WO1996016334A2

WO1996016334A2 - Screening technique

Info

Publication number: WO1996016334A2
Application number: PCT/GB1995/002732
Authority: WO
Inventors: Frank Steven
Original assignee: The Victoria University Of Manchester
Priority date: 1994-11-23
Filing date: 1995-11-23
Publication date: 1996-05-30
Also published as: AU3878495A; WO1996016334A3

Abstract

A method of detecting tumour cells in a cell sample from a human or animal comprises treating the sample with a first competitive inhibitor which binds to isoenzymic forms of guanidinobenzoatase on the surfaces of tumour cells and other cells having distinct isoenzymic forms of guanidinobenzoatase. Subsequently the first inhibitor is selectively displacing from either the tumour cells or said other cells by a second competitive inhibitor which is distinguishable from the first inhibitor whereby the presence of tumour cells in the sample may be detected. For preference, it is the first inhibitor which is displaced from the tumour cells and the second inhibitor is fluorescent. Non-tumour cells do not fluoresce and the tumour cells are easily distingued by their fluorescence.

Description

SCREENING TECHNIQUE

The present invention relates to a method of testing for the presence of tumour

cells or potential tumour cells in a sample of cells taken from a human or animal patient.

The detection of carcinoma or potential carcinoma in a human (or animal) relies

extensively on the examination of one or more samples of cells taken from that human

(or animal). Such examinations are used, for example, in the diagnosis of a person

suspected of having cancer and also in mass-screening programs (e.g. cervical

screening) in which cell samples from many people are screened to detect any cases of

carcinoma or potential carcinoma.

Examples of the types of cell samples which are examined for this purpose

include, but are not limited to, sputum (for detecting lung carcinoma), colonic tissue (for

detecting colonic carcinoma), smears and spreads prepared from scrapings taken from

the uterine cervix (for detecting cervical cancer in women), and breast aspirates (for

detecting breast tumours).

The detection of tumour or potential cells in cell samples is currently effected by

conventional cytological techniques. Briefly such techniques involve staining of the

sampled cells (to give different coloured staining of the cytoplasm and nuclei) and then

examination of the stained cells by a highly trained cytologist who is required to discern

abnormal cells (by the shape and size of their nuclei) within a substantially greater

number of normal cells. The cells in which the cytologist is particularly interested are tumour cells or

potential tumour cells which may develop to tumour cells ('preneoplastic cells'), particularly the early stages of such cells. For example, in the case of cervical epithelial

cells, the cytologist is most interested in locating (a) metaplastic cells, (b) dyskaryotic

cells, and (c) inversion tumour cells for cytological classification by nuclear

morphology.

For convenience, the term 'tumour cells' as used in the subsequent description is intended to cover both tumour cells and potential tumour cells. Such cells are also

referred to herein as 'abnormal cells' or 'cells or interest'.

The current cytological techniques are highly skilled and also labour intensive if a large number of samples are to be analysed such as in a screening program. It is also

known that human error can result in an incorrect diagnosis and there is the further

difficulty that two trained cytologists may come to different conclusions as to whether or not a particular sample contains abnormal cells.

It is known that tumour cells possess a trypsin-like enzyme known as guanidinobenzoatase ("GB") which is a membrane bound enzyme on the surface of the

cells. The enzyme is characterised by its ability to cleave the active site titrants 4- methyl-umbelliferyl-p-guanidinobenzoate ("MUGB") and p-nitrophenyl-p-

guanidinobenzoate ("NPGB"), hence its name. The enzyme is not either not present on the surface of normal mature epithelial cells (as in the case of cervical cells) or is present as a different isoenzymic form (as in the case of colonic cells). GB possesses an affinity (i.e. to attract and hold) for various substrate

compounds. For example, GB will bind 9-aminoacridine (a fluorescent compound) and

it has been previously proposed in J. Enzyme Inhibition 1990 Vol 3 pp 317-312 that

cervical smears may be probed with 9AA so that tumour cells may be located by

fluorescent microscopy. Thus such tumour cells are easily discerned from the

background of (non-fluorescing) normal cells and may be readily subject to

conventional nuclear examination.

There is however a disadvantage in that whilst the 9AA does not bind to normal

mature epithelial cells which lack GB it does nevertheless bind to other non-tumour cell

types which may be present in the sample (e.g. normal cells with isoenzymic forms of

GB, infiltration lymphocytes, phagocytic cells etc) and which each have an isoenzymic

form of GB on their surfaces. Therefore such other cells also bind 9AA and thus

fluoresce when the probed sample is examined using fluorescent microscopy. Whilst

such fluorescing, non-tumour cells are (to the trained cytologist) readily distinguishable

from tumour cells, they do nevertheless represent an interference in the diagnostic

procedure, particularly in an automated diagnostic procedure.

It is therefore an object of the present invention to provide a method of detecting

the presence of carcinoma cells which obviate or mitigates the abovementioned

disadvantages.

According to a first aspect of the present invention there is provided a method of

detecting tumour cells in a cell sample from a human or animal, the method comprising

treating the sample with a first competitive inhibitor which binds to an isoenzymic form of guanidinobenzoatase on the surface of tumour cells and other cells having an

isoenzymic form of guanidinobenzoatase, and selectively displacing the first inhibitor

from either the tumour cells or said other cells by a second competitive inhibitor which

is distinguishable from said first inhibitor whereby the presence of carcinoma cells in the

sample may be detected.

Preferably it is the tumour cells from which the first inhibitor is displaced. In

this embodiment the first inhibitor may be non-fluorescent and the second inhibitor may

be fluorescent so that the presence of carcinoma cells in the sample may be detected by

fluorescence.

The present invention is based on the fact that isoenzymic forms of GB will have

different structures, different functions, and different rate constants for binding and

releasing active site directed competitive inhibitors. Consider the events which take

place when a molecule of a competitive inhibitor is attracted to the active centre of the

enzyme. There are two rate constants for each isoenzyme. namely one for the binding of

the inhibitor and the other for its release from the active centre. Both rate constants are

concentration dependant, the equilibrium being shifted towards binding at higher

concentration of inhibitor and reversed by reducing the concentration of inhibitor.

During competitive inhibition the normal substrate for the enzyme would be used to

displace the inhibitor by increasing the concentration of the normal substrate. However

in the method of the present invention a second competitive inhibitor is used. There is a

second set of rate constants for the binding of the second inhibitor and its release from

the active centres of the isoenzymic forms. Again these new constants are concentration dependant and involve a time factor. One other factor must also be considered - the

status of the active centres of the isoenzymic forms after exposure to the first

competitive inhibitor. This last factor can be regulated by the experimental conditions

simply by changing the concentration of the two competitive inhibitors in the medium

and the time allowed for exchange of inhibitors. It would be difficult to arrange these

changes for enzymes in a solution but it is easy to do so for membrane bound enzymes

on the surface of cells within frozen sections of tissue or on the surfaces of cells on a

slide.

Therefore by challenging the different isoenzymic forms of GB sequentially by

two different competitive inhibitors under the appropriate conditions, different cell types

(each possessing isoenzymes of GB) may be differentiated by their differing abilities to

bind the second inhibitor when they already possess first inhibitor molecule.

Explained in an alternative way, the conditions for treatment with the first

competitive inhibitor are such that this inhibitor binds to all of the various isoenzymic

forms of GB present on the surfaces of the different cell types. Under the conditions of

treatment with the second inhibitor, the latter selectively displaces the first inhibitor

from one isoenzymic form of GB but not (in the time scale of the treatment) from other

forms. Thus if the second inhibitor is fluorescent and displaces the first inhibitor from

the isoenzymic form of GB present on preneoplastic or even tumour cells then such cells

will be the sole fluorescent cells in the sample and may be readily detected, even

amongst several layers of normal cells and other components (e.g. mucous) which may

be present in the sample. The properties required for the first inhibitor are that it will bind to all

isoenzymic forms of GB but, under conditions of treatment with the second inhibitor, is

selectively displaced from one isoenzymic form of GB, preferably the isoenzymic GB of

the surface of the tumour cells. The first inhibitor preferably comprises N-substituted

guanidino compounds. The preferred first inhibitor is p-guanidinobenzoate. Other

possibilities include MUGB, a-N-acetyl-agmatine, and benzamidine.

The second inhibitor is One which, in the conditions and time scale of the treatment with the second inhibitor, displaces the first inhibitor from one of the

isoenzymic forms of GB. If the first inhibitor is guanidinobenzoate then it is preferred that the second inhibitor is 9-aminoacridine which may be used selectively to displace guanidinobenzoate from tumour cells thereby leaving the tumour cells as the sole cells which are capable of binding the fluorescent 9AA. The tumour cell surfaces fluoresce

and can be used to locate those cells of cytological interest. Other possibilities for the second inhibitor include Rhodamine-a-N-Agmatine and Dansyl-a-N-Agmatine.

Examples of cell samples on which the method of the invention may be effected include sputum, colonic tissue, thin needle aspirates of breast tissue and scrapings taken from the uterine cervix. The samples may be in the form of multilayers of cells (e.g. smears) or monolayers (spreads) or sections of tissue.

Prior to the treatment with the first inhibitor, the cell sample should be treated to

remove inhibitor proteins from the cell surface GB isoenzymes. This may be achieved by standard methods, e.g. by the use of formaldehyde in isotonic saline. Furthermore, if desired or necessary the samples may be stained for the purpose of subsequent nuclear analysis of tumour cells detected by the method of the invention. Staining may, for

example, be by means of haematoxylin. Since the latter is an acid stain, it is desirable to

adjust the pH of the cells to _& 7.5 prior to application of the first competitive inhibitor.

The conditions for treatment with the first inhibitor are not critical but should be

such that the first inhibitor will bind to isoenzymic forms of GB present on various

different cell types. Typically the first inhibitor will be used in solution at a

concentration of 10^" 1 to 10^" 1 M and the cells will be treated with this solution for a period of 2-3 minutes until all cells possessing the active GB have the first inhibitor bound thereto.

In the case where the first inhibitor is p-guanidinobenzoate and the second

inhibitor in 9AA, the following procedure may be used to check whether or not a

particular set of conditions under which the cells are treated with the first inhibitor lead to saturation of all isoenzymic forms of GB with p-guanidinobenzoate. The cells are

initially treated with the p-guanidinobenzoate under the particular conditions. Subsequently the cells are challenged, with 10^'3M 9AA for 20 seconds, washed for 10 seconds, and then checked for 9AA yellow fluorescence. If all cells appear negative

then this confirms that all binding sites for 9AA are already occupied with the p-

guanidinobenzoate. If necessary, the time and concentration of treatment with the first inhibitor can be varied to achieve this aim. Similar procedures may be used for other

combinations of first and second inhibitors.

Treatment conditions with the second inhibitor are more critical since the treatment must only be effected so that the second inhibitor displaces the first inhibitor from one isoenzymic form of GB. This can generally be achieved by effecting the

treatment with the second inhibitor only for a limited period of time so that the second

inhibitor is not present for a sufficient length of time to displace the first inhibitor from

all isoenzymic forms of GB but only from one of the selected form. Typically the

second inhibitor will be used in a solution of lower concentration than that of the first

inhibitor and the treatment times will be lower. Thus for example, the concentration of

the second inhibitor in its solution may, for example, be of the order of 10^"4 to 10^'3 M

and the treatment time may be for 1-2 minutes. The exact treatment conditions

(concentration and time) required for the second inhibitor may be optimised by routine

experiment.

As a result of the treatment of the invention, any cells of cytological interest in

the sample may be detected due to the detectable difference between the binding

affinities of the first and second competitive inhibitors. As indicated, it is preferred that

the second competitive inhibitor displaces the first competitive inhibitor from the

tumour cells. It is also preferred that the first competitive inhibitor is non-fluorescent

and that the second competitive inhibitor is fluorescent. As such, any tumour cells

present in the sample may be detected by virtue of their cell surface fluorescence. If the

cells were stained they may be destained prior to use of the present technique, then cells

of interest in the sample may be readily located by virtue of their fluorescence and

characterised by nuclear analysis.

A particular advantage of the method of the invention is that it may be used to

identify tumour cells on samples of cells which are several years old. It is conventional practice to retain cell samples which have been examined by conventional screening

techniques. If a query about the original diagnostic analysis of the cells arises (e.g. in a

medical negligence case) then the cell sample may be retrieved from the archive and re-

examined by the present technique which provides a new parameter for diagnosis,

namely the presence of an active cell surface GB isoenzyme. We have been able to

demonstrate that the technique of the present invention may be used on archive samples

which are at least 10 years old and still identify the cells of interest. Such archive

samples are usually PAP stained and may be used in the method of the invention after

de-staining, e.g. with alcohol/HCl for three hours. After application of the present

method to the samples, they may be re-stained for example with PAP. Such re-stained

samples regenerate the original archive material in which the nuclei are found to be

more clearly stained than previously.

In carrying out the method of the first aspect of the invention the cells to be

examined are generally applied to a suitable support (e.g. a microscope slide or similar),

treated with the inhibitors as described more fully above, and then overlaid with a cover

element to await examination. Generally the method is conducted under aqueous

conditions and is perfectly satisfactory provided that the cell samples are analysed

relatively soon (e.g. within a few hours) after their preparation. If however the samples

are to be kept for a longer period of time, e.g. days or weeks then there are number of

disadvantages, a) the inhibitors bound to the cell surface guanidinobenzoatase are, in effect,

extracted into the aqueous phase remaining under the cover element.

b) the aqueous phase and the cells tend to dry out unless a sealant is applied around

the edges of the cover element; and

c) if inhibitor bound to any tumour cells in the sample is fluorescent, then free

radicals in the aqueous phase may quench the fluorescence.

According to a second aspect of the present invention there is provided a method

of preparing a sample of cells from a human or animal for subsequent examination to

detect the possible presence of tumour cells in said sample, the method comprising

applying the cells to a support, treating the cells with a first competitive inhibitor which

binds to isoenzymic forms of guanidinobenzoatase on the surface of tumour cells (if

present) and other cells having an isoenzymic form of guanidinobenzoatase, selectively

displacing the first inhibitor from either the tumour cells or said other cells by a second

competitive inhibitor which is distinguishable from said first inhibitor, applying a non-

drying organic liquid to said cells, and overlying the cells with a cover element.

We have found that the presence of a non-drying organic liquid at the interface

between the cells and the cover element ensures that there is no extraction of inhibitor

bound to tumour cells (if present). Additionally, the liquid prevents drying-out of the

cells without the need for a separate sealant provided around the edges of the cover element. Furthermore, if the inhibitor bound to tumour cells (if present) is fluorescent,

then the organic liquid prevents quenching of the fluorescence.

Typically, cell samples prepared using the method of the second aspect of

present invention may still be successfully examined to determine the presence (or

otherwise) of tumour cells after years and can be stored for archival purposes as

compared to less than 2-4 hours for samples produced without the use of the non-drying

organic liquid.

The non-drying organic liquid is preferably an oil. particularly an Immersion

Oil. We have found Cargille's non-drying Immersion Oil type B (Code No. 1248) to be

particularly suitable.

The supports to which the cells are applied may, for example, be a microscope

slide and the cover element may be a standard cover slip.

A non-limiting procedure for carrying out the method of the second aspect of the

invention is given below

1. A smear is formed on a slide from a sample of cells which are then stained with

thionin using a standard procedure.

2. The slide is placed in formaldehyde solution (10% v/v in phosphate buffered

saline) for twenty minutes to release inhibitor proteins from cell surface GB isoenzymes.

3. The slide is washed in water and then equilibrated in PBS for 5 minutes. 4. The GB on the cell surfaces is treated with a solution of guanidinobenzoate (10^" ²M) for 1 minute.

5. Fluid is drained from the slides and the cells then gently blotted or alternatively dipped in water to remove excess guanidinobenzoate.

6. The slides are placed in a tank of 9-aminoacridine ( 10^' M) for 3 minutes.

7. The slides are drained and gently blotted.

8. The cells are stained with propidium iodide (10 M) for 1 minute.

9. The slides are rinsed and washed for 1 minute in PBS.

10. The slides are drained and dried at room temperature.

11. A drop of oil is placed on a glass cover slip which is placed over the cells so that

the oil forms a film between the coverslip and the slides.

The slides can now be examined by fluorescence microscopy for cells with

yellow/orange fluorescence (i.e. cells of interest to cytologists). The analysis for fluorescence may be effected in an automatic analyser. The nucleus of each of these cells may be analysed by a Xillix automated cervical smear analyser (now called Cyto-

Savant imager).

The cells at stage 9 also exhibit good fluorescent properties. However at this stage the cells are immersed in an aqueous medium and the fluorescence begins to fade due to extraction of the bound 9-AA/PI into the aqueous phase. After 2-4 hours there is

marked fading - this would be undesirable for automated analysis since the 'fading' would differ between each slide especially if slides were held overnight prior to

analysis. Furthermore the slides dry out.

On the other hand, the dry cells placed under a non-drying oil as of step 1 1

remain well stained with no loss of fluorescence.

The slides as obtained at step 1 1 can be stored in archive records for future

reference (e.g. legal matters) and retain their fluorescence and nuclear staining.

It will be appreciated that the fluorescence of the cells and the thionin (or other

nuclear) staining procedure can be linked to provide two parameters of abnormality for

each cell, i.e. cell surface fluorescence and nuclear morphology.

Screening of different cell types may require different staining protocols for

optimum results. For example cells of cytological interest in thionin prestained sputum

samples can be stained as follows:-

1. Thionin staining of all nuclei

2. Treatment with 10% formaldehyde in PBS for 20 min. to remove inhibitor

proteins from GB

3. Nuclear staining with Gill's Haematoxylin one min. to block nuclei against

fluorescent probes

4. Wash in water

5. Equilibrate pH in PBS for 5 min.

6. Drain and dry slides in air

7. Place under 10-20 μl guanidinobenzoate solution (10^"2M in PBS) for 12 min. 8. Drain

9. Place in 9AA ( 10^'3M in PBS) for 1 min.

10. Drain

11. Wash for 10s in water

12. Air dry

13. Cover with Cargille Immersion oil type B

14. View in LEITZ Diaplan Microscope with cube I₃ (513-719) in place.

a. Cells of no interest lack yellow surface fluorescence and can be ignored.

b. Cells of cytological interest, possess yellow cell surface fluorescence and thionin stained nuclei.

These cells can then be selected for Cyto-Savant nuclear analysis.

Typical cell of interest are: (i) Dyskaryotic and metaplastic epithelial cells, (ii) Carcinoma cells

(iii) Squamous cells which have normal moφhology but are intensely cell surface

fluorescent.

These are of interest because they may be MACs (iv) Some but not all of the histiocytes are negative, but in some sputum intensely

stained histiocytes are seen. The method of the first and second aspects of the invention as described above is

perfectly effective for detecting the presence of tumour cells in a cell sample.

Nevertheless we do recognise that cytologists have been trained to make a final

diagnosis as to the presence of tumour cells and/or the classification of such cells by

nuclear moφhology. As such, cytologists may not feel able wholly to rely on the

method of the aforementioned application.

Therefore according to a third aspect of the present invention there is provided a method of detecting the possible presence of tumour cells in a plurality of cell samples from humans or animals, the method comprising

(i) screening the cell samples using the method of the first aspect of the invention

so as to select those samples in which tumour cells are detected, and (ii) examining stained nuclei of said selected samples to make a diagnosis as to the

presence of said cells and/or a classification thereof.

The third aspect of the invention has the advantage that step (i) provides a

preliminary screening operation (which may be automated) so that only those samples in which tumour cells are detected need be subject to detailed examination of nuclear

features. The third aspect of the invention therefore combines the advantages of step (i) (which may be readily automated) with conventional nuclear examination (on which

cytologists currently rely heavily) for the piuposes of final diagnosis. A further major

advantage of the third aspect of the invention is that it provides (for an abnormal cell)

two parameters of abnormality on the same cell, namely that provided by the method of the kind defined and an abnormal nuclear moφhology. This greatly enhances the degree of certainty in the final diagnosis.

In one embodiment of the third aspect of the invention, all of the cell samples to be examined are treated to stain their nuclei prior to being subjected to step (i). As such any sample selected from step (i) for further, nuclear analysis will be ready for analysis

in step (ii) without further treatment. Alternatively the cell samples may be subject to

step (i) without prior staining of their nuclei, in which case only the samples selected form that step need be treated with a nuclear stain prior to further analysis in step (ii).

For all embodiments of the third aspect of the invention it is preferred that, in step (i), the second inhibitor is fluorescent and displaces the first inhibitor from any tumour cell present in the cell sample whereby such tumour cells may be readily detected by virtue of their fluorescence.

A variety of nuclear stains may be used in the method of the invention. Examples include non-fluorescent nuclear stains such as PAP (as conventionally used for examination of cervical smears), Haematoxylin, Thionin. Neutral Red, Light Green, Acid Fuschin, and Safronin. It is also possible to use fluorescent probes for nuclear

staining, e.g. DAPI, Ethidium Bromide and Calcefluor. Most conveniently the stain is

PAP, Thionin or Haematoxylin.

If the cell samples are to be stained prior to step (i) then the preferred nuclear

stain is Thionin or Haematoxylin since neither stain the cytoplasm of the cells. If PAP is to be used as the stain then it is preferably employed after the cell samples have been

screened in step (i) as it stains the cytoplasm and can therefore interfere with step (i). The method of the invention may be applied to archival samples which have

previously had their nuclei stained. Thus such samples may be treated in accordance

with step (i) and the selected samples then further analysed for their nuclear detail.

If the archival slide has been stained with PAP then, if desired, the

may be selectively destained using ethanol followed by treatment of the sample in

phosphate buffered saline prior to step (i). Alternatively the archival slide may be

completely destained using HC1 and ethanol prior to a first nuclear staining operation

effected either before or after step (i).

The method of the third aspect of the present invention lends itself readily to

automation. Consider for example that step (i) is effected to locate tumour cells by

means of a fluorescent probe. All of the samples (which may for example be on

microscope slides) are examined in an automatic screening apparatus to detect

fluorescence in any of the samples. Those samples which exhibit fluorescence may then

be presented to a cytologist for further examination (after nuclear staining). In a

development of this procedure, an image of the cells of interest may be obtained and the

image of those cells (showing their nuclear detail) presented to the cytologist for

examination.

A suitable apparatus for use in an automated embodiment of the present

invention is PAPNET, an automated PAP analyser used for cervical smears which

examiners nuclear moφhology and presents a cytologist with a panel of selected

abnormal cells based on nuclear parameters. According to a fourth aspect of the present invention there is provided a kit of

reagents comprising, in separate vessels,

(a) a first competitive inhibitor which is capable of binding to isoenzymic forms of

guanidinobenzoatase on the surface of tumour cells and other cells having an

isoenzymic form of guanidinobenzoatase; and

(b) a second competitive inhibitor which is capable of selectively displacing the first

inhibitor from said isoenzymic forms of guanidinobenzoatase and which is detectably

different from said first inhibitor.

Preferably the first inhibitor is an N-substituted guanidino compound, most

preferably p-guanidinobenzoate. Further examples include 4-methyl-umbelliferyl-p-

guanidinobenzoate (MUGB), a-N-acetyl-agmatine, and benzamidine.

The first inhibitor is preferably provided as an aqueous solution (preferably in

isotonic saline) having a concentration of 10^" M to 10^* M.

The second inhibitor is preferably fluorescent. The preferred second inhibitor is

9-aminoacridine. Other examples include Rhodamine-a-N-Agmatine and Dansyl-a-N-

Agmatine.

The second inhibitor may be provided as an aqueous solution (e.g. water,

isotonic saline or phosphate buffered saline) having a concentration of 10^"3 M to 10^"4 M.

The kit may further comprise, in separate vessels, one or more of the following

additional reagents. (c) a formaldehyde solution for use in releasing inhibitor proteins from cell

surface GB isoenzymes. The formaldehyde solution may. for example, be provided as a

5%-10% v/v solution in PBS solution:

(d) a nuclear blocking agent to prevent the first or second inhibitor binding

to cell nuclei. A preferred nuclear blocking agent is haematoxylin (Mayers or Gills

variety) provided as an aqueous solution; and

(e) propidium iodide which may for example be used as a solution (e.g. in

water, isotonic saline, or PBS) having a concentration of 10^"4 M to 10^" M.

The manner in which the above reagents (a)-(e) are used is as disclosed above.

The kit may further comprise a non-drying liquid (e.g. an Immersion Oil) for use

as disclosed above.

The kit may further comprise a nuclear stain (e.g. Thionin. or PAP) for use as

disclosed above.

The invention will be further described by way of example only with reference

to the following non-limiting Examples and accompanying drawings, in which:

Figs. 1 (a)-(f) illustrate the results of Example 1; and

Figs. 2-4 illustrate the results of Example 3.

Example 1

Cervical smears were first treated with 10% v/v formaldehyde in isotonic

solution for 2h in order to release inhibitor proteins from the cell surface GB isoenzymes

followed by placing the slides in Meyer's haematoxylin for 1 min. to stain the nuclei so that nuclear moφhology could be observed by conventional microscopy employing

white light. The pH of the cells was then adjusted to 7.5 by placing the slides in 2%

NaHCO₃ for 5 min. and the excess fluid drained from the surfaces of the slides.

The slides were dried in air and 10-20ml of a solution containing

guanidinobenzoate (10^"2M) (ex Sigma Chemical Company) placed over the surface of

the cells beneath a glass cover slip, the treatment with guanidinobenzoate could equally

well be carried out in a glass staining trough filled with the reagent. After 1 min. the

excess guanidinobenzoate was rinsed from the slides placed in 9-aminoacridine (10^'3 M

in isotonic saline) (ex Sigma Chemical Company) for 2 min; the excess 9AA was rinsed

off the slides and each slide washed for 10s in a tank of fresh isotonic saline.

A glass coverslip was placed over the cells (covered with isotonic saline) and the

cells examined by fluorescence microscopy using a Leitz Diaplan fluorescence

microscope with barrier filter K470 and cube [G] (Leitz catalogue No. 513609). Under

these conditions those cells which lack active GB do not bind 9AA and do not fluoresce.

Cells which possess active GB on their surfaces exhibit yellow fluorescence, and these

cells could be clearly distinguished from the mass of non-fluorescent cells present in

cervical smears. The yellow fluorescence due to bound 9AA was enhanced by treating

the cells with 10⁵M propidium iodide for 1 minute and washing the excess propidium

iodide from the cell surfaces for 10 seconds in isotonic saline. The propidium iodide co-

stacks on bound 9AA and as a result the cell surfaces exhibit an orange/yellow

fluorescence. An automatic camera and Kodak ASA 400 colour film were used to record the results which are shown in Figs. 1 (a)-(f) (see below) as black and white

copies of the original colour photographs (magnification x500).

The fluorescence is generally seen in the copy photographs as being the lightest

region. The nucleus of the cell is usually out of focus if the cell surface is in focus.

The nuclei of cells which exhibited cell surface orange/yellow fluorescence were

examined in white light by switching the optics on the Leitz microscope. Epithelial

cells with cell surface GB were shown to be CIN cells, according to their nuclear

moφhology.

A summary of the results shown in Figures l(a)-(f) is given below.

Fig. la Normal mature epithelial cells lack GB and do not fluoresce.

Fig. 1 b Two linked GB positive cells which were observed to have enlarged

nuclei. The nucleus of one cell can just be seen in this focal plane (arrow).

Fig. lc Single GB positive cell with two nuclei (arrows).

Fig. Id Small clumps of positive cells with enlarged nuclei.

Fig. le Three individual positive cells with a clump of positive cells on the edge

of the picture. The nucleus of one cell is clearly enlarged (arrow). These cells are part

of a thick layer of normal cells and the orange/yellow fluorescence of the GB positive

cells clearly defines these abnormal cells.

Fig. If A clump of abnormal cells with surface GB. The nuclei of several of

these cells can be seen as a dark region within the cell and are obviously enlarged

compared to those in normal epithelial cells (see Fig. la). As shown in Fig. la, normal cells lack cell surface orange fluorescence: note the

presence of normal epithelial cells in most of the other prints even though these contain

examples of abnormal cells. The abnormal cells possess cell surface GB in an active

form, which binds 9AA and these cells exhibit orange/yellow fluorescence. The

haematoxylin stained nuclei of these cells could be visualised by switching d e optics to

conventional light microscopy enabling their classification as metaplastic. dyskaryotic,

and carcinoma cells etc., by a cytotechnician or a cytologist. The puφose of the present

technique is to provide a rapid method of selecting cells of interest to the cytologist for

nuclear analysis. It can be seen from the photographs that the use of a fluorescent probe

does in fact achieve this purpose. This study is an example of how the principle of

differential competitive inhibition of cell surface GB isoenzymes can be used to locate

abnormal epithelial cells in cervical smears.

Example 2

In this study we examined sputum samples obtained from 40 patients presenting

at a hospital chest clinic. At the time of the fluorescent analysis we had no knowledge

of the patients history or the state of their lungs. In all cases the fluorescent indication of

malignant cells was confirmed by conventional cytological analysis and by subsequent

biopsy of the patient.

(i) Fluorescent location with Rh-Agm (comparative^

Conventional sputum smears were placed under 15-20ml of an isotonic saline

solution containing 10^"5M Rh-Agm (Rhodamine-Agmatine) for 5 min. beneath a glass cover slip. The excess reagents were washed from the slide by a quick rinse in isotonic

saline before the surface of the slide was protected with a cover slip and examined by

microscopy. We employed a Leitz Diaplan microscopy with cube {N} (Leitz catalogue

No. 513609) or cube {A} (Leitz catalogue No. 513596) in place. Cells with GB bound

Rh-Agm and exhibited orange cell surface fluorescence with cube {N} in place. Cells

lacking GB (most cells of a normal smear) did not bind Rh-Agm and appeared brown

like the surface of the glass slide when viewed with cube {N} in place. When cube {A}

was in place, the nuclei of all cells exhibited blue auto-fluorescence enabling cells to be

classified according to their nuclear details and moφhology. (ii) Fluorescent location with 9AA

In this technique we used the method of the invention inhibition.

The slides with sputum cells were placed in 10% w/v formaldehyde for 2h,

washed and stained for 1 min. with Meyers haematoxylin. The pH of the cells on the

slides was adjusted to 7.5 by placing the washed slides in a beaker of 2% NaHCO₃ for

10 min. The slides were allowed to dry in air and then covered with 10-20ml of

guanidinobenzoate (10^"2M dissolved in water), covered with a glass coverslip for 1 min.

and then the excess reagent rinsed from the slide. The treated slides were then placed in

a tank of 9AA (10^" M in isotonic saline) for 1 min., and finally washed for 10s. in

isotonic saline. The slides were examined by fluorescence microscopy with cube {G}

(Leitz catalogue No. 513602) and barrier filter K570 in place. Cells with active GB

exhibited yellow cell surface fluorescence. Results 2

Staining with Rh-Agm

In a typical sputum sample obtained by a deep cough there are a mass of

different cell types derived from the oral cavity, lungs and bronchial passageways.

These cells are usually thickly coated in a gelatinous layer of mucus and bacteria, the

cells often being in clumps or in layers several cells thick. Cells with GB on their

surfaces bind Rh-Agm and their surfaces exhibit orange fluorescence when viewed in

the microscope with filter cube {N} in place.

Thus etaplastic cells of epithelial origin bound Rh-Agm on their surfaces and

fluoresced orange when viewed with cube {N} in place. These cells have enlarged

nuclei with a characteristic moφhology.

The mature epithelial cells (such as those derived from the oral cavity of a

normal subject) lacked GB, failed to bind Rh-Agm and exhibited no orange cell

fluorescence. On the other hand, normal sputum contained many histiocytes which

possessed GB and bound Rh-Agm resulting in orange cell fluorescence. Histiocytes are

easily recognisable by their nuclear composition and foamy appearance which can be

observed with both cubes {N} and {A} in place.

Staining ___h 2ΔΔ after differential competitive inhibition with guanidinobenzoate.

Cells which retained active GB bound 9AA and exhibited intense yellow surface

fluorescence. Histiocytes were ignored in this study. The only other cells which

exhibited yellow surface fluorescence were epithelial cells with typical dyskaryotic

nuclei cells. The nuclei of these cells (which had previously been loaded with haematoxylin) were clearly visible in white light microscopy and enabled their cell

characterisation on the basis of nuclear moφhology.

Example 3

For each of Figs. 2-5, the cell samples were prepared as disclosed in Example 1 save that the step of treating with haematoxylin was omitted.

Fig. 2 shows a cell sample (glandular cells and mature epithelial cells of the cervix) which have been pre-stained with Thionin and then treated in accordance with the method of the first aspect of the invention such that abnormal cells are caused to fluoresce. Fig. 2a shows the sample as viewed using fluorescence microscopy. The abnormal cells have bright surfaces as viewed in Fig. 2a whereas normal cells do not fluoresce. Cell samples which do not display any fluorescence may be passed as "clear" and only those cells which display fluorescence presented to a cytologist for further examination of nuclear moφhology. Alternatively the cells exhibiting fluorescence may be submitted to nuclear analysis in an automated procedure, those cells with abnormal nuclei being recorded and presented to the cytologist for diagnosis. Fig. 2b shows the same sample as in Fig. 2a but viewed simultaneously in white light and fluorescence microscopy. The enlarged nuclei of the cell, on which the cytologist would make a diagnosis, are clearly visible.

Fig. 3 shows a cell sample (dyskaryotic cervical cells) which has been pre¬ stained with Thionin and treated in accordance with the method of the first aspect of the invention such that abnormal cells are caused to fluoresce. Fig. 3a shows the sample as viewed using fluorescence microscopy and shows dyskaryotic cells which have light coloured surfaces by virtue of the fluorescence which they exhibit. Fig. 3b shows the sample as viewed under fluorescent and white light, whereas Fig. 3c shows the sample as viewed under white light only. The enlarged nuclei of the cells can clearly be seen in Figs. 3b and 3c by virtue of the Thionin nuclear staining.

Fig. 4 again shows a cell sample (invasive carcinoma of cervix) which has been pre-stained with Thionin and treated in accordance with the method of the first aspect of the invention such that abnormal cells are caused to fluoresce. Fig. 4a shows invasive carcinoma cells which are identified by their fluorescence (light coloured surfaces). Fig. 4b shows the same sample as seen under both fluorescence microscopy and white light. Again the enlarged nuclei, on the basis of which a diagnosis may be made, are clearly seen.

It should be understood that the present technique is designed to speed the

preselection of smears which contain potentially interesting epithelial cells for the

cytologist to examine and diagnose, thus reducing the tedium of selection and delay for

the patients awaiting their diagnosis.

We believe that this method can be further modified by the inclusion of

fluorescent DNA specific stains (e.g. DAPI, propidium iodide, etc.) to stain cell nuclei

stoichiometrically. In this way, nuclear features such as nuclear size and shape, ploidy

and foremost texture features describing the distribution of the DNA in the cell nuclei,

can be measured. On those cells with active GB on their surfaces. The above

modifications would require the cell surface GB fluorescent probe to have distinct

spectral properties from the fluorescent stain used to define the nuclear detail.

Quantitative data of atypical cells can be obtained which can then be used for objective

classification of the pre-malignant or malignant lesions for more precise diagnosis

and/or prognosis. The above procedure can be fully automated for clinical use.

Claims

1. A method of detecting tumour cells in a cell sample from a human or animal, the

method comprising treating the sample with a first competitive inhibitor which binds to

isoenzymic forms of guanidinobenzoatase on the surface of tumour cells and other cells

having distinct isoenzymic form of guanidinobenzoatase. and selectively displacing the

first inhibitor from either the tumour cells or said other cells by a second competitive

inhibitor which is distinguishable from said first inhibitor whereby the presence of

tumour cells in the sample may be detected.

2. A method as claimed in claim 1 wherein the first inhibitor is displaced from the

tumour cells.

3. A method as claimed in claim 2 wherein the second inhibitor is used at a

concentration of the order of 10 to 10^" M and the treatment time with the second

inhibitor is for 1-2 minutes.

4. A method as claimed in claim 2 or 3 wherein the first inhibitor is non-

fluorescent and the second inhibitor is fluorescent.

5. A method as claimed in claims 2 to 4 wherein the first inhibitor is p-

guanidinobenzoate, MUGB, a-N-acetyl-agmatine, or benzamidine.

6. A method as claimed in claim 5 when dependent from claim 4 wherein the

second inhibitor is selected from 9-aminoacridine, rhodamine-a-N-agmatine and dansyl-

a-N-agmatine.

7. A method as claimed in claim 4 wherein the first inhibitor is p-

guanidinobenzoate and the second inhibitor is 9-aminoacridine.

8. A method as claimed in any one of claims 1 to 7 wherein the cell sample

comprises sputum cells, colonic cells, or cervical cells.

9. A method as claimed in any one of claims 1 to 8 wherein the cells are provided as frozen sections of tissue or on a slide.

10. A method of preparing a sample of cells from a human or animal for subsequent examination to detect the possible presence of tumour cells in said sample, the method comprising applying the cells to a support, treating the cells with a first competitive inhibitor which binds to isoenzymic forms of guanidinobenzoatase on the surface of tumour cells (if present) and other cells having an isoenzymic form of guanidinobenzoatase. selectively displacing the first inhibitor from either the tumour cells or said other cells by a second competitive inhibitor which is distinguishable from said first inhibitor, applying a non-drying organic liquid to said cells, and overlying the cells with a cover element.

1 1. A method as claimed in claim 10 wherein the non-drying liquid is an Immersion Oil.

12. A method as claimed in claim 10 or 1 1 wherein the support is a microscope slide or the like and the cover element is a cover slip.

13. A method as claimed in any one of claims 10 to 12 wherein the first. inhibitor is displaced from the tumour cells.

14. A method as claimed in claim 13 wherein the second inhibitor is used at a concentration of the order of 10 to 10^" M and the treatment time with the second inhibitor is for 1 -3 minutes.

15. A method as claimed in claim 13 or 14 wherein the first inhibitor is non- fluorescent and the second inhibitor is fluorescent.

16. A method as claimed in any one of claims 13 to 15 wherein the first inhibitor is p-guanidinobenzoate. MUGB, a-N-acetyl-agmatine. or benzamidine.

17. A method as claimed in claim 16 when dependent from claim 6 wherein the second inhibitor is selected from 9-aminoacridine. rhodamine-a-N-agmatine and dansyl- a-N-agmatine.

18. A method as claimed in claim 15 wherein the first inhibitor is p- guanidinobenzoate and the second inhibitor is 9-aminoacridine.

19. A method as claimed in any one of claims 10 to 18 wherein the cell sample comprises sputum cells, colonic cells, or cervical cells.

20. A method of detecting the possible presence of tumour cells in a plurality of cell samples from humans or animals, the method comprising (i) screening the cell samples using the method as claimed in any one of claims 1 to 9 so as to select those samples in which tumour cells are detected, and

(ii) examining stained nuclei of said selected samples to make a diagnosis as to the presence of said cells and/or a classification therefor.

21. A method as claimed in claim 20 wherein, in step (i) the second inhibitor is fluorescent and displaces the first inhibitor from any tumour cells present in the cell sample.

22. A method as claimed in claim 20 or 21 wherein the cell samples are subject to a nuclear staining procedure before step (i).

23. A method as claimed in claim 20 or 21 wherein samples selected from step (i) are then subjected to a nuclear staining procedure prior to examination in step (ii).

24. A method as claimed in any one of claims 20 to 23 wherein the nuclear stain is PAP, Haematoxylin, Thionin, Neutral Red, Light Green, Acid Fuschin or Safronin.

25. A method as claimed in claim 24 wherein the stain is PAP, Thionin or Haematoxylin.

26. A kit of reagents comprising, in separate vessels,

guanidinobenzoatase on the surface of tumour cells and other cells having an

isoenzymic form of guanidinobenzoatase; and

different from said first inhibitor.

27. A kit as claimed in claim 26 wherein the first inhibitor is an N-substituted

guanidino compound.

28. A kit as claimed in claim 27 wherein the first inhibitor is p-guanidinobenzoate.

29. A kit as claimed in claim 26 wherein the first inhibitor is 4-methyl-umbelliferyl-

p-guanidinobenzoate (MUGB). a-N-acetyl-agmatine or benzamidine.

30. A kit as claimed in any one of claims 26 to 29 wherein the first inhibitor is

provided as an aqueous solution having a concentration of 10^' M to 10^" M.

31. A kit as claimed in any one of claims 26 to 30 wherein the second inhibitor is

fluorescent.

32. A kit as claimed in claim 31 wherein the second inhibitor is 9-aminoacridine.

33. A kit as claimed in claim 31 wherein the second inhibitor is Rhodamine-a-N-

Agmatine or Dansyl-a-N-Agmatine.

34. A kit as claimed in any one of claims 25 to 33 wherein the second inhibitor is

provided as an aqueous solution having a concentration of 10^" M to 10 M.

35. A kit as claimed in any one of claims 25 to 34 further comprising a

formaldehyde solution.

36. A kit as claimed in any one of claims 25 to 35 further comprising a nuclear

blocking agent.

37. A kit as claimed in claim 36 wherein the nuclear blocking agent is haematoxylin.

38. A kit as claimed in any one of claims 25 to 37 further comprising propidium

iodide.