CA2149784A1 - Ligand assay using interference modulation - Google Patents

Abstract

An optical detection method for detecting specific ligands in an immunoassay is formed by diffracting a beam of light through slits formed in a mask and thereby forming an interference pattern.
The diffracted light of one of the slits is disturbed by an assay having a ligand which when reacted with anti-ligand changes the optical characteristics of the assay thereby changing the interference pattern in a concentration dependent manner.

Description

J - ~
'" Ult~ nA ~ QQ~
~v~.,~0~ PCT~S92110072 214~784 LIGAND ASSAY USING_INTERFERENCE MODULAT~ON

Background of the Invention It is desirable in certain circumstances to measure very 19w concentrations of certain organic compounds. In ~edicine, for example, it i5 very u~s~ful to determine the concentration of a yiven kind of molecule, usually in solution, which either exists naturally in physiological fluids ~e.g. blood or urine) or which has been introdu~ed into the living system (e.g. drugs or contaminants). Because of the rapidly adv~cing state of understanding of the ~olecular basis of both the nor~al and diseased states of living systems, there is an increasing need for methods of detection which are quantitative, specific to the molecule of interest, highly sensitive and relatively simple to implement. Exampies of molecules of interest : in a medical and/or biological context include, but are not limited to, drugs, sex and adrenal hor~ones, biologically active peptides, circulating hormones and antigens associated with tumors or infectious agents.
In the cace of drugs, for example, the safe and efficacious use of a particular drug requires that its concentration in the circulatory system be held to within relatively narrow bounds, referred to as the therapeutic range.
One broad approach used to detect the presence of a particular compound, referred to as the analyte, is the immunoassay, in which detection of a given molecular species, referred to generally as the ligand, .
is accomplished through the use of a second molecular species, often called the antiligand, or the receptor, W094/~U~Z PCTNS~/lOOn

2~4~7~

which specifically binds to the first compound of interest. The presence of the ligand of interest is detected by measuring, or inferring, either directly or indirectly, the extent of binding of ligand to antiligand. The ligand may be either ~onoepitopic or poly epitopic and is generally defined to be any organic molecule for which there exists another molecule (i.e. the antiligand) which specifically bonds ~ to said ligand, owing to the recognition of some portion of said ligand. Examples of ligands include ~acro~olecular antigens and haptens ~e.g. drugs). The antiligand, or receptor, i~ usually an antibody, which either exists naturally or can be prepared artificially. The ligand and antiligand together form a ho~ologous pair. Throughout the text the terms antigen and antibody, which represent typical examples, are used interchangeably with the t~erms ligand and antiligand, respect~vely, but such usage does not signify any loss of generality. In some cases, the antibody would be the ligand and the antigen the antiligand, if the presence of the antibody is to be detected.
Implementation of a successful immunoassay requires a detectable signal which is related to the extent of antigen~an$ibody; binding which occurs upon the reaction of the analyte with various assay reagents. Usually that signal is provided for by a label which is conjugated to either the ligand or the antiligand, depending on the mode of operation of the immunoassay. Any label which provides a stable, conveniently detectable signal is an acceptable candidate. Physical or chemical effects which produce detectable signals, and for which suitable labels ~- WO941L~82 --- PCT~S92110072 exist, include radioactivity, fluorescence, ~hemiluminescence, phosphorescence and enzymatic activity, to name a few.
Broadly speaking, immunoas~ays fall into two general categories - heterogeneous and ho~ogeneous. In heterogeneous assays, the purpose of the label is simply to establi~h the location of the ~olecule to which it conjugates - i.e. to establish whether the - labeled molecule is free i~ solution or is part of a bound complex. Heteroyeneous assays generally function by explicitly separating bound antigen-antibody co~plexes from the remaining free antigen andlor antibody. A ~ethod which is frsquently employed consists of attaching one of the members of the ho~ologous pair to a solid surface by covalent binding, physical absorption, or some other means. When antigen-antibody binding occurs, ~he resulting bound complexes remain attached to this solid surface (composed of any suitable inert material such as p~astic, paper, glass, metal, polymer gel, etc.), allowing for ~eparation of free antigen and/or antibody in the surrounding solution by a wash step. A
~ariation on this method consists of using small (typically 0.05 to 20 microns) suspendable particles to provide the solid s~rface onto which either antigen or antibody is immobilized. Separation is effected by centrifugation of the solution of sample, reagents and suspendable beads at an appropriate speed, resulting in selective sedimentation of the support particles - 30 together with the bound complexes.
In the homogeneous assay, the signal obtained from the labeled ligand or antiligand is modified, or ~odulated, in some systematic, recognizable way when W09Vl~82 PCT~592/l0072 ~ j 2~7~ 1 ligand-antiligand binding occur~. Consequently, separation of the labeled bound complexes from the free lakeled molecules is no longer required.
There exist a number of ways in which immunoassays can be carried out.
In the competitive ~ode, the analyte, a~sumed to be antigen, is allowed to compete wlth a known concentration of labeled antigen (provided in reagent ~ for~ in the assay kit) for binding to a limited number of antibody molecules which are attached to a solid ~atrix. Following an appropriate incubation pericd, the reacting solution is washed away, ideally leaving just l~beled antig~n-antibody complexes attached to the bindin~ surface, thereby permitting the signal from the labels to be quantitated~
In another method, called the sandwich mode, the analyte, again assumed to be antig~n, reacts with an sxcess of surface-immobilized antibody molecules.
After a suitable incubation period, an excess of label-conjugated antibody is added to the system to reactwith another bonding site on the antigen. After this reaction has gone to e~sential completion, a wash step removes unbound labeled antibody and other sources of contamination, permitting measurement of the signal produced by labels which are attached to antibody-antigen-antibody complexes. Any non-specific bonding of the labeled antibody to the surface will, however, contribute to the signal.
In yet another approach, called the indirect mode, the analyte, this time assumed to consist of specific antibody, is allowed to bind to surface-immobilized antigen which is in excess. The binding surface is then washed and allowed to react with label-conjugated W094/~U82 PCT~S92110072 21497~ l i antibody. After a ~uitable incubatian period the surface is washed again, removing free labeled antibody and permitting measurement of the signal due to ~ound labeled antibody. The resulting signal strength ~aries inversely with the concentration of th~ starting (unknown) antibody, ~ince labeled antibody can bind only to those immobilized antigen molecules which have not already complexed to the analyte.
One of the most ~ensitive immunoassays develo]ped thusfar is the radioimmunoassay ~RIA), in which thle label is a radionuclide, such as It~ conjugated to either member of the homologous (binding) pair.
Fluorescence provides a potentially attractive alternative to radioactivity as a suitable label for i~munoassays. For exa~ple, fluorescein (ucually in the form of fluorescein isothiocyanate, or ~FITC") and a variety of other fluorescent dye m~lecules can be attache~ to ~ost ligands and receptors without significantly impairing their bindin~ properties.
Fluorescent molecules have the property that they absorb light over a certain range of wavelengths and (after a delay ranging form 10-9 to lO-~ seconds) emit light over a range of lon~er wavelengths. Hence, through the use of a suitable light source, detector and optics, including excitation and emission filters, the fluorescence intensity originating from labeled molecules can be determined.
U~e of an enzyme as a label has produced a variety of useful enzyme immunoassays (EIA), the most popular of which is known as E~ISA. In the typical heterogeneous format a sandwich-type reaction i~
employed, in which the ligand of interest, assumed here .

WO94tL~2 PCT~S92110072 ~1~97~

to be antigen, binds to surface-immobilized specific antibody and then to an enzyme-antibody conjugate.
After suitable incubation, any remaining free enzyme conjugate is eliminated by a wash or centrifugation tep. A suitable subctrate for the enzyme is then brought into contact with the surface containing the bound complexes. The enzyme-substrate pair is cholsen to provide a reaction produet which yields a readily detectable signal, such as a color change or a fluorescence emi~sion. The use of an e~zyme as a label ~erves to effectively amplify the contribut~on of a single labeled bound complex to the measured ~ignal, because ~any ~ubstrate mol~cules can be converted by a single enzyme molecule.
~s ~ay be een from the above background summary many of the immunoassay techniques rely on optical methods to detect specific ligands~in a sample matrix and it may be said that these techniques have fallen generally into three optical classes:
1) Methods based on the molecular absorbance o~
~ight, which comprise all standard spectroscopic method~ such as absorbance, fluorescence, phosphorescence, fluorescence polarization, circular dichroism, Raman, and infrared spectroscopies;
2) Methods based on the generation of light by a chemical reaction, which is known as chemiluminescence, or biolumineæcence when the reaction is catalyzed by an enzyme; and ~- 3) Methods based sn the change in the direction of propagation of a lightwave such as refractometry, I optical rotary dispersion, and methods based on the i scattering of light.
With all of the above the signal is collected and ~, WO9411~2 PCT~S92/10072 ~1~.97~a .:

measured by a detection apparatu~ which monitors the phenomenon as it occur~ within the sample's ~olecular population. The measurement i~ a function of the intensity or degree of change occurring within ~he S sample. Although most of the~e methods produce isotropic signals, even the ani~otropic methods ~uch as fluorescence polarization, produce si~nals which are fundamentally ~easured ~ intensities.
- Emerging now are optical detection techniques which may compri~e a new and fourth class of optical ~ethods based on spatial patterns produced by the interaction of radiation with the ligand or anti-ligand. In this clas~ of optical detection techni~ues the optical signal is collected Qutside of the ~ample, lS a~ a pattern or an intensity mea~ured at more than one geometrically defined spatial position. The following descriptions are representative of~this new class of optical detection method~.
Nicoli in U.S. Patent No. 4,647,544 issued 3 March 1987 entitled "Immunoassay Using Optical Interference Detection" describes optical detection of a binding reaction between a ligand and an antiligand wherein a pattern is formed on a substrate by a ~patial array of microscopic dimensions of antiligand material immobilized to a substrat~. Upon exposure to the sample the ligand binds to the substrate to form a physical pattern. A source of optical radiation is directed to the pattern at a particular incidence angle to produce an optical interference pattern in accordance with the binding reaction and with a strong scattering intensity at one or more Bragg scattering angles. An optical detector is located relative to the pattern and aligned with the Bragg scattering angle to W094/l~82 PCT~S92/10072 ~ ;

detect the 6trong scattering intensity and produce a signal repr~sentative of the binding reaction.
Another patent, "Diffraction Immunoassay and Reagents" EP0 OZ76968 publi~hed 3 Augu6t 1988 describes a "biograting~ based assay in which a biological diffraction grating con~i~ting of lines of active binding reagent is formed on a ~ilicon substrate - surface. After contacting the assay surface with the ~ sample and separating the sample from the assay, the surface is illu~inated and the binding of analyte to ~urface in a uniform manner generates a diffraction pattern. An optical detector, or array of detectors, positioned at predetermined angles is u~ed to measure the diffracted light.
PCT Application No. PCT/GB85/00427 describes the use of fluorescently tagged molecules binding to a substrate pattern so that the fluo~escence emission is organized into a narrow cone of angles instead of being uniform in all directions.
In each of these examples the detected signal is produced by light interacting with analytes that have been entrapped on a surface în a geometric manner, and the detectable signal is characterized by both amplitude and pattern formation.
These spatial based optical detection immunoassay methods are controlled by the spatial geometry, foxmed by either the ligand or antiligand, which must be of extremely fine detail and small dimensions;
sophisticated immobilization technology will be

3~ re~uired which may be difficult to reproducibly ~ implement for many useful ligands or antiligands.
,~
~ Summar~ of the Invention :

. 1:
- W094/L~8~ 2~97~4 PCT~S92/10072 _g_ The invention compri es a method of modulating an interference pattern created by interference between two or more light beamc by reacting a ligand with an antiligand in an as8ay to cause a change in the optical S tharacteri~tic~ o~ the assay. This change in optical characteristics is u~ed to di~turb the interference pattern by placing the reacted assay in the path of one or more of the light beams creating the interferel-ce pattern. The resultant change in the interference lO pattern i~ dependent on the concentration of the ligand and can be used to detect the presence of t~e ligand and its concentration in the assay.
In a first e~bodiment, an intexference pattern is for~ed by projecting light through a mask haYing at 15 l~ast two narrow parallel slits, having a width which can be very s~all but remains larger than the wavel~ngth of the source light, and separated by a di~tance which exceeds the slit width. The light is thereby diffracted into two or more light beams which 20 interfere and form an interference pattern, which may, for axample, be visualized by projection on a screen.
one or more of the slits may then be blocked to product a different interference pattern. An assay surface with a ligand which when reacted with an anti-ligand 25 reduces the transparency of the assay surface is then placed on or over a slit or slits, causing the 3 interference pattern to change to a pattern which is intermediate between the standard pattern formed by the uncovered ~ulti-slit and the standard pattern formed 30 when the slit or slits relevant to the assay surface are completely obscured. The c~ange in the ~t interference pattern compared to the standard is mea~ured and related to the concentration of the assay WO94/L~82 PCT~S92110072 ~

. .

10-- ~
- ligand.
Preferably a ~ource of monochromatic light, such as a laser, is used to project light through ~ulti-slit openings for~ed on a foi} ~ask. An as~ay slide i~
mounted adjacent to the foil to obstruct one or ~ore slit~.
A lens may be used to magni f y the interferenc:e pattern ~o that is can be projected onto a screen and photographed. The resultant photographs are then analy~d by a densitometer to determine the intensity of ~he light at certain points on the pattern before and after the slit is covered, thereby to determine the presence of he ligand and its concentration in the assay.
Alternatively, the intensity of the pattern can be analyzed in real time using an array of pixels formed by photodetectors to form electri8al signals corre~ponding to each pixel and processing these signals in a computer.

Brief Descri~tion of the Drawinas Fig. 1 is a diagram of light interference produced by a two-slit ~ask to illustrate the theory behind the inYention .
Fig. 2 is a diàgram illustrating in curve (a) the diffraction pattern when only one slit (Sl) is open and in curve (b) the interference pattern wben both slits (S~ & S2) are open.
Fig. 3 is a schematic diagram of a preferred embodiment of the invention.
Fig. 4 is a flow chart of the important steps in forming a colloidal gold-labeled antibody (Au-antibody) ~ I`
WO94/L~82 PCT~S92J10072 21497~g sandwich assay for use in the apparatus of the invention.
~ ig. 5 is a plot of th~ intensity of two adjacent ~axima of double-slit interference pattern ~howing the change caused by varying the transmission through one slit.
Fig. 6 are plots of intensity I versus transmission T for changes in I~ (curYe Io~) and changes in transmission (curve ~T~.
Fig. 7~a) is a plot of the visibility V2 of the fringe pattern as a function of the percent of tran~ sion ~I2%) through the second slit S2.
Fig. 7(b) i~ a plot of the -log V2 x l,OOO versus ~% I2.
Fig. 8 is a plot of the changs in Im~ and I~o as a function of change in transmission I2 through slit S2.
Fig. 9a-9d is a schematic flo~ diagram of an alternate embodiment of the invention.
Fig. lO is a schematic of yet another alternate embodiment.
Fig. 11 is a schematic of an alternate embodiment in which the double-slit pattern is produced by a virtual slit S2 ~nd an actual slit Sl and a mirror M.
Fig. 12 is a schematic of a Fresnel's mîrror embodiment of the invention.
Fig. 13 is a schematic of a bi-prism embodiment of the invention.
Fig. 14 is a schematic of a Billet' 8 split lens embodiment of the invention.
I `30 Fig. 15 is a schematic of a flow cell embodiment.
¦~ Fig. 16 is a schematic of an alternate flow cell embodiment.

L~l~
W094l~U~2 , PCT~S92/10072 ~

21~978~

Fig. 17 is a schematic illustrating an alternate embodiment wherein N slit~ are formed and N-X slits covered by the assay and wherein N=6 ~nd X=5.
Fig. 18 is the same a~ Fig. 17 except that N=6 and X=3.
Fig. 19 is a plot of the intensity of the principle ~axima Io versu~ the number of slits.
Fig. 2~(a)-20(b) is an alternate embodiment of a two or more slit embodiment in which the a~say substrate is formed with slits.
Fig. 21(a) i~ an alternate e~bodiment in whic~l the slit opening is replaced by a square opening.
Fig. 21(b) is ~n alternate e~bodiment of Fig.
20(a) wherein the ~lit opening is triangular in ~hape.
Figs. 21(c) and (d) illustrate formation of an assay on a diagonal half of a square ~lit to form a triangular based diffraction patte~n.

Desc~ ion of the P~çfer~ed Embodiment A. ~ackoround ~heo~Y
In 1801 Thomas Young demonstrated the phenomenon of interference bands produced by a light shining through two identical and closely spaced slits. A
diagram of the optical configuration used by Young is shown in Figure 1~ !A light source (S) is incident on two narrow, parallel slits (S~ and S2). Each slit has a width (D) which can be very small but remain larger than the wavelength of light. The slits are separated by a distance ~a), which exceeds D. The light passing through the slits is diffracted and emerges as two wavefronts with identical wavelength and phase. The two wavefronts interfere and produce a ~isible WO94/~B2 PCT~S92/10072 ~1497~4 interference pattern on the ~creen 20. The pattern consists of a series of bright and dark parallel bands which are typically referred to as fringes.
The ovarall light amplitude at any given puint in the interference pattern i~ ~he result of the superposition of the two wave amplitude~ from S~ and S2.
Two waves that add constructively produce a bright fri~g~ ~maxi~a) while two wave~ that add destructively produce a dark fringe (minima)~ The distance betw~sen two adjacent Daxi~a is calculated by:
y ~ ~ Eq. t~) where X is the distance from the slits to the screen, i~ the wavelength of the light, and m is an integer used to designate the position of the fringe. The central fringe, located on the ~ymmetry axi~, is called .~ the zero-order maximum. The first maximum on either ~ side (m = + 1), is called the firs~-order maximum, and -; the maxima continue in either direction, through the higher orders. The minima are located exactly halfway between the maxima and are completely dark only when the two slits provide equal intensities.
When working with Eq. 1 the quantities X and a are both controlled by the experimental design and the value for Y can be ~easured from the fringe pattern, leaving only ~ as~an unknown value.` Thus, one can use this equation to determine the wavelength of light, which was another historical achievement of the original double-slit experiment.
When only one of the two slits is open a broad central diffraction peak is obtained centered on the open slit as shown in Fig. 2, curve a. With two slits open, the overlapping wavefronts of the diffraction WO94/LU82 PCT~592/l0072 21~9784 bands produced by the individual slits combine to forman interference pattern located within the boundaries of the diffraction bands; the interference fringes are visible where each of the diffraction bands would have S been observed.

B. ~rinc~ple of ~ssaY based on ~ou~ls-Sli~
Interference Pattern The assay of the invention is based on a tran~duction mechanism of double-slit or multi-slit interference modulation~ As stated previously, complete loss of transmission through only one slit in a double slit interference configuration will produce a collap~e o~ the fringe pattern to the single-slit diffraction band. A partial 106s of transmi~sion produces an intermediate pattern where the amplitudes of the interference maxima and mi~ima are moving towards the condition of the diffraction band envelope.
Therefore, any method of reducing the transparency of a slit can be related to a loss of fringe pattern.
One way to convert this process to an assay is to provide a chemistry which produces a localized coverage over one ~lit which is opaque enough to reduce the transmission of light through the slit. The reduction of light through~the sli~ decreases the intensity of light available for interference with light emanating from the other slit, and thus decreases the formation of the resulting fringe pattern. In practice, the surface coverage due to the assay is equated with a predictable change of fringe pattern.
-~ 30 The chemistry for producing loss of transmission through a slit can involve the temporary or permanent localization on the surface of the slit of any ~ .`! , :.. . : ~ ''. ' ~ W0941L~82 2 1 4 ~ 7 8 4 PCT~S9211~072 optically dense molecule or ~aterial which will cause a scattering or absorbance of the incident radiation from a source. Any physical phenomena that can alter or attenuate radiation transmi~sion can af~ect the interference fringe formation. There~ore, any chemical phenomena producing one or ~ore of the~e physical pheno~ena can potentially be a~sayed by this technology. The physical phenomena may comprise a change in reflectance, absorbance, phase, refraction, or polarization of light impinging on an assay. Items capable of behaving thusly include colloidal gold-labeled molecules, colorimetric labels and pigments, bacteria, polymeric particles, cells, L~ngmuir-Blodgett films, as well as other polymeric films.
Por loss of trans~ission by means of light absorbance, materials must be used that absorb at tbe frequency of the incident light. ~Table 1 below lists the spectral color and associated wavelengths of a ~ariety of colored compounds. For example, using a HeNe laser which emits a light of wavelength 630 nm, one would predict that blue/blue-green materials would absorb the energy. If one wanted to detect, for example, a yellow compound then the laser source would be changed to one working in the 430 nm range.
An absorban~e-modulated interference system may~
also be used to detect multiple analytes in a sample with a variable, incoherent source, such as a Xenon arc $
lamp and a scanning monochromater. As a scanning 3 spectrophotometer measures an absorbance spectrum as a function of the frequency of the incident radiation it nay also be possible to measure the interference patterns produced by the transmissions at different frequencies. In this way, one could analyze a sample ~ l~s W094/L~2 PCT~S92110072 .

21~9784 containing a variety of molecule~ with non-overlapping 'J~
absorbance ~pectra; at a radiation wavelength coinciding with the ab~orbance maxi~um of a particular molecule in the sample the decrease of the interference pattern would be due to the attenuation resulting from the concentration of the ~olecule. Each molecule in the sample would cause attenuation of transmission at a discrete wavelength, and in this way ~ultiple analytes could be detected in the region of a single slit (,S2).

TABkE 1 ~: COLORS OF CQM~ S. S~EÇ~R~$ C~LORS ~D_WAVE~ENGTHS

A~prox.
Color of Spectral color wavelength Compoun~ absorbed (nm~
~: 15 Colorless Ultraviolet <400 emon Yellow Violet 410 ~: Yellow Indigo 430 Orange Blue 480 Red Blue-green 500 20 Purple Green 530 Violet Lemon yellow 560 Indigo Yellow 580 Blue Orange 610 Blue-green Red 680 25 Green Purple-red 720 Colorless Infrared >720 .
C. Prefer~ed Embodiment of A~paratus Qf the Invention Turning now to Fig. 3 the apparatus for a first emkodi~ent of the invention will be described in . I

~. I
~. F `'i~
~ W094/~82 PCT~S9~ 072 ~1497~tl connection t~erewith.
A light source lO i~ dispo~ed on a track (not shown) of an optical bench adjacent and in line with a foil mask 24 which is also mounted on the track. Ma~k 24 is formed with at least two narrow vertical ~lits S~
and S2. An assay is formed on assay slide 14 which i8 positioned on mask 24 so as to cover ~lit Sl.
An optional filter 12 is disposed on the track between -iource lO and mask 24. If ~ource lO iB Zl laser emitting highly intense ~onochromatic light, filt:er 12 may be a neutral density filter used to adjust the inten~ity of the light. If source lO i~ a broad band ~oderately intense liqht ~ource, filter 12 may be u~ed to fi~lter all but the desired wavelength for projection on:the ~ask 24.
Light passing through the s~ts S1 and ~ is diffracted ~nd the resulting wavefronts of diffracted light interf~re to create a fringe pattern 22 which is magnified by lens 16 and projected onto screen 20.
Note: Optionally, a second lens (not shown) may be disposed on the side of the mask nearest the source to increase the distribution of the pattern at the origin o. Camera 18 is used to photographically record the ~ i~age of the pabtern. ~This image may then be processed .
by converting the fringe patterns to intensity graphs using a densitometer and analyzing the intensity qraphs produced with and without the slit covered by the assay to determine the concentration of the ligand in the ~r.~
W094/~82 PCT~S92/10072 21~97~

-~8-assay.
The double-slit mask component ~4 i~ a critical item in this process and ~ust be of high ~uality.
Preferably the mask i~ cut out of foil by a laser proce~s. Three different 8iZ~8 have been experimented with as shown in Table 2 below, Tho~e skilled in the art will ~ppreciate that the dimension~ indi~ated have been arbitrarily selected from a continuum of suitable values, and ~re exemplary rather than li~iting on the invention.
TABL~_~

~I~E~IONS OF ~IR SLI~S

Slit width Slit separ. Slit length Foil ~_ Lum) (~ml _lEm) 1 100 5~0 10 3 50 lO~O 2 Foil ~2 produces the most intense and distinct fringe patterns in the current format. The fringe separation (Y) for this foil when placed at X distance from a screen are determined from Eq. 1 and are listed in Table 3 below.

i; .. , - . . . - .

W094/~2 PCT~Ss2/10072 ~1~978~

CALCULATED FRINGE ~EP~R~TIQ~_EOR FO~L ~_A~_VAR~LNG
DIS~C~S ~ETWEEN ~LIT~ L~E$N

~cm) a (uPI~ ~ (u~) Y (um) ~:: 5 S0 5S0 0. 633 57S. 4 550 0 . 633 460 . 4 550 0 . 633 345 . 2 550 0. 633 230. 2 :~ : 10 550 0. 633 llS. 1 -- .
S 550 0 . 633 57 . 5 The foil mask is secured on a magnetic mount (not shown) which is then place,d on a magnetic strip attached to an X-Y-Z positioned in line with the light ~ source 10. The assay requires positioning of the assay - ~ 15 slide over one slit ~ and this must be done so that the ~ slide edge remains between two slits and does not aause ,, ~ any diffraction of the light beam transmitting through ;:: the uncovered slit.
In an experimental model the light source compris~d a HeNe las~r tUniphase, 2mW), geherating : monochromatic light at ~33 nm. Neutral density filters were used to reduce the ~aser intensity.
, .
-~ The fringe patterns were recorded on film ~.
;~: (Polaroid film #667, 3000ASA) and subsequently 25 converted to intensity graphs by a video densitometer i, ' .

WO~4/LU~2 PCT~S9211007~

-2~-(PFÇplus 640-3-U-RT PCVision Plus Frame ~rabber with a 64Qx480 Display, OP~IMAS - Optical ~easurement &
Analysis Software, PV~-l34~Q Sony Trinatron l3" color ~onitor with fflultiple inputs, Logimouse Plus - Logitech ~echanical ~ousé, SYS/386-80-color-l IBM Compatible 38 20 ~hz with E~A Color Monitor and 80 Megabyte Hard Drive) using a video ca~era (8l5-2000-0000 Cohu Series 4800 Higb Resolution CCD Monochrome Camera) focused on - ~he illuminated photographs~ To make a tracing the cursor was po~itioned at the top of a fringe pattern or diffraction envelope with the single-~lit, and then drawn thrnugh the center of the pattern and terminated at the oppo~ing side.
.
D. Assay ChemistrY
A flow chart of a first embod~ment of the assay chemistry of the invention followed by the optical test is ~ummarized in Figs. 4a-g. In this as~ay tbe ligand is rabbit IgG which is incubated with a coverslip 14 (Fig. 4a) coated with anti-rabbit IgG, followed by incubation with a colloidal gold-labeled anti-ligand, such as, goat anti-rabbit IgG 40 (Fig. 4b). After a silver enhancement step, wherein the gold colloid nucleates precipi'tationiof silver from a solution 42 as it comes into contact with the labeled antibody in the assay (Fig. 4c), the coverslip exhibits a deposition of darkly stained antibody-antigen complex. Insertion of this treated assay coverslip over one slit S2 of a WO 941~882 pcTrus92lloo72 7 ~ ~;

dou~le-slit coniguration produce~ 10~8 0~ trarlsparency through ~lit S~ tFig. 4d) re~u~ting in lo8~ of intensity I~ of llght fro~ S2 (Fig~ 4e) and ~n observable, mea~;urable 1068 of interf erence f ringes S (F~ g. 4f ) . Thi8 loss in a~plitude of frinqes i~ then related to analyte concentratlon (Fig. 4g) in the - quantification procedure de~ ribed later in Section E.
The above general descr~ption of the ~etbodology of the assay preparation is followed by the specific experi~ental example below:
The ob~ective of the i~mobilization che~stry 1~
co~alent attachmen~ of the anti~ and, in thi~ case a protein, which will selectively bind the ~arget liga~d or analyte. To prepare the gl~ss for the ~5 im~obilization che~istry the cove~r81ip i8 acid treated ` to produce reactive ~ilanol groups:

~s~ r S;4 '~s;/ U~s ~ s~

In a first example, this was accomplisbed as follows: Glass coverslips (8lmm2) were soaked in a l:l mixture of concentrated HCl and methanol, followed by several rinses with di~tilled H2O. Coverslips were t~en soaked in hydrosulfuric ac1d for 30 ~in., followed W094/L~82 - PCT~S92/1007t 21497~4 by ~veral rln~e~ with dl~tllled ~0. Cover~lips were bolled ~n di~tilled H20 for 30 minutes ~nd then air-dried on low lint paper. (~ee Bhai~, S.K., Shri~r-Lake, L.C., Prlor, K.J., Georger, J~., C~lvert, J.M., BredebOr8t~ R., Ligler, F.S., Analytical Biochemlstry, 178, 4~8-413, 1989~) Next the glass i8 reacted with a~inoethYlpropyltriethoxysilane (APTES), to produc:e a react~ve, amine-derivatized surface:

10 ¦ Siolt ArrFS~ ~S;-C)-S;--N~J1 .

Specifically, a 2~ solution (v/v) of APTES was prepared ~y ~xing 200 ul of APTES in 9.8 ~1 toluene (dried over ~olecular sieves). T~e cover~lips were : soaked for 2 h, and rinsed in dry toluene. (see Weetal, H.H., "Methods of Enzymology~, Mosach, K., Ed., Acade~ic Press: New York, 19~6, Vol. XLIV, pp 139).
The next step ~s,the reaction of the amine derivatized glass w~th ~uccinic anh~dride to produce a carboxylic acid group on the surface:

~s;_O_s; ~ ~ S;-O~s~v~ ~~ ' ._ . . - - - . - ~

j .
41L~82 21~978~ PCT~S92/10072 A olution of 0.20 q (2 mmol) succin~c anhydridQ wa~
combined with 40 mg (0.33 mmol) 4-dimethyla~ino-pyridine in 6 ml anhy~rous pyridine and dlvided ~nto test tube~. The amine-deriv~tlzed glass coversl~p~
were placed into the tube~, which were then sealed and ~haken at room temperature for 16-~0 hours. Coverlslips were removed, rinsed with pyridine, ~ethylene chlo:r~de and ether. If stored, t~ey were placed in a desic,cator over P20~ ee Damha, H.J., P.A., Zabarylo, S.V., Nucleic Acid Research, 18 tl3), 33813, 1990.) The glas i~ now ready for the carbodiimide chemi~try whlch ~egins w~th activation of the succi~ylæted ~urface ~ith 1 ethyl-3- (3-dimethylaminopropyl) car~odii~ide hydrochloride (EDC): R
, . ~
. N
111 S; H~ V ~,~ =C~N Q~ ~ 0~ 11 R, The prscedure and reagents were those recommended with the Car~odiimide Kit for Carboxylated Microparticles (Polysciences, Inc., tl9539). Carboxylated coverslips were soa~ed in carbonate buffer (Bottle tl) for 5 min, and then in phosphate buffer (Bottle t2) for 5 min. A
0.6 ml phosphate buffer was placed in test tu~es with prepared coverslips~ A 2% solution of carbodiimide solution was prepared by mixing 75.0 mg EDC wit~ 3.75 ml phosphate buffer. 0.6 ml EDC solution was added, dropwise, to each of the test tubes containing the W094/t288~ PCT/U59~1111~72 1 4 9 rl 8 ~

cQv~rslips, and capped tightly. The solutions were shaken f or 3 . 5 ~ 4 hours ~t room temperature . The coverslip~ were then reD~oved and shaken in borate bufîer (~ottle ~4) for 5 min. followed by two illore 5 rin~ings with borate buffer, The activated ~urface i then re~c'ced with the alkylamine functional groups of the prote~ n (anti-rabbit IgG) to be iD~mc~bilized, forming an amide linlcage:

o ll N~ --n O-c 5~ / J
v I
. ~ .

O ~
1--S~ /~\NI~

~< ~.
' l 9~ .

.
`;

W094/L~82 - PCT~S92/10072 ~ or the prot¢in reaction ~ 0~05 mg/ml ~olution of anti rabbit IgG wa~ prepared by adding 73 ul o~ anti-rabbit IgG (4.1 ~g/~l; commercially a~ailable fro~
Sigma) to 6 ~1 borate buffer, which was then aliquotte~
into 6 test tubes. ~he coYer~lip~ wer~ ersed, the test tubes were ~ealed and ~ocked gently overnight:.
The coverslips were removed and the ~upernatant W~IB
retained for absorbance ~easurements. For the blocking reaction the çoverslip6 were then added to test tubes containing 1 ml 0.1 ~ ethanola~ine (Bottle ~5) and g~ntly ~ixed ~or 30 ~in. The coverslip~ were then transf~rred to 1 ~1 of ~SA solution ~Bottle ~6) and gently ~ixed for 30 uin. The cover~lips were then rinsed in PBS and stored in storage buffer (Bottle ~7) or PBS, 4 - 6 degrees.
The protein used as the antigligand can a~so be immob~lized to ~ tran~parent pl~stic substrate t~at has been coated with a polymer containing residual CarboXylatQ group8 (Sera-Coat, Seradyn) for carbodiimide coupling.
For the assay, rabbit IgG was incubated with the anti-rabbit IgG coverslips, rin~ed wit~ PBS, and t~en treated with gold-labeled anti-rabbit IgG, and the silver enhancement'step.

~ U ~3 ~J

1.:
wo 94l12882 - -- PCTnJSg2110072 1-2I 49 78~ ` ---2 6-- ;
Spec:if ically, thQ CoVIE~rRlipB wera rinsed with PBS
two t~mes. For ~a o~ handllng, a eoverslip W~8 pos~tioned in the outer groove of ~ di6posabl~ cuv~tte ~nd 20-40 ul of eolloidal gold(30nm~ d Goat ~nti-5 rabbit IgÇ; (AuroPro~ E~S GAR G30, Am~rsbam) wa~pipetted onto th~ surface ~nd left for 60 ~in. Thl~
- eover~l~p was then plaeed insid~ thQ ~a~ cuvette ~md r~nsed thorougbly three time~ with PBS followed by three r~nses with H20.
A ~lver en2~ancement ~tep follows whieh produee~ a ~or~ dramat~e deereas~ ~n the tran~pareney o~ th~
reaeted Z1~8ay. In thi8 ~tep the gold nuel~ates preeip~tation of silver fro~ a ~olution in the region ~: of ~e l~b~led ~ntibody:

A~ SC>~

:

Thi~ was achieved using the IntenSE Silver Enhance~ent Xit ~PN.491, A~er~ha~. A ~olution was prepared by combining equal number of drops from ~olutions A and B. Tbe enhancement ~olut~on wa~
pipetted ~nto cuvette~ containing individual gold-labeled antibcdy treated coverslips and reacted for 6-~8 min. at roo~ temperatuxe~ The coverslips were then rinsed with di~tilled H20. The reaction can be repeated to increase tha effect.

` WO94/1~2 2149784 PCT~S9211~0~2 E. Quanti~ation of the A~s~y E~uations have been derived for describing the dependency of the interference fringes on the lu~inescence intensitie6 produced by the slits, and th~se equations can be employed in this detection method for quantifying the assay.
The equation for describing a ~uality termed the - visibility (Y~) of the fringe pattern can be calculated by the rela'tive intensities (I1 and I2) of the &ource light transmitted through the two identi~al slits (S
and S2), as shown in Figure 1 and can be calculated t~usly:

2~I1 ~2 V, = ~ Eq. (2) Il + I2 where I~ and I2 are the intensities at point P from the radiation passing through the slits l and 2 when viewed independent of each other, and ¦ ~12, = l is a coherency limit for a laser source.
The visibility (V2) can also be calculated by the intensity of the fringes:

Iou - Im~
V2 ~q. (3) I0~ + I ~

where I~ and I~o are the intensities corresponding to a maximum and an adjacent minimum in the fringe system.

W094/~82 PCT~S92/10072 ~1497~

I~ and I~ are function& of the light intensities transmitted through the slits S~ and ~:

I,~ = I~ + I~ + Z ~I2 '~' Eq. (4) I = Il ~ I2 - 2 ~ I~2l Eq. (5) Since I,~ and I~o are determined by I~ and I2 then it follows that the two visibility equations ~hould produce identical values (~ 2~.
I~ the intensity for I~ is held constant while the intensity of I2 is reduced one can oalculate the change in the fringe pattern. Table 4 contains values for a ~odel syste~ where the intensity (Il and I2) through each slit are assigned a value of~lO. When the intensity is identical through eac~ slit the upper value of 1.0 ~xi~ts for V2; as the intensity, I2, is : decreased the V2 decreasas towards 0.O. When V~ = l.O, ~: the fringe pattern has the maximum amplitude of IQ~ and a complete loss of amplitude at I~. As V2 moves towards o.O the fringe pattern collap~es to the diffraction envelope.
, WO94lL~8t PCT~S92~10072 2149~

T~B~ 4 THE C~ANGES IN ~ N~_ IMN WITH L0SS OF
TRANSMISSION_~ROUGH ON~L

~ Io~
5 I~ V2= ~ Q~e 40.0 0 1.0 A
9.~ 38.99 0.006 0~999 9.0 3~97 0~026 0.998 1~ 10 8.0 35.88 0.111 0.993 7.0 33.73 0.266 0.S84 6.0 31.4g 0.508 0O968 5.0 29.14 0.86 0.94 B
2.0 20.94 3.05 0.74 C
1.0 17.32 4.6~ 0.57 D
0.1 12.10 8.10 0.19 E
0.01 1.64 9.38 0.063 F

A graph of two adjacent fringes is shown in Figure 5. The line labeled A describes the intensities of I
and I~n formed by two equal, unobstructed slits. The lines B through F show the change in amplitudes as the inte~sity through one slit is decreased. At the Im~
for line F the intensity amplitude is almo~t 10, which is the intensity of light passing through t~e unobstructed slit, or I~.
It is i~portant to appreciate that the summation of the two slit intensities tI~ + I2) without tbe phenomenon of constructive interference, would be, in Figure 5, a value of 20. Thus, the intensity of I~u observed in lines A through C, where the amplitude is W094/LU82 ' pcT~s92llno72 ~reater than 20, is due to con~tructive interference.
Thus, the amplitude values between 20 and 40 represent an increased signal content for t~e correlation of assay infor~ation.
The intensity of I,~ at point P can al~o be ~uantified from the attenuation of I2 due to a change in absorbance ~A = ~bc) of the sample. The existing f~rmulas which govern absorbance spectroscopy can be linked with the formulas used for evaluating interference pheno~2na since both ~ethodologies are based on trans~ission intensity of a source and the xesultant intensity due to passing the radiation through a sa~ple.
Transmission, T, is defined as the ratio of the intensity of unabsorbed radiation, I, to the intensity of the incident radiation~ I/Io. 9From Beer' 8 Law, : exists the relationship between ~bc and T, where ~ =
the extinction coefficient of the absorbing species, b = path length of cell, and c = concentration of species:
I

= 1~ = loA Eq. (6) Io In the proposed configuration for measuring interference phenomena the I2 transmission intensity can be viewed as a ratio of the intensity of unabsorbed radiation to the intensity of incident radiation, I~/Io: 4 .
Il = unmodulated beam = Io Eq. (7) I~
I2 - assay beam = ~ q- (8) ~W094/LU~e 2 1 4 9 7 8 4 PCT~592/1~72 From Eq. 6:
I~
--- = 10~

I~ - Io ~ 10~ Eq. (9) Substitution of the preceding identities into the I~ equation results in an I~ expression as a function of ab~orbance defined by ~bc for a particular ~pecies:

I,~ I2 + 21 I, I, Eq. (10) I~ = Io + -- + 2~Io -- Eq. (11) Io Io I~
. I~u = Io + __ + 2~ Eq. (12) Io I~ = Io + 10~ + 2~Io 10~ Eq. (~3) Of interest is a comparison of the relative signal cbange of both ~T and ~I~u for identical sample concentration and volume. The transmission values for an absorbance ~eries (~bc) as measured in a standard spectrophotometer are shown in Table 5, along with the I~ valu~s calculated from Eq. 13 for the~same series.
Both functions; ~T and ~ImU, are plotted in Figure 6.
For any given concentration change the ~l~U function has a greater dynamic range than the ~T function for the ~ame variation in ~bc~ For example, tbe change in concentration between 0.25 and 0.50 absorbance units will produce a ~ of 0.55 units of intensity change and a ~T of 0.23 units of intensity change. Therefore, WO94/L~ 49~ 8 ~ PCT~S92/10~72 the ~ function would be more than 2 times as sensitiv~ as the ~ function. The increa~ed dynamic range results form the fact that Eq. 13 contain~ two terms which change as a function of absorbance ~dulation. Since 6ensitivity of spectroscopic ana~ysis is defined by the magnitude of absorptivi.ty as well as the ~ini~um ab~orbance which can be measured with the required degree of certainty, the ~Io~
function would provide greater spectrophotometric sensitivity.

~A~

~ux VA~UES CALCUI~TED ~OM A~ A~SOR~ANCE SERIES

~bc T = 10~ 2~Io 10~ Io Iou .o~ 0.977 1.976 1.0 3.95 .1 0.794 1.782 1.0 3.57 .2 0.630 1.587 1.~ 3.21 .3 0.501 1.415 1.0 2.91 .4 0.398 1.261 1.0 2.66 .5 ~0.316 1.124 1.0 2.44 1.0 0.100 0.632 1.0 1.73 1.5 0.031 0.352 1.0 1.38 2.0 0.01~ ~ 0.200 1.0 1.21 W094/~U~2 ~ -. . PCT~S92110072 2l4~8~

F. Dvnamic Ranqe of the Optical Det~~iQn_~yst~m The dyna~ic range of a signal tran~duction ~echani~m is understood to be the degree of change of signal p~r unit of change in concentration of the analyte:
Dynamic range = ~ S-~nalT--concentratlon ~ As the degree of ~ignal change per concentration change increases so does the detectabili~y of he signal as-well as the accuracy or resolution of the concentration ~easure~ents.
The V2 (Eq. 3) is plotted in Figure 7(a) as a function of the p~rcent of trans~ission II2%) through the ~econd slit (S2) (Table 4). At 100% transmission, equal intensities pass through two identical slits and the fringe pattern has the most de~ined fringes, where I~ is at its highest value and I~ i5 zero. At 0~
complete loss of the f ringe pattern occurs and only the ~0 diffractioll envelope remains. On the graph shown in Figure 7(a) one can arbitrarily assign 40 units on the visi~ility scale. In the assay, the change in trans~ission through S2 is a function of the concentration of analyte bound to the coverslip, and therefore the I2% isithe cbncentration component. The dynamic range is therefore:
Dynamic range = ~-V-2_ Eq. (14) ; 30 One can see by the graph of Fig. 7(a) that this relationship is not fixed but changes rather ~:- drastically as the ~I2% approaches zero. Of the 40 ;- units of signal division a loss of 50% transmission in WQ94/L~2 21~ ~ ~ 8 4 PCT~S92t10072 I2 produces a visibility change from 1~0 to 0.94, or 2.5 units of signal. The dynamic range would be o.05, calculated thusly:

~ I2~- _5-0 0-05 A loss of transmisæion in I2 from ~o~ to 0% produces a visibility change from 0.94 to 0.0, or 37.5 units of signal, and an increase of dynamic range to 0.75.
Thus, one can s~e that the dynamic range i5 clearly a function of the degree of transmission lo~s and it increase significantly as the transmission approa~hes zero. A more linear plot of the Visibility (V2) data can be generated by plotting -logV2 x 1000 against ~%I2, which is ~hown in Figure 7(b).
However, the change in fringe~visi~ility is not the only u~eful function for calculating a dynamic response. Using t~e model and values from Table 4, the change in ImU and I~ as a function of ~I2% is shown in ~o Figure 8. Where there is no loss of transmission the is at its highest amplitude, 40.0, and the I~ is at its lowest amplitude, 0Ø With complete loss of transmission both values approach identical values of 10.0, which is t~e intensity through one open slit. If the I2% chanqes ~orm 100 to 50% the dynamic range in ter~s of I~ is 0.22;

Dynamic Range = ~ 1_ 0.22 Eq. (15) which is ~ignificantly greater than the dynamic range calculated in terms of ~V2, as plotted in Figure 7(a).

.,,-,j.
WO941L~2 PCT~S92/10072 21~978~

The dynamic response of the I~ over the same range is 0.012. Now, if the transmission loss creates a ~I2~ of 50% o 0~ the dynamic range in terms of In~ becomes ~.3~.
Again, these calculations are ~ade over large, arbitrary segment~ of th~ graph and are gro~s averages.
But they do indicate that an appreciation of the dynamic range must be associated with a specified form of the sîgnal measurement (V or I,~j ~nd must be ~arefully calibrated over the full pereentage of transmission lo~s. Nevertheless, both of the6e graph~
indicate that a ~ignificantly changing function can be correlated with a change in percent of I2 transmi~sion and its contribution to fringe pattern values can be subseguently correlated with analyte concentration.
A very interesting aspect of ~his detection method should be considered at this point. Also plotted in Figure 8 is the function C which represents the loss of intensity o f I2 i f monitored independently. ~ decrease from 100 to 50% transmission produces only 5 units of signal change and has a dynamic range of O.lO. The change from 50 to 0% transmission has the same number - of signal units and thus the same dynamic range.
Therefore O.l is th~ limiting dynamic range for a linear loss of transmission measurement and this is conciderably less than the dynamic range produced by a transduction ~echanism based on a changing nmplitude of an interference fringe, as described in the preceding analyses of the graphs in Figures 7 and 8. -¦
Therefore, the ~lit format, in principle, i5 more sensitive than straight optical density. This enhanced .
i . .... . ~ .

~` I
WO94/L~82 PCT~S92110072 21~978~ . ~

sensitivity can be appreciated by co~paring the relative resolutions of the mea~urements. Resolution is conventiGnally taken ag the inver~e of the dynamic range. For a dynamic range of O.lO, as calculated for the linear loss of trans~i~ion, the resolution, or number that can be measured with certainty, would be lO.00. For a dynamic range of 0.22, as calculated for the change in I~ from lO0 to 50% loss of transmi~ion, the resolution would be 4.54, meaning that the certainty of the measured value would be to a smaller num~er than the preceding case, and thus more accurate.
A dynamic range of 0.75 has a resolution of l.33, providing even ~ore accuracy.
The sensitivity of this assay is defined by the num~er of target molecules required to o~struct the ~: transmi~sion through S2 to produce ~ measurable change in the fringe pattern. Several factors will contribute to the ultifflate sensitivity:

l. The size of the slit.
2. The limiting concentration of immobilized anti-ligand on the cQverslip.
3. The limiting concentration of analyte and gold labeled-antibody coupled to the assay surface from the solution.
:: 25 4. The contribution of the silver precipitation step to the transmission loss.
5. The ability to make a kinetic measurement ~ with the silver precipitation reaction.

¦~ An anælysis of potential sensitivity was based on .

`:?`~
WO 94/12882 P~TIUS92l1007~
214g~84 a paper (Bhatia et al . Anal . Biochem. 178, 403, 1989 ) that optimized covalent i~mobilization of antibody to a glass surf ace . The concentration~ of bound antibody to both coverslip and optical f iber subs~rates are shown 5 below~

COMPA~ISON OF ANTIBQ~Y BINDING To_CovE:p~sLIp A~r2 l~ER

Cover~lip Optical Fiber Area _ 9 6 8 ~Im2 _ 7 4 mm2 Antibody - immobilized 638+134ng 71+3ng (o. 66 ng/mm~, (o. 96 ng/mm2,

4 fmol/mm2) 6 fmol/mm2) Specific antigen 350+50 ng 27+4nq binding (O. 36 ng/mm2;~ (O. 36 ng/mm2) - ~ 15For a slit of size S ~m x lOO ~m, a sensitivity of 6 x 105 molecules may be calculated as follows:

Slit size: 5 ~m x 100 ~m = 500 ~m2= 500 x 106 nm2 Maximum antibody immobilized on slit area:
; ~ ~

2010~ ~m~ 500 ~m2 . x - 0.002 fmol Antigen capture (50%) = O.ool fmol = 6 X 105 mol~cules 25Labeled antibody aapture (50S) = 0.0005 fmol = 3 X 105 molecules ~ ~
W094/1æ~2 PCT~S9~/10072 -~8-The area of the colloid gold label of ~O nm:

Area - ~r2 z 7~6.5 nm2 The theoretical number of molecules r~quired to cover the entire slit can be ~sti~ated by:

Axea of slit 500_x_10~ nm2_ = 7 x 105 molecules Area of gold colloid 706.5 nm/molecule Ratio of e~eerimentally_ca~tured antibodx = 3 x lOs molecules =
Theoretlcal requlrement for ~llt coverage 7 x 10 ~olecules : 0.43 However, this theoretical treatment does not mclude the eff~ct of the silver p~ecipitation step, which appears to make the most significant oontribution in the change of intensity of the coverslip. If 3 x 105 molecules of gold-labeled antibody covers 43~ of : the slit, a red~ction of the transmis~ion to 57% should reduce the intensity of Im~ by ca.25~ (Figure 8). The silver precipitation may extend the sensitivity well below this estimation of 3 x 105 molecules~ In some of the experiments no gold-labeled antibody was observable :~ by ~he naked eye on the coverslip until after t~e precipitation step, at which point the coverslip became significantly opaque.
~ The detection method of the invention possesses a ; numker of substantial advantages. The signal output is in the form of both an amplitude parameter and a patte~n formation which offers greater information WOg4/~2 PCT~S92110~72 21497~4 -3g-content for data analysis over a system based only on intensity measurement. For an interference assay, the signal, in the form of the entire interference pattern, may be ~tored in aomputer memory and compared with a comp~lter standard ~o that a microscopic alteration ts the assay ~tandard pattern, d~tectable only throus~h co~puterized treatment~ could deliver ~n exqui~tely sensitive diagnostic.
Another advantage of the transduction method of the invention is that the amplitude component in the form of I,u has been increased fourfold over a direct measurament of intensity, due to the contribution of the two individual wa~efronts as well as the constructively interf~ring electromagnetic wave~.
Again, this increases the information content of t~e potential ~ignal which should increase the dyna~ic range and thus the sensitivity of the assay. A
sensitivity of less than 105 molecules i possible which would equal or exceed the sensitivity of isotopic methods.
Another underlying quality which should make this ~ethodology very competitive with other detection techniques is the simplicity of its optical desiyn and principles aslwell.as ease of assay procedure. Such simplicity can lead to ruggedness as well as economy.
A compact bench top instru~ent should be easily manu~acturable, with the assay chemistry incorporated into a disposable dipstick. For use with gold~labeled antibody reagents, the dipstick would require incubations with sample, gold-labeled antibody, and silver enhancement solutions, which could be ~ -WO941LU~2 PCT~S92t10072 21~9~84 accomplished in 1 - 2 hours. For u~e with a bacterial cell, simple incubation ~ith the sample, followed by rinsing would complete the procedure. (It ~ay ~e possible that as few a~ 2 - 3 bound cell~ would suffice S to alter the fringe pattern~.
And finally, ~his ~ a generi~ ~ystem which could be used for any clinical or environmental ~amples participating in ~pecific, recognition events. As~ays based on im~unochem~stry and DNA would ~e the initial areas of development. Eventually other-molecular recognition systems, ~uch as che~ical receptors~
cyclodextrins, or ion-selective membranes, may be incorporated into analytical applications. This system could also be used in the study of material science processes, such as monitoring crystal growth, photo-induced polymerizations, and surf~e film depositions.

G. Alternate Embodiments Numerous variations of the basic e~bodiment previously described are contemplated. For example, the slit nu~ber and shape may be varied. The present preferred e~kodiment is a double-slit formed of two identical, parallel slits, but predictable interference phenomens can be producediby other numbers of identical slits. With more than two slits it would be possible to assign individual assays to a different slit or tes~
assays in each slit would be for multiple analytes with a single procedure.
: Although tha visibility equations previously described are not derived for non-identical parallel slits, it may be possible that this assay would be enhanced wit~ non-identical slits.

f~ i:
wo g4/LU~2 ~ 1 4 9 7 8 4 PCTt~S92/10072 ~f Also, hole~ may be u~ed in place of slit to form a circular interference pattern. For example, two pinholes will al~o pr~duce interference fringes and can be precision made to about 2 ~m in diameter, which would give a s~aller total ~urface area than the proposed slits ~and thu~ enhance sensitivity). The assay chemi~try could be perfor~ed a~ drops on the pinhole, therefore using very minute sample volumes.
The assay occurring on the glass coverslip 50 could al~o ~terve to pr~duce two slits S~ and S2 from a single slit S0 as shown in Fig. 9a-9d. In thi~ ca~e the anti-ligan~ ~t is immobilized otn an edge of the coverslip (Fig. 9(b)). The assay would darken the edge of~ the c~verslip, Fig. 9(c), whi~h would then be . 15 positioned over the single slit SO Fi~. 9(a) as shown -~ in Fi~. 9(d) forming two slits S1 and ~.
The as~ay ohemistry could be centralized on a glass slit 9O ~o that gold-labeled antibody assay forms a doubl2 slit 92 as shown in Fig. lO.
~ 20 It is also possible, as shown in Fig. 11, to -~ produce double-slit interference fringes by optical ~- configurations that create virtual slits. Although the principle of the assay would be the same as heretofore described, the al~eration of transmission would require interference with the geometry producing the ~irtual slit. ~
In this arrangement light rays from a single slit ~t S~ pass to the screen 80 via two paths, one of which is direct, A-A, and the other of which is an indirect path, B-B, created by reflection from a mirror M placed in the center line C/L so as to produce a reflection ': t~

WO94/LU~2 . PCT~S92l10072 .
,., ; .
2149~8~

which would appear to emanate ~rom a ~lit S~ locatedequidistant from center line C/L. The ~ource i~
therefore viewed by two paths, one direct and one reflected at a glancing ~ngle in a mirror~ The reflection rever es the phase 180, but otherwise the analysis and the fringes are the same. The area of reflection on the mirror would be the location of the assay chemistry, which would cause appearance or loss of rsflection.
The previous configuration involves use of a Lloyd' 8 mirror. A Fresnel' 8 mirror can be used a. in Fig. 12 to produce two virtual slits which are formed with a point ~ource S projected on two pl~ne ~irrors, Mj and ~, ~utually inclined at a small angle.
The wo resulting virtual images, S~ and S2, act as coherent sources. The separation bf S~ and S2 is defined:

d = 2b ~in where ~ is the angle between t~e mirrors. Again th~
assay would be placed on one of the mirrors and the resultant change in reflectivity used to correlate the analyte concentr~tion.
Two equal prisms when placed toget~er, base to base as in Fig. 13, with their refracting edges parallel, form a Fresnel bi-prism and will divide a cone of light from source S into two overlapping cones.
The prisms thus form two virtual slit images, S~ and S2.
A so-called Billet's split lens arrangement formed by two convex lens can form two images, both of which are WO~4~L~2 2 1 ~ 9 7 8~ PCT~S92l10072 real a~ shown in Fig. 14.
With all of these optical arrangements, interference fringe~ are ~ormed in the region defined by the divergin~ cones fro~ the source~ S~ and S2.
As~ays could be developed w~ich altered the optical arrangement, thus alteri~g fringe pa~tern~. For example, application to the Lloyd' 8 mirror tFig. 11) or Fresnel's mirror (Fig. 12) configurations would require location of the a say chemistry in the region of reflection on the ~irror, which produces the virtual sl~t. `If the assay chsmistry resulted in a lo~s of refl~tivity of the mirror' 8 surface the intensity of reflected ligh~ would decrease and the interferen~e pattern would be modulated. Application to the Fresnel bi-prism ~Fig. ~3) or Billet's split lens (Fig. 14) ~ would require location of the assa~ chemistry on one of - the prisms or lenses so that the post-assay intensity of diffracted radiation would no longer equal the intensity diffracted by the other optical unit in the 20 pair.
As shown in Fig. 15, reactions could be monitored by use of a transport flow-cell microcell 94 which would build up an interfering absorbance on screen 96.
In Fig. 15, B is a r(eference cuvet~e, A i8 an assay cuvette, and they are divided by the cuvette separation, C. The progress of the reaction produces a ~olecular population that would either remain in the flow cell A or become entrapped on its surface, and thu~ ~ake the transmitted light non-identical to that passing through cell B.
Another configuration would have 81it5 S~ and S2 WO94lL~2 21~ 9 7 8 4 PCT~S92/10072 . . ~

incorporated into a transport flow cell 8~ as in Fig.
16. The cell 86 is dispo~ed opposite a Source 85.
Antiligand would be immobilized on S~ and buildup of ligand on it~ surface would alter the fringe pattern as seen by the detector 87.
The dynamic ranqe of the detection ~ystem ~ay~ be enhanced not only by increasing the number of slits, but by increasing the n~mber of slits covered by the assay. Table 6 below shows how the intensity of t:he principle ~axi~a increases as a function of ~2 times the intensity Io of one ~lit.

IN$E~SITY QE PRI~CIPAL MAXI~A FROM N Shr~S IS N2 X I~

15Slits I 4 Io = 10 Io 10 2 4Io 40 ~3 9Io 90 4 16Io 160 25Io 250 6 36Io 360 7 49Io 490 Fig. 19 is a plot of the data from Table 6. As shown in Fig. 17, if the assay slide 100 covers 5 out of 6 slits 104 of the mask 102 the intensity range incxeases by 350 units; instead of 1 of 2 in which case as follows: 6 slits, I = 360~ 1 slit I = 10, the change = 350 whereas in the two slit system ~ = 30 as follows 2 slits I = 40, 1 slit I z 10, and ~ = 40-10 or 30.

W094lLW~2 21 ~ 9 7 ~ ~ PCT~S92/10072 If only 3 slits are co~ered on a 6 slit ~ask a~ in Fig. 18 the resultant Q = 270 computed as follows:

6 ~lits, I z 360 3 ~lits, I = 90 ~ = 270 Note: The slits dedicated to the as~ay need not be adjacent.

The assay ~ubstrate 60 could al50 contain the slit~ Sl and S, (Fig. 20(a)) with the ~nti-ligand 62 immobilized to one of the slits (i.e. S~). The assay reaction causes slit S~ to become less transparent (Fig. 20b). In this embodiment the optical radiation is passed through identical slit material, preserving the symmetrical phases of the two wavefronts. Also, any nonspecific binding of protein from the sample would occur ~qually in th~ regions of Sl and S2, again preserving the symmetry of Il and I2 since both intensities would be identically altered by the contaminating material.
Fig. 21(a~ and 21(b) illustrates alternate embodi~ents in which the mask or assay slit 70/70' is square shaped (Fig. 21a) or triangular tFig. 21tb)) in shape. A single 5quare ~lit will produce an interference pattern 72 as shown in Fig. 21ta).
Likewise, a single triangular shaped slit 70' produces the pattern 72' of Fig. 21(b).
By immobilizing anti-ligand 74 on a diagonal half of ~quare slit 70 as in Fig. 21(c) and reacting the :

~ ~:WO94/1~82 PCT~S92110072 ~
21~978~

anti-ligand with an appropriate ligand in an a~say, the square slit 70 becomes a triangular slit Fig. 21(d), and a triangular based diffraction pattern 72' occur~
Fig. 2~(b). A detector (~ot ~hown) monitors the appearance of the new diffraction pattern 72' to deter~ine the pre~ence and concentration of the ligand.

~q~ivalents Those skilled in the art will reco~nize, or be able to ascertainS u~ing no more than routine experimentation, many equivalents to ~he specific embodi~ents of the invention described herein.
The~e and all other equivalents are intended to be encompassed by the following claims. For example, while the lnvention has been described in connection with light waves in the visible sp~ctrum, the invention is applicable to any electro~agnetic waves throughout the entire spectrum.

Claims (35)

1. A method of optical detection of the presence of a ligand in an assay in which a ligand is reacted with an antiligand to cause a change in the optical characteristics of the assay comprising the steps of:
a) forming an interference pattern by projecting light through openings in a mask disposed in the path of the light so that the diffracted beams emerging from the openings interfere with one another to create said pattern;
b) disturbing at least one such split beam by introducing the assay in the path of the beam thereby changing the interference pattern; and c) observing the change in the pattern to determine the presence of the ligand.

2. The method of Claim 1 wherein the change in optical characteristics results from a phenomena from the group comprising scattering, absorbance, phase change, refraction or polarization of the incident light.

3. The method of Claim 1 wherein the change in the pattern is dependent on the concentration of the ligand in the assay and includes the step of correlating the change to determine such concentration.

4. A method for providing an optical indication of the relative concentration of a ligand and an antiligand in an assay comprising the following steps:
a) forming an interference pattern by projecting light through a mask having N slits wherein N is an integer greater than one;

b) covering N-X slits with an assay comprised of said ligand and antiligand which when reacted results in change in an optical characteristic of the covered slits and a resultant change in the interference pattern, and wherein X is an integer less than N;
c) determining the concentration by measuring the change in the optical characteristic of one or more points on the interference pattern before and after the N-X slits are covered.

5. The method of Claim 4 wherein the points in step (c) are substantially all points within a range of points within the interference pattern.

6. The method of Claim 4 wherein either the ligand or anti-ligand is an optically dense molecule which causes a scattering, absorbance, phase, refraction or polarization change of incident light.

7. The method of Claim 4 wherein N = 6 and X = 5.

8. The method of Claim 4 wherein N = 2 and X = 1.

9. The method of Claim 4 wherein the concentration is measured by determining a visibility parameter and comparing the determined visibility to a reference curve of visibility versus concentration.

10. The method of Claim 4 wherein the assay is comprised of a ligand and a metal-labeled anti-ligand.

11. A method of providing an optical indication of the presence of a ligand or an antiligand in an assay comprising the following steps:

a) forming an interference pattern by projecting light through a mask having N slits wherein N is an integer greater than one.
b) covering N - X slits, where X is an integer less than N, with an assay comprised of said ligand and antiligand; and c) reacting the ligand and antiligand to produce a loss in transparency through the N - X covered slits indicating the presence of the ligand or antiligand.

12. Apparatus for detecting the presence or concentration of a ligand in an assay comprising :
a) a light source for emitting a beam of light;
b) a beam splitter for causing the beam of light to split into at least two split beams, which split beams interfere with one another to form an interference pattern and c) an assay for modulating one of the split beams to change the interference pattern in a manner related to the concentration of ligand in the assay;
d) detection means for quantifying the pattern change and determining therefrom the concentration of ligand.

13. The apparatus of Claim 12 wherein the beam splitter comprises a mask having at least two slits for diffracting the beam of light.

14. The apparatus of Claim 13 wherein the light source is a laser.

15. The apparatus of Claim 14 wherein the detection means comprises an optoelectronic detector.

16. A method of determining the concentration of a ligand in an assay comprising the steps of:
a) generating an interference pattern between two or more electromagnetic waves, said waves resulting from the passage of a beam of electromagnetic energy through a mask containing openings transparent to the electromagnetic energy;
b) reacting the ligand with an antiligand in the assay to cause a change in the response of the assay to the waves;
c) modulating the interference pattern by disposing the reacted assay in the path of one of the waves;
d) determining the ligand concentration by the change in the interference pattern.

17. The method of Claim 16 wherein the electromagnetic waves comprise light waves.

18. The method of Claim 17 wherein the change in response is a change resulting from scattering, absorbance, polarization or refraction.

19. The method of Claim 17 wherein the interference results from passing the waves through N slits formed in a mask wherein N is an integer.

20. The method of Claim 19 wherein the slits are square shaped.

21. The method of Claim 19 wherein the slits are triangular shaped.

22. The method of Claim 19 wherein the assay is comprised of an assay substrate containing the slits.

23. The method of Claim 22 wherein N is greater than one and X is an integer less than N and X slits are covered by the assay.

24. The method of Claim 16 wherein the change in the interference pattern is determined by a densitometer.

25. The method of Claim 16 wherein the change in the interference pattern is determined by calculating the visibility parameter of the pattern and comparing the visibility parameter with a reference curve of visibility versus ligand concentration.

26. The method of Claim 17 wherein the interference pattern is produced by one or more virtual slit(s).

27. The method of Claim 16 wherein the virtual slit(s) are produced by a device from the class comprising Lloyd's mirror, a Fresnel's mirror, a Fresnel biprism, or a Billet's split lens.

28. The method of Claim 19 wherein the slits are incorporated into a flow cell in which antiligand is immobilized.

29. Apparatus for determining the concentration of a ligand in an assay comprising:
a) an interferometer for generating an interference pattern, said interferometer comprising a mask with one or more openings, whereby waves passing through said openings are diffracted to produce an interference pattern;
b) an antiligand, which when reacted with the ligand in the assay causes a ligand concentration dependent change in the response of the assay to the waves;
c) said assay being disposed in the path of one of the waves so as to modulate the interference pattern and d) detector means for detecting changes in the interference pattern caused by the modulations and correlating the detected changes with ligand concentration.

30. The apparatus of Claim 29 wherein the waves are light waves.

31. The apparatus of Claim 29 wherein the modulation results from scattering, absorbance, polarizat:ion or refraction of the light waves.

32. The apparatus of Claim 29 wherein the openings are formed on a substrate which supports the assay.

33. The apparatus of Claim 29 wherein the assay is an immunoacsay.

34. The apparatus of Claim 29 wherein the ligand is taken from the group comprising: molecules, bacteria, polymeric particles, cells and films.

35. The apparatus of Claim 29 wherein the ligand absorbs light, in a prede,termined frequency range and a light source generates the light waves within that range.