JP2001514375A - Drug discovery support using multiple membrane pseudo-affinity - Google Patents

Abstract

  (57) [Summary] Use of the methods and compositions useful in determining multiple membrane affinities of test compounds, obtaining data indicative of the affinities, and using the data alone or in combination with other molecular descriptors leading to predictive biological activity A method and apparatus are disclosed. A value indicative of the biologically relevant interaction of the test compound with the membrane mimetic is compared to the corresponding value of one or more control compounds having known biological activities. The expected biological activity of the test compound is identified using a control compound. The multi-membrane interaction value of the control compound is the one most closely correlated with the test compound.

Description

DETAILED DESCRIPTION OF THE INVENTION

FIELD OF THE INVENTION [0001] The present invention relates to predicting biological properties, and more particularly, to measuring multi-membrane affinities of test compounds, methods and compositions useful for collecting data indicative of such affinities, and methods of analyzing such data. Methods and systems for use in predicting biological activity, alone or in combination with other molecular descriptors.

BACKGROUND OF THE INVENTION Many research and development efforts have been directed at defining screening methods that allow the identification of drug induction. The goal of such efforts is to provide an effective methodology for predicting biological activity in vivo using experimentally definable (or calculated) descriptors that predict biological activity without the need for time-consuming and costly animal studies. Is to define. Over the years, most drug screening methods have been based on traditional bioactivity assays. More recently, assays have been developed that predict the cell membrane transport properties of test compounds. Generally, drug induction results from comparing the biological and physical properties of a test compound with a known compound having recognized biological activity in vivo. There is significant literature directed at predicting biological activity, which is based on comparisons of physical, chemical and biological descriptors, and pattern recognition analysis on such descriptors as part of a drug screening protocol.

SUMMARY OF THE INVENTION [0003] The present invention is directed to a method of screening compounds for promising biological properties, the method comprising, in principle, numerical values characterizing the interaction with two or more membrane pseudo-surfaces. Based on correlation with corresponding values for control compounds of known biological activity / function. This method is based on the premise that a set of compounds with similar membrane binding properties will have similar pharmacological properties and / or biological activities. The membrane binding properties of the test compound can be calculated or, for example, liposomes, immobilized artificial membranes, US Pat.
, 931, 498 (the disclosure of which is incorporated herein by reference);
It can be determined experimentally using Langmuir-Blodgett membranes, computer chips or similar devices with immobilized lipids, capillary zone electrophoresis columns coated with membrane lipids and the like.

[0004] In one embodiment of the present invention, biological properties are predicted using vector calculus on the pattern signature related to membrane interaction values of test and control compounds. This pattern matching protocol can be applied to datasets containing only values indicative of multi-membrane interactions, and such datasets can include other biologically unique molecular descriptors such as molecular surface area, molecular weight, dipole moment, Octanol-water partition coefficient, molecular volume,
It can include membrane diffusion coefficients, metabolic rates, cell efflux rates, and the like.

In another embodiment, membrane binding data for test and control compounds used in the present invention is obtained using an immobilized artificial membrane chromatographic substrate in a high pressure liquid chromatography system using an aqueous mobile phase. Thermodynamic and kinetic related data on compound / membrane interactions are reflected in retention times and peak widths, respectively. The data are preferably normalized relative to all standard compounds or compound sets, eg, compound sets having a common biological activity or function.

The present invention provides novel carboxyl-functional, head-group protected phospholipids useful for preparing immobilized artificial membrane structures useful for collecting membrane interaction data. These phospholipids are prepared by a novel, high yield transphosphatidylation of phosphatidylcholine derivatives using phospholipase D in the presence of a protected alcohol.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS In accordance with the present invention, an in vitro membrane binding model has been established that predicts the biological activity of a compound regardless of both receptor binding assays and in vivo experiments. The membrane-bound model was developed based on measuring the membrane-bound constant of a bioactive compound on a membrane simulated surface. Collectively, the membrane binding data show that therapeutic class compounds show similar binding profiles on amphipathic / membrane mimetic surfaces and, concomitantly, a similar tendency to partition within the tissue lipid pool from which activity is elicited. Prove that.

[0008] The heterogeneity of lipids in cell membranes functions to control the distribution and activity of many drugs. Drugs that have an affinity for a particular lipid accumulate in a lipid pool enriched in that lipid. Since phosphatidylserine (PS) and phosphatidylinositol (PI) concentrate on the cytoplasmic side of the cell, examples include vinblastine and chlorpromazine, which accumulate in the inner leaflet of the plasma membrane. Drug-lipid interactions have been used to describe biological activity, drug distribution, drug binding to membrane receptors, and intramembrane conformation of drugs. Boundary lipids proximal to membrane receptors play an important role in chelating compounds and presenting membrane-related compounds to receptors, or directly participate in the regulation of receptor function. This is not true for all membrane receptors, but has been documented for several receptors. The results of the experimental studies on which the invention is based demonstrate that compounds with similar biological activities exhibit similar membrane binding properties.

[0009] Due to chemically stringent ligand-receptor binding conditions, the process of drug discovery has traditionally focused on the structure of a compound as the dominant factor controlling biological activity. However, it has now been observed that a wide variety of structurally similar compounds exhibit similar membrane binding properties within a given therapeutic class, suggesting a cooperative role between the receptor and its proximal membrane. You. Thus, the in vitro membrane binding model proposed in accordance with the present invention provides new insight into the conditions required for a compound to elicit a given biological activity. This in vitro membrane binding model coupled receptor binding assay offers a synergistic approach to screening for potential therapeutic compounds, as both the receptor and the surrounding membrane contribute to the elicitation of compound-specific biological activity .

According to one embodiment of the present invention, there is provided a method for screening a test compound for potential biological properties. The method comprises the step of identifying two or more membrane pseudo-surfaces, each having a unique amphiphilic composition. Select a set of control compounds for comparison. Each control compound has a known biological property, such as a biological activity or an interaction with a known receptor type.

[0011] Control compounds can have common properties or can be selected to exhibit a wide variety of biological properties. For each control compound, an ordered set of numerical values is defined that characterizes the biologically relevant interaction (eg, affinity) of the compound with each selected membrane mimetic surface. An ordered set of numerical values for each control compound or each set of control compounds (the "training set" described below) is denoted <C1, C2 ... Cn>, where n is identified and used in the screening method. Represents the number of the film pseudo surface. For each test compound, an ordered set of similar values <T1, T2... Tn> that characterize the biologically relevant interactions for each membrane pseudo-surface are defined. The set of values for the test compound is then compared to each set of values for the control compound, and the control compound having the ordered set of values that most closely matches the ordered set of values for the test compound. Identify biological properties. Numerical values C1, C2... Cn and T1, T2
... Tn can be determined computationally or experimentally by one or two of a variety of industry-accepted methods for assessing membrane interactions.

In one embodiment of the invention, the numerical value characterizing the membrane affinity is a chromatograph using a water-soluble mobile and stationary phase consisting of a membrane mimetic surface, such as US Pat.
31, 498 (hereby incorporated by reference) as determined by high performance liquid chromatography. The term "membrane mimetic surface", which describes and defines the present invention, refers to any surface having immobilized amphipathic molecules (i.e., capable of exhibiting a selective affinity or of a solute (e.g., , Test or control compounds) that have both a lipophilic moiety and a hydrophilic moiety that can interact with it. The term is intended to cover a wide range of detailed stationary phases that are commercially available for chromatography. Suitable membrane pseudo-surfaces are those described in U.S. Pat. No. 4,931,498, which is already incorporated herein.

[0013] Accordingly, in one embodiment of the invention, the use of a test compound is characterized by comparing the membrane binding properties of a test compound of unknown biological activity with the membrane binding properties of a compound of known biological activity in vivo. Assay for potential for more than one species of biological activity. The analysis is performed using the pattern-matching membrane binding constant measured with an immobilized artificial membrane column of a high performance liquid chromatograph system. Two or more immobilized artificial membrane chromatographic columns are used. Each test compound is injected and chromatographed for each column, and the peak width and peak time of the test compound are recorded and normalized to an internal or external standard. Thus, experimentally determined selectivity values (α _{k ′} = k ′ / k ′ standard) and normalized standard deviations (α _sd = σ
/ Σ standard) is calculated for both the test compound and the compound with known biological properties.
Pattern matching using multivariate or principal component analysis of vector calculus, α _{k ′} and α _sd values of these test and control compounds allows for a comparison of the membrane binding properties of the test compound with the respective control compound, Alternatively, if the control compounds all have a common biological activity / property, an arithmetic or arithmetic mean comparison of the membrane binding values of the control compound set for each membrane surface is possible.

Important conditions for optimal utilization of immobilized artificial membrane substrates that give reliable values for characterizing membrane interactions are stable membrane mimetic surfaces, preferably with the same or similar mobile phase composition and sample concentration. The use of at least one common external standard. The external standard allows compensation for variations between daily measurements on the membrane surface and lot-to-lot variations inherent in the preparation of membrane mimics.

The external standard must be stable and soluble in a water-soluble mobile phase (typically a modified organic phase), have an appropriate retention time using the mobile phase, and have a high extinction coefficient. Must have. External standards are usually injected into HPLC / IAM systems, which often reflect transient fluctuations in retention time. Normalizing the data eliminates variations between synthesis lots, variations due to column degradation, and other factors that cause temporary variations in retention times and peak widths. The mobile phase is 0.01M PBS (p
H7.4) One external standard which has been found to give good results when 85% / acetyl nitrile 15% is 4-methylanisole. In one embodiment, one or more sets of compounds, members of each set having a common biological activity or function, are combined and loaded as a mixture into a liquid chromatograph-mass spectrometry system,
The binding properties of the test and control compounds are compared directly in a chromatographic system.

The membrane coupling constant (k ′ value) is expressed by the following equation: k ′ = (t_r-T₀) / T₀Is calculated according to _r Is the retention time, t₀Represents the dead time obtained from each injection, but the denominator
t₀Is the value determined by arithmetically averaging all infusions on each data collection day.

Using chromatographs on the immobilized artificial membrane, data characterizing the equilibrium binding to each membrane pseudo-surface (k ′ value) and mass transfer kinetics between the immobilized artificial membrane surface and the mobile phase Both data (δ value) characterizing the target are obtained. As with all chromatographic columns, columns with immobilized artificial membranes exhibit column-to-column and lot-to-lot variability. This unavoidable variation in experimentally determined k 'and δ values for each of the test and control compounds was due to the pattern matching, where the k' and δ values were chromatographed on each immobilized artificial membrane column. It can be removed by normalizing to an internal or external standard compound. Therefore, the selectivity value and normalized standard deviation α _{k ′} , α _sd are calculated by dividing the k ′ and δ values of the test compound by the respective k ′ and δ values of each of the standards, respectively. In the numerical value k ', α means the selectivity of the surface of the immobilized artificial membrane chromatograph with respect to the molecular recognition of the analyte, and α is the thermodynamic property corresponding to the ratio of the membrane partition coefficient.

[0018] Numerical _T 1 of the each test _compound, T 2 ..., corresponding numeric _{C 1, C 2 ..., C} n is retrieval and subsequent analysis of _{T n} (α _{k 'or} alpha _sd) and the respective reference compound Can be stored in a convenient database management system. Comparison of each set of test compounds with each set of values or control compounds can be performed using any of a number of mathematical analysis techniques.

In one embodiment of the invention, the comparison is performed using a multidimensional vector analysis of the data, wherein an ordered set of values for each test compound is compared to a corresponding ordered set of values for each control compound. Or all control compounds that have a common biological activity or function, arithmetic or arithmetic mean of a selected set of control compounds. Thus, in one embodiment, the control compound which value is most consistent with a test compound _{formula, cosΘ = (T 1 C 1} + T 2 C 2 + ... T n C n) / (T 1 2 + T 2 2 + ... + T n 2 ^{_{) 1/2 (C 1 2 + C}} 2 2 + ... C n 2) is given by ^1/2 angle Θ
Is less than about 20 compounds. The best matching membrane binding pattern between the test compound and the control compound is the pattern with the smallest angle の in the above formula. If the number of membrane surfaces used in this method is two or three, an ordered set of values for each of the test compounds and the values for each of the respective control compounds will be graphed as vectors in a two- or three-dimensional coordinate system, respectively. This makes it easy to compare the value of the test compound with the value of the control compound.

In one embodiment, the test compound data is displayed in a uniquely identifiable vector quantity, ie, in a different color than the control compound. When the control compound constitutes a set of compounds having a predetermined biological activity, the numerical value for each member of the above set of control compounds is clearly distinguishable as a member of the above set of control compounds and from the test compound. It is displayed as a uniquely identifiable vector quantity. Alternatively, control compound data in a set of control compounds having a predetermined biological activity is mathematically processed to define arithmetic or arithmetic mean membrane interaction values for each membrane type. The arithmetic mean or arithmetic mean vector amount calculated for the control compound set may be used in the pattern code analysis described above, or may be graphically displayed as a uniquely recognizable vector amount together with the vector amounts of one or more test compounds. it can.

Thus, in one embodiment of the present invention, the membrane-binding properties of the six hallucinogens are phosphatidylcholine (IAM.PC), phosphatidylserine (IAM.PS), phosphatidylethanolamine (IAM.PE) and sphingomyelin. (IA
M. It was measured on an immobilized artificial membrane (IAM) prepared from SM). Arithmetic mean vectors were calculated for several treatment classes of compound. The calculated arithmetic mean (ie, the arithmetic mean vector representative of a group of compounds with similar biological activity) is shown in Table 1 for hallucinogens.

[0022]

[Table 1]

As shown in the last row of Table 1, the average vector representing the membrane binding characteristics of the hallucinogen is (17.08, 5.441, 113.276, 19.574), and
IAM. PC, IAN. PE, IAM. PS, IAM. SM). The arithmetic mean vectors for several other training (control) sets were calculated and shown in Table 2 for hallucinogens. The goal is to compare the membrane binding constant of the unknown compound with a representative arithmetic mean vector for each class. As shown in Table 2, the unknown compound (
In this example, when the angle between 3,4-methylenedioxyamphetamine (MDA) was calculated based on the arithmetic mean vector for each treatment class of compound, the closest match was the hallucinogen, and the closest match The receptor involved was serotonin.
These numbers indicate that the test compound has similar biological properties to hallucinogens and acts at the serotonin receptor, which is indeed the case for MDA.

[0024]

[Table 2]

In one embodiment of the invention, the membrane binding constant is determined by injecting the drug into a high performance liquid chromatograph (HPLC) using the following IAM stationary phase using an immobilized artificial membrane (IAM) column: did.

[0026] ^{ester IAM.PC C10 / C3, ester IAM.PE} C10 / C3, e ster IAM.PS C10 / C3, ester IAM.SM C10 / C3

The peak position of the drug in the chromatograph is a measure of the affinity of the compound for the surface of the immobilized membrane. Using the retention time (peak time), the capacity factor, k ', of the compound on the IAM surface was calculated. The capacity factor is proportional to the equilibrium membrane coupling constant K in the following equation. k '= φK where φ is the phase ratio of the column. Biological function is the mean k 'value set of similar compounds in practice, constituting the membrane affinity fingerprint of the biological function (MAF ^mu). Therefore,
MAF ^mu means arithmetic mean vector, thus vector components biological activity or function, such as oral absorption, is the arithmetic mean membrane binding constants of similar compounds (training set).

Establishing a Training Set The term “training set” as used in the description and claims of the present invention is a set of compound parts with a common biological property used to define / calculate the MAF ^μ value of that biological property. Membrane / membrane pseudo-binding affinities normally distributed in N-dimensional space occur when the bivariate log (log) plot of αk ′ (ie, PC vs. PE, PC vs. PS, etc.) has an elliptical shape. The oval indicates that the affinity data for a given set of compounds with a known therapeutic effect or function is normally distributed in a two-dimensional space. If a compound falls outside the quartile 0.95 (3 standard deviations from the mean), it is considered an outlier.
Deleting outliers increases the ellipticity of the bivariate data and improves multivariate normality. For this reason, the general procedure used to establish a training set of compounds continues to (1) remove outliers, and (2) if the compound has a quartile 0
. Redraw the elliptic plot until it is within 95. Thus, the elliptic plot was used to remove outliers and to confirm the normal distribution of membrane binding data in N-dimensional space. Using all possible combinations of membrane binding constants, the respective training set: PC vs PE, PC vs PS, PC vs SM, PC vs PS, PE
Paired SM and PS versus SM were defined.

Thus, in another embodiment of the present invention, there is provided a composition comprising a mixture of compounds having a common biological activity or function, or a therapeutic effect. Typically, the molar ratio of the compounds in the training set composition is predetermined. In one embodiment,
The control compounds of the training set composition are in substantially equimolar ratio, optionally associated with one or more external standards. Preferably, the compounds of the training set are selected such that the elliptic plot for the membrane pseudo-binding data for the compounds of the training set composition includes all compounds within the 0.95 quartile. A training set typically includes at least 2, preferably at least 3, and more preferably 5 or more control compounds. The training set composition can be combined with a test compound (and optionally a training set composition for other biological properties) to prepare a mixture for use in LC / MS analysis for membrane / membrane pseudo-binding properties. The training set composition according to the present invention provides an ideal internal control for directly comparing the affinity of the test compound to the affinity exhibited by the training set members in the LC / MS analysis protocol.

Determination of the arithmetic mean vector The average membrane binding affinity for each training set is PC, PE, PS
And SM are arithmetically averaged and are illustrated in Table 3 for selected hallucinogen compounds. Multivariate analysis of variance (MANOVA) confirmed that mean membrane binding affinities greater than 90% for all treatment classes and mechanisms of action differed at a confidence level of 0.05. In other words, the average vector of each training set developed for use according to the present invention was unique compared to all other training sets.

Table 3. Mean vector or membrane affinity fingerprint of hallucinogens (MAF ^mu) data are to Ln weighting.

[0032]

[Table 3]

Prediction of activity of unknown compound D ² and Q value In addition to the vector analysis method described above for calculating the value of angle Θ, D ² and Q value are also vector of binding data (and other quantitative physiochemical properties). It can be used for classification / comparison of compounds by analysis. The membrane binding data for the unknown compound is the observation vector y, and the average membrane binding data for the ith training set of compounds is referred to as the vector y _i . The classification of unknown compounds include mean square distance D ² between the y and y _i.

D ² _i = (y−y _i ) ^I S _i ⁻¹ (y−y _i ) where _Si is the covariance matrix of the ith group of compounds. Unknown compound is assigned to the treatment class or receptor type with a minimum D ² _i values.

Another method of classifying compounds utilizes the Q value. Q_i(Y) =-1 / 2ln | S_i| -1/2 (y-y_i)^lS_i ^-1(Yy _i ) Where | S_i| Is the determinant of the covariate matrix of the i-th group of compounds. Compound
When Q value is used for classification, unknown compounds are assigned to treatment class with maximum Q value.
It is.

[0036] with both D ² and Q for each of the active compounds of all the training set to classify. Compounds if the top of its known therapeutic class or mechanism of action is based on D ² and Q within No. 2 was considered to be correctly classified. Table 4 shows the ^{D 2} and Q values for classification of the MDA (hallucinogen).

[0037] Table 4: ^{D 2} and Q values of all the compared MDA compounds in the treatment class

[Table 4]

Elution Order The relative order of elution is a measure of the relative affinity of each compound for each membrane simulated surface. Unfortunately, the Θ, D ₂ , Q values for compounds of unknown activity do not necessarily reflect the order of elution. Thus, even if the relative elution orders of the compounds exemplified by the mean vector of most contrast training sets are not the same, low Θ and D ² or large Q values may result. Thus, in vivo activity can be predicted using the 溶出, D2 and Q values, as well as the LC elution order of each solid phase substrate for the unknown compound relative to the mean vector elution order of the known training set.

Another parameter to consider when predicting the in vivo activity of a standard deviation compound is the peak width associated with each of the membrane pseudo-surfaces. Peak width measures the on-off kinetics of drug membrane interactions. For example, the unknown compound BDFA was evaluated for hallucinogenic activity. Its Θ values in contrast to hallucinogens MAF ^mu 3.6, the elution order is hallucinogens MA
It was consistent with the order of F ^μ. However, the peak width of BDFA about SM and PS is much larger than the peak width of the corresponding MAF ^mu. In vivo studies are BD
FA was shown to be inactive. Θ values and elution order both indicated that BDFA was active in vivo. Only the difference in peak width demonstrated that BDFA was inactive in vivo.

According to another embodiment of the present invention, test compounds are tested for promising biological properties.
A cleaning system is provided. Is this system more than one kind of membrane
Each having a unique amphiphilic composition. System controls test compound
Means for quantifying the interaction of the compound with each surface, the test compound and each
Means for assigning a numerical value characterizing the quantified interaction of the membrane surface of the sample. standard
Typically, the screening system will monitor the selected control compound or control compound.
Data that characterize the quantified interaction between the
Database. At least a portion of the selected control compound is
Has scientific characteristics. The system includes the values determined for the test compound as control compounds.
The control compound with the value that best matches the value of the test compound compared to the corresponding value of
Includes an analyzer (eg, a programmable computer) to identify things
It is. In one embodiment, the system comprises a test compound number and a control compound count.
Image for displaying at least some numerical values as visually identifiable vector quantities
Algorithm included. In another embodiment, the analyzer is a control compound or a test compound.
Vector calculus to identify the compound with the number that best matches the compound
Contains the algorithm used. The system is thus, for example, Θ (see above)
All the control compounds having a temperature of less than 15 °, more preferably less than 10 ° C., or D ² Report all control compounds with a minimum and / or maximum Q value
Can be programmed as

Generally speaking, certain biologically active compounds exhibit a relative membrane affinity such that when presented as a vector in “membrane space”, the membrane binding data forms an easily identifiable grouping. It turned out to show. See FIGS. 1 and 2. FIG. 1 shows buspirone, sinanserin, mescaline, methysergide, pI amphetamine,
FIG. 3 shows a vector plot of three-dimensional membrane space for compounds acting at serotonin receptors, such as psyrosine, quipazine, metelogulin. FIG. 2 shows apomorphine, clozapine, domperidone, haloperidol, pridinol, SCH23390, SK
Figure 3 shows a three-dimensional membrane space vector plot for compounds that act at dopamine receptors, such as F38393, spiperone, sulpiride, chlorpromazine, dihirexidine, dopamine, perphenazine, prochlorperazine.

This phenomenon (grouping of membrane pseudo-binding data as membrane space vectors) forms the basis of the present invention and allows multiple membrane binding constants to function as predictors of biological activity. Figures 1 and 2 are examples of the results of pattern matching by vector analysis, but are tested by applying other technically recognized mathematical analysis methods such as multivariate analysis and principal component analysis to membrane binding data. The compound can be compared to a control compound of known biological function. Vector analysis is particularly useful for pattern matching according to the present invention, and this method does not present an image, but can be analyzed in three or more dimensions. Such methods can be applied to data characterizing multiple membrane affinities, as well as other datasets containing biologically specific molecular descriptors. further,
The collection of membrane-bound data has been exemplified using immobilized artificial membrane chromatographic supports in high performance liquid chromatography systems, but other methods, such as computer chips or similar devices with immobilized lipids, coated with membrane lipids A capillary zone electrophoresis column, a Langmuir-Blodgett membrane, a liposome, or a monolayer of lipid adsorbed on any surface, such as an AFM tip, can also be used.
The change in vibration of the tip in the presence of the test compound and the control compound is evaluated. In addition to such experimental data collection methods, numerical values that characterize both MAF and non-MAF parameters can be computed. Thus, the data set used in accordance with the present invention includes calculated non-MAF parameters and calculated MAF parameters, including simulated MAF characteristics. “Calculated non-MAF parameters” as used in describing the present invention refers to numerical quantities derived from known chemical structures. Examples include surface area, polar surface area, number of hydrogen bonds, phase index, solubility, and the like. These parameters alone do not predict the overall membrane binding properties, ie, the MAF of the compound. The term “calculated MAF parameter” means a combination of calculated non-MAFs, with or without individual membrane binding constants, that can be used to predict membrane affinity fingerprints (MAFs). Since the HPLC retention time is used to calculate the membrane binding constant, the method used to calculate the elution retention time actually calculates the MAF parameters. Examples of retention times calculated for compound 12 are Ami, D. Davidovic-Am
ie), D. Trinastick, N.M. , "J. Ch.
em. Inf. Comput. 1995, 35, 136-139 ". The term “simulated MAF” refers to an IAM, a two-layer membrane, another membrane pseudo-surface,
Alternatively, it refers to calculated MAF parameters obtained from molecular dynamics simulations of compounds having the force field properties of these films.

According to another embodiment of the present invention, there is provided a novel carboxyl-functional, head-group protected phospholipid represented by the formula HOOC-W-OPO ₂ OZ. In the structural formula, Z is protected glyceryl, 2- (protected aminoethyl), 2-protected carboxy-2-
W represents aminoethyl, 2-protected carboxy-2-protected aminoethyl or an easily-cleavable protecting group; Alternatively, W consists of lipid residues of other amphipathic molecules found in natural biofilms, including but not limited to lecithin, lysolecithin, cephalin, sphingomyelin, cardiolipin, glycolipid, ganglioside, cerebroside. Such compounds are useful for preparing immobilized artificial membranes that are useful for comparing membrane interactions and predicting the biological activity of a test compound.
Another option to, W is a general structural formula W-OPO lipid residues phospholipid represented by 2 _OB, W in the formula Ashiruguriseriru, diacyl glyceryl or N- acyl 3-O-(protection, ) Sphingoshi-1-yl, wherein "acyl" is C8-C
24 alkanoyl or C8-C24 alkenoyl.

In another embodiment of the present invention, a novel carboxyl-functional, head-group protected phospholipid of the formula HOOC-W-OPO ₂ OZ is prepared by reacting a structural formula ZOH in the presence of an excess of the protected alcohol of the formula ZOH It is synthesized in high yield by _{HOOC-W-OPO 2 OCH 2} CH 2 N + (CH 3) phospholipase D (PLD) trans phosphatidyl di configuration of ₃ compounds. In the above formula, Z represents protected glyceryl, 2- (protected aminoethyl), 2-protected carboxyl-2-aminoethyl, or a residue of an acidic protecting group, for example, 4-nitrobenzyl or 4-nitrophenylethyl. W represents a lipid residue as defined above, provided that the starting compound is selected to act as a substrate for phospholipase D activity. Typically, about 3 to 1 stoichiometric
A 5-fold excess of alcohol is used and the reaction is carried out in a water-soluble organic buffer such as chloroform or ethyl acetate in the presence of a water-soluble calcium salt. The resulting carboxyl functional compound is useful for the synthesis of a novel head group-protected immobilized artificial membrane on a solid substrate having a carboxyl-reactive functional group such as a hydroxy, amino, or thiol group. Removal of the protecting group on the immobilized phospholipid results in the corresponding deprotected membrane structure.

As used in describing the present invention, the term “protected” or “protecting group” is used to (1) temporarily bind a reactive functional group (eg, phospho, carboxyl, hydroxy or amino) to attach that functional group (2) a chemical moiety that inhibits an undesired reaction of a reactive functional group that follows during another chemical modification of a compound having the functional group (one of the functional groups (one of Chemical moiety removed from the group (s). Protecting reactive groups of amino, carboxyl and hydroxy functionality is a well known technique for both its synthesis and removal (deprotection) conditions. Protected alcohols useful in the synthesis of the carbokiryl-functional, head-group protected phosphorus material according to the present invention include isopropylidene glycerol, allylserine, 2- (t-
Butoxycarbonylamino) ethyl, 4-nitrobenzyl alcohol and 2-
(4-nitrophenyl) ethanol.

Another artificial membrane stationary phase useful for assessing membrane interactions and predicting biological properties according to the present invention is a novel immobilized ceramide-based membrane. In one embodiment, the ceramide stationary phase is N- via ω-carboxyl group (via ceramide-imidazolide).
Synthesized by covalent attachment of (13-carboxyltridecanoyl) -D-erythro-sphingosine to silica propylamine, followed by C-3 and C-10 terminal capping of the propylamine residue. The resulting stationary phase can be used in an HPLC system to predict skin permeation constant.

The immobilized artificial membrane used in accordance with the present invention to predict biological activity is synthesized using one or more ω-carboxyl substituted lipids. The artificial membrane structure, which is useful as a stationary phase for HPLC chromatographic determination of membrane interactions and has a predetermined or stoichiometric ratio of phospholipid components, provides a known ratio of lipid / phospholipid present in a given biofilm. To be simulated. Thus, the binding affinity of the test compound and the control compound for the same or different membrane substrates can be determined,
It can be compared to the evaluated / predicted biological function.

[0048] Membrane binding data (values associated with interactions with membranes or membrane mimics) were obtained using HPLC on a number of membrane mimics, preferably immobilized artificial membranes (IAM). The membrane mimic used to quantify membrane lipid interactions was phosphatidylcholine (PC)
, Phosphatidylserine (PS), phosphatidylethanolamine (PE),
Synthesized from analogs of sphingomyelin (SM). These membrane lipids were chosen because of the asymmetry established in their plasma membrane. Synthetic methods for preparing immobilized membranes have already been established. Experimental, high-performance liquid chromatography (HPLC)
Injected into the column. The location of the chromatogram peak gives the affinity of the compound on the surface of the immobilized membrane lipid.

To compare the membrane binding properties between a large number of compounds, a pattern matching process was required, whereby the matching patterns became a set of membrane binding data. Pharmacologically similar compounds were found to form clusters in the membrane space. This supports the hypothesis that pharmacologically similar compounds have similar membrane binding properties.
The vector variance between the therapeutic class of compounds defines the region of the membrane space to which the activity of the therapeutic class belongs. When outliers in membrane space are observed for class members,
Other factors in addition to membrane binding are important for outlier activity. The overlap in membrane space between CNS compounds is consistent with these compounds, indicating the essential membrane-binding properties required for drug penetration across the blood-brain barrier. Although it has recently been emphasized that active transport and efflux mechanisms control permeation through the CNS, all C
Since NS compounds have unique membrane binding properties compared to non-CNS compounds,
It should have greatly contributed to transport control. The area occupied by the non-CNS compound is different from the area occupied by the CNS compound in both the direction and magnitude of the vector. Thus, the set of membrane binding data is used to determine whether a given compound can cross the blood-brain barrier. As CNS compounds cross the blood-brain barrier and are freely distributed in different areas of the brain, the unique membrane-binding properties of each treatment class control the localization of the compounds within the CNS. Specifically, if the localization of compounds in the CNS is uniquely controlled by membrane-bound properties, drug vectors within a treatment class are expected to occupy a particular area of membrane space. This clustering is observed for all compounds tested, suggesting that the in vivo membrane binding model can be effectively used to model the distribution of compounds within the CNS.

Vector plots in membrane space are useful to demonstrate that the angle between two drug vectors is small when the compounds exhibit similar biological activities. However, the calculation of the angle is not limited to three membrane coupling constants. Good grouping by therapeutic activity can be achieved when three or more membrane association constants match the pattern. For example, the diazepam pattern better matches the membrane binding properties of anxiolytic compounds when four membrane binding constants are used in the analysis. As shown in Table 5, when only three membrane-binding parameters were used for the diazepam pattern in the database, only two kinds of anxiolytics were included in ten compounds, but four membrane-binding parameters were used. And five anxiolytics were identified. When four parameters are used for three membrane parameters, the calculated angle is large (Table 5), but the pattern sign of the membrane binding constant in N-dimensional membrane space is good for grouping compounds for biological activity. Seems to give results. The finding of compounds with different biological activities than diazepam in Table 5 emphasizes that duplication of membrane binding properties is inevitable between different CNS treatment classes.
In addition, the absence of non-CNS compounds in the first 10 compounds of Table 5 indicates that CN
This strongly suggests that there is a clear difference between the membrane spaces exhibited by S and non-CNS compounds.

[0051]

[Table 5]

Table 5: Pattern code membrane binding constants in three and four dimensions. The angle was calculated using the membrane-bound data and based on the membrane-bound data of diazepam. The membrane space for each analysis is shown.

Synthesis of Ligand for Preparation of Immobilized Artificial Membrane 1-Myristoyl-2- (13-carboxytriamine) is a common chemical intermediate for preparing an immobilized artificial membrane (IAM) surface for measuring drug-membrane interaction. Decanoyl)-
sn-3-phosphatidylcholine (PC-COOH) was used. PC, PG, P
Conventional methods for preparing IAMs containing A, PE and PS from ether lipids involve:
Although 30 steps were required, enzymatic synthesis of ω-carboxyl lipid ligand (PL ^{2ω-COOH / P} ) protected from phospholipase D (PLD) was
By utilizing C-COOH, the total number of synthesis steps could be reduced to ~ 20. The five synthesized PL ^{2ω-COOH / P} ligands are phosphatidylglycerol (PG ^{2ω-COOH / P} , 1a) and phosphatidylserine (P
S ^{2ω-COOH / P} , 1b), phosphatidylethanolamine (PE ^{2ω- COOH / P} , 1c), phosphatidic acid (PA ^{2ω-COOH / P} , 1d)
And an analog of sphingomyelin ( ^{SM2ω-COOH / P} , 2).
PLD transphosphatidylation reactions using PC-COOH were typically performed to ９０90-95% complete. It was an interesting discovery that serine allyl ester was an excellent substrate for PLD and that trans-phosphatidylation of PC-COOH with serine allyl ester was obtained quantitatively. This is the first report describing the synthesis of a suitable SM analog that binds to the molecular surface. The PL ^{2ω- COOH / P} protecting group was removed after the ligand was attached to silica propylamine and end-capped with linear anhydride. After deprotection, PG, PA, PE, PS and S
Five immobilized artificial membrane surfaces containing the same interfacial functional groups as the M membrane lipid were synthesized.

Although single-chain ether PL has been used in the synthesis of first-generation IMA surfaces, two factors have hindered the use of ether lipids in IMA synthesis for measuring drug-membrane binding constants. Was. The first problem is that the optimal prediction requires an immobilized membrane similar in structure to the endogenous lipids found in cell membranes, and some interfaces where single-chain ether lipids are commonly found in most membranes. The functional group was deleted. The second problem is that ether PL requires many synthesis steps, and the synthesis cost is high. Thus, in another embodiment of the present invention, P
An efficient method for synthesizing carboxyl analogs of immobilized membrane phospholipids with the same interfacial functional groups as found in L was developed.

Scheme 1. Embedded image Protected diacetylated phospholipid analogs of PG, PA, PE and SM. After immobilization and subsequent deprotection, these PLs have the same interfacial functional groups as the intrinsic membrane phospholipids (PG, PA, PE or PS).

[0056]

Embedded image

Scheme 1 illustrates the protected diacetylated PL ^{2ω-COOH / P} ligand prepared in this study. The protecting group R _p shown in Scheme 1 is the same protecting group used for the preparation of the immobilized single-chain ether PL, except for the serine analog. The ether PL containing the serine carboxyl was protected with tertiary butyl ether, but the synthesis procedure of the present invention used an allyl ester group. This is because the allyl ester group does not interfere with the PLD transphosphatidylation reaction. Scheme 2 shows the synthesized SM analog (2) and the structural differences between SM ^{2ω-COOH / P} and PL ^{2ω-COOH / P} ligands. A model of drugs that bind to IAM surfaces, SM-enriched intrinsic membranes using ^{SM2ω-COOH / P} ligand was prepared. This example is a renal brush border cell, where 80% of the plasma membrane SM is located outside the membrane leaflet.

Scheme 2. Embedded image Structure of an ω-carboxyl sphingomyelin analog.
The major structural differences from the ester PL are labeled.

[0059]

Embedded image

Scheme 3. (Formula 3)

Embedded image

A general route for synthesizing a double-chain ester PL having an ω-carboxyl group instead of a single-chain ether PL is shown in Scheme 3. DM for obtaining lysolecithin (5)
PLA ₂ hydrolysis of PC was found to be quantitative, but the reaction conditions were enzyme-specific. In addition, PLA ₂ from different production lots contains different types of contaminants that interfere with product formation. In this study, Cobra Musambaik Musambaik (N
aja mocambique mocambique) PLA ₂ (pI = 8.
8), tested several PLA ₂ containing Cobra Musanbiku-Musanbiku PLA ₂ acidic isoenzyme and bee venom PLA2 (Apis mellifera), all of the enzymes could be used to quantify the hydrolysis of DMPC, enzymes Each time, it was necessary to optimize the reaction conditions. Bee venom PLA ₂ has low water - were reported to have high activity in organic solvent systems (1.7% buffer / chloroform V / V). 1.7 in the preliminary survey
The goal was to determine if% V / V buffer / CHCl ₃ was sufficient to scale up the reaction and if several grams of product 5 could be obtained in a single reaction. The first reaction was performed on a 100-fold scale of the published procedure, using Tris buffer instead of HEPES. 1.7% V / V Tris buffer (50 mM, pH 7.2) / CH
DMPC (1 unit) incubated with bee venom PLA ₂ (1000 units) in Cl ₃
. 0g) did not react for more than 4 hours, but the aqueous layer was 5% V / V buffer /
Reaction be prepared in CHCl ₃ did not proceed. Aqueous solution condition 10% V / V
The reaction proceeded when in buffer / CHCl ₃ and was completed by 6 hours. Precipitation of the product in acetone always resulted in a large amount of buffer reagent containing product 5. Therefore, a reaction was investigated to identify an acetone-soluble buffer that did not interfere with PLA2 cleavage of the phospholipid. It was found that the triethanolamine buffer was acetone-soluble, and the lysolecithin (5) was produced by quantitatively progressing the PLA ₂ hydrolysis reaction of DMPC. In triethanolamine buffer,
Cobra Musanbiku-Musanbiku _PLA 2 (pI = 8.8) Cobra was tested Musanbiku-Musanbiku PLA ₂ isoenzyme and bee venom PLA2 (Ap
was found to have the highest efficiency with respect to DMPC hydrolysis among enzymes including ismelifera). For purification of lysolecithin, three precipitation operations from acetone were required. Several PCs are hydrolyzed using PLA ₂ in quantitative yields,
It has been shown that this procedure can be used as a general synthetic method for obtaining lysophosphatidylcholine intermediates.

PC-COOH (4) was synthesized from product 5 as described above. Early studies have shown that PLD from either plants or microorganisms is active in organic / aqueous systems, accepts artificial substrates (including primary and secondary alcohols), and produces phosphate ester products. Indicated. Therefore, we extended the PLD transphosphatidylation reaction to synthesize the final ligands 1a-1d (Scheme 1). Ligands 1a and 1b were synthesized directly from PLD and did not require an additional head group protection step prior to ligand immobilization. In contrast, ligands 1c and 1d required a synthetic step to protect each primary amine of each of these compounds such that the primary amine did not interfere with the immobilization of the PL ^{2ω-COOH / P} ligand. The main by-product in the PLD reaction is the formation of PA, which is the water
This is because it efficiently competes with the alcohol used for the D reaction. Therefore, PLD
One of the primary goals in the reaction is to minimize the production of PA by-products without regard to PC analogs undergoing PLD transphosphatidylation. According to TLC and PLD reaction standards using Compound 4, PA-COOH was produced either in a very small amount or not at all.

PG ^{2ω-COOH / P} (1a) was synthesized by PLD transphosphatidylation using a commercially available IPG alcohol. The reaction proceeds easily and yields ~ 8
3.2% was obtained after purification of the product by flash chromatography. PL
The D reaction was over 90% complete and the product recovery was a decrease from the chromatographic column. Under the reaction conditions used for the conversion of PLD compound 4 to ligand 1a, no PA by-product was formed. Compound 4 is less soluble in the buffer / ethyl acetate solvent, but the yield of PG ^{2ω-COOH / P} is higher in this mixed solvent system compared to the buffer / CHCl ₃ solvent system and the reaction time is much shorter. Was. Similar observations were found for the PLD transphosphatidylation reaction using compound 4 and ethanolamine (PE ^{2ω-COOH / P} (1c) the alcohol used in the synthesis).

The most important point to consider when using PLD enzymology for the synthesis of PS ^{2ω-COOH / P} (1d) is that a serine carboxyl group and an ω-carboxyl group located at the terminal of an acyl chain used for immobilizing a ligand are used. Is the presence of a group. The serine carboxyl group is converted to PLD so that the ω-carboxyl group remains unbound and used for immobilization.
It must be protected before the reaction. However, after immobilization, it is necessary to remove the protective group of serine carboxy using a solution condition that does not deteriorate the surface of the immobilized film.
PLD reactions using serine methyl ester are not available. This is because the deprotection conditions for serine methyl ester hydrolyze the interfacial ester of the immobilized lipid.

Thus, under mild deprotection conditions that identify the serine carboxyl-specific protecting groups to allow PLD transphosphatidylation to proceed effectively and release the serine carboxyl groups after immobilization. It was necessary to remove the protecting group. Therefore, strategies for synthesizing ligand 1d in PLD enzymology were tested with serine benzyl esters, phenacyl esters and allyl esters. All of these esters could be removed under mild solution conditions. The steric hindrance of benzyl alcohol and 2-hydroxyacetophenone makes these alcohols poor substrates for PLD. However, serine allyl ester underwent quantitative PLD transphosphatidylation and the chemical intermediate was not purified. That is, unreacted serine amine was directly protected by Boc, and a ligand 1d suitable for immobilization was produced. After immobilization, the allyl ester is converted to tributyltin hydride (
(Bu ₃ SnH) and the catalyst bis- (triphenylphosphine) palladium chloride (II)
) Can be removed using PdCl ₂ (PPH ₃ ) ₃ . A one-pot, two-step reaction from compound 4 to PS ^{2ω -COOH / P} (1d) gave an overall yield of 87.1% after product purification by column chromatography. The reason that serine amine is not protected prior to the PLD reaction is that serine requires a free amine as a good substrate for the PLD enzyme.

The synthesis of the SM ^{2ω-COOH / P} (2) (see Scheme 4) ligand involved aminolysis of the endogenous SM, producing free amine and sphingosine-1-phosphochlorin. The allylic alcohol was protected with tert-butyldiphenylsilane (TBDPS) before the free amine reacted with the large ring of the cyclic anhydride. T
The BDPS protecting group can be removed with TBAF, which does not degrade the silica surface after ligand 2 has bound to the silica.

The ester ligand was attached to the silica surface by the same procedure previously developed in our lab. See U.S. Pat. No. 4,931,498. Representative IAM
Is synthesized from three sequential processes. In the first binding process, the PL ^{2ω-COOH / P} ligand is linked to the surface using CDI as the activator and the ET
IR spectroscopy is used to verify that the protecting group of the head group remains intact during the immobilization process. The amine remaining on the silica surface is end-capped with an alkyl anhydride.

Scheme 4. (Formula 4)

Embedded image

A detailed study on the end-capping reaction was performed, and the results and other findings showed that both long-chain and single-chain anhydrides could be used for end-capping. However, an important concept is that the end capping reaction is performed only when the surface is ninhydrin negative. Typically, decane (10) anhydride is the first end capping agent, followed by propion (C3) anhydride. However, depending on the use of the IAM packaging material, C3 or C10, or other combinations of double end capping can be performed. For example, IAM
NMR studies require that the IAM be suspended in D ₂ O, which can only be achieved by omitting end capping of the C10 anhydride. Thus, end capping can be used to adjust hydrophobicity, wettability and other final surface properties, and end capping is not a routine operation from a synthetic point of view.

The end capping is various and the protecting group is also PL^{2ω-COOH / P}Change between
In the chemical surface generated throughout the multi-step bonding process
Nomenclature to identify the interstitium is required. PL^{2ω-COOH / P}The ligand is always
It first binds to the silica surface, followed by end capping and deprotection.^ester IAM. PS^{C3 / C3}Life through a sequential bonding process to synthesize surface (10)
The names were as follows; SPA->^esterIAM. PS^p->^esterIAM. PS^{p / C3}-> ^ester IAM. PS^{PC3 / C3}->^esterIAM. PS^{C3 / C3}(
10) where SPA is a silica propylamine starting material and PS^{2ω-COOH / P} After immobilization of^esterIAM. PS^pBecomes^esterIAM. PS ^p Is then followed by two-step sequential endcapping with C3 anhydride^esterIAM
. PS^{PC3 / C3}Becomes Subscript p indicates that the serine protecting group has not been removed
. After removing all protecting groups,^esterIAM. PS^{PC3 / C3}Is the final generation
It becomes thing 10. This nomenclature is^esterIAM. PS^{C3 / C3}Surface synthesis
It is used for all IAMs.

The method of attaching the PL ^{2ω-COOH / P} ligand to the silica surface is the same as the method of PC-COOH which does not require a protecting group. However, the PL ^2ω- COOH ^{/ P} ligand is less soluble in CHCl ₃ than PC-COOH, which was responsible for the long CDI activation time. The CDI-activated PL ^{2ω-COOH / P} ligand was extremely soluble and easy to bind to silica propylamine. When ether PL was used for IAM synthesis, Boc, tertiary-butyl ester and isopropylidene (IPG) were used because the immobilized ether lipid was stable under these reaction conditions.
An aqueous HCl solution was used for deprotection of the groups. However, the ester PL described herein is susceptible to acid hydrolysis and the inventors have used anhydrous TFA to remove acid-sensitive protecting groups.

It has been found that the only way to monitor trace functional groups on the IAM surface after the PL ^{2ω-COOH / P} ligand has been immobilized is to use an IR microscope. However, compared to all of the protecting group of other tested, ^{este r} IAM by IR microscope. Direct monitoring of the deprotection of the allyl ester group on the PS ^{C3 / C3} was not possible because the allyl ester group does not have an IR band in a unique region of the IR spectrum. Deprotection of the allyl group from peptide chemistry proceeds quantitatively using tributyltin hydride. However, monitoring of deprotection is essential because surface chemistry has a significant effect on many chemical reactions, and deprotection of allyl esters is not easy on IAM surfaces.

In the IR spectrum under acidic or basic conditions, the IR band was changed to the ^ester IAM. The degree of deprotection was monitored using shifting on PS ^{PC3 / C3} . Thus, Bu ₃ SnH and PdCl ₂ (P
Ph ₃ ) ₃ in the ^ester IAM. Acid IAM after deprotection of PS ^{PC3 / C3}
Surface (serine -COOH) and basic IAM surface (serine -COO ^-) 17
Between 80 and 1680 cm ^-1 , a reduction of the integrated area of the ester of up to 1.3 times was shown. This decrease in strength was expected as the surface contained two interfacial esters from the immobilized phospholipid. Serine COOH serine COO ^- when ionized species, serine carboxyl from 1710 cm ^-1 functional groups shifted by protecting group comes off the 1550 cm ^-1 moved to a new area of the IR spectrum, whereby deprotection reaction Monitoring becomes possible.

Several synthetic methods could be developed for synthesizing and synthesizing the PL ^{2ω-COOH / P} ligand shown in Scheme 1. For example, PG, PA, PE,
PLA ₂ hydrolysis of endogenous phospholipids including PS is achieved following protection of the polar head group and linking of the long chain acyl group with the ω-carboxyl group. This type method of synthesizing each of the ^{PL 2ω-C OOH / P} ligand is synthesized from endogenous precursor has several advantages, the most obvious advantage is get all the ^{PL 2ω-COOH / P} ligand There are several additional synthetic steps to take place.

However, it is important to consider at the same time that (1) the endogenous starting materials such as PG, PA, PE and PS are expensive, (2) PG, PA, PE and P
That is, the lysophospholipid of S is hardly dissolved in the aprotic solvent, so that it is hard to undergo a subsequent chemical reaction. Therefore, the choice of PC-COOH (4) as a common intermediate (Scheme 3) for synthesizing the ester PL ^{2ω-COOH / P} avoids the use of expensive starting materials PG, PA, PE and PS, Eliminate lysophospholipid intermediates. The synthetic methods described herein are efficient on both small and large scales and are generally useful for synthesizing membrane surfaces. Ligands can bind to any molecular surface.

General Examples 1,2-Dimyristoylcholine-sn-glycero-3-phosphocholine (DMP
C) was purchased from Genzyme Pharmaceuticals (Cambridge, MA).

Bee venom PLA ₂ (Apis mellifera), Naja
mocambique mocambique PLA ₂ (pI = 8.8), N
aja mocambique mocambique PLA ₂ acidic isoenzyme, PLD (obtained from Streptomyces actinomycetes), CaCl ₂ · 2H ₂ O and ethanolamine were procured from Sigma Chemical Company (MO York St.Louis standing). Triethanolamine (TEA), tri (hydroxymethyl) aminomethane (Tris), ethylenediaminetetraacetic acid disodium salt (EDTA
), Trifluoroacetic anhydride (TEA), 1, -1-carbonyldiimidazole (
CDI), propionic anhydride (C ₃ ), decanoic anhydride (C10), tributyltin hydrate (Bu ₃ SnH), bis (triphenylphosphine) palladium (
II) Chloride [PdCl ₂ (PPh ₃ ) ₃ ], anhydrous chloroform containing no alcohol (stabilized with amylene), methylene chloride, 4-dimethylaminopyridine (DM
AP), sodium acetate, di-tert-butyl dicarbonate (Boc anhydride)
, Para-nitrophenyl alcohol (NPEA), triethylamine, tert-butyldiphenylsilyl chloride (TBDPSCL), imidazole, normal butanol, 1,12-dodecanedicarboxylic acid, tetrabutylammonium fluoride (TBAF) and allyl bromide From Aldrich (Milwalky, Wis.). Sphingomyelin (brain) and egg PC are from Avanti
Procured from Polar Lipids, Inc. (Alabaster, AL). Deuterated solvents for NMR spectroscopy, including methanol and chloroform, were from the Cambridge Institute for Isotope Research (Andover, MA), chloroform (without alcohol), acetone, isopropyl alcohol, methanol, acetone , Hexane, ethyl acetate, acetic acid, hydrochloride, and sodium hydroxide were from Mallincrodt (Paris, NY). Experimental motors, stirring shafts and stirring blades were purchased from ACE Glass, Inc., Louisville, KY. Phosp for detecting phosphorus-containing organic compounds
The ninhydrin assay for the detection of ray reagents and amine-containing compounds is available from Supelco
, Inc. (Belfonte, PA), silica propylamine (
SPA) was provided by Rockland Technology, Inc. (Newport, DE). Using SPA (RN 39-94), the ester-IAM. The PL surface is finished, this surface is trifunctional and has a surface area of 180 m ² / g and 80
It has an A pore size. The density of propylamine groups on RN39-94 is 3.09 μmol / m ² . SPA (RN38-94) and IAM. SM
A face was prepared, which was monofunctional and showed the same surface area and pore size as RN39-94. The density of propylamine groups on RN38-94 is 2.0
6 μg / m ² and all infrared spectra were recorded on a Nicolet Magna-IR spectrometer coupled to an IR Plan I microscope. About 10-15 mg of I for elemental analysis
Using AMs, Perkin at the Microanalytical Laboratory at Purdue University's Faculty of Chemistry
-Elmer PE 240 was used.

1-Myristoyl-sn-2-hydroxyphosphatidylcholine (5, lysolecithin) DMPC (10.0 g) was weighed and dissolved in chloroform (300 ml). Then 50 mM TEA containing 2.5 mM Ca ⁺⁺ and 0.25 mM
(PH 7.2) 32.0 ml of buffer was added. Naja mocambi
An enzyme solution of que mocambique PLA ₂ (1 ml of 1,000 units / ml in 50 mM TEA buffer) was pipetted into the DMPC suspension. Viscous white reaction product at 100 rpm for 7.5 hours (for basic isoenzyme)
Or 11.0 hours (for acid isoenzymes) with paddle agitation. Prior to the rotary evaporation operation, isopropyl alcohol (100 ml) was added to the reaction mixture to avoid foaming upon removal of the solvent. After the rotary evaporation operation, the residue was suspended in methanol (100 ml) (1-2 minutes) and sonicated to dissolve the residue. The methanol solution containing the product is gently pipetted into acetone (~
500 ml), which resulted in an immediate white precipitate (product).

In order to completely form a precipitate, it was necessary to leave the mixture at 4 ° C. overnight at low temperature. Settling
The product was collected after filtration using a glass fiber-sintered filter. TLC (C
HCl₃/ CH₃OH / H₂O = 65: 35: 4) contains a trace amount of myristic acid
Contains no trace myristic acid to contaminate lysolecithin products with TEA
To obtain compound 5, a second precipitation had to be obtained. Adjust the pH to 9.7
TEA was completely removed from the product in the third precipitation operation, and pure product 5 (5
. 16g, 73.1%) was obtained under dry vacuum overnight. For additional precipitation
Thus, the recovery of lysolecithin from the third furnace liquid was increased to 80.0%.
This reaction operation was performed up to 50 times to obtain 285 g of lysolecithin.
Following the established procedure, it was used for the preparation of PC-COOH (4). CHCl₃/ CH ₃ OH / H₂TLC in O (65: 25: 4) showed pure product (Rf = 0.3
4), FTIR (Fourier transform infrared spectroscopy); (thin film, cm ^-1 ) 2955.24, 2917.03, 2849.68, 1734.22, 1
467.66, 1234.61, 1086.39, 1056.77. Indicates that¹ In H NMR (H nuclear magnetic resonance), (300 MHz, CDCl₃, Δ ppm)
AB-MS solid

[M + ^{H _+],} C _{22 H} 46 _O 7 PN Calculated 468.6, were Found 468.5.

1-Myristoyl-2- (13-carboxytridecanoyl) -sn-3-phosphatidylcholine (4, PC-COOH) L-serine allyl ester (3) Using a round bottom flask (1000 ml) N in anhydrous ethyl acetate (600 ml)
-(Tert-butoxycarbonyl) -L-serine (25 g) was suspended and the
Allyl chloride (60 ml) and trimethylamine (24 ml) were added. 24:00
After stirring for a while, the reaction mixture was filtered, rotary evaporated and the filtrate concentrated to give a yellow oil
. The crude product was treated with hexane / ethyl acetate (4: 1, 1000 ml), hexane / acetic acid
Ethyl (2: 1, 1000 ml) and hexane / ethyl acetate (1: 1, 100
0 ml) followed by hexane / ethyl acetate from hexane (500 ml)
Purification was performed by flash chromatography using a gradient. In a vacuum oven
After drying overnight with 27.2 g of intermediate, N-tert-butoxycarbonyl
As a result, ru-L-serine allyl ester was obtained (yield: 93%). Ethyl acetate / hexane
Pure product (Rf = 0.1) by TLC (thin layer chromatography) using (1: 1).
39) is shown. In FTIR (Fourier transform infrared spectroscopy), (thin film,
cm^-1), 3415.86, 2978.14, 2933.60, 2886.
66, 1743.43, 1715.75, 1512.33, 1392.96, 1
370.85, 1342.40, 1165.13, 1060.54, ¹ 1 H NMR (300 MHz, CDCl₃, Δ ppm); 5.87 (1H
, M), 5.54 (1H, d, J = 7.79 Hz) 5.28 (1H, d, J = 1)
7.17 Hz), 5.20 (1H, d, J = 10.40 Hz), 4.61 (2H
, D, J = 5.62 Hz), 3.87 (2H, m), 1.40 (9H, s). But
Admitted. CI-MS

In [M + H ⁺ ] analysis, the calculated value of C ₁₁ H ₁₉ NO ₅ was 246.3, and the measured value was 246.0.

N-tert-butoxycarbonyl-L-serine allyl ester (27.2 g
) Was placed in a 500 ml round bottom flask and cooled to 0 ° C. N2 to the original positive pressure was added the yellow oil substance 50% _{TFA / CH 2 Cl 2 (100ml} ). The reaction mixture was stirred at 0 ° C. (30 minutes), warmed to room temperature and stirred for another 2 hours. TF
A and CH ₂ Cl ₂ are removed by rotary evaporation to give a yellow oil. This yellow oil was dissolved in ether (200ml). The product precipitated completely when kept at 4 ° C. for 3 hours. The precipitate was collected by filtration and washed with ether (50 ml) and hexane (2 ml).
0 ml). When the white precipitate is dried in a vacuum oven overnight,
The dried L-serine allyl ester (3) was obtained (15.6 g, yield 10).
0%). TLC using CH ₂ Cl ₂ / CH ₃ OH / H ₂ O (65: 25: 4)
Gives a pure product (Rf = 0.58). Measurement results of FTIR (thin film, cm ⁻¹ ). 2994.54, 290.64, 2891.49, 1747.28
, 1674.06, 1201.52, 1145.92. ¹ H NMR (300 M
Hz, CDCl ₃ , δ ppm). 5.86 (1H, m), 5.30 (
2H, m), 4.59 (2H, d, J = 6.43 Hz), 4.09 (2H, m)
, 3.77 (1H, m). FAB-MS

[M + H⁺】Measurement result. C₆H₁₁NO ₃ , 146.2, found 146.0.

3-O- (tert-butyldiphenylsilyl) -sphingosine-1-phosphocholine (6) Sphingosine-1-phosphocholine was prepared by hydrolyzing SM with minor modifications. Briefly, SM (10.5 g, 15 mmol), n-BuOH (
150 ml) and a solution of HCL (45 ml, 6.0 N) at 100 ° C. for 1 hour.
And then vacuum dried by rotary evaporation. The sphingosine-1-phosphocholine product is not purified, but the formation of the product is FAB-MS m / z 465.
. 2

[M + H ⁺ ]. Sphingosine-1-phosphocholine is THF
(500 ml) and added in the order of TBDPSCl (23.4 ml, 90 mmol) and imidazole (12.2 g, 90 mmol). After the reaction was stirred (40 ° C., 10 h), it was rotary evaporated and purified by column flash chromatography. After column chromatography operation, a white solid (5
. 64, 53% overall yield). CHCL ₃ / CH ₃ OH / H ₂ O (65
: 25: 4), a pure product was obtained (Rf = 0.25%).
). The measurement results of FTIR (thin film, cm ⁻¹ ) were 3390.56, 3071.5
8, 3046.65, 3015.28, 2956.99, 2925.18, 28
54.48, 1963.80, 1898.78, 1830.70, 1620.9
8, 1428.53, 1239.36, 1059.06, 973.34. And the measurement result of ¹ H NMR (300 MHz, CDCl ₃ δ ppm); 7.66 (
4H, m), 7.32 (6H, m), 5.43-5.09 (2H, m), 4.4
5 (3H, m), 4.12 (4H, m), 3.40 (9H, s), 3.39 (1
H, m), 1.64 (2H, m) 1.24 (22H, m, br s), 1.00
(9H, s), 0.86 (3H, tr, J = 6.45 Hz), and FAB-M
S

As a result of [M + H ⁺ ], calculated value of C ₃₉ H ₆₇ N ₂ O ₅ 703.1,
Obtained value 703.2.

3-O- (tert-butyldiphenylsilyl) -sphingomyelin (2, Sω ^{-COOH / P} ) “6” (5 g, 7.1 mmol) and 1,12-dodecanedicarboxylic anhydride [ Prepared from 1,12-dodecanedicarboxylic acid (5.0 g, 19.35 mmol)] and DCC described in [Pidgeon, 1989 # 8] (4.0 g, 1
(39.39 mmol) was stirred at room temperature for 10 hours. The reaction solvent was removed by a rotary evaporation operation, and CH ₂ Cl ₂ (500 ml), CH ₂ Cl ₂ / T
HF (1: 1, 500 ml), CH ₂ Cl ₂ / CH ₃ OH (1: 1, 1000 m
l) and CH ₂ Cl ₂ / CH ₃ OH / H ₂ O (65: 35: 4, 3000 ml)
The product was purified using a step gradient mobile phase consisting of The product-containing chromatographic fractions are pooled, rotary evaporated and vacuum oven dried to give 5
. 4 g of SM ^{ω-COOH / P} (2) were obtained with a yield of 80.6%. TLC using CHCL ₃ / CH ₃ OH / H ₂ O (65: 25: 4) showed pure product. Its R factor was 0.47. Measurement results of FTIR (thin film, cm ⁻¹ ) are 3312.05, 3071.70, 3046.10, 2923.61.
, 2855.04, 1960.21, 1899.87, 1831.31, 174
2.70, 1703.39, 166.42, 1546.28, 1466.74
, 1428.35, 1065.80, 967.63, and ¹ H NMR (50
0 MHz, CDCl ₃ , δ ppm), 7.64 (4H, m), 7
. 34 (6H, m), 5.34-5.04 (2H, m), 4.39 (1H, dd
), 4.24 (2H, m), 4.02 (2H, m), 3.68 (3H, m), 3
. 21 (9H, s), 2.30-1.92 (4H, m), 1.76-1.37 (
6H, m), 1.24 (38H, br s), 1.04-1.00 (9H, 3s
), 0.84 (3H, tr, J = 6.56 Hz). FAB-MS

[M +
As a result of H ⁺ , the calculated value of C ₅₃ H ₉₁ N ₂ O ₈ PSi was 943.4, and the measured value was 943.0.

1-myristoyl-2- (13-carboxytridecanoyl) -sn-3-phosphatidylglycenol-2′-3′-isopropylidene (1a, PG ^{2ω-COOH / P} ) PC-COOH ( 1.01 g, 1.43 mmol) was weighed, and ethyl acetate (5.
6 ml) and sonicated (5 minutes). 10.62 ml of NaOAc
Buffer (CaCl ₂ of NaOAc and 50ml of 100 ml, composed pH 6.5) and (R) - isopropyl retentivity 1.6ml glycerol (IPG) (
The suspension was agitated (30 ° C, 1 .10 ml) before pipetting.
5 minutes). Upon addition of IPG, the two-phase system immediately becomes suspended. PLD (1
00 μl, 1.0 unit / μl of NaOAc buffer) was added and the reaction state was determined by TLC
(CHCl ₃ : CH ₃ OH: NH ₃ H ₂ O = 65: 25: 4). After 2 hours, more than 60% of the PC-COOH was converted to PG ^{2ω-COOH / P.} PLD (40 μl) was added and the reaction was continued for another 5 hours (reaction completion rate by TLc was 90%). The reaction mixture was KHSO ₄ (1 M, 1.0 mmol)
And concentrated by rotary evaporation, and the residue formed was purified using flash chromatography with CHCl ₃ / CH ₃ OH / H ₂ O = 65: 25: 3. The product-containing chromatographic fractions are pooled, concentrated by rotary evaporation and dried under vacuum. The yield after column chromatography was 1.74 g as a white solid (yield 83.2%). CHCl ₃ / CH ₃ OH / H ₂ O (6
5: 25: 4) showed that a pure product was obtained (Rf = 0).
. 40%). FTIR (thin film, cm ⁻¹ ) results are 2982.51, 292
3.36, 2852.97, 1741.51, 1710.84, 1556.64
, 146.17, 1415.60, 1379.37, 1370.00, 124
4.68, 1219.69, 1160.35, 1106.00, 1069.15
And ¹ H NMR (500 MHz, CDCl ₃ , δ ppm) showed 5
. 24 (1H, br, m), 4.37 (1H, br, m), 4.29 (1H, b
r, m), 4.17 (2H, br, m), 4.04 (3H, br, m), 3.9
0 (1H, br, m), 3.80 (1H), 2.29 (6H, m), 1.57 (
6H, m), one 1.42 and _{1.35 (3H, s, -CCH 3} , two adaptations), 1.40 and _{1.33 (3H, s, -CCH 3} , two adaptations One), 1.26 (36H, m), and 0.86 (3H, t, J = 7.00 Hz). FAB-MS

For [M + H ⁺ ], the calculated value for C ₃₇ H ₆₉ O ₁₂ P was 737.9, and the measured value was 737.5.

1-Myristoyl-2- (13-carboxytridecanoyl) -sn-3-phosphatidyl-p-nitrophenylethanol (1b, PA ^{2ω-COOH / P} ) PA ^{2ω-COOH / P} was prepared by PG ^Although similar to that of ^{2ω-COOH / P} , the difference is that p-nitrophenethyl alcohol was used for IPG for the transfer reaction of PLD phosphatidyl group, and the reaction rate was increased 1.5 times. It is. The required optimum PLD (1.0 unit / μ) yield is sequentially added in three stages. 100 μl,
Culture for 2 hours: 37 μl, culture for 2 hours followed by 50 μl-PLD culture for 12 hours. All of the products were phospray positive and UV positive. The reaction mixture was acidified using KHSO ₄ (1.042 ml, 1M), concentrated by rotary evaporation, and purified by flash chromatography using silica gel. As the solvent system, 800 ml of CHCl ₃ / CH ₃ OH / H ₂ O (65: 25: 3)
The accompanying 700 ml CHCl ₃ / CH ₃ OH (9: 1). The product-containing chromatographic cuts were pooled, rotary evaporated and dried in vacuo. The yield by chromatography was 1.12 g (yield 75.1%) of a yellow solid oil. TLC with CHCl ₃ / CH ₃ OH / H ₂ O (65: 25: 4) shows the pure product as Rf = 0.42. FTIR (thin film, cm ^-1 ) results are 2923.70, 2852.81, 1740.44, 1705.13, 16
03.81, 1550.54, 1522.11, 1465.95, 1345.8
8, 1220.07, 1108.70, 1070.67, 1034.59, and FAB-MS

The result of [M + H ⁺ ] is 7 calculated for C ₃₉ H ₆₆ NO ₁₂ P.
72.9, found 777.5.

N-tert-butyloxycarbonyl-1-myristoyl-2- (13-carboxytridecanoyl) -sn-3-phosphatidylethanolamine (1C, PE ^{2ω-COOH / P} ) PE ^{2ω-COOH /} Although ^P preparation is similar to the PG ^{2 [omega-COOH / P,} different from a mixed solvent system of ethyl acetate / chloroform (4.6: 1) lies in using. Chloroform does not contain alcohol. CHCl ₃ : CH ₃ OH
: _NH 3 _H 2 O based on the operation concentration meter operation devised TLC plate (65: 25: 4), up to 88% of PC-COOH starting material was converted to the corresponding PE carboxyl similar compounds. The reaction mixture is concentrated by a rotary evaporation operation and the first PE
The intermediate was not purified before the amine was protected with Boc. Na ₂ CO ₃ (20 ml, 1.0 M) and dioxane (40 m
l) was added. The reaction mixture was cooled to 0 ° C. prior to the addition of (Boc) ₂ O (1.98 g, 9.07 mmol). The reaction was stirred (0 ° C.? For 1 hour) and then acidified with KHSO ₄ (40 ml, 1.0 M) at room temperature for 3 hours. Addition of chloroform (40 ml) produces an emulsion reaction mixture which separates into two layers. The organic layer was collected and combined with the chloroform washing solution (2 × 40 1) in the liquid layer. The chloroform extract was concentrated by rotary evaporation using silica gel flash chromatography before purifying the product. The composition of the solvent system is CHCl ₃ / CH ₃ OH (9
; 300 ml), followed by CHCl ₃ / CH ₃ OH / H ₂ O (65:
The product-containing chromatographic fractions were pooled using 25: 3, 800 ml), concentrated by rotary evaporation and dried under vacuum. The yield after the chromatographic operation was 0.96 g (88.8%) of a white solid. Pure product Rf = 0 by TLC operation with CHCl ₃ : CH ₃ OH: NH ₃ H ₂ O (65: 25: 4)
. 44 obtained. The measurement result of FTIR (thin film, cm ⁻¹ ) was 3384.
65, 2923.00, 2852.32, 1742.15, 1715.33, 1
690.96, 1553.86, 1517.93, 1464.04, 1365.
55, 1244.64, 1221.96, 1171.99, 1108.70, 1
071.16, ¹ H NMR (500 MHz, CDCl ₃ , δ ppm)
The measurement results were 5.20 (1H, br, m), 4.36 (1H, d, J = 10.5)
Hz), 4.13 (1H, br, m), 3.93 (2H, br, m), 3.86
(2H, br, m), 3.33 and 3.20 (2H, br, m), 2.25 (
6H, m), 1.55 (6H, m), 1.40 (9H.s), 1.25 (36H, m)
, M), 0.85 (3H, t, J = 6.75 Hz). FAB-MS

[M +
According to the measurement results of Na ⁺ ], the calculated value of C ₃₈ H ₇₁ NO ₁₂ P was 789.0.
, Measured value 788.5.

N-tert-butyloxycarbonyl-allyl ester-1-myristoyl-
2- (13-Carboxytridecanoyl) -sn-3-phosphatidylserine (1d, PS ^{2ω-COOH / P} ) PC-COOH (7.0 g, 9.90 mmol) using a 500 ml round bottom flask. Was dissolved in anhydrous chloroform (200 ml). To the reaction mixture was added serine allyl ester (7.2 g, 49.50 mmol) followed by (10
33. 0 mM NaOAc, consisting of 50 mm CaCl ₂ exhibiting a pH of 6.5)
Added with 4 ml of buffer solution. PLD (500 μl, 1.0 unit / NaOA
c buffer) was added and the reaction was stirred at 30 ° C under a stream of N ₂ for 8 hours. The chemical intermediate, a carboxylPS analog obtained from the reaction of PLD with PC-COOH, was concentrated by rotary evaporation, but was in an unpurified state before protecting serineamine with Boc. The primary serine-amine was Boc protected as described above in the preparation of PE ^{2ω- COOH / P.} KHSO ₄ (16
(0 ml, 1.0 M), the product was extracted with diethyl ether (4 × 100 ml) after oxidizing the protection reaction mixture. The diethyl ether extracts were pooled and the solvent was rotoevaporated to obtain a yellow oil, which oil contained C
Purify using a silica gel flash chromatograph using H ₂ Cl ₂ / CH ₃ OH / H ₂ O (65: 25: 4). From the final product (white solid) obtained by column chromatography, a total yield of 87% (7.32 g) was obtained, but the TLC of CH ₂ Cl ₂ / CH ₃ OH (1: 1) was obtained. The operation showed it to be a pure product (Rf = 0.81). The measurement results of FTIR: (thin film, cm ⁻¹ ) were 3364.26, 2925.48, 2855.87, 1741.78, and 17
13.99, 1510.08, 1461.08, 1372.02, 1185.0
2, 1080.51, and ¹ NMR measurement results (300 MHz,
Under conditions of CDCl ₃ , δ ppm), 5.90 (1H, M), 5.34-5.20 (
4H, m), 4.62-4.00 (8H, m), 2.27 (6H, m), 1.5
7 (6H, m), 1.41 (9H, s), 1.25 (36H, br, s), 0.
87 (3H, tr, J-6.65 Hz), MALDI-MS

[M
-H] ^- _For, Calcd for C ₄₂ H 75 _{NO 14} P is 848.1, shown as the measured value 848.0.

Ester IAM. PG ^{C10 / C3} (7) PG2 ^{ω-COOH / P} (1a) (1.0 g, 1.35 mmol) was suspended in anhydrous CHCl ₃ (50 ml) using a round bottom flask (100 ml), and
I (0.380 g, 2.35 mmol) was added. The 1a ligand is not very soluble in CHCl ₃ , but the CDI-activated PG ^{2ω-COOH / P} ligand (ligand) is very soluble. After stirring at room temperature overnight, TLC-CH ₂ C
Both l ₂ / CH ₃ OH / H ₂ O (65: 25: 4) and IR spectroscopy
DI was completely consumed, indicating that only ６０60% of the starting material was activated. Vacuum-dried SPA (RN39-94) (10.59 g) was added and the suspension was stirred using an orbital stirrer (150 rpm, 24 hours continuous).
The above ester-IAM. The PG-P silica is now filtered through a glass fiber sinter funnel, with chloroform (2 × 20 ml), methanol (3 × 20 ml),
After washing with acetone (10 ml), vacuum drying was performed overnight at 40 ° C.
Here, the coupling reaction was repeated under reaction conditions suitable for 0.25 equivalents of PG ^{2ω-COOH / P} ligand. The ester-IAM. To obtain ^{PGP / C10 / C3} , the end capping of C10 and C3 described above was performed. The above IAM. Protective removal (or deprotection) of isopropylidene on the PG surface is a necessary operation to obtain a fixed artificial membrane containing PG. Simply put, ester-
IAM. ^{PG P / C10 / C3 (~10.9g} ) was _CH 2 _CL 2 Anhydrous suspended (24 ml). Thus, anhydrous TEA (8 ml) was added to the suspension.
After 2 hours of stirring the silica was filtered, CH ₂ Cl ₂ ( ₂ × 20 ml), CH ₃ OH
(3 × 20 ml) and acetone (20 ml), then (40 ° C.
Overnight). This protection removal operation was repeated for 4 hours using 50% TFA / 50% CH ₂ Cl ₂ (V / V, 24 ml). The silica was then immersed in 50% 1M Na ₂ CO ₃ /50% CH ₃ OH (80 ml, 6 minutes) under constant stirring, followed by water (600 ml), CH ₃ OH (2 × 40 ml) and acetone (2 × 40 ml).
40 ml). The surface was finally vacuum dried overnight at 40 ° C. FTIR microscopy of surface (7) showed that the IPG groups were quantitatively removed because at 1370.0 cm 2 the IPG peak feature was completely lost as a result of deprotection. It depends. Ester-IAM. Binding ligand density on the PG ^P

[Ong, 1994 # 14] (1a) is 134.67 μmo.
1 / g-silica, based on the carbon content (C, 2.13%) before ligand 1a binding to the silica surface and the carbon content (C, 7.46%) after binding.
) Based on the chemical analysis results.

Ester-IAM, PA ^{C10 / C3} (8) Coupling operation for immobilization of PA ^{2ω-COOH / P} (1b) and Ester-IA
M, C10 and C3 endcapping operations to prepare PAP ^{/ C10 / C3} are described in Ester-IAM. There was no change from the operation of PGP ^{/ C10 / C3} . Ester-IAM. ^P -Nitrophenylethyl on PA ^{P / C10 / C3} (N
The protection removal of the PE) group was performed as described. Briefly, ester-IAM
. PAP ^{/ C10 / C3} (〜15.0 g) was suspended in anhydrous CH ₂ Cl ₂ (50 ml), and anhydrous DBU (20 ml) was added. After 2 hours of stirring, the silica is filtered, washed with brine (1.0 M NaCl, 500 ml), H ₂ O (500 ml) and acetone (50 ml) and subsequently dried in vacuo (40 ° C., overnight). Continuous I of NPE between 1333-1358 / cm ^-1 before and after protective removal
A first deprotection based on the R wavelength band removes up to 70% of the P-nitrophenylethyl groups. The second deprotection with benzene removes up to 81%. After the deprotection operation, the silica was filtered and CHCl ₃ (200 ml), CH ₃ OH (10
0 ml), brine (1.0 M NaCl, 500 ml), H ₂ O (500 ml) and acetone (100 ml), followed by vacuum drying (overnight at 40 ° C.). Ester-IAM. The binding ligand density (1b) on PA-P was determined by elemental analysis of the carbon content (C before binding, 2.13%, C after binding, 6.34%) before and after binding of the binding ligand 1b to the silica surface. Based on this, it was 96.50 μmol / g-silica.

Ester-IAM. PE ^{C10 / C3} (9) PE ^{2ω-COOH / P} (1b), C10 and C3 endcapping immobilized and ester-IAM. The coupling procedure for preparing PEP ^{/ C10 / C3} is described in Ester-IAM. Identical to the procedure for PGP ^{/ C10 / C3} . Ester-IA
M. Deprotection of Boc from the PE-P / C10 / C3 face was performed as described. Briefly, the ester IAM. PEP ^{/ C10 / C3} (~ 8.9 g) was converted to anhydrous C
Suspended in H ₂ Cl ₂ (20 ml). Then anhydrous TFA (20 ml) was added to the suspension. After 2 hours of stirring the silica was filtered, CH ₂ Cl ₂ ( ₂ × ₂₀ ml),
Washed with CH ₃ OH (3 × 20 ml) and acetone (20 ml) followed by vacuum drying (40 ° C. overnight). The first deprotection was ８０80% complete and the second deprotection step was completed in 50% TFA / 50% CH ₂ Cl ₂ over 4 hours, this operation giving the Boc IR wavelength. The band 1369.0 / Cm disappeared completely. Ester-IAM. The bound ligand density (1c) on PE-P is the element of carbon content before binding (C, 2.13%) and carbon content after binding (C, 6.64%) of ligand 1c on the silica surface. Based on analysis, it was 109.33 μmol / g-silica.

Ester-IAM. PS^{C10 / C3}(10) Ester-IAM. PS^{P / C3 / C3}The coupling procedure for preparing
Except when using end-capping, the ester-IAM. PG^{P / C10 / C3} There was no change from the preparation procedure. Ester-IAM. PS^{P / C3 / C3}Up
The surface protection removal of the allyl ester group of is determined by the deprotection conditions of the solution. simply
In other words, ester-IAM. PS^{P / C3 / C3}(22.44 g) in anhydrous CH ₂ Cl₂(50 ml), remove nitrogen for 5 minutes and sonicate for 10 seconds
Nitrogen was released for 1 minute before further deprotection was added. PdCl₂(PPh₃) ₃ (96.07 mg, 0.14 mmol) acetic acid (600.4 μl, d = 1.04
9, 10.50 mmol) and Bu₃SnH (4.53 ml, d = 1.082
, 16.84 mmol) are sequentially ester-IAM. PS^{P / C3 / C3}In suspension
Was added. After 2 hours of stirring, Bu₃Add SnH (1.0 ml, 3.72 mol)
The suspension was stirred for a further 30 minutes. Filter the surface to CH₂CL₂(50ml),
CH₃OH (5 × 50 ml), hexane (30 ml), ethyl acetate (30 ml)
And acetone (40 ml), followed by vacuum drying (at 40 ° C. overnight).
Drying was performed. Protective removal of allyl ester is performed repeatedly before deprotection of Boc
Was. For removal of Boc protection, ester-IA was used except for one deprotection.
M. The above conditions were used for the preparation of the PE-C10 / C3 surface. FT
Based on IR microscopic analysis, quantitative deprotection could occur. Ester-IAM. P
The bound ligand density on SP is determined by the carbon content before and after binding of the ligand to the silica surface.
Amount, based on elemental analysis of C, 2.13% before bonding, C, 6.99% after bonding
104.86 μmol / g.

The IAM. SM ^{C10 / C3} (11) SM ^{ω-COOH / P} (2) Immobilizing the end capping of C10 and C3, The conjugation procedure for preparing ^{SMP / C10 / C3} is similar to that of ester IAM. The conjugation procedure used for the preparation of ^{PGP / C10 / C3} was unchanged. The procedure for removing the TBDPS protecting group attached to the SM head group is performed as described. Briefly explaining this, IAM. Up to 2.1 g of SMP ^{/ C10 / C3} was added to TBAF (20 ml, 1.0 ml).
MTHF solution) was suspended in THF (20 ml) prior to addition to the suspension. After 2 hours of stirring, the silica was filtered and THF (2 × 20 ml), CHCl ₃ (3 ×
(20 ml), CH ₃ OH (3 × 20 ml) and acetone (20 ml), followed by vacuum drying at 40 ° C. overnight. FTIR microscopic analysis showed that deprotection was quantitative. The reason is 3074.
This was due to the complete disappearance of the TBDPS IR peaks at 23 / cm and 3056.10 / cm. IAM. Binding ligand density on SM ^P (2) is based on the carbon content of elemental analysis of which is shown before and after ligand binding to a silica surface (2), before binding C, 2.04%, after binding the C 90.53 from the result of 7.35%
μmol / g silica.

Synthesis and Immobilization of Ceramide Analogues on Silica Particles Skin is the largest organ in the human body, its main body is in contact with the environment, and the pathway through which many chemicals are taken into the body. Is also equivalent. Research has shown that drug delivery through the skin is extremely easy for many simple yet powerful drug molecules of 1000 Da and below for delivery of drugs through the skin. Seven drugs marketed in the United States, including clonidine, estradiol, fuentanyl, nicotine, nitroglycerin, scopolamine and testosterone, are signals of skin penetration. Steady-state and stable plasma concentrations of the drug are ensured via this skin-penetrating drug route, and the high peak levels associated with oral treatment are also avoided. Avoiding this high peak level can reduce the side effects of some drugs.
In any case, while exposure of the skin to toxic substances poses a major occupational hazard, a successful risk assessment can also reduce chronic illness and environmental problems.
Interest in the skin should facilitate research in this area and lead to an understanding of the biological mechanisms of the skin. According to the research results, the permeability that is the main component of human skin depends on the outer layer, the stratum corneum (SC) of the epidermis. The stratum corneum is made up of dead cells surrounded by lipid-containing extracellular tissues. The major lipid in this extracellular tissue is ceramide, which is composed of 50% total lipids and six heterogeneous ceramides. Ceramide 2 accounts for 40% of the total ceramide. Ceramide also binds to sphingosine and dihydrosphingosine, mainly 24 to 28.
Made of carbon fatty acid amide.

There is an ongoing search for models that can predict plasma levels, with the goal being to facilitate the selection of transdermal drugs or to predict exposure to toxic chemicals. By immobilizing the skin lipid of ceramide 2 on a chromatographic carrier, a model for predicting the chemical action of the skin can be provided. This model system would be much better than an octanol / water partitioning system using HPLC. Ceramide ligand (N-

[13-Carboxyltridecanoyl] -D-erythrosphingosine 2) is used to obtain a ceramide silica surface, and has a basic ceramide 2 semi-function and an ω free carboxyl group. This carboxyl group is 1,
It is believed that via the 1'-carbonyldiimidazole (CDI) activity function, it serves as a terminal to the silica propylamine particles. An anhydride is used to endcap excess amine groups on the silica propylamine particles. Scheme 5
)reference.

[0095]

Embedded image

Scheme 6 (Synthesis 6) Synthesis route of stationary phase of silica-fixed ceramide

Embedded image

Scheme 6 (Scheme 6) outlines a typical synthetic route for the constant face of ceramide. Treatment of N-Boc-L-serine with methyl iodide gives N-Boc-L-
Serine methyl ester 3 is obtained, which is reacted with 2,2-dimethoxypropane using p-toluenesulfonic acid monohydrate as a catalyst to give acetal protected compound 4, which is (diisobutylaluminum hydride). (Reacted at −78 ° C. using (DIBALH) as a reactant) and reduced to aldehyde 5 under mild reduction conditions. Meanwhile, 1-pentadecine 6 is synthesized from 1 bromotridecane and lithium acetylide ethylenediamine complex in dimethyl sulfoxide (DMSO) solution. Here, the basic structure of sphingocin 7 is obtained by flash chromatography with oxazolidine aldehyde 5 and 1-pentadecine 6 and lithium acetylide from n-butyllithium at −23 ° C. in a yield of 84%. In this reaction, high erythroselectivity of diastereoselectivity (ds) -89% is observed. Compound 7 was treated with mild acid Amberlys in methanol at room temperature.
Treatment at t15 results in selective cleavage of the acetal protecting group to give 1,3-diol 8, which is converted to N-Boc-sphingosine 9 by the selective reduction of Red-Al and the carbon triple bond in ether. Was. Under these reducing conditions, the triple bond was converted to a trans double bond. Cleavage of half of the carbamate by 1.0N HCl in dioxane at 100 ° C. yields D-erythro-sphingosine 10, which protects the tert-butyl diphenylsyl chloride (TBDPSCL) two hydroxyl groups. The TBDPS group is tertiary to acid hydrolysis
It is about 100 times more stable than the t-butyldimethylsilyl (TBDMS) group and is therefore suitable for the immobilization reaction. TBDPS protected sphingosine 11 and dodecanedicarboxylic anhydride 12 (diacid and dicyclohexylcarbodiimide (D
CC) and subsequently reacted with 4-dimethylaminopyridine (DM) by flash chromatography.
AP) N- [13-Carboxyltridecanoyl] -D-erythro-sphingosine 2 was obtained under a catalyst in a yield of 87%. The ω-carboxyl group of ceramide homolog 2 was activated with CDI in chloroform. Following this activation,
Silica propylamine (SPA) while shaking ceramide-imidazolide for 24 hours
) Bound to particles. After work inspection, a ceramide-based silica stationary phase was obtained, and C10 and C3 were end-capped for residual amine on the silica particle surface. The TBDPS protecting group is tetrabutylammonium fluoride (TBAF) in THF.
) And removed. The course of the reaction was monitored using a Nicolet Magna 550 FT-IR with a Spectratech IR-Plan I microscope.
After protection removal, a monolayer of ceramide lipid membrane was formed and the same two-sided group was retained by endogenous ceramide lipid. Bond density was calculated from elemental analysis. Finally, these silica particles were packed in a stainless steel column and used for high performance liquid chromatography. Our preliminary data show that ceramide-based stationary phases are suitable for predicting the skin permeability constant of various chemicals. The coefficient of performance (k ') of the solute measured on the column was in good agreement with the skin permeability coefficient (kp) or the transdermal absorption rate (% Abs) of the excised human or subject.

Correlation between Oral Absorption and Drug Affinity Table 6 shows eight compounds for inducing low oral absorption, which are used as a training set for low MAF ^μ calculation, which represent low drug absorption. Low MA in (last column of Table 6)
Small average D ² value corresponding to F ^mu, namely 3.5 ± 1.24 indicates that the compound group of the training set is statistically sufficiently set. 1 showing complete oral absorption
For three compounds, ^{D 2} value based on the low MAF ^mu is very high as shown in Table 7.
The large average D2, 96.26 ± 96.01, found in the last column of Table 7, indicates that compounds with high absorption have statistically different membrane binding properties than compounds with low oral absorption. Is shown.

Table 6: Training set showing low rates of oral absorption

[Table 6]

Table Note: The above 8 compounds were used to determine the low MAF ^μ vector / (1) This value was calculated for the low MAF ^μ vector.

The test compounds in Table 7 show separate D ² values compared to the low MAF ^μ vector,
The compounds in the table were used as the expected training set resulting in a high MAF ^μ vector.

[0102]

[Table 7]

[0103] Table 8 shows the high MAF μ ^(h igh MAF ^μ) related ^{D 2} values for each compound in the training set of high drug absorption. 8 compounds exhibiting low oral absorption D ² value based on the high MAF ^mu respect (Table 9) is shown higher (Note in the table pursuant to Table 6)
.

Table 8: Training set representing high oral absorption

[Table 8]

Note to the table: * The above 8 compounds were used to determine the high MAF ^μ vector / ** D ² values were calculated for the high and low MAF ^μ vector.

[0106]

[Table 9]

Data Acquisition Using Liquid Chromatography / Mass Spectroscopy (LCMS) —and General Experimental Records A mass spectrometer (MS) can be used as a compound identification and detection device eluting from a liquid chromatograph (LC) device. . The only requirement for detection is:
The point is that the compound is ionized even if it shows no composition. Therefore, one of the preferable methods for measuring the membrane binding constant is to use MS as a device for detecting a compound eluted from an LC column. The general advantages of using an MS detector are: (1) compounds that do not have a chromophore can be easily detected; (2) multiple compounds can be uniquely identified by a single injection operation; (3) binding constants Requires only a small amount of compound, in which case these constants are measured and displayed on immobilized artificial membranes (IAMs) or another solid-phase chromatographic support. These three advantages allow for the simultaneous injection of very small amounts of multiple compounds for the purpose of detecting any eluted compounds from the instrument column. This procedure allows one to directly compare the binding affinities of large numbers of test compounds with multiple compounds that incorporate one or more training sets, thereby resulting in unique data variability resulting from the diversity of the columns. And eliminate the variability of other experimental factors. An example of obtaining data using both LC and MS will be described below.

Test Mixture A mixture containing 0.100 μmol of 155 compounds was prepared. At least some of the compounds have known clinical significance, and the compounds serve as an internal standard for measuring the binding affinity of a test compound. 6.0 m of the mixture
l, dissolved in 15% acetonitrile and 30 mM 85% ammonium acetate, 20 μL of this test mixture, pH 7.4, is loaded onto a 3 × 0.46 cm IAM column. This corresponds to a １６16 μ load of the test mixture and 〜320 pmol of each compound on the column.

LC Conditions A 3 × 0.46 cm IAM column was used. The mobile phase is 15% acetonitrile and 85%, 30 mM ammonium acetate, pH 7.4. The flow rate was 1.0 ml / min and the flow path splitter was not used. The operation time was set to 2 hours, followed by washing with 100% acetonitrile for 10 minutes.

Conditions for MS Bruker Esquire-LC was used. Each operation was performed twice. In the first operation, no automatic MS / MS was used. In the second operation, the automatic MS /
MS was used. With this automatic MS / MS, compounds having the same molecular weight can be identified.

A standard MS instrument can handle a flow rate of 2 ml / min, at least one of which has a washing mechanism for salts from the inlet of the ion source. As a standard, the mobile phase cannot contain salts, because the evaporation of the mobile phase in the ion source causes excessive accumulation of salts and impedes the flow of ions to the ion source.
Finnigan has recently developed an LC / MS instrument in which salts are washed away during chromatographic operations. Affinity data is collected and output as an output from the MS detector, electronically stored in a storage device, and further processed and used, usually using computer data which can be accessed and communicated with a data storage device, and programmed using a vector analysis algorithm. The processed, pre-programmed format provides the user with readable output, making it easier to compare the binding affinity of the test compound with the affinity of the control compound, or the average vector quantity for one or more training sets.

[Brief description of the drawings]

FIG. 1 is a three-dimensional vector plot of membrane affinity data for compounds known to act at serotonin receptors.

FIG. 2 is a three-dimensional vector plot of membrane affinity data for compounds known to act at dopamine receptors.

[Procedural Amendment] Submission of translation of Article 34 Amendment of the Patent Cooperation Treaty

[Submission date] February 22, 2000 (2000.2.22)

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name] Claims

[Correction method] Change

[Correction contents]

[Claims]

2. The method of claim 1, wherein at least a portion of said numerical value is calculated.

3. The method according to claim 1, wherein a numerical value relating to the interaction of the compound with each of the membrane pseudo-surfaces is determined by a chromatographic apparatus using a mobile phase and a stationary phase provided with the membrane pseudo-surfaces.

4. comprises a head group and at least one is generated in the biofilm in a natural state phospholipid compound of the film pseudo-surface The method of claim 1.

Each value in the ordered set of numbers that the wherein each of the compounds corresponds to the retention time of the compound of the chromatographic the apparatus using a predetermined stationary phase method according to claim 3.

6. A method according to claim 1, wherein each value in the ordered set of values for each compound is
4. The method of claim 3 , wherein the method corresponds to the peak width of a compound in the chromatographic apparatus using a predetermined stationary phase.

7. liposome said film pseudo-surface, Rangumiyua - selected from Blodgett film and the fixed artificial membrane, the method according to claim 1.

8. The method of claim 1, wherein the values for T and C for each membrane surface are respectively normalized to a common reference for the membrane surfaces.

Claims9The best control compound has the formula cosΘ = (T₁C₁+ T₂C₂ +. . . T_nC_n) / (T₁ ²+ T₂ ²+. . . T_n ²)^1/2(C₁ ²+ C₂ ² +. . . C_n ²)^1/2Corresponding to a compound in which the angle Θ is less than about 20 °
The method of claim 1.

Wherein said angle Θ is less than about 15 °, The method of claim 9.

Wherein said angle Θ is less than about 10 °, The method of claim 9.

12. The method according to claim 3 , wherein each membrane pseudo-surface is a fixed artificial membrane.

At least one of 13. A film pseudo-surface comprises a mixture of lipid compounds,
The method of claim 1.

14. The species and stoichiometry of each lipid compound in said mixture,
Those are raw Ji a predetermined biofilm species The method of claim 13.

15. quantified with two or more films pseudo surfaces each face has its own composition, the interaction between the test and control compounds with each of the film pseudo-surface,
An expectant organism comprising: means for assigning a numerical value indicative of the quantified interaction of the compound to each membrane mimic surface; and a programmable computer for comparing the numerical value of the test compound with the numerical value of the control compound. Screening device for test compounds for chemical characteristics.

16. The test device of claim 15 , further comprising a printer, display, or other means for reporting the control compound with the value that most closely matches the test compound value.

17. A method according to claim 17 , further comprising a database comprising numerical values indicating the interaction of each membrane pseudo-surface with the selected control compound, wherein at least a portion of the selected control compound has predetermined biological properties. The test apparatus according to claim 15 ,

18. The apparatus for determining is a chromatographic device, the film pseudo-surface is a stationary phase for the chromatographic apparatus, the test apparatus according to claim 15.

19. Choose at least one training compositions each face has one or more control compounds selected two or more film pseudo-surface provided with a normal biological properties with unique composition, the test Preparing a test mixture by combining the compound with the training composition;
Contacting at least a part of the test mixture with each membrane pseudo-surface, defining a numerical value indicating the interaction between the test compound and the control compound and the respective membrane pseudo-surface, comparing the numerical values, Identifying a test compound having a value that best matches the value of the control compound.

20. The method of claim 19 , wherein said contacting said test mixture with said membrane simulated surface is performed on a chromatographic apparatus wherein each membrane simulated surface is the stationary phase therein.

21. The method of claim 20 , wherein said chromatographic device is a liquid chromatograph device utilizing a mass spectrometric detection device.

22. comprising the comparison step of the numeric calculates the average vector for the reference compound in the training composition, method of claim 20.

23. A training set composition for use in screening test compounds for biological properties in vitro, comprising at least two compounds having a common biological activity, biological function or clinical efficacy.

24. The compound having a common biological activity is present in a substantially equimolar ratio in the mixture, training composition according to claim 23.

25. Each of the compounds in the training composition exhibits an affinity for a plurality of membrane pseudo-planes, and an elliptic plot of the membrane pseudo-plane binding data for the compound in the training composition shows that all compounds have 24. The training composition of claim 23 , wherein the training composition is in the 0.95 quartile.

26. including three or more control compounds, training composition according to claim 23.

27. including three or more control compounds, training composition according to claim 23.

28. including three or more control compounds, training composition according to claim 25.

29. A phospholipid compound represented by the formula _HOOC-W-OPO 2 OZ, wherein Z is protected glyceryl, 2- (protected amino ethyl), 2-protected carboxyl-2-aminoethyl, 2 - protected carboxyl-2-protected amino ethyl or easily torn easily ester protecting group, and W comprises the lipid residue of both amphiphilic compound, and wherein the above, the phospholipid compound.

30. The compound according to claim 29 , wherein said amphiphilic compound is an amphiphilic compound found in natural biofilms.

31. Lecithin said amphiphilic compound, lysolecithin, cephalin, sphingomyelin Miyu phosphorus, cardiolipin, glycolipids, selected from the ganglio Cid and Sereburoshiddo A compound according to claim 30.

32. A process for producing a protected phospholipid represented by the formula HOOC-W-OPO ₂ OZ, comprising a starting compound of the formula HOOC-W-OPO ₂ OCH ₂ CH ₂ N ⁺ (CH ₃ ) ₃ , Contacting phospholipase D (PLD) with an excess of a protected alcohol of formula ZOH, wherein Z is protected glyceryl, 2- (protected amino) ethyl, 2-protected carboxyl-2-aminoethyl or acid protected Group, and W is the lipid residue as a biologically significant amphipathic compound, where W is the protected phospholipid wherein the starting compound serves as a substrate for phospholipase D activity Manufacturing method.

33. Using the alcohol from about 3-fold to about 15 fold excess of stoichiometry, carried out in an aqueous organic solvent buffer in the presence of water soluble calcium salt the reaction method of claim 32, .

34. A stationary phase for use in a chromatographic apparatus, comprising N- (13-carboxyltridecanoyl) -D-erythro-sphingosine covalently bound to a solid support via a 13-carboxyl group.

35. The solid support is silica propylamine, stationary phase of claim 3 4.

──────────────────────────────────────────────────続 き Continuation of front page (81) Designated country EP (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, IT, LU, MC, NL, PT, SE ), OA (BF, BJ, CF, CG, CI, CM, GA, GN, GW, ML, MR, NE, SN, TD, TG), AP (GH, GM, KE, LS, MW, SD, SZ, UG, ZW), EA (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM), AL, AM, AT, AU, AZ, BA, BB, BG, BR, BY, CA, CH, CN, CU, CZ, DE, DK, EE, ES, FI, GB, GE, GH, GM, HR, HU, ID, IL, IS, JP, KE, KG, KP , KR, KZ, LC, LK, LR, LS, LT, LU, LV, MD, MG, MK, MN, MW, MX, NO, NZ, PL, PT, RO, RU, SD, SE, SG, SI, SK, SL, TJ, TM, TR, TT, UA, UG, US, UZ, VN, YU, ZW Apartment K 218 (72) Inventor Hauer, Kimberley United States, California 91367 Woodland Hills Candle Lulight Drive 2881 (72) Inventor Inn, Jamin 91367, Woodland Hills Grade Avenue 6301 Apartment K 324, United States of America (72) Inventor Tsai, Soon, J. United States of Americaー Jersey 08832 ・ Kesby Sunnyview Oval 810 F-term (reference) 2G045 AA40 BB02 BB52 FB05 FB06 FB15 JA01 JA04 JA07 4B063 QA18 QA19 QQ61 QQ70 QR41 QR45 QR84 QR84 QS02 QS17 QS39 QX01

Claims (37)

[Claims] 1. A method comprising: identifying two or more membrane pseudo-surfaces, each surface having a unique composition; providing a set of control compounds wherein each control compound has known biological properties; , Define an ordered set of numerical values that are related to the interaction of each membrane pseudo-surface with each control compound, whereby the ordered set of numerical values is represented by the formulas {C ₁ , C ₂ . . . C _n ｝, where n is the number of membrane faces, and an ordered set of numerical values for each test compound interacting with each membrane pseudo-face 面T ₁ , T ₂ . . . Defining T _n } and comparing the set of values for each test compound with the respective set value of the control compound to best match each value in the ordered set of test compounds. Identifying the biological properties of a control compound comprising an ordered set of numerical values. 2. The method of claim 1, wherein said numerical value is defined experimentally. 3. The method of claim 1, wherein at least a portion of said numerical value is calculated. 4. The method according to claim 1, wherein a numerical value relating to the interaction of the compound with each of the membrane pseudo-surfaces is determined by a chromatographic apparatus using a mobile phase and a stationary phase having the film pseudo-surface. 5. The method of claim 1, wherein at least one of the membrane mimetic surfaces comprises a head group of a phospholipid compound that naturally occurs in a biological membrane. 6. The method of claim 4, wherein each value in the ordered set of values for each compound corresponds to a retention time of the compound in the chromatograph using a predetermined stationary phase. 7. Each value in the ordered set of values of each compound is
5. The method of claim 4, wherein the method corresponds to the peak width of a compound in the chromatographic apparatus using a predetermined stationary phase. 8. The method according to claim 1, wherein the membrane simulated surface is selected from liposomes, Langmuir-Blodgett membranes and immobilized artificial membranes. 9. The method of claim 1, wherein the values for T and C for each membrane surface are respectively normalized to a common reference for the membrane surfaces. 10. The said best control compound is of the formula cosΘ = (T₁C₁+ T₂C ₂ +. . . T_nC_n) / (T₁ ²+ T₂ ²+. . . T_n ²)^1/2(C₁ ²+ C ₂ ² +. . . C_n ²)^1/2Corresponds to compounds in which the angle Θ is less than about 20 °
The method of claim 1. 11. The method of claim 10, wherein said angle Θ is less than about 15 °. 12. The method of claim 10, wherein said angle Θ is less than about 10 °. 13. The method of claim 4, wherein each membrane pseudo-surface is an immobilized artificial membrane. 14. At least one of the membrane mimic surfaces comprises a mixture of lipid compounds.
The method of claim 1. 15. The species and stoichiometry of each lipid compound in said mixture,
15. The method of claim 14, wherein the method corresponds to the species and stoichiometry generated in a given biofilm species. 16. Quantifying two or more membrane simulated surfaces, each surface having a unique composition, and the interaction of a test compound and a control compound with each of said membrane simulated surfaces;
An expectant organism comprising: means for assigning a numerical value indicative of the quantified interaction of the compound to each membrane mimic surface; and a programmable computer for comparing the numerical value of the test compound with the numerical value of the control compound. Screening device for test compounds for chemical characteristics. 17. The test device of claim 16, further comprising a printer, display, or other means for reporting the control compound with a value that most closely matches the test compound value. 18. The method of claim 18, further comprising a database comprising numerical values indicating the interaction of each membrane pseudo-surface with the selected control compound, wherein at least a portion of the selected control compound has predetermined biological properties. 17. The test apparatus according to item 16. 19. The test device according to claim 16, wherein the quantitative device is a chromatographic device, and the membrane pseudo-surface is a stationary phase for the chromatographic device. 20. selecting at least one training composition with one or more control compounds having normal biological properties, selecting two or more membrane pseudo-faces, each face having a unique composition; Preparing a test mixture by combining a compound and the training composition; contacting at least a portion of the test mixture with each of the membrane simulated surfaces; and allowing the test compound and the control compound to interact with the respective membrane simulated surfaces. A method of screening a test compound for a biological property, comprising defining a numerical value indicative of an effect and comparing said numerical values to identify a test compound having a numerical value that best matches said control compound numerical value. 21. The method of claim 20, wherein the step of contacting the test mixture with the membrane simulated surface is performed on a chromatographic apparatus, wherein each membrane simulated surface is the stationary phase therein. 22. The method of claim 21, wherein said chromatographic device is a liquid chromatograph device utilizing a mass spectrometric detection device. 23. The method of claim 21, wherein said comparing said numerical values comprises calculating an average vector for said control compound of each training composition. 24. A training set composition for use in screening test compounds for biological properties in vitro, comprising at least two compounds having a common biological activity, biological function or clinical efficacy. 25. The training composition according to claim 24, wherein said compounds having a common biological activity are present in said mixture in substantially equimolar ratio. 26. Each of the compounds in the training composition exhibits an affinity for a plurality of membrane pseudo-planes, and an elliptic plot of the membrane pseudo-plane binding data for the compound in the training composition shows that all compounds have 25. The training composition of claim 24, wherein the training composition is in the 0.95 quartile. 27. The training composition according to claim 24, comprising three or more control compounds. 28. The training composition according to claim 25, comprising three or more control compounds. 29. The training composition according to claim 26, comprising three or more control compounds. 30. A phospholipid compound represented by the formula HOOC-W-OPO ₂ OZ, wherein Z is protected glyceryl, 2- (protected aminoethyl), 2-protected carboxyl-2-aminoethyl, The phospholipid compound as described above, which comprises -protected carboxyl-2-protected aminoethyl or an easily cleavable ester protecting group, and W is a lipid residue of a biologically significant amphiphilic compound. 31. The compound of claim 30, wherein said biologically significant compound is an amphipathic compound found in natural biofilms. 32. The compound of claim 31, wherein said amphiphilic compound is selected from lecithin, lysolecithin, cephalin, sphingomyelin, cardiolipin, glycolipid, ganglioside, and cerebroside. 33. The amphiphilic compound is a phospholipid represented by the general formula W-OPO ₂ OB, wherein W is acylglyceryl, diacylglyceryl, or N-acyl3-O- (protected) sphingosine-1. -yl, also the acyl is _C 8 _{-C 24} alkanoyl or _C 8 _{-C 24} alkenoyl, a compound according to claim 31. 34. A process for producing a protected phospholipid represented by the formula HOOC-W-OPO ₂ OZ, comprising: a starting compound of the formula HOOC-W-OPO ₂ OCH ₂ CH ₂ N ⁺ (CH ₃ ) ₃ ; Contacting phospholipase D (PLD) with an excess of a protected alcohol of formula ZOH, wherein Z is protected glyceryl, 2- (protected amino) ethyl, 2-protected carboxyl-2-aminoethyl or acid protected Group, and W is the lipid residue as a biologically significant amphipathic compound, where W is the protected phospholipid wherein the starting compound serves as a substrate for phospholipase D activity Manufacturing method. 35. The method of claim 34, wherein said alcohol is used in about 3 to about 15 fold stoichiometric excess and said reaction is carried out in an aqueous organic solvent buffer in the presence of a water-soluble calcium salt. . 36. A stationary phase for use in a chromatographic apparatus, comprising N- (13-carboxyltridecanoyl) -D-erythro-sphingosine covalently linked to a solid support via a 13-carboxyl group. 37. The solid support of claim 3, wherein the solid support is silica propylamine.
7. The stationary phase according to 6.