide was confirmed to be 1:1 in each case.

Competition binding with radiolabelled endothelin and galanin indicated that the peptides had maintained their biological activity and retained a high affinity for their respective receptors. The IC₅₀ for unlabelled galanin, as compared to fluorescent galanin, was 1.75 nM versus 8.86 nm. The K^ for this compound was 1.06 nM versus 5.36 nM. These results indicate a high degree of retention of biological activity. IB - Fluorescent Somatostatin

Synthesis of fluorescent somatostatin is described in U.S.S.N. 08/475,751 entitled "Fluorescent Somatostatin", filed June 7, 1995, which is a continuation-in-part of U.S.S.N. 08/416,007, having the same name and filed April 4, 1995, the contents of which are incorporated by reference.

To synthesize fluorescent somatostatin, 1 mmole NHS FTC (an isothiocyanate analog of fluorescein; Molecular Probes) was diluted in 200 ml of acetonitrile (Pierce) and then incubated with an somatostatin analog (0.6 mmole of D-Trp8-ssl-14, Neosyste s, Inc.) in a final volume of 1 ml Borate/Phosphate Buffer (50mM/50 mM) at pH 6.5 for 3 hours at 4°C. The solution was purified on an HPLC C18 column (10 mm X 250 mm, Ultrosphere, ODS, Beckmann Instruments) and eluted in 0.1% TFA with a linear gradient of acetonitrile from 20% to 70% during 100 minutes at a debit rate of 1 ml/min. Elution of the compound was monitored by observing optical density profiles at 213 nm.

Collected fractions were screened by analytical HPLC using both UV and fluorescence detection, with the excitation wavelength being 338 nm, and the emission wavelength at 425 nm. The collected fractions were then lyophilized.

Once synthesized, the emission peaks from fluorescent somatostatin were monitored. The fluorescent peptide was concurrently subjected to Edmann's degradation to determine whether the α-amino function present on the amino terminal amino acid was free or blocked. The first and largest peak was found to be unchanged when the compound was exposed to phenylisothiocyanate. This peak therefore corresponded to FTC-[Alal]-D-Trp8-ss. The amino acid composition of this fragment was assessed by quantitative amino acid analysis after acidic hydrolysis in vacuo (6N HCl, 110°C, 18h) and carboxypeptidase Y digestion (6U/0.3mmole, 37°C, 48h) .

The site of attachment between the NHS fluorescein and the somatostatin analogue was confirmed to be at the N-terminus, and was identified as Nα-Alal. Isolation of this peak yielded a compound with a molar ratio of NHS fluorescein to D-Trp8-somatostatin of 1:1. FTC-[Alal]-D- Trp8-somatostatin. The fluorescent peptide was evaluated to be pure (as indicated by a single elution peak from reverse phase HPLC) , was freely soluble in water or aqueous buffer, and was stable if protected from light and maintained in a lyophilized form at 4°C. IC - Fluorescent Qpioids As described above, fluorescent opioid peptides can include linker groups to maintain their biological activity. In a particular example, dermorphin, a mu opioid receptor agonist, was originally isolated from the skin of a frog (Phylo edusa sauvagei) . An analogue of this peptide, [Lys7]dermorphin (NH₂-Tyr-Dala-Phe-Gly-Tyr- Pro-Lys-C NH₂) , has a very high affinity for the mu opioid receptor, and has a high specificity for the mu receptor when compared to the delta opioid receptors (Negri et al., 1992. Proc. Natl. Acad. Sci. USA 89:7203). Another peptide, deltorphin 1 (NH₂-Tyr-Dala-Phe-Asp-Val- Val-Gly-C NH₂) , also isolated from the skin of another frog (Phylomedusa bicolor) is an agonist with high affinity and selectivity for delta opioid receptors. Structure-activity studies of opioid peptides suggest that the amino terminus of the peptide is necessary to maintain receptor binding and for conferring overall opioid specificity.

The deltorphin I and [Lys7]dermorphin peptides were modified by adding an linker molecule; this process enabled the coupling of a fluorophore to a previously unreactive or less reactive carboxy terminus amino acid site. The C-terminal amide function (-CONH₂) was substituted by an aliphatic chain ending with a primary aminopentyl group (-CONH-(CH₂)₅-NH₂) . The corresponding peptides were then labeled, respectively, on their single amino group with fluorophores.

1C.1 Preparation of Opioid Precursors

The modified opioid peptides described below were prepared by solid-phase synthesis on a standard Merrifield resin including reticulated polystyrene with 1% divinyl benzene. The aminopentyl group was grafted on the C-terminal amino acid by aminolysis of the peptide resin with 1.5 diaminopentane as described by Goldstein et al., Proc. Natl. Acad. Sci. 85:7375 (1988), the contents of which are incorporated herein by reference. First, the amino acids which are side-chain protected and activated on their carboxylate groups are attached to the resin in a sequential fashion starting from the C- terminus and ending on the N-terminus. The bond between the resin and the peptide is an ester linkage formed from an OH group of the resin and the carboxyl group of the last amino acid.

Using acid hydrolysis, the ester bond between the peptide and resin is then broken in order to introduce the aminopentyl linker group. This process liberates a peptide with a free COOH at the C-terminus. Acid hydrolysis includes incubating the peptide with 10% (v/v) 1.5 diaminopentane in methanol for 60-72 hours at 25°C. This cleaves the ester linkage and replaces it with an amide bond including the C=0 moeity of the peptide and one of the amino groups from the 1.5 diaminopentane (either group is acceptable since the molecule is symmetrical) . The amino acid side chains of the peptide are then deprotected by treating the peptide with liquid hydrogen fluoride. Ten milliliters of hydrogen fluoride per gram of peptide was used at 0°C for 45 minutes in the presence of p-cresol as scavenger. After evaporation of the hydrogen fluoride, the crude reaction mixture is precipitated with diethylether, washed, dried, dissolved in AcOH and lyophilized. The product was then purified by HPLC using a silica C₁₈ column (5-25um, 100A) 10 nm x 250 nm (Althosphere, ODS, Beckman) eluted with 0.1% TFA with a linear gradient of 25% to 70% CH₂CN/A for 30 minutes at a flow rate of 40 ml/min, and optically detected at 210 n using ultraviolet absorption spectroscopy. All fractions containing the target compound were individually analyzed by HPLC. Fractions demonstrating sufficient purity were polled and lyophilized. The final products were identified and monitored for purity and identity using analytical HPLC and electron-spin mass spectral analysis. After filtration of the resin, the protected peptide is dissolved in DMF and concentrated with a rotary evaporator. The product is extracted in DCM, with the excess diaminopentane eliminated with KHSO^ (0.01M) washings. After evaporation, the peptide is precipitated with diethylether and dried. 1C.2 Preparation of Fluorescent Opioids

Opioid precursors described above were reacted with the N-hydroxysuccinimide (NHS) esters of the following fluorophores: Fluorescein, BODIPY 503/512, and BODIPY 576/589 (Molecular Probes, Eugene Oregon) . These three reagents were mixed in different solutions (2 μmoles in 400 μl of dimethyl sulfoxide) and were individually incubated with the deltorphins DLT-1 5APA and [K7] DRM 5APA (2 μmoles) in a final volume of 1 ml of Borate/Phosphate buffer (pH 8.5) for 3 hours at 4°C. The different derivatives were purified by HPLC on a C₁₈ column (10 mm X 250 mm, Ultrosphere, ODS, Beckman) eluted in TFA (0.1%) with a linear gradient from 20% to 60% over an 80-minute time period. Several fluorescent peaks were identified and tested for their ability to displace the specific binding of iodinated analogues of DLT-1 5APA and [K7]DRM 5APA to the delta and mu opioid receptors, respectively.

The peptides were purified and then subjected to mass spectroscopy for the determination of the molar ratio of the fluorophore-to-peptide labeling. These studies indicated that all peptides were labeled with a single fluorophore. The sequence of each peptide, as well as the position of the fluorophore along the amino acid chain, were determined by Edman's degradation.

Peptide sequencing indicated that none of the peptides were labeled on the α-amino group. In addition, no lysine residue could be detected in position 7 for peptide peaks 3 and 5, indicating that these peptides were labeled by a fluorescent group on the lysine side chain. All other peptides were labeled on the amino group of the carboxy-terminus aminopentyl linker.

The biological activity of the peptides was examined by measuring their ability to displace specific binding of iodinated DLT-I 5APA and [K7]DRM 5APA to rat mu and delta opioid receptors expressed by transient transfection in COS cells. Iodinated analogues of DLT-1 5APA and [K7]DRM 5APA (50 000-200 000 cpm) were incubated with delta and mu opioid receptors (5-20 μg) respectively during 30 minutes at 25°C in 0.25 ml of 0.2% BSA and 50 mM Tris-HCl (pH 7.5). The incubation medium was then diluted in 3 mis of ice-cold incubation buffer and filtered under reduced pressure on GF/C filters presoaked in 0.3% PE1-HC1 (pH 7.5). Filters were rinsed by 2X3 ml of ice-cold incubation buffer and counted in a gamma counter. The results of these pharmacological binding experiments are summarized in Table 1.

Fluorescent analogues were also tested for their ability to selectively label mu and delta opioids receptors on COS cells by fluorescent microscopy. Cells in suspension were incubated for 30 minutes in Earle's buffer containing 0.2% BSA and 10 nM BODIPY 513/512 DLT-1 5APA. Cells were spun down, air dried, and then examined by confocal microscopy. Labeling with the fluorescent derivative appeared as punctate labeling of the cell surface membrane and cytoplasm.

Table l — Binding Properties of Fluorescently Labelled Opiod Peptides

Fluorescent Peptide Molecular IC50 IC50 Weight (nM) (nM) delta mu

DLT-I (Deltorphin) 2.3 1200 ω-BODIPY 503/512 DLT-I 5APA 1084 2 1400 ω-BODIPY 576/589 DLT-I 5APA 1121 2 450 ω-FL DLT-J 5APA 1168 26 >1000

[Lys7]DRM (Dermorphin) 0.9 740 ω-BODIPY 503/512 [K7]DRM 5APA 1177 10 0.6 ω-BODIPY 576/589 [k7]DRM 5APA 1214 21 1 ω-BODIPY 503/512 [R7]DRM 5APA 1205 7.2 0.6 ω-FL [K7]DRM 5APA 1261 116 4.8 ω-FL [R7]DRM 5APA 1289 56 3.6 e-FL [K7]DRM 5APA 1261 436 11 e-BODIPY 503/512 [K7]DRM 5APA 1177 2.6 0.7 e-BODIPY 576/589 [K7]DRM 5APA 1214 14 1.8 Example 2 - Sorting of rat mesencephalic neurons using fluorescent neurotensin

Synthesis of fluorescent neurotensin is described in the patent application entitled "Marker for Fluorescent Neurotensin" U.S.S.N. 08/402,777, filed

March 9, 1995, the contents of which are incorporated herein by reference.

2A. Cell Dissociation

Ventral mesencephalic tissue was obtained from 14-day-old embryos from anesthetized Sprague-Dawley rats. The embryos were removed by cesarian section and placed in cold (4°C) Hank's balanced salt solution (HBSS; Gibco BRL, Burlington, Ontario, Canada) . The ventral mesencephalic tissue was dissected under aseptic conditions, followed by incubation of the tissue pieces in 0.1% trypsin (Gibco BRL) at 37°C for 20 min. After rinsing five times in HBSS containing 10% fetal calf serum for 5 minutes, the tissue was dissociated by titration through the tip of a fire-polished Pasteur pipette into a single-cell suspension. The resulting cell suspension was centrifuged at 300g for 3 minutes. The pellet was resuspended in warm DMEM/F12 containing vitamin B27 supplements, 2 mM glutamine, 500U/I fungizone, penicillin, and streptomycin, and passed through a 62-μm Nitex filter (Becton Dickinson; Mountain View, CA) .

Cells were counted on a hemocytometer and then diluted. The cells were then plated at a density of 3 x IO⁵ cells/cm² on 22-mm diameter poly-L-lysine-coated glass coverslips and/or 100-mm plastic petri dishes. Cells were then grown in a humidified atmosphere of 90% air and 5% C0₂ at 37°C.

Following a period of neuronal growth and differentiation in culture, mesencephalic cells were dissociated mechanically from the petri dish or glass coverslip and centrifuged for 6 mins. at 900 rpm. The resulting pellet was then resuspended in Earle's buffer.

The effect of temperature on the intensity of fluorescent peptide binding (expressed as mean fluorescent intensity, MFI) was then examined. Fluorescence from the samples was analyzed on a Becton-Dickinson Facscan flow cytometer (Mountain View, California) equipped with an argon ion laser operating at 200 mW and 488 nm. Fluorescence from the light-emitting moiety (FITC) of the fluorescent peptide in the green spectral region was collected through a DF 530/30 band-pass filter. Red fluorescence from peptides attached to DAPI was collected through a 585/42 nm filter. In most cases, 10,000 events/sample were collected, stored, and analyzed by a Consort-30 computer program. Debris and dead cells were excluded from the analysis by conventional scatter-gating methods.

The results of this study are shown in the graph of Fig. 5, which plots MFI as a function of time. Data derived from individual cell suspension profiles are expressed as the MFI for each cell suspension. Specific binding was calculated by subtracting the MFIs of the non-specific-binding samples (i.e., incubated with 100-fold excess unlabeled neurotensin) from those of the specific binding samples (i.e, no excess unlabeled neurotensin) . The MFI was determined by integrating the fluorescence intensity over the time required for a particular cell to pass through a single laser beam of the flow cytometer. A minimum of 10,000 cells were used for this calculation. These values are themselves the mean of ten independent preparations.

As is clear from Fig. 5, incubation of rat mesencephalic neurons at 22°C resulted in peak fluorescent neurotensin binding intensities at 15 minutes. After this time period, the fluorescence decreased steadily until about 80 minutes, where it reached initial values (i.e., the MFI at time 0) of fluorescence intensities. At 4°C, fluorescent neurotensin binding peaked at about 20 minutes and remained close to this level for an additional 60 minutes before falling to initial levels after 120 mins.

2B. Viability of Dissociated Cells

As indicated in Fig. 6, cell viability of mesencephalic neurons was assessed before, during, and after cell sorting using propidium iodide (PI) labeling as an indicator of cell death. PI is a well-known DNA stain which is impermeable to viable cell membranes but which penetrates the "leaky" cell membranes of dead and dying cells. During the experiment, a stock solution of PI (3.2 mM, 20 μl) was added to each sample with rapid mixing to give a final concentration of 32 μM immediately before introduction into the flow cytometer. Specific neurotensin binding was also examined in these same cells in order to determine the time frame of maximal fluorescent neurotensin binding. Fig. 6 indicates how mesencephalic neurons dissociated from tissue cultures and maintained at 37°C for 1 hour exhibit a loss of about 25% of all neurons. Another 10% of the total neurons were lost upon changing the temperature to 4°C, although cell survival was maintained at a fairly steady level of 65% for the ensuing hour. Specific fluorescent neurotensin binding (expressed as MFI) rose rapidly in the first 20 minutes of incubation with fluorescent neurotensin and thereafter remained high for the next hour.

These experiments demonstrate that there is a period of approximately one hour where cells are maximally labeled with fluorescent neurotensin and maintain good cell viability (this time period is indicated by the hatched lines in the graph) . These experiments also demonstrate that a recovery period of 60-90 minutes is necessary after dissociation of the cells from the petri dish. The recovery period is required to obtain maximum specific fluorescent peptide binding, presumably because it helps to reestablish the cell membrane integrity or reexpression of receptors on the cell surface. 2C. Controls for specific high-affinity neurotensin binding

Several controls were performed to ensure that the fluorescently labeled cells measured by flow cytometry indeed represented specific fluorescent neurotensin binding. Chinese Hamster Ovary (CHO) cells transfected with the cloned neurotensin receptor were initially used for the positive control. Prior to transfection, the cells were grown in alpha essential medium (MEM) without nudeoside or deoxyribonucleoside (Gibco-BRL, Quebec, Canada) . The MEM was supplemented with 10% decomplemented fetal calf serum, 2mM glutamine, and 80 μg/ml of gentamicin. Cells were grown under a humidified atmosphere of 90% air and 5% C02 at 37°C. The CHO cells were then transfected with NTR cDNA plasmid CDM8 and neomycin-resistant plasmid pRSVneo using a standard calcium phosphate precipitation technique. The transfectants were isolated in 500 μg/ml of G418.

Cultures of rat cortical astroglial cells were also used in the control experiments. These cells were obtained from cerebral hemispheres of newborn rats (1-2 days old) . Briefly, the meninges were removed in aseptic conditions and the hemispheres passed through a sterile nylon sieve (82-μm pore size) immersed in a nutrient-containing medium. The basal nutrient medium consisted of Dulbecco's Modified Eagle's Medium (GIBCO, Grand Island, NY) containing 10% fetal calf serum (GIBCO, Grand Island, NY) , 2 mM glutamine, 5 ml penicillin, and 5 μm streptomycin. Cells were seeded into 35-mm plastic petri dishes (Corning, New York, NY) or 6-well plates (Costar, Cambridge, MA) at a plating density of 0.5-1.0 x io⁵ cells/cm². The cultures were incubated at 37°C in a humidified 5% C02/95% 0₂ atmosphere. The culture medium was changed after 8 days and then every 4 days for 3 months.

Figs. 7A-7F show the results of the control experiments. First, unlabeled cells were used to measure the level of cell auto-fluorescence. Cellular auto¬ fluorescence yielded a weak fluorescence peak (not shown in the figure) . The highest value of thiε peak was, in all caseε, used to normalize the zero value on the fluorescent intenεity scale (x axis) of the flow cytometry hiεtogramε.

For positive controls, flow cytometry was used to analyze SN17 cells (a cell line known to contain neurotensin receptorε) . Measurements were taken from cells incubated with fluorescent neurotensin in the presence (Fig. 7A) and absence (Fig. 7D) of 100-fold excess unlabeled neurotensin. The SN17 cells were found to express relatively low levels of fluorescent neurotensin-labeled receptors. This iε indicated by the slight shift to the right of fluorescent neurotensin- labeled cells in conditions of specific binding (Fig. 7D) as compared to non-specific binding (Fig. 7A) .

CHO cell lines transfected with the cloned neurotensin receptor were incubated with fluorescent neurotensin in the presence and absence of excess unlabeled neurotensin (Figs. 7B and 7E, respectively) . Specific binding in these cells is indicated by the shift to the right of the fluorescence peak in Fig. 7E (as compared to Fig. 7B) . From the graph it was determined that specific binding was restricted to approximately 60% of the entire cell population. This value represents a fraction equal to that representing the fraction of cells successfully transfected with the neurotensin receptor. As a negative control, CHO cells which were not transfected with the neurotensin receptor were examined under the same conditions (Figs. 7C and 7F, respectively) . As indicated by the peaks in these figures, virtually no specific fluorescent neurotensin binding was observed. A small subset of cells nonetheless seemed to specifically bind fluorescent neurotensin perhaps due to endogenous neurotensin receptors in the CHO cellε.

Glial cells, which are known to possess only low-affinity neurotensin receptors, were also examined in the presence and absence of excess unlabeled neurotensin. As shown in Figs. 8A and 8C, in the abεence of levocabaεtine (an antagoniεt to the low-affinity neurotensin receptor) moderate specific fluorescent neurotensin binding was detected in a subpopulation of glial cells. Figε. 8B and 8D show that a complete absence of specific high affinity neurotensin binding was observed in the preεence of levocabaεtine.

In another serieε of controls, COS-7 cells transfected to expresε the somatostatin and neurotensin receptor were examined. As shown in the figureε, only the COS-7 cells transfected with the neurotenεin receptor (Figs. 9C and 9F) displayed significant specific fluorescent neurotensin binding. Cells which were not transfected (indicated by the data in Figs. 9A and 9D) or transfected to expresε the εomatostatin receptor (Figs. 9B and 9E) did not display significant specific binding of fluorescent neurotensin. 2D. Cell sorting using flow cytometry

Fluorescence-associated cell sorting (FACS) was used to isolate the neurotensin receptor-bearing cells described above. A mesencephalic cell culture consisting of 85% dopamine neurons was generated by labelling a cell population with fluorescent neurotensin, followed by purification with FACS. Before each sort, the cell-sorting apparatus

(Becton Dickinson, FacScan flow cytometer, Mountain View, California) was aligned with glutaraldehyde-fixed chicken erythrocytes and calibrated with fluorescein-labeled microspheres (QuickCal Quantum, Becton Dickinson, Missisauga, Ontario, Canada) . A standard 70-μm nozzle was used. The sorting was performed at a rate of 1000 events/sec to reduce aborted and missorted events. The charging pulse εhape waε set to deflect 3 dropletε/event. Raw data for each εorting experiment were collected in list mode on a Macintosh computer FACSstation (Becton Dickinεon Canada Inc., Miεεisauga, Ontario). The data were presented as single-parameter frequency histograms. Data included histogram statiεticε such as the total number of events (cellε) measured, the proportion of cells gated as negative (Ml) , the proportion of cells gated as positive (M2) , the mean, median and coefficient of variation for the profile within each gate. These histograms were used to compare various εa ples and correlate sorting parameters. Gates (i.e., Ml and M2) for sorting were created by comparing relative fluorescent intensity (RFI) of controls for non-specific binding (i.e in the presence of 100-1000-fold excess unlabeled neurotensin) to fluorescence of samples treated with fluorescent neurotenεin alone. Only thoεe eventε from specific binding samples that exhibited fluorescent intensities greater than those from the non-specific samples were sorted as positive. Mean fluorescence intensities (MFI) measured for the population of non-specifically labeled cells were excluded from the analysiε of εpecifically labeled cells. The specific binding of fluorescent neurotensin thus served as the critical marker for establishing appropriate gates for the sorting of dopamine neurons. 2E. Conditions for mesencephalic dopamine cell sorting As shown in Figs. IOA and 10D, after 0 days no significant difference between non-specific and specific fluorescent neurotensin binding was observed in embryonic mesencephalic cells incubated with fluorescent neurotensin. These data indicate an absence of neurotensin receptors on the mesencephalic cells immediately following tisεue dissociation. Figs. 10B, IOC, 10E, and 10F show flow cytometric profiles of non¬ specific and specific binding of fluorescent neurotensin to mesecenphalic neurons after 7 days in the absence

(Figs. 10B and 10E) and presence (Figs. IOC and 10F) of levocabastine. Specific neurotensin binding was observed in mesencephalic cells cultured for 7 days in the absence of levocabastine. In this case, a population of neurons exhibiting specific neurotensin binding resulted in a single fluorescence peak (Fig. 10E) . This peak partially overlapped with the fluorescent intenεities observed in populations of cells exhibiting non-specific fluorescent neurotensin binding. this single profile of specific neurotensin binding became bimodal when levocabastine was included with the incubation medium (Fig. 10F) .

Since low-affinity neurotensin receptors are known to be present on glial cells, only cells correlated to the strongest fluorescent peak were eventually sorted. Sorted cells were collected in 1.5 ml sterile Eppendorf tubes (tubes #1-3) containing culture media. The percentage of total events sorted positively was determined for each sample and varied from 10-20% depending on the quality of the initial dissection of the mesencephalon. As a representative example, asεay #10 reportε the sorting of 60,000 cells out of a total of 265,000 for tube #1, 60,000 positive cells sorted out of 310,000 cells for tube #2, and 44,000 cellε out of 250,000 cells for tube #3. This yielded an average of 20% of the cells being sorted.

2F. Cell Culture of Sorted Cells

Positive fractions of selected samples were concentrated by centrifugation at 3OOg for 5 minutes and resuspended in plating medium. Cells were plated at a minimal density of 400 cells/mm² on glass coverslipε. After 4 days, the plating medium was replaced with a maintenance medium containing the εame ingredientε, and the cultureε maintained at 37°C in a humidified atmoεphere of 90% air and 5% C02. These cells were maintained for at least 12 days in culture.

Cells were proceεεed for tyrosine hydroxylase (TH) and glial fibrillary acidic protein (GFAP) immunocytochemistry to identify dopamine neurons and glial cellε both immediately before and after cell sorting. Theεe immunocytochemical experiments were done using well-known techniques for cells in εuεpenεion and cellε in culture. Briefly, after waεhing cover slips twice with HBSS, the cells were rinsed twice in phosphate buffered saline (PBS) and incubated in PBS containing 5% normal goat serum (NGS) and 0.1% Triton-X (Sigma,

St-Louis, MO) at room temperature for 30 minutes. The cells were then incubated in the same buffer with an anti-TH polyclonal antibody (Incstar, Stillwater, MN) at a dilution of 1/750 overnight at 4°C, then washed again and incubated with Texas Red-conjugated secondary antibodies (Bio/Can, Missisaugua, Ontario, Canada) dissolved at 1/200 for 1 hour at room temperature. Coverslips were then mounted in Aquamount (Polysciences, Warrington, PA, USA) and examined on a Leica epifluorescence photomicroscope operating with a high-pressure mercury lamp and the appropriate filter combinations for excitation and emission of Texas Red (596/630) . The purity of the sorted cell population was determined using confocal microscopy. Cells were examined under a Leica confocal laser scanning microscope (CLSM) configured with a Leica Diaplan inverted microscope. The device was equipped with an argon ion laser (488 nm) having an output power of 2-50 mW. All image generating and procesεing operations were performed with the Leica CLSM software package. Micrographs were taken from the image monitor using a Focus Imagecorder (Foster City, California) . Images of embryonic cells were acquired as a single 0.42-μm thick optical section and averaged over 32 scans/frame. Double fluorescence images were acquired in two passes (fluorescein first and then Texas Red) to avoid bleeding from one channel to the other. The results of these experiments confirmed that

84.8% of all cellε counted immediately following εorting were dopaminergic and 81.5% of cellε were dopaminergic after 5 dayε in culture.

Example 3 - Sorting of cortical cell cultureε uεing fluoreεcent εomatoεtatin

Cortical tissue was obtained from 18-day-old

Sprague-Dawley rat embryos and used as a source of cells including the somatoεtatin receptor. The embryos were removed by cesarian section and placed in cold (4°C) HBSS (Gibco BRL, Burlington, Ontario, Canada) . Cortical tissue was disεected under aseptic conditions and then incubated in 0.1% trypsin (Gibco BRL) at 37° for 15 minutes. The tissue was rinsed five times in HBSS containing 10% fetal calf serum for 5 minutes and then dissociated by titration through the tip of a fire- polished Pasteur pipette into a single-cell suspension. The resulting cell suεpenεion waε paεεed through a 62-μm Nitex filter (Becton Dickinεon, Mountain View, CA) . The cells were counted on a hemocytometer and were diluted in warm DMEM/F12 containing 2 mM glutamine, 500U/I fungizone, penicillin, and streptomycin before plating at density of 3 X IO⁵ cells/100 mm plastic petri dishes. Cells were then grown in a humidified atmosphere of 90% air and 5% C0₂ at 37°C.

After 7 days in culture, the culture medium was removed and the cells waεhed twice with HBSS, dissociated mechanically from the petri dish, and centrifuged for 6 minutes at 900 rpm. The pellet waε resuspended in a solution of HBSS and 5 nM fluorescent εomatoεtatin at 4°C. A εignificant difference between εpecific and non- εpecific binding of the fluoreεcent peptide waε observed in the cultured cells. Thiε indicated the preεence of somatostatin receptors on these cells, a result that has been previously reported (Gonzales et al., 1989, Neuroscience 29:629-644) . As indicated by Figs. IIA and 11B, a single peak of non-εpecific binding of fluorescent somatostatin to cortical cells is significantly shifted to the right under conditions of specific binding. In addition, the unimodal curve present in non-specific conditions becomes bimodal under specific conditions, with a significant proportion (45%) of the population exhibiting strong fluorescent labeling with fluorescent somatostatin. Dead cells and debris were excluded by prior labeling with propidium iodide, and are not included in this flow cytometry profile.

The cells labeled selectively and specifically with fluorescent somatostatin (defined by gate Ml in the figures) represent a population of εomatostatin-receptor- positive cortical neurons. These cellε can be sorted using the same parameters described above for sorting of mesencephalic neurons following labeling with fluorescent neurotensin.

Other Embodiments Other embodiments are within the scope of the following claimε. In general, a biologically active fluorescent peptide containing peptide and light-emitting moieties and synthesized as described above can be used in the cell-sorting method of the invention. Likewise, the method can be used with any types of cells, including neurons and pancreatic, ovarian, liver, testicular, kidney, lung, adrenal, and intestinal cells.

The method for synthesizing biologically active fluorescent peptideε (e.g., εomatostatin and mu and delta opioid peptides) can differ from the method described above and shown in Fig. 3. For example, the mixture of labelled biologically active and inactive fluorescent peptides can be separated by column chromatography (e.g., HPLC) . Various HPLC peaks corresponding to the fluorescent peptides can then be observed and isolated. Each peak is tested in a pharmacological binding assay to determine if the peak is displaced by peak corresponding to a radiolabelled peptide (i.e., IC₅₀ determination). Compounds posεeεsing an IC₅₀ close to 1 indicate good retention of biological activity. Cells can also be sorted using techniques other than flow cytometry. For example, magnetic beads coated with anti-fluorophore (fluorescein) antibodies can be used to purify the mixture of labeled cells. In this case, the antibodies on the beads indirectly attach to receptors on the surface of the desired cells by binding to the light-emitting moiety of the fluorescent peptides. The cell- containing solution is then placed near magnets which retain the magnetic beads. The desired labeled cells are isolated by washing away the unwanted cells and then removing the sample from the influence of the magnet.

Negative selection of unlabeled cells is also possible using this method. For example, if the cells desired for purification do not possess peptide receptors which can be labeled with fluorescent peptides, then it is the washed-away, unlabeled cells that are collected. Negative selection can also be used with flow cytometry, i.e., cells which do not fluoresce are selected. Other methods for sorting cells using the fluorescent peptides may include panning or affinity chromatography with anti-fluorophore antibodies.

Still other embodiments are within the scope of the following claims.

Claims

What is claimed is:

1. A method for sorting a first group of cells from a cell population which contains said first group of cells, as well as other, different cells, wherein the cells of said first group bear a receptor which the other cells in the population lack, said method comprising: labelling the first group of cells by contacting the cell population with a biologically active fluorescent peptide of the formula:

X R_χ C R₂

wherein R_j is a light-emitting moiety; R₂ iβ a peptide of between 2 and 200 amino acids which is not neurotensin and which specifically binds to said receptor on said first group of cells but not to receptors or the other cells in the population; and,

X is selected from the group consisting of =0, =S, -OH, *=C=0, =NH, -H, -OR, -NR, -R, and -R₃R₄, wherein each R, R₃, and R , independently, is H or a C1-C6 branched or unbranched, substituted or unsubstituted, alkyl; and sorting either the labelled first group of cells or unlabelled other cells of the cell population.

2. The method of claim 1, wherein said sorting comprises: irradiating the cell population to induce fluorescence in the light-emitting moieties attached to the first group of cells by way of the peptides; and, sorting the cells attached to fluorescing, light- emitting moieties from the cell population.

3. The method of claim 1, wherein said sorting compriseε sorting the first group of cellε uεing flow cytometry.

4. The method of claim 1, wherein the biologically active fluorescent peptide is attached to a electrically or magnetically active material, and said sorting compriseε: exposing the electrically or magnetically active material attached to the first group of cells to an electric or magnetic field; and, sorting the cells attached to materials which exhibit a magnetic or electric response.

5. The method of claim 2, wherein a time period of between 5 and 120 minutes elapses between said labelling and said irradiating.

6. The method of claim 1, wherein R₂ comprises a peptide selected from the group consisting of adrenocorticotrophic hormone, amylin, an amyloid beta- fragment, an atrial natriuretic peptide, bombesin, bradykinin, cadherin, calcitonin, a casopmorphin, a morphiceptin, cholecystokinin, corticotropin-releasing factor, a dermorphin, dynorphin, an endorphin, endothelin, enkephalin, fibronectin, galanin, a gonadotropin-associated peptide, a gonadotropin-releasing peptides, a growth factor or growth factor-related peptide, gastrin, glucagon, growth hormone-releasing factor, somatostatin, a GTP-binding protein fragment, inhibin, insulin, interieukin, luteinizing hormone- releasing hormone, againin, melanocyte-stimulating hormone, a morphiceptin, a neurokinin, a neuromedin, neuropeptide-Y, an opioid peptide, oxytocin, pancreatic polypeptides, a parathyroid hormone, vasoactive inteεtinal polypeptide, Peptide YY, substance P, thyroid- releasing hormone, a toxin, vasopressin, and fragments, derivatives, and analogues thereof.

7. The method of claim 1, wherein R₂ is selected from the group consisting of endothelin, galanin, and somatostatin.

8. The method of claim 1, wherein C is bonded to R₂ through an amino residue of an alpha carbon atom.

9. The method of claim 1, wherein R_χ is bound, through C, to a region of said R₂ peptide which is not involved in said biological activity.

10. The method of claim 1, wherein said R₂ peptide binds to a human receptor.

11. The method of claim 7, wherein R₂ is endothelin and R_χ is bonded, through C, to the ninth amino acid residue of R₂.

12. The method of claim 11, wherein R--_^ is bonded, through C, to the e-amine group of Lys.

13. The method of claim 7, wherein R₂ is galanin and R-^ is bonded, through C, to the fifth amino acid residue of R₂.

14. The method of claim 11, wherein R is bonded, through C, to the e-amine group of Lys.

15. The method of claim 7, wherein R₂ is somatostatin or an analog thereof, and R_λ is bonded, through C, to a N-terminus -Ala- residue of R₂.

16. The method of claim 1, wherein R_χ is selected from the group consisting of fluorescein, FTC, Texas Red, phycoerythrin, rhodamine, carboxytetramethylrhodamine, DAPI, indopyra dyeε, Cascade blue coumarin, NBD, Lucifer Yellow, propidium iodide, a porphyrin, Bodipy, CY3, CY5 and derivatives and analogues thereof.

17. The method of claim 10, wherein R is attached to C-X through a linking moiety selected from the group consisting of indoacetamide, maleimide, isothyocyanate, succinimidyl ester, sulfonyl halide, aldehyde, glyoxal, hydrazine, and derivatives thereof.

18. A substantially pure population of cells sorted by the method of claim 1.

19. A substantially pure population of dopamine cells sorted by the method of claim 1.

20. A method for sorting a first group of cells from a cell population, which contains said first group of cells, as well as other different cells wherein the cells of said first group bear a receptor which the other cellε in the population lack, said method comprising: labelling the firεt group of cellε by contacting the cell population with a biologically active fluorescent peptide of the formula:

X R-_L C R₇

wherein R is a light-emitting moiety;

R₇ is a peptide of between 2 and 200 amino acids bound to C by a binding moiety selected from the group consisting of the residueε Ala, Arg, Asn, Asp, Cys, Gin, Gly, Hiε, Ile, Leu, Lyε, Met, Phe, Pro, Ser, Thr, Trp, Tyr, and Val; and,

X is selected from the group consisting of =0, =S, -OH, =C=0, =NH, -H, -OR, -NR, -R, -R₃R , wherein each R, R₃ and R₄, independently, is H or a C1-C6 branched or unbranched, subεtituted or unεubεtituted, alkyl; and sorting either the labelled first group of cells or unlabelled other cells of the cell population.

21. A method for sorting a first group of cells from a cell population, which contains said first group of cells, as well as other different cells wherein the cells of said firεt group bear a receptor which the other cells in the population lack, said method comprising: labelling the first group of cells by contacting the cell population with a biologically active fluorescent peptide of the formula:

X

wherein R_τ is a light-emitting moiety; R₂ is a peptide of between 2 and 200 amino acids;

X is selected from the group consiεting of =0, -OH, =C=0, =NH, -H, -OR, -NR, -R, -R₃R₄, wherein each R, R₄ and R₃, independently, iε H or a C1-C6 branched or unbranched, substituted or unsubstituted, alkyl; and, sorting either the first labelled group of cells or unlabelled other cells of the cell population.

22. A method for sorting a first group of cells from a cell population, which contains said first group of cells, as well as other different cells wherein the cells of said first group bear a receptor which the other cells in the population lack, said method comprising: labelling the first group of cells by contacting the cell population with a biologically active fluorescent peptide of the formula:

X xs.-] *^

wherein iε a light-emitting moiety; R₂ is neurotensin or an analogue thereof; and,

X is selected from the group consisting of =0, =S, -OH, =C=0, =NH, -H, -OR, -NR, -R, and -R₃R₄, wherein each R, R₃, and R₄, independently, is H or a C1-C6 branched or unbranched, substituted or unsubstituted, alkyl; and, sorting either the labelled first group of cells or unlabelled other cells of the cell population.

23. The method of claim 22, wherein R-^ is bound, through C, to a region of said R₂ peptide which is not involved in said biological activity.

24. The method of claim 23, wherein R is bound, through C, to an Arg-Pro-Trp moiety of R₂.

25. The method of claim 22, wherein R_j is bound, through C, to an Arg-Pro-Tyr moiety of R₂.

26. A method for sorting a first group of cells from a cell population, which contains said first group of cells, as well as other different cells wherein the cells of said first group bear a receptor which the other cells in the population lack, said method comprising: labelling the first group of cells by contacting the cell population with a biologically active fluorescent peptide of the formula:

X

wherein R_λ is a light-emitting moiety; R₂ is a peptide of between 2 and 200 amino acids which is not neurotenεin;

R₇ is a C1-C6 branched or unbranched, substituted or unsubstituted, alkyl; and,

X is selected from the group conεiεting of =0, =S, -OH, =C=0, =NH, -H, -OR, -NR, -R, and -R₃R₄, wherein each R, R₃, and R₄, independently, is H or a C1-C6 branched or unbranched, subεtituted or unεubεtituted, alkyl; and sorting either the labelled firεt group of cells or unlabelled other cells of the cell population.

27. The method of claim 26, wherein R₂ is an opioid peptide and R₇ is -NH(CH₂)₅NH-.