FIELD OF THE INVENTION

The present invention relates generally to techniques for detecting a solute of interest in a fluid sample and more particularly to on-line techniques for detecting a solute of interest in an effluent sample produced by a differential migration separation technique.

BACKGROUND OF THE INVENTION Differential migration separation techniques are well-known in the biological and chemical arts and are commonly used to separate a mixture of solutes present in a fluid sample. One example of a differential migration separation technique is chromatography. Chromatography typically involves the separation of solutes according to their differential partitioning between two or three phases, the phases frequently being solid and liquid phases. Solute partitioning results from the differing mobilities of the various solutes in the sample through a matrix of solid particles in the presence of a flowing liquid phase, solute transfer through the solid matrix typically being driven by a pressure gradient. Based on the specific type of solid matrix used, chromatography can be used to separate solutes by any number of characteristics, including size, electrical charge, hydrophobicity, hydrophilicity and/or a specific affinity for the matrix or a ligand bound thereon.

Another example of a differential migration separation technique is electrophoresis. Electrophoresis typically involves the separation of solutes based on their differential electrophoretic mobility through a matrix. Solute transfer in such a system is driven by a voltage gradient from an applied electric field. As can readily be appreciated, because differential migration separation techniques exploit the difference in speeds at which various solutes migrate through a particular type of matrix, the product of a differential migration separation technique is typically a continuous effluent stream exiting the matrix within which the various solutes of the mixture are spatially separated from one another. Determination of the location of the various solutes, or of a particular solute of interest, within the effluent stream has typically been achieved by collecting the

effluent stream as a series of discrete fractions and then sampling the fractions by any of a number of means known in the art to identify their contents. Examples of techniques known in the art for detecting various analytes in solution include U.S. Patent No. 4,281 ,061 , inventors Zuk et al., which issued July 28, 1981; U.S. Patent No. 4,207,075, inventor Liburdy, which issued June 10, 1980; U.S. Patent No. 3,998,943, inventor Ullman, which issued December 21 , 1976; European Patent Application No.290,269, published November 9, 1988; European Patent Application No. 242,527, published October 28, 1987; Huang et al., "Direct and Homogeneous Immunoassay for IgG Analyses," Biotechnology and Bioenqineering. Vol. 40, pp. 913-8 (1992); Haidukewych, "Therapeutic Drug Monitoring By Automated Fluorescence Polarization Immunoassay," Immunoassay Technology, Vol. 2, pp. 71-103 (1986); and Li et al., "Automated Fluorometer/Photometer System for Homogeneous Immunoassays," Clinical Chemistrv. Vol. 29, No. 9, pp. 1628-34 (1983), all of which are herein incorporated by reference. One problem with the above-described method of solute detection is that, typically, the entire effluent stream exiting the matrix (or a large portion thereof) must be collected in the form of a series of collection samples, and, then, each collection sample (or a sizeable subset thereof) must be individually tested for the presence of the solute. As can readily be appreciated, such a testing procedure is frequently time-consuming, tedious and expensive. Moreover, it can readily be appreciated that the aforementioned process results in the sample no longer existing as an effluent sample since, for testing purposes, the effluent sample must be converted into a series of discrete collection samples. In addition, because each collection sample typically contains the solutes exiting the matrix over a fairly broad period of time, the above-described procedure provides only limited information regarding the precise distribution of the desired solute within the effluent stream. Such information as to the precise distribution of the solute within the effluent stream is often useful in determining whether any parameters of the differential migration separation technique should be modified. One approach that has been taken in response to problems of the type described above is described in U.S. Patent No. 5,234,586, inventors Afeyan et al.,

which issued August 10, 1993, and which is incorporated herein by reference. The aforementioned Afevan et al. patent relates to an on-line method for rapidly identifying the presence and location of a preselected solute or subset of solutes in an effluent stream and comprises first passing a mixture through a system capable of separating the solutes in the mixture so that they are temporally and spatially separated from one another to some degree as they exit the system in a fluid phase (effluent stream). The first system may be a liquid chromatography matrix, e.g., a column, or other means for separating solutes, such as an electrophoresis module. The effluent stream from this first solute separation system is then passed through a UV absorbance detector to produce a first output depicting the elution profile of all UV absorbing solutes (e.g., proteins, nucleic acids, etc.) exiting the column. Identification of a particular UV absorbing solute of interest within the many UV absorbing solutes is then determined by passing at least a portion of the effluent stream through a second system capable of selectively removing the solute of interest from the fluid phase, preferably using immunoadsorbents and immunoaffinity matrices. The effluent stream exiting the second system is then passed through the same or a similar UV detector to produce a second output depicting the elution profile of all UV absorbing solutes exiting the second system. Because the second system selectively removes the solute of interest without altering the temporal and/or spatial arrangement of the other solutes in the effluent stream, the difference between the first and second outputs can be used to determine the presence and location of the solute of interest in the effluent stream.

According to one embodiment disclosed in the Afevan et al. patent, IgG is the solute of interest and the second system contains a perfusive Protein A affinity matrix.

The present inventor has identified several shortcomings with the Afevan et aL technique. One such shortcoming is that the second system, which is used to remove the solute of interest from the effluent stream, can quickly become saturated with the solute of interest and, therefore, has to be periodically washed or replaced to prevent the solute from going undetected. A second shortcoming

with the Afevan et al. technique is that the range of solutes capable of being removed by the second system may be limited by the type of affinity sorbents capable of being immobilized on a matrix. Moreover, even in those instances in which immobilization of the affinity sorbent can be performed, the costs of immobilization can be high, particularly for some affinity sorbents like Protein A and Protein G. A third shortcoming with the Afevan et al. technique is that the identification of a difference between the first and second outputs may be difficult to make in those instances in which one or more solutes co-elute from the column with the solute in question. A fourth shortcoming with the Afevan et al. technique is that UV absorbance is incapable of detecting quantities of biomolecules less than about 0.1 mg/ml, with even larger quantities being required in the presence of interfering solutes. A fifth shortcoming with the Afevan et al. technique is that the affinity matrix used therein might cause some interaction with other solutes present in the effluent sample and thus change the spatial distribution of solutes exiting the affinity matrix, thereby causing some difficulty in identifying the solute of interest.

SUMMARY OF THE INVENTION

It is an object of the present invention to provide a novel technique for detecting a solute of interest in an effluent sample.

It is another object of the present invention to provide a technique as described above that is an on-line detection technique.

It is still another object of the present invention to provide a technique as described above that overcomes at least some of the problems described above in connection with existing techniques.

Additional objects, as well as features and advantages, of the present invention will be set forth in part in the detailed description which follows, and in part will be obvious from the detailed description or may be learned by practice of the invention. In the description, reference is made to the accompanying drawings which form a part thereof and in which is shown by way of illustration specific embodiments for practicing the invention. These embodiments will be described in sufficient detail to enable those skilled in the art to practice the invention, and it is to be understood that other embodiments may be utilized and that changes may be made without departing from the scope of the invention. The following detailed description is, therefore, not to be taken in a limiting sense, and the scope of the present invention is best defined by the appended claims. According to one aspect of the invention, there is provided a method for detecting the presence of a solute of interest in an effluent sample, said method comprising the steps of: (a) adding to the effluent sample a fluorescent agent which preferentially binds to the solute of interest and whose fluorescence is different when bound to the solute than when not bound to the solute; (b) exciting the effluent sample in such a way as to cause the fluorescent agent to fluoresce in accordance with whether or not the fluorescent agent is bound to the solute; and (c) detecting the fluorescence emitted from the effluent sample, whereby detected fluorescence consistent with the binding of the fluorescent agent to the solute of interest is indicative of the presence of the solute of interest in the effluent sample. The effluent sample subjected to the above-described method may be produced by a differential migration separation technique, such as chromatography

or electrophoresis. The fluorescent agent added to the effluent sample preferably comprises a conjugate of a moiety having an affinity for the solute of interest and a fluorescent moiety that does not give a large positive response to bovine serum albumin (BSA). Most fluorescein dyes give a manageable positive response to BSA and, therefore, are suitable as the fluorescent moiety. Moreover, to further reduce any positive response to the fluorescent agent by BSA, free fluorescein dye may be added to the sample, together with the fluorescent agent. Where the solute of interest is immunoglobulins, such as IgG, the fluorescent agent is preferably a conjugate of (i) a fluorescein dye and Protein A, (ii) a fluorescein dye and Protein G or (iii) a mixture thereof, such as a 1:1 mixture. Where the solute of interest is antithrombin III, 8-hydroxypyrene-1 ,3,6-trisulfonic acid trisodium salt (HPTS) may be used, by itself, as the fluorescent agent or may be conjugated with heparin to form the fluorescent agent.

According to another aspect of the invention, there is provided a method for determining the distribution of a solute of interest within a length of a fluid stream, the fluid stream preferably being made by a differential migration separation technique, such as chromatography or electrophoresis, said method comprising the steps of: (a) adding throughout a length of the fluid stream a fluorescent agent which preferentially binds to the solute of interest and whose fluorescence is different when bound to the solute than when not bound to the solute; (b) exciting the length of the fluid stream in such a way as to cause the fluorescent agent to indicate by its fluorescence whether or not the fluorescent agent is bound to the solute; (c) detecting the fluorescence emitted from the length of the fluid stream as a function of location within the length of the fluid stream; and (d) using the results of said detecting step to ascertain the distribution of the solute of interest within the length of the fluid stream.

According to yet another aspect of the invention, there is provided a method of determining the presence and location of a solute of interest within a fluid stream, said method comprising the steps of: (a) splitting an effluent sample, the effluent sample preferably generated by a differential migration separation technique, such as chromatography or electrophoresis, into a first fluid stream and

a second fluid stream, said first and second fluid streams having a proportionate chemical composition; (b) determining the presence and location of a solute of interest within the first fluid stream using fluorescence spectroscopy; (c) determining the time-correlation between the first and second fluid streams; and (d) using the results of said solute presence and location determining step and said time-correlation determining step to determine the presence and location of the solute of interest in the second fluid stream.

As can readily be appreciated, once the location of the solute of interest in the second fluid stream has been determined, one may collect only that portion of the second fluid stream containing the solute of interest.

According to still another aspect ofthe invention, there is provided a method of determining the distribution of a solute of interest in a fluid stream, said method comprising the steps of: (a) splitting an effluent sample, the effluent sample preferably generated by a differential migration separation technique, such as chromatography or electrophoresis, into a first fluid stream and a second fluid stream, said first and second fluid streams having a proportionate chemical composition; (b) determining the distribution of a solute of interest within the first fluid stream using fluorescence spectroscopy; (c) determining the time-correlation between the first and second fluid streams; and (d) using the results of said distribution determining step and said time-correlation determining step to determine the distribution of the solute of interest in the second fluid stream.

As can readily be appreciated, once the distribution of the solute of interest in the second fluid stream has been determined, one may collect only that portion of the second fluid stream containing the greatest concentration of the solute of interest. Moreover, where the effluent sample is a product of a differential migration separation technique, the distribution of the solute in the second fluid stream illustrates the efficacy ofthe separation technique employed. Consequently, the aforementioned method provides an on-line tool for rapidly assessing whether any parameters of a separation technique being employed should be altered. According to still yet another aspect of the invention, there is provided a method of determining the concentration of a solute of interest at a location within

a fluid stream, said method comprising the steps of: (a) splitting an effluent sample, the effluent sample preferably generated by a differential migration separation technique, such as chromatography or electrophoresis, into a first fluid stream and a second fluid stream, said first and second fluid streams having a proportionate chemical composition; (b) determining the concentration of a solute of interest at a location within the first fluid stream using fluorescence spectroscopy; (c) determining the time-correlation between the first and second fluid streams; (d) determining the volume-correlation between the first and second fluid streams; and (e) using the results of said concentration determining step, said time-correlation determining step and said volume-correlation determining step to determine the concentration of the solute of interest at a corresponding location in the second fluid stream.

According to a further aspect of the invention, there is provided a system for monitoring a fluid stream for a solute of interest, said system comprising: (a) means for splitting an effluent sample into a first fluid stream and a second fluid stream, said first and second fluid streams having a proportionate chemical composition; (b) a fluorescent agent which preferentially binds to the solute of interest and whose fluorescence is different when bound to the solute of interest than when not bound to the solute of interest; (c) means for mixing the fluorescent agent into the first fluid stream; (d) means for monitoring the first fluid stream for the solute of interest based upon fluorescence of the fluorescent agent; and (e) means for time- correlating the first and second fluid streams so that the results obtained for the first fluid stream can be applied to the second fluid stream.

According to still a further aspect of the invention, there is provided a method of purifying a solute of interest from a mixture of solutes, said method comprising the steps of: (a) subjecting the mixture of solutes to an appropriate differential migration separation procedure which results in the production of an effluent sample in which the solute of interest is separated to some extent from the remaining solutes; (b) splitting the effluent sample into a first fluid stream and a second fluid stream, said first and second fluid streams having a proportionate chemical composition; (c) determining the location of a concentration peak of the

solute of interest within the first fluid stream using fluorescence spectroscopy; (d) determining the time-correlation between the first and second fluid streams; (e) using the results of said concentration peak determining step and said time- correlation determining step to determine the location within the second fluid stream of a corresponding concentration peak for the solute of interest; and (f) collecting that portion of the second fluid stream corresponding to the concentration peak for the solute of interest.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings, which are hereby incorporated into and constitute a part of this specification, illustrate the preferred embodiments of the invention and, together with the description, serve to explain the principles of the invention. In the drawings wherein like reference numerals represent like parts:

Fig. 1 is a schematic diagram of one embodiment of an automated on-line system for monitoring an effluent sample for a solute of interest, said on-line system being constructed according to the teachings of the present invention;

Figs. 2(a) and 2(b) are graphic representations of the fluorescence intensity signals to 5 μg HIgG, a mixture of 500 μg BSA and 5 μg HIgG, and 500 μg BSA obtained using Protein G-DTAF without MES and Protein G-DTAF with MES, respectively, as set forth in Example 5;

Figs. 3(a) and 3(b) are graphic representations of the fluorescence intensity signals to 5 μg HIgG, a mixture of 500 μg BSA and 5 μg HIgG, and 500 μg BSA obtained using Protein A-HCCS and Protein G-HCCS, respectively, as set forth in Example 6;

Figs. 4(a) and 4(b) are graphic representations of the fluorescence intensity signals to 5 μg HIgG, a mixture of 500 μg BSA and 5 μg HIgG, and 500 μg BSA obtained using Protein A-MCCS and Protein G-MCCS, respectively, as set forth in Example 7;

Fig. 5 is a graphic representation of the fluorescence intensity signals to HIgG, BSA, mlgG, RlgG, BlgG, goat anti-mlgG and goat anti-RIgG produced using a 1:1 mixture of Protein A-CFSE and Protein G-CFSE at a flow rate of 2 ml/min, as set forth in Example 8; Fig. 6 is a graphic representation of the fluorescence intensity signals to

HIgG, BSA, mlgG, RlgG, BlgG, goat anti-mlgG and goat anti-RIgG produced using a 1 :1 mixture of Protein A-FHSE and Protein G-FHSE at a flow rate of 2 ml/min, as set forth in Example 9;

Fig. 7 is a graphic representation of the fluorescence intensity signals to HIgG, BSA, mlgG, RlgG, BlgG, goat anti-mlgG, goat anti-HIgG and goat anti-RIgG produced using 76-day old Protein G-DTAF, as set forth in Example 10;

Fig. 8 is a graphic representation ofthe fluorescence intensity signals to goat anti-rabbit IgG, goat anti-mlgG, and HIgG produced using RIgG-CFSE, as set forth in Example 11 ;

Fig. 9 is a graphic representation ofthe fluorescence intensity signals to goat anti-HIgG and BSA produced using HIgG-CFSE conjugate, as set forth in Example 12;

Fig. 10 is a graphic representation of the fluorescence intensity signals to goat anti-HIgG, Protein A and BSA produced using HlgG-FITC conjugate, as set forth in Example 13; Fig. 11 is a graphic representation of the fluorescence intensity signals to goat anti-HIgG, Protein A and Protein G produced using HlgG-DATF conjugate, as set forth in Example 14;

Figs. 12(a) and 12(b) are graphic representations of the fluorescence intensity signals to HIgG using goat anti HlgG-DATF conjugate at 0.0002 mg/ml in PBS and at 0.0008 mg/ml in PBS, respectively, as set forth in Example 16;

Fig. 13 is a graphic representation of the fluorescence intensity signals to goat anti-RIgG and HIgG produced using a complex of Protein A-CFSE and RlgG, as set forth in Example 17;

Fig. 14 is a graphic representation of the fluorescence intensity signals to goat anti-RIgG and HIgG produced using a complex of Protein G-CFSE and RlgG, as set forth in Example 18;

Fig. 15 is a graphic representation of the fluorescence intensity signals to RlgG, HIgG, BlgG, mlgG and BSA produced using a complex of Protein G-DTAF and goat anti-RIgG, as set forth in Example 19; Fig. 16 is a graphic representation of the fluorescence intensity signals to

RlgG and HIgG produced using a complex of Protein G-FHSE and goat anti-RIgG, as set forth in Example 20;

Fig. 17 is a graphic representation of the fluorescence intensity signals to RlgG, HIgG, mlgG and BSA produced using a Kappa Lock-DTAF conjugate, as set forth in Example 21 ;

Fig. 18 is a graphic representation of the fluorescence intensity signals to RlgG, HIgG, mlgG and BSA produced using a complex of Kappa Lock-DTAF and goat anti-RIgG, as set forth in Example 22;

Fig. 19 is a graphic representation of the fluorescence intensity signals to RlgG, HIgG, mlgG, BlgG, goat anti-mlgG, goat anti-RIgG and BSA produced using Kappa Lock-FHSE, as set forth in Example 23;

Fig. 20 is a graphic representation of the fluorescence intensity signal to concanavalin A using a conalbumin-CFSE conjugate as set forth in Example 24;

Fig. 21 is a graphic representation of the results obtained in Example 25; Fig. 22 is a graphic representation of the results obtained in Example 26; and

Fig. 23 is a graphic representation of the results obtained in Example 27.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

Referring now to Fig. 1 , there is shown a schematic diagram of one embodiment of an automated on-line system for monitoring an effluent sample for a solute of interest, said system being constructed according to the teachings ofthe present invention and being represented generally by reference numeral 11.

As illustrated by Fig. 1 , one possible application of system 11 is in the monitoring of an effluent sample 13 produced by running a fluid sample 15 containing a mixture of solutes through a differential migration separation system 17 capable of physically separating a solute of interest from the remaining solutes in the mixture. System 17 may be, for example, a chromatography column, an electrophoretic gel or the like.

System 11 includes a UV absorbance detection system 21 comprising means (not shown) for irradiating a sample with UV light and means (not shown) for measuring the UV light absorbed by the sample. Effluent sample 13 is passed through system 21 to produce a continuous output corresponding to the spatial distribution of all UV-absorbing solutes (e.g., proteins, nucleic acids, etc.) present in sample 13. The spatial distribution of such solutes, when graphically depicted, typically includes a series of peaks, each peak representing a heightened concentration of one or more solutes at a specific location within sample 13. System 11 also includes a stream-splitting valve 25 located downstream from system 21. Valve 25 splits effluent sample 13 into a major fluid stream 27 and a minor fluid stream 29, minor stream 29 having the same relative chemical composition as major stream 27 but preferably having a much reduced volume as compared to stream 27. For example, the ratio of volumes of stream 29 to stream 27 may be in the range of 1:10, respectively.

System 11 further includes a mixer 31 located downstream from valve 25 along the path of minor stream 29. Mixer 31 is used to continuously mix an analyzing solution 33 to be hereinafter described into minor stream 29 as minor stream 29 flows through mixer 31. Mixer 31 may be a micromixer capable of mixing together analyzing solution 33 and minor stream 29 in μ\ volumes, preferably as low as about 5 μ\. Alternatively, mixer 31 may be a dynamic mixer

capable of mixing together analyzing solution 33 and minor stream 29 in ml volumes, typically about 1 ml. One advantage to using a micromixer, as opposed to a dynamic mixer, in system 11 is that a micromixer may enable one to detect a spatial distribution of a solute of interest with greater resolution than may obtained using a dynamic mixer. This is because a micromixer has a relatively small volume and is less likely than a dynamic mixer to intermingle solutes that have previously been spatially separated by great volumes within an effluent sample. As can readily be appreciated, this benefit to a micromixer will be greatest where the flow rate is comparatively high. One advantage to using a dynamic mixer, as opposed to a micromixer, is that a dynamic mixer, due to its design, may mix analyzing solution 33 and minor stream 29 together more thoroughly than a micromixer may. However, as can readily be appreciated, the specific differential in mixing efficiencies between a micromixer and a dynamic mixer will depend upon the specific embodiments compared. System 11 also includes a pump 35 for pumping analyzing solution 33 into mixer 31 , preferably at a rate of about 1 to 8 ml/min.

Analyzing solution 33 includes a fluorescent agent which preferentially binds to the solute of interest and whose fluorescence is different in one or more respects (e.g., intensity, decay time, polarization, etc.) when bound to the solute of interest than when not bound to the solute of interest. One type of fluorescent agent is a conjugate of a fluorescent dye and a moiety having an affinity for a solute of interest. Where the mixture of solutes contains or is likely to contain bovine serum albumin (BSA) and a solute other than BSA is the solute of interest, fluorescein dyes are preferred as the fluorescent dyes because they typically exhibit a manageably small, if any, positive response to bovine serum albumin (BSA). (Moreover, the aforementioned positive response to BSA can be effectively counteracted by additionally including in analyzing solution 33 a small quantity of free hydrolyzed fluorescein dye, for example, at a concentration of about 10 ^"7 mg/ml to about 10 ^"4 mg/ml.) Examples of fluorescein dyes suitable for use in the present invention include fluorescein isothiocyanate (FITC), 5-bromomethylfluorescein ( B M F) , 5-(a nd 6-)-iod oaceta m id ofl uorescei n ( IAF) , 5-(4 , 6-

dichlorotriazinyl)aminofluorescein (DTAF), 5-(and 6-)-carboxyfluorescein succinimidyl ester (CFSE) and 6-(fluorescein-5-(and-6)-carboxamido)hexanoic acid succinimidyl ester) (FHSE). Of the foregoing fluorescein dyes, CFSE and FHSE are preferred because they do not exhibit any positive response to BSA. Additional fluorescent dyes include 1-Alkyloxy-pyrene-3,6,8-trisulfonic acid sodium salt (Cascade Blue) and 5,7-Dimethyl-4,4-difluoro-4-bora-3a,4a-diaza-s-indacene (BODIPY FL). Other fluorescent dyes which are small, hydrophilic and lack amino groups also may be suitable as the fluorescent dye where BSA is not the solute of interest but may be in the mixture of solutes. Protein A and Protein G each bind to a wide variety of immunoglobulins

(e.g., IgG, IgA, IgD, IgM, etc.) in many animal species, the most notable exceptions being chicken IgG and human IgD. Accordingly, where immunoglobulins other than chicken IgG and/or human IgD are the solute of interest, suitable fluorescent agents include (i) a conjugate of a fluorescein dye and Protein A, (ii) a conjugate of a fluorescein dye and Protein G or (iii) a mixture of fluorescein dye/Protein A and fluorescein dye/Protein G conjugates. A mixture of Protein A and Protein G conjugates is particularly preferred where the solute of interest is a diverse collection of immunoglobulins since the affinities of Protein A and Protein G are complementary for many immunoglobulins. (For example, Protein G binds to bovine IgG more strongly than does Protein A.) The synthesis of conjugates between fluorescein dyes and Protein A or Protein G may be carried out in substantially the fashion disclosed by M. Brinkley in Bioconiugate Chem., Vol. 3, pp. 2-13 (1992), which is incorporated herein by reference. The working concentration of the Protein A(or G)/fluorescein dye conjugate in analyzing solution 33 is in the range of 0.0001 to 0.1 mg/ml.

Where the solute of interest is antithrombin III, the fluorescent agent may either be (i) 8-hydroxypyrene-1 ,3,6-trisulfonic acid trisodium salt (HPTS) or (ii) a conjugate of heparin and HPTS. Surprisingly, the present inventor has discovered that HPTS possesses its own affinity for antithrombin III and, therefore, does not need to be conjugated to an affinity moiety.

The fluorescent agent of the present invention also may take the form of a complex between a fluorescein dye/Protein A(or G) conjugate and a moiety having an affinity for both Protein A(or Protein G) and the solute of interest.

Analyzing solution 33 also includes a buffer solution in which the fluorescent agent is dissolved. The function of the buffer system is to maintain the fluorescent agent at a fairly constant pH. This is because the fluorescence intensity of fluorescein dyes tend to increase as pH increases. As can readily be appreciated, however, the maintenance of a constant pH in the present invention is complicated by the fact that analyzing solution 33 is mixed with minor stream 29, which itself also typically contains a buffer (e.g, a sample buffer or an elution buffer) that may or may not be at the same pH as the buffer in solution 33. Nevertheless, it is believed that, by using a strong buffer, such as a sodium phosphate buffer (preferably 0.05-0.2M, pH 7-8.9), and/or by keeping the pH of the sample buffer in the range of slightly lower than the elution buffer to about 2 pH units above the elution buffer and/or by using a good micromixer 31 , the types of problems alluded to above will be minimized. It should also be noted that antigens do not bind as well to Ig in the presence of salt concentrations higher than 0.5 M.

System 11 also includes a fluorescence detection system 41 located downstream from mixer 31 along the path of minor stream 29. Fluorescence detection system 41 includes means (not shown) for exciting the fluorescent agent added to minor stream 29 in such a way as to cause the fluorescent agent to fluoresce in accordance with whether or not the fluorescent agent is bound to the solute of interest and means (not shown) for measuring the resultant fluorescence emitted from the exciting portion of minor stream. In the present embodiment, the measuring means measures fluorescence intensity; however, the measuring means could alternatively measure decay time, polarization, or any other distinguishing fluorescence feature. By passing minor stream 29 through system 41 , a continuous output is produced corresponding to the spatial distribution of the solute of interest in minor stream 29. The spatial distribution of the solute of interest, when graphically depicted, typically includes a peak, the peak representing a heightened concentration of the solute at a specific location within stream 29. As can readily

be appreciated, the shape of the peak typically imparts certain information regarding the effectiveness of the separation performed by system 17, a broad peak generally indicating a poorer separation than a sharp peak.

System 11 further includes a computer 51. Computer 51 is electrically connected to and receives the respective outputs from UV absorbance detection system 21 and fluorescence detection system 41. Computer 51 processes these outputs and graphically displays them on a display 53. In addition, computer 51 uses the outputs from systems 21 and 41 , together with data inputted thereto regarding the flow rate of the system (which can be in the range of 0.1 to 30 ml/min), to determine the corresponding location within major stream 27 of the peak concentration of the solute of interest. Computer 51 then uses this information to control a valve 55 located in the path of major stream 27 to selectively collect only that portion of major stream 27 containing the peak concentration of the solute of interest. In the embodiment shown, computer 51 is also electrically connected to valve 25 so that sample 13 is split only after a UV absorbing solute has been detected by system 21. In addition, computer 51 is electrically connected to mixer 31 so that mixer 31 is actuated only when valve 25 splits sample 13 into major stream 27 and minor stream 29. As can readily be appreciated, by providing computer 51 with appropriate standards interrelating fluorescence intensity and concentration, system 11 could be used to determine the concentration of a solute of interest in sample 13.

One advantage of system 11 , as compared to exclusively UV-based on-line detection systems, is that system 11 can detect solutes of interest at as low a concentration as 0.01 mg/ml whereas the aforementioned UV systems have a detection limit of 0.1 mg/ml. Furthermore, system 11 is capable of detecting as little as 5 μg IgG in the presence of 500 μg BSA whereas UV-based systems cannot.

The following examples are merely illustrative of various aspects of the present invention and should in no way limit the scope of the present invention.

EXAMPLE 1

Protein A-FITC conjugate was synthesized as follows: 0.2 ml of 20 mg/ml Protein A was combined with 0.5 ml of 0.2 M sodium bicarbonate buffer at pH 9. 1 mg of FITC was dissolved in 0.1 ml DMSO by vortexing. The FITC solution was then added to the Protein A solution, and the resulting mixture, after vortexing for 10 seconds, was stirred on a rocker at room temperature for 7 hours. The resultant Protein A-FITC conjugate was purified by gel permeation chromatography (Ultrogel AcA44, BioSepra, Marlborough, MA) using PBS containing 0.02% sodium azide as the loading and elution buffer. The first band (10 ml) was collected for the conjugate. The concentration ofthe conjugate was approximately 0.4 mg/ml, which was then diluted to 0.001 mg/ml in PBS.

EXAMPLE 2 Protein G-FITC conjugate was synthesized in the same fashion as described above in Example 1 for Protein A-FITC. EXAMPLE 3

Protein A-DTAF conjugate and Protein G-DTAF conjugate were synthesized in the same fashion as described above in Example 1 for Protein A-FITC, except that the reaction time was 2.5 hours.

EXAMPLE 4 Protein A-FITC, Protein A-DTAF and Protein G-DTAF were each used in an on-line detection system similar to system 11 to detect human IgG (HIgG) in a mixture containing 5 μg HIgG and 500 μg BSA. All three conjugates showed a relatively small peak for BSA, as well as a relatively large peak for HIgG. However, the small BSA peak was reduced to the noise level by the addition of about 2 μ\ of a 0.1 mg/ml solution of hydrolyzed free FITC to 200 ml of a 0.001 mg/ml solution of the conjugate, the final concentration of the free FITC being about 10 ^"6 mg/ml.

The free FITC was found to exert a negative response on BSA while exerting a positive response on IgG. In this manner, the negative BSA peak from the free

FITC cancelled out the positive BSA peak from the conjugates, without adversely affecting the HIgG peak.

Protein A-FITC and Protein G-DTAF were also used in an on-line detection system similar to system 11 to detect 5 μg quantities of mouse IgG (mlgG), bovine IgG (BlgG) and rabbit IgG (RlgG), respectively. Both conjugates resulted in the generation of easily detectable peaks for all three immunoglobulins. A positive peak for rabbit IgG was observed when Protein G-DTAF was used for a mixture containing 5 μg rabbit IgG and 500 μg BSA, and a positive peak for mouse IgG was also observed when Protein A-FTIC was used for a mixture containing 5 μg mouse IgG and 500 μg BSA. However, when Protein A-FITC was used to detect bovine IgG, a weaker peak was observed than was observed when Protein A-FITC was used to detect HIgG, mlgG and RlgG. Without wishing to be limited by any theory, the present inventor believes that this weaker peak was the result of the fact that Protein A binds somewhat weakly to BlgG. In an effort to improve the response to BlgG and based on the fact that Protein G binds more strongly to BlgG than does Protein A, a mixture of Protein A-FITC and Protein G-DTAF was then used to detect BlgG. An increase in the peak resulted.

EXAMPLE 5 After the conjugation reaction of Protein G (4 mg) in 0.7 ml of 0.2 M bicarbonate at pH 9 with 2 mg DTAF in 0.1 ml of DMSO at room temperature for 2.5 hours, 0.1 ml of 1 M MES (2-mercaptoethane sulfonate sodium salt) was added to the reaction mixture to stop the reaction and to put a negative charge on Protein G-DTAF in an effort to reduce the BSA peak. The resulting mixture was stirred at room temperature for 2 hours. The Protein G-DTAF conjugate was then isolated using GPC (Ultrogel AcA 44) and PBS as an eluent. MES could be attached to the conjugate through the reaction with the residual chloro groups of DTAF already attached to Protein G. As seen in Figs. 2(a) and 2(b), the Protein G-DTAF conjugate containing some MES groups (and therefore having more negative charges) showed a somewhat lower response to BSA than did a Protein G-DTAF conjugate lacking MES. Nevertheless, the BSA peak was not completely eliminated. Protein A-DTAF, with and without MES, behaved in a similar fashion.

EXAMPLE 6

A Protein A-HCCS (7-hydroxycoumarin-3-carboxylic acid, succinimidyl ester) conjugate was synthesized as follows: 0.2 ml of 20 mg/ml Protein A solution (4 mg Protein A) in 0.5 ml of 0.2 M sodium bicarbonate at pH 8.3 was reacted with 2 mg HCCS in 0.1 ml DMSO at room temperature for 3 hours. The reaction was stopped by adding 0.1 ml of 1 M NH ₂ OH followed by stirring the resulting mixture at room temperature for 1 hour. The resultant Protein A-HCCS conjugate was then isolated using GPC and PBS as an elution buffer. A Protein G-HCCS conjugate was made in a comparable fashion. As seen in Figs. 3(a) and 3(b), each of the Protein A-HCCS and Protein G-

HCCS conjugates produced a fairly strong BSA peak, as well as an HIgG peak. Without wishing to be limited by any theory, the present inventor believes that the strong BSA peak may attributable to a lack of sufficient hydrophilicity in the coumarin moieties. EXAMPLE 7

Protein A-MCCS (7-methoxycoumarin-3-carboxylic acid, succinimidyl ester) and Protein G-MCCS were synthesized in a manner corresponding to that described in Example 6 for Protein A-HCCS and Protein G-HCCS, respectively. As seen in Figs.4(a) and 4(b), Protein A-MCCS and Protein G-MCCS, respectively, gave even bigger peaks for BSA than did Protein A-HCCS and Protein G-HCCS, respectively. Moreover, Protein A-MCCS and Protein G-MCCS did not respond well to HIgG.

EXAMPLE 8 Conjugates of Protein A-CFSE (carboxyfluorescein succinimidyl ester), Protein G-CFSE, Protein A-FHSE (6-(fluorescein-5-(and-6-)-carboxamido)hexanoic acid succinimidyl ester) and Protein G-FHSE were synthesized in a manner corresponding to that described in Example 6, except that 2 mg of CFSE and FHSE were used. As seen in Fig. 5, a 1 :1 mixture of Protein A-CFSE (0.0003 mg/ml) and Protein G-CFSE (0.0003 mg/ml) running at a flow rate of 2 ml/min. each pump did not produce any peaks to BSA, but did produce good signals to

HIgG, mlgG, RlgG and BlgG and detectable peaks to goat anti-mlgG and goat anti- RIgG.

EXAMPLE 9 Referring to Fig. 6, it can be seen that a 1:1 mixture of Protein A-FHSE and Protein G-FHSE did not result in any response to BSA but gave responses to the IgG's similar to that produced by the mixture of Protein A-CFSE and Protein G- CFSE discussed above in Example 8, except that the signals to BlgG, goat anti- mlgG and goat anti-RIgG were weaker than in Example 8.

EXAMPLE 10 Protein G-DTAF, which had been stored for 76 days in a refrigerator, was used as a peak-tracking agent in a dilute (0.0003 mg/ml) PBS solution. As seen in Fig. 7, the Protein G-DTAF was still effective in producing peaks to HIgG, RlgG, mlgG, BlgG, goat anti-RIgG, goat anti-HIgG and goat anti-mlgG. A small peak to BSA was also detected. EXAMPLE 11

6.3 mg of rabbit IgG (RlgG) was conjugated with 2.1 mg of CFSE in the same manner as set forth in Example 8. As can be seen in Fig. 8, the RIgG-CFSE conjugate (0.0005 mg/ml in PBS) produced weak signals to goat anti-rabbit IgG and mlgG, but did not produce a peak to HIgG. EXAMPLE 12

10 mg of HIgG was conjugated with 1 mg of CFSE in the same manner as set forth in Example 11. As can be seen in Fig. 9, the HIgG-CFSE conjugate (0.0005 mg/ml in PBS) produced a weak negative signal to goat anti-HIgG, but produced no signal to BSA. EXAMPLE 13

HlgG-FITC conjugate was purchased from Jackson Immuno Research, West

Grove, PA. A diluted solution of the conjugate (0.0005 mg/ml in PBS) was used in the peak tracking of goat anti-HIgG. As can be seen in Fig. 10, the conjugate produced a very weak negative peak to goat anti-HIgG, a very weak positive peak to Protein A and no peak to BSA.

EXAMPLE 14

15 mg of HIgG was conjugated with 2.3 mg of DTAF in the same manner as set forth in Example 3. As shown in Fig. 11 , HIgG-DTAF conjugate (0.0005 mg/ml in PBS) produced no signal to goat anti-HIgG, a positive signal to Protein A and a negative signal to Protein G.

EXAMPLE 15 Goat anti-RIgG (1 ml, 2.4 mg/ml) in 0.5 ml of 0.2 M sodium bicarbonate at pH 8.3 was reacted with 1 mg of CFSE in 0.1 ml of DMSO at room temperature for 3 hours. The reaction was stopped by adding 0.1 ml of 1 M NH ₂ OH. The reaction mixture was stirred at room temperature for 1 hour. The resultant goat anti RIgG- CFSE conjugate was isolated using GPC. The conjugate (0.0005 mg/ml in PBS) produced no response to RlgG or HIgG.

EXAMPLE 16

Goat anti HIgG-DTAF conjugate was purchased from Pierce, Rockford, Illinois, and used in the peak tracking of HIgG. As shown in Fig. 12(a), the conjugate (0.0002 mg/ml in PBS) showed detectable positive peaks to HIgG and no peak to BSA. Increasing the concentration of the conjugate to 0.0008 mg/ml improved the peak intensity (see Fig. 12(b)) to HIgG.

EXAMPLE 17 Protein A-CFSE conjugate (0.1 ml, 0.3 mg/ml) was mixed with 0.1 ml of 2.3 mg/ml RlgG, which binds Protein A-CFSE well. The resulting mixture was allowed to stand at room temperature for 10 minutes and then was transferred to 100 ml of PBS. The resultant solution was then used in the peak tracking of goat anti- RIgG and HIgG. As can be seen in Fig. 13, no peak to goat anti-RIgG was produced. In addition, no peak to HIgG was produced, indicating that there was no free Protein A-CFSE since all the Protein A-CFSE should have bound to RlgG, which was in excess.

EXAMPLE 18 The experiment of Example 17 was repeated, except that Protein G-CFSE was used instead of Protein A-CFSE. As can be seen in Fig. 14, no peaks were detected.

EXAMPLE 19

Protein G-DTAF (0.1 ml, 0.3 mg/ml) was incubated with 0.5 mg goat anti- RIgG in 0.5 ml of PBS at room temperature for 10 minutes. The incubated solution was then added to 100 ml of PBS and used to detect RlgG, HIgG, BlgG, mlgG and BSA. As can be seen in Fig. 15, no peaks were detected for RlgG, HIgG, BlgG and BSA. A peak was detected, however, for mlgG. It is believed that this peak is the result of mlgG binding to Protein G-DTAF so strongly that it can compete with RlgG in excess.

EXAMPLE 20 Protein G-FHSE (0.1 ml, 0.3 mg/ml) was incubated with 1 mg goat anti-RIgG in 0.5 ml of PBS at room temperature for 10 minutes. The resulting solution was added to 100 ml of PBS and used in the peak tracking of RlgG and HIgG. As can be seen in Fig. 16, a broad negative peak to RlgG was detected. No peak to HIgG was detected. EXAMPLE 21

4.3 mg of Kappa Lock (Aaston, Natick, MA), a light chain Fab fragment capable of binding to the Fab region of IgG, in 0.7 ml of 0.2 M sodium bicarbonate at pH 9 was reacted with 2.1 mg of DTAF in 0.1 ml of DMSO at room temperature for 3 hours. The reaction was stopped by adding 0.1 ml of 1 M NH ₂ OH to the mixture, which was stirred at room temperature for an additional hour. The Kappa Lock-DTAF conjugate was isolated in PBS solution using GPC. The conjugate was diluted to 0.0003 mg/ml in PBS and used to detect RlgG, HIgG, mlgG and BSA. As can be seen in Fig. 17, the conjugate produced weak peaks to RlgG, HIgG and BSA and produced a strong peak to mlgG. EXAMPLE 22

A sandwich (i.e. complex) between Kappa Lock-DTAF and goat anti-RIgG was formed by combining 0.1 ml of 0.3 mg/ml Kappa Lock-DTAF with 0.5 ml of 2.3 mg/ml goat anti-RIgG and incubating for 10 minutes at room temperature. The resultant sandwich was then transferred to 100 ml of PBS to be used as the working solution (0.0003 mg/ml) to detect RlgG, HIgG, mlgG and BSA. As can be seen in Fig. 18, the sandwich produced peaks similar to those obtained using

Kappa Lock-DTAF. In fact, the peak intensity using the sandwich was somewhat worse than that obtained using Kappa Lock-DTAF.

EXAMPLE 23 Kappa Lock-FHSE conjugate was made by the same technique used in Example 21 , except that 4.2 mg of Kappa Lock in 0.7 ml of 0.2 M sodium bicarbonate at pH 8.3 was reacted with 2 mg of FHSE in 0.1 ml of DMSO at room temperature for 3 hours. The conjugate was then used to detect RlgG, HIgG, mlgG, BlgG, BSA, goat anti-mlgG and goat anti-RIgG. As seen in Fig. 19, Kappa Lock-FHSE did not give a good response to most IgG's, except for MlgG. EXAMPLE 24

2.3 mg of concanavalin A in 0.7 ml of 0.2 M sodium bicarbonate at pH 8.3 was reacted with 1 mg of CFSE in 0.1 ml of DMSO at room temperature for 3 hours. The reaction was stopped by adding 0.1 ml of 1 M NH ₂ OH. The mixture was stirred at room temperature for an additional hour and then purified using GPC and PBS as an eluent. Conalbumin-CFSE was made in the same fashion, except that 4.3 mg of conalbumin and 2.2 mg of CFSE were used. The concanavalin A- CFSE conjugate (0.001 mg/ml in PBS) did not give any signal to conalbumin, and the conalbumin-CFSE conjugate (0.001 mg/ml in PBS) showed a very weak peak to concanavalin A (see Fig. 20). EXAMPLE 25

Without using a column, a sample of 100 μ\ 0.05 mg/ml AT3 (Miles) in a mixture of 20 mM Tris and 150 mM NaCI pH 7.45 was pumped in a stream of a mixture of 70% PBS and 30% water at a flow rate of 2 ml/min into a static mixer where it was mixed with a stream of 0.0001 mg/ml HPTS in PBS pumped into the mixer at a flow rate of 2 ml/min. The resulting mixture flowed into a fluorescence detector which showed an increase in the fluorescence intensity due to AT3 over the blank (20 mM Tris and 150 mM NaCI) as shown in Fig. 21. This is an example of the fluorescence detection of a non-lgG protein using a fluorescent dye alone, without using an antigen such as protein A or protein G.

EXAMPLE 26

A 100 / I sample containing 5 mg/ml BSA and 0.05 mg/ml HIgG in 0.2 M NaOAc, pH 5.1 , was passed through S-HyperD ^™ media, a silica oxide/polystyrene composite support with hydrogel filled pores having a sulfopropyl ion exchange functionality (commercially available from BioSepra, Mariborough, MA and disclosed in U.S. Patent No. 5,268,097, incorporated herein by reference), and separated using 0.2 M NaOAc, pH 5.1 , and 1 M salt gradient. The effluent sample exiting the chromatography media was then monitored using a system similar to system 11 , the analyzing solution comprising a mixture of Protein A-CFSE and Protein G-CFSE in 0.2 M Na2HP04, pH 8.9. As can be seen in Fig. 22, the first UV peak, corresponding to BSA, gave no response to the fluorescence monitoring whereas the second broad UV peak, corresponding to HIgG, produced a good fluorescence peak.

EXAMPLE 27 A 100 μ\ sample containing 5 mg/ml BSA and 0.05 mg/ml HIgG in 0.2 M

Tris, pH 6, was passed through Q-HyperD ^™ media, a silica oxide/polystyrene composite support with hydrogel filled pores having a quartenary amine ion exchange functionality (also commercially available from BioSepra, Mariborough, MA), and separated using 0.2 M Tris, pH 5.5, and 1 M NaCI gradient elution. The effluent sample exiting the chromatography media was then monitored using a system similar to system 11 , the analyzing solution comprising Protein A-FHSE in 0.2 M Tris, pH 8.4. As can be seen in Fig. 23, the first peak (HIgG) gave a good fluorescence peak and the second peak (BSA) produced no fluorescence signal, it being noted that the fluorescence baseline decreased during salt gradient elution. The embodiments of the present invention described above are intended to be merely exemplary and those skilled in the art shall be able to make numerous variations and modifications to it without departing from the spirit of the present invention. For example, UV absorbance testing unit 21 could be located after valve 25 along the path of major stream 27. All such variations and modifications are intended to be within the scope of the present invention as defined in the appended claims.