The present invention relates to human monoclonal antibodies to the Rh(D) antigen of human red blood cells. In particular, it relates to such antibodies of the IgGl sub-class which may be used to detect ^{^"} not only the normal Rh(D) antigen on either D-positive or "weak D" or D ^u cells, but also important variants of the Rh(D) antigen.

Of the antigens of the so-called Rh blood group system, the Rh(D) antigen is responsible for some of the most severe reactions following transfusion to a patient with corresponding antibody. Since an Rh(D-) individual with anti-Rh(D) who receives Rh(D+) blood is liable to suffer substantial red blood cell (RBC) destruction due to the Rh(D) phenotype incompatibility, blood of donors and blood transfusion recipients is routinely classified as Rh(D+) or Rh(D-) by agglutination tests with anti-Rh(D) antibody. The Rh phenotype of RBCs is commonly further defined with reference to the Fisher-Race system, which is based on the assumption that the inheritance of the Rh antigens is determined by three pairs of allelic genes, C-c, D-d and E-e, acting at very closely linked loci. According to this theory, a person may inherit a set of three Rh genes from each of his parents (i) C or £, (ii) D or ά _f (iii) E or e (no d antigen has as yet been identified, but the symbol *d' is used to indicate the presence of a gene, allelic to the D gene, which does not produce D antigen) . For example, an Rh(D+) person may inherit CDe from one parent and cde from the other. The frequencies of the commonest Rh gene combinations as determined with

reference to the Fisher-Race system for an English population, together with the 'short symbols' which are used, particularly in speech, are given in Table 1 below.

Table I Frequency of common Rh genes for an English Population

Short symbol CDE nomenclature Frequency (%)

R~ CDe 40.8

cdE 1.2

0.01 rarer

Despite expansion over the years, the Fisher- Race system has not been adequate to account for all the reactions that have been observed with the Rh system (Mollison _r P.L. (1983) Blood Transfusion In Clinical Medicine, 7th edn. , Blackwell Scientific, Oxford) o Nevertheless, the World Health Organisation has recommended that in the interest of simplicity and uniformity this nomenclature should be universally adopted and all Rh genotypes given hereinafter are defined on the basis of the conventional Fisher- Race system.

In addition to the need for anti-Rh(D) antibody for Rh-typing of RBCs, such antibody is also importantly required for passive immunisation of Rh(D-) mothers to prevent haemolytiσ disease of the newborn (HD ) . This condition arises in newborn Rh(D+) infants of Rh(D-) mothers previously sensitized to Rh(D)

antigen as a result of IgG anti-Rh(D) antibodies crossing the placenta during pregancy and causing foetal RBC destruction. Sensitization of the Rh(D-) mother to Rh(D) antigen may have occurred at the birth of an earlier Rh(D+) child due to some foetal RBCs entering the maternal circulation and being recognised by the maternal immune system. To reduce the incidence of HDN, it is routine practice in the United Kingdom and many other countries to give anti-Rh(D) antibodies to Rh(D-) mothers immediately after the birth of an Rh(D+) infant so that any Rh(D+) RBCs which have entered the maternal circulation are rapidly removed (Mollison, P.L. (1983) loc. cit.; Laros Jr., R.K. (1986), "Erythroblastosis Fetalis" in "Blood Group Disorders In Pregnancy", Ch. 7, p. 103) .

At ^' the present time, anti-Rh(D) antibody for use in both Rh-typing of RBCs and passive immunisation of Rh(D-) mothers is largely obtained directly from female donors immunised during pregnancy or from immunised male volunteers. The success of the programme of post-partum prophylactic administration of human anti-Rh(D) immunoglobulin to Rh(D-) women has, however, resulted in a dramatic reduction in the number of naturally alloimmunised women (Urbaniak, S.J., "RhD haemolytic disease of the newborn: the changing scene", Br. Med. J. (1985) 291, 4-6) . Also, deliberate immunisation of individuals with Rh(D+) RBCs carries the risks common to receiving any transfusion of RBCs,e.g. risk of transmission of hepatitis viruses and HIV. Hence, there is much interest in obtaining human monoclonal anti- Rh(D) antibodies for both diagnostic and therapeutic purposes. As stated above, in routine blood testing, blood types are divided into Rh(D+) and Rh(D-) on the basis of the apparent presence or absence

of Rh(D) antigen on the RBCs as indicated by agglutination tests with anti-Rh(D). However, a small number of persons with apparently Rh(D-) blood have RBCs that are not directly agglutinated by anti-Rh(D) during such routine testing, but that do react when the D-typing test is performed using selected anti-R (D) reagents by the indirect antiglobulin test. Cells thus identified are designated D ^u . The frequency of the D ^u phenotype is about 0.2% overall, 0.6% among Caucasians, and about 1.5% of all Rh(D-) gravid women. At least three different mechanisms may be responsible for the expression of the D ^u phenotypes (1) hereditary absence of a portion of the complete Rh(D) antigen, (2) gene interaction with suppression of D by C in the trans position, and (3) a D gene producing a weak antigen.

In the early 1950s, reports first appeared of the presence of anti-Rh(D) in individuals of the D ^u phenotype following blood transfusion with Rh(D+) blood o pregnancy resulting in the birth of a Rh(D-f-) infant. It later became apparent that in some individuals whose blood is classified Rh(D+) parts of the Rh(D) antigen are missing from the RBCs. When exposed by transfusion or pregnancy to Rh(D+) RBCs carrying the complete Rh(D) antigen, persons carrying an incomplete Rh(D) antigen on their RBCs are capable of making alloanti-D against the Rh(D) antigen portion they lack. The blood of such individuals is called D variant when the RBCs react directly with routine anti-Rh(D) reagents - or Du variant when the cells react only by the indirect antiglobulin teehniσue _*

The observation that allo anti-R (D) can be produced in patients who have Rh(D+) RBCs has led to common usage of the term "D mosaic" to describe the Rh(D) antigen in its complete native form. Routine anti-Rh(D) reagents generally cannot differentiate

those RBCs that lack part of the D mosaic from those that have all the D components.

The D variant phenotypes have been categorised by Tippett and Sanger (Vox. Sang. (1962)2, 9-13). This system is based on the interaction of RBCs and serum from D- and D ^u variant individuals. The six categories (see Table II below) allow for expansion; subdivisions are already recognised in categories III, IV and V. Categories I and II have been found to have so many similarities that they are now generally considered as a single sub-group.

Table II

An alternative, but lesser used, classification by Wiener uses letters A,B,C,D instead of Roman Numerals. Although there is no direct correlation between the two systems, it is often considered that D ^B and D ^Vl are interchangeable. Althouqh the frequency of D and Du variant individuals within the human population is relatively

low, the total number of individuals of these blood types who potentially have some risk of effective anti-Rh(D) formation as a result of exposure by blood transfusion or pregnancy to non-variant Rh(D+) cells is far from insignificant. Moreover, in addition to Rh(D-) women who give birth to Rh(D+) or D ^u infants, D ^u variant women who give birth to an Rh(D+) infant may also benefit from post- partum anti-Rh(D) treatment to reduce the risk of HDN (White, CA. et al. (1983) Am. J. Obstet. - Gynecol. 145, 1069-1073) . Anti-sera capable of distinguishing D and D ^u variant RBCs are not widely available. Hence, provision of anti-Rh(D) monoclonal antibodies with a range of binding specificities for D and D ^U variant RBCs is seen as useful in enabling the more ready identification and categorisation of individuals possessing such cells (especially D or Du variant pregnant females who are suitable candidates for prophylactic anti-Rh(D) treatment) as well as for providing further structural information on the Rh(D) antigen complex.

Human monoclonal anti-Rh(D) antibody production has previously been achieved b s- (a) directly cloning Epstein Barr virus transformed B lymphocyte cell lines (hereinafter referred to as EBV-transformed LCD derived from B lymphocytes of anti-Rh(D) positive donors (see GB-A 2127434; Crawford et al. (1983) Lancet 1, 386=388 and Paire et al (1986) Immunol. Lett. .13, 137-141) , (b) cloning hybridoma cell lines formed by fusing anti-Rh(D) producing, EBV-transformed LCL with mouse, mouse-human or human myeloma cell lines (see co-pending British application no= 8709748, - Thompson et al (1986) Immunol. 5_8 _f 157-160 and EP-A-0162918) , or

(c) by fusion of a human LCL with immune B cells (Lowe et al (1986) Vox. Sang. 51, 212-216) .

By cloning EBV-transformed LCL from 3 anti- Rh(D) positive donors, we have been able to obtain, however, monoclonal anti-Rh (D antibodies of the IgG class which have a particularly useful binding specificity spectrum not shown for any previously disclosed anti-Rh(D) monoclonal antibody reagent.

According to the present invention, we provide human monoclonal antibodies having the following essential characteristics: (a) exhibiting activity against Rh(D) antigen, but not C, c, E or e antigens of the Rh blood group system;

(b) being IgGl proteins;

(c) having kappa light chains; (d) being Glm (3) or Glm (1, 17) allotype;

(e) exhibiting activity against D ^u cells by an indirect antiglobulin test;

(f) eexxhhiibbiittiinngg aaccttiivviittyy aa<gainst D , D ^v and D ^Tar variant antigens; and

VI B (g) being inactive against D or D variant antigens, and antigen-binding fragments thereof.

Such monoclonal antibodies can be employed as routine anti-Rh (D) reagents to classify RBCs as Rh(D+), D ^u or Rh(D-).

For this purpose, a monoclonal antibody of the present invention may be employed either alone or in combination with one or more further anti- Rh (D) monoclonal antibodies having one or more additional binding specificities. Thus, for example, .a monoclonal antibody of the present invention may be advantageously blended with a further monoclonal

VI antibody capable of binding the D variant to provide an anti-Rh (D) reagent of broader specificity

VI capable of classifying D variant RBCs as D-positive.

In such an anti-R (D) reagent, an IgGl antibody of the present invention may, for example, be combined

with an IgGl monoclonal antibody of the type disclosed in our co-pending International application of even date herewith claiming priority from GB- A 8722020, having the following binding characteristics: (a) exhibiting activity against Rh(D) antigen, but not C, c, Ξ or e antigens of the Rh blood group system; (b) exhibiting activity against D ^V , D ^Tar

D VT and D "__ variant antigens; and (c) being substantially non-reactive with non-papain treated D IV cells in an

IAG test.

Amongst such antibodies, particularly preferred for use in combination with an anti-Rh(D) monoclonal antibody of the present invention is the monoclonal antibody designated B7 deposited at the European

Collection of Animal Cell Cultures, Porton Down,

U.K= under accession No. ECACC 86091603 on 16th

September 198 _ If monoclonal antibodies of the present invention are used for Rh-typing in parallel with an anti- Rh(D) reactive against the D VI variant, eg. an appropriate polyclonal anti-Rh(D) serum, those blood samples giving a positive result in an agglutination test with the latter, but negative results with a monoclonal antibody of the invention can be predicted to be mainly or entirely of the D VI category (since this is virtually the only D variant antigen against which the new monoclonal antibodies are inactive) . It has been established that amongst individuals classified as Rh(D+) or D by a conventional agglutinatio test, but who are capable of making anti-Rh(D), a high percentage have the D VI variant antigen

(Mollison, PoL. (1983) in "Blood Transfusion In Clinical Medicine",. Ch. 8, p 339).. One use of the monoclonal antibodies of the invention is, indeed, in investigating the incidence of individuals of the D VI type in the population.

A monoclonal antibody of the present invention may also be of particular value for use in an anti- Rh(D) typing reagent to supplement the specificity of an anti-Rh(D) with no or only weak anti-D ^u activity, i.e. insufficient activity against D ^u cells to be able to reliably distinguish such cells from D-negative cells in a conventional agglutination test. Indeed, under FDA regulations in the U.S.A. governing commercial anti-Rh(D) typing reagents, it is obligatory for such a reagent to be able to distinguish D ^u RBCs from truly D-negative RBCs. Especially preferred amongst combination anti-Rh(D) reagents of the present invention are such reagents satisfying the above condition wherein an IgG anti- Rh(D) of the invention is employed together with an IgM anti-Rh(D) with only weak D ^u activity, e.g. an IgM monoclonal anti-Rh(D) selected from the monoclonal IgMs of the deposited hybridoma cell lines MAD-2 (ECACC 86041803) and FOM-1 (ECACC 87021301) , which form inter alia, the subject matter of published European Patent Application 0251440. Such a reagent will advantageously be further complemented with at least one further IgG monoclonal anti-Rh(D) antibody which individually exhibits activity with D ^u red cells by the indirect antiglobulin test, such that the blended reagent reacts by the same test with D ^u , D ^IV , D ^V and D ^VI cells. When Rh typing is carried out with such a reagent, D-positive cells will firtly be directly agglutinated by the IgM anti-Rh(D). The remaining non-agglutinated cells (apparently D-negative) may then be subsequently divided into truly D-negative and Du cells by addition of conventional Coomb's reagent for an indirect antiglobulin test, whereupon D ^u cells binding IgG antibody will be agglutinated and thus distinguished.

The monoclonal anti-Rh(D) antibodies of the invention can be made by conventional methods known

for the production of monoclonal antibodies and in particular by the culture of EBV-transformed human B-lymphocytes selected on the basis of secretion of anti-Rh(D) immunoglobulin having the characteristics set out above for the required antibodies. The culture supernatants so produced constitute a further feature of the present invention.

We have now investigated in detail 9 cloned EBV-transformed LCL which produce IgGl anti-Rh(D) monoclonal antibodies as defined above. All these cloned cell lines were obtained by starting with peripheral B lymphocytes from 2 chosen anti-Rh(D) donors and employing the procedure described in GB-A 2145113 or a substantially similar procedure to establish and clone EBV-transformed LCL producing a monoclonal antibody of the desired specificity (see Example 1) . In continuous culture using RPMI- 1640 medium supplemented with 10% (v/v) myeoplasma free-fetal calf serum. Go2 mg/ml arginine and antibiotics to prevent myeoplasma growth, they have been found to be highly stable and to provide a culture supernatant having an anti-Rh(D) titre, as determined by an indirect antiglobulin (IAG) assay in low ionic strength saline versus R _j ^ (CD /CDe) RBCs, in the range 2000-8000. Such a culture supernatant is suitable for use in Rh-typing without the need for concentration and indeed may be diluted for use. Four of the selected cell lines (Al _p A2 and A3 derived from a donor A) have been shown

to maintain a stable anti-Rh (D) titre in continuous culture for over 2 years. Four further cell lines capable of providing a culture supernatant as above (Bl, B2, B3 and Bll derived from a donor B) have been maintained in continuous culture for over

8 months without substantial decrease of anti-Rh(D) titre. Antibody production characteristics of the above-mentioned specific clones in continuous culture are summarised in Tables Ilia and Illb below.

Table Ilia

IAG titre (3% RBCs in low ionic strength saline) R _χ r cells 64-512

Table Illb

culture Al A2 A3 Bl B2 B3 Bll B12 B Supernatant IAG reactivity of supernatant with 3% RBCs in low ionic strength saline. (Grade: 0 to 6)

It was further found that those monoclonal antibodies of the invention that were tested reacted by IAG with RBCs of phenotypes R _g rG-, hr ^s - _f R__,R and ₂ ^R _Z ^{DUt are} negative with r ^re Gr, r , r , r ^,s r, hr ^B -, r ^lW r and Rh33+. 5 G allotyping of the antibodies of clones

Al, A2, A3, Bl, B2, B3, Bll, B12 and B13 has shown that they fall into two allotype groups. Using the oHoO. (1974) notation, the IgGl anti-Rh(D) antibodies of clones Al, A2 and A3 were found to

10 be of the Glm (1, 17) allotype, whereas the antibodies of the remaining above-mentioned clones derived from donor B were found to be of the allotype Gl (3). ^~ The latter antibodies are of particular interest from the point of view of providing a monoclonal

15 antibody preparation for use in post-partum immunisation of Rh (D-) mothers. In general, the antibodies in the Glm (1, 17) allotype group were active against the S ^" antigen, whereas those of Gl (3) allotype were not.

20 . IgGl anti-RhD antibodies are generally poor promoters of phagocytosis of Rh (D+) RBCs by monocytes

and macrophages. However, we have found that IgGl anti-Rh(D) antibodies of the Glm(3) allotype, e.g. the IgGl antibodies of clones Bl, B2, B3, B l, B12 and B13, unlike IgGl anti-Rh(D) antibodies of the Glm(1,17) allotype, are highly effective in mediating lysis of sensitised RBCs by K lymphocytes in an antibody-directed cell cytoxicity (ADCC) assay. This is in keeping with the previously reported observation of Parinaud et al. that foetal haemolysis is more severe with maternal IgGl antibodies of the Glm(4) allotype, i.e. Glm(3) allotype according to the W.H.O (1974) nomenclature (Am. J. Obstet. Gynecol (1985) 1111-1115) .

Thus, according to a further aspect of the present invention, we provide a monoclonal IgGl antibody of the present invention, preferably a monoclonal antibody of the present invention having the allotype Glm(3) , for use in passive immunisation of an Rh(D-) or D- or D ^u - variant mother after the birth of an Rh(D+) child to prevent sensitisation of the mother to Rh(D) antigen. A sterile solution of such an antibody for human injection may be formulated in any physiologically acceptable aqueous medium, for example isotonic phosphate buffered saline or serum. Alternatively, the antibody may be supplied in a freeze-dried formulation ready for reconstitution prior to use. To provide a highly efficient prophylactic preparation for use in the prevention of HDN, a monoclonal anti-Rh(D) of the present invention, especially such an antibody of the Glm(3) allotype, may desirably be employed with one or more further anti-Rh(D) antibodies, for example one or more further anti-Rh(D) antibodies promoting phagocytosis of Rh(D+) RBCs _irι vivo, e.g. an anti-Rh (D) monoclonal.antibody of the IgG3 sub-class, such as an lgG3 monoclonal anti- Rh (D) of our co-pending International Application

of even date herewith claiming priority from GB-A 8722019, the contents of which are incorporated herein by reference. The aforementioned lgG3 monoclonal anti-Rh(D) antibodies are exemplified by the monoclonal antibody of the deposited cell line ECACC 86091606. For routine use, ideally a prophylactic pharmaceutical composition of the present invention will include an anti-D antibody.

According to a still further aspect of the present invention, we provide a method of Rh-typing of RBCs wherein an aqueous solution of a monoclonal anti-Rh(D) immunoglobulin of the present invention is employed. The monoclonal immunoglobulin is preferably contained in a culture supernatant which may be used directly or, more usually, after dilution. Particularly preferred for use in Rh-typing are culture supernatants containing monoclonal anti- Rh(D) immunoglobulin according to the present invention which will agglutinate at high dilution (e.g. 1:1000 dilution) enzyme-treated RBCs carrying the Rh(D) antigen and in the IAG test will agglutinate D ^u RBCs at, for example, Is10 dilution.

As hereinbefore indicated, it may be desirable to blend an IgGl antibody of the present invention with one or more further anti-Rh(D) monoclonal antibodies of different specificity, e.g. a further IgGl antibody having anti-n VI activity. Suitable diluents include physiological saline or phosphate buffered saline, advantageously containing bovine serum albumin and a surfactant or suspending agent such as Tween 80 or methyl cellulose.

The cell lines A3 and B2 were deposited on

16th September 1987 at the European Collection of Animal Cell Cultures, Po ton Down, TJ„K. under accession numbers ECACC 86091605 and ECACC 86091604 respectively.

Further details of the preparation of the above-noted deposited cell lines of the present

invention and the identifying characteristics of the culture supernatants obtainable by continuous culture of these are provided in Example 1 of the following non-limiting examples.

EXAMPLE 1

(i) Establishment and Cloning of Anti-Rh(D) Producing EBV-transformed LCL

(a) Sources of B Lymphocytes

Donor A: female, immunised during her first and only pregnancy (which resulted in delivery of a normal Rh(D+) infant), boosted with 0.5 ml packed Rh(D+) (R ₂ ^r ) RBCs 4 years after parturition, and a peripheral blood sample obtained 8 days after boosting when her serum anti-D level was 60IU/ml. Donor B: male, initially immunised by transfusion in 1966, boosted 6 times since and last boosted 13 days before donating a "buffy coat" fraction (white cells) in 1985 when his serum anti-D level was 318IU/ml.

(b) Establishment of cell lines

Peripheral blood mononuclear cells from donor B were separated on Lymphoprep (Nyegaard and Co) , incubated in the presence of EBV (1 ml culture supernatant from filtered myeoplasma free B95-8 cell line per 10 ⁷ cells) at 37 ^β C for 1 hour and washed in phosphate-buffered saline (PBS) . Aliguots were either (a) plated at 0.5x10 cells/ml in 2 ml wells using lymphoblastoid cell culture medium (RPMI- 1640 medium containing 10% (v/v) myeoplasma free foetal calf serum (FCS) _p 0.2 mg/ml arginine, 100 lϋ/ l penicillin (Glaxo) , 50 pg/ml streptomycin (Glaxo) , 25 lU/ml polymixin (Glaxo) , 25 μg/ml kanamycin (Gibco) , 20 μl/ l fungizone (Squibb) , 25 μg/ml gentamycin sulphate (Sigma) and 20 pg/ml trobicin (UpJohn) ) , supplemented with either 1% (v/v) phytohaemag-

glutinin (PHA) or 0.5 μg/ml cyclosporin A (CsA) ; or

(b) enriched for surface anti-D positive lymphocytes by rosetting with bromelain-treated OR^ _j fCDe/cDE) RBCs prior to plating out as above.

Four cell lines from donor B were set up: LCL and enriched LCL with either initial PHA or CsA supplementation.

Isolated peripheral blood mononuclear cells from donor A were similarly infected with EBV.

After washing with PBS, the cells were resuspended in lymphoblastoid cell culture medium supplemented with 1% (v/v) PHA and dispensed into 2 ml wells above a feeder layer of mouse peritoneal macrophages. All the cultures were subsequently incubated at 37°C in 5% C0 ₂ , 95% humidified air. Medium changes were performed every 3 to 4 days and, after 3 weeks culture, the cells were transferred to 50 ml flasks. All the cell lines (except that giving rise to clone B3) were enriched by rosetting at 3-4 weekly intervals.

(c) Cloning

Cells were plated out at limiting dilution at 5 and 10 cells per well in flat bottomed 96- well plates over a feeder layer of mouse peritoneal macrophages (Doyle et al. (1985) Human Immunology 13, 199-209) . Cultures were fed once a week and after 3-4 weeks cloned cells positive for anti- D were grown up.

(d) Derivation of clones

Three clones producing anti-Rh(D) antibody of the IgGl sub-class (designated Al, A2 and A3)

were derived from B lymphocytes of donor A. The polyclonal cell line giving rise to these clones was initiated with PHA and sequentially enriched for anti-Rh(D) positive cell by rosetting 13 times before cloning.

Five further clones producing anti-Rh(D) antibody of the IgGl sub-class (designated B^, B ₂ , B,, Bll and B12) were derived from B lymphocytes of donor B via a polyclonal cell line initiated with PHA. Bl and B2 were derived from cell lines repeatedly (6x) selected by rosetting after establishing an LCL. Bll and B12 were derived from a cell line resetted only twice and B3 was produced from an LCL maintained for 1 year without rosetting before cloning. [It should be noted that Bll was erroneously designated Cl in the priority application (GB-A 8722018) ].

A further anti-Rh(D) IgGl producing clone (B13) was derived via a donor B polyclonal cell line supplemented with CsA. This cell line was established from lymphocytes selected for anti- Rh(D) immediately after transformation by EBV and was subsequently re-rosetted four timeso

The results of tissue type and karotype analysis of the above selected cell lines are set out in Table IV below.

Table IV

Cell Line Al A2 A3 Bl B2 B3 Bll B12 B13 r-τz y

B Male

1,3

8,35

1,3

xy most cells diploid

(e) Quantitation of anti-Rh(D) activity and IgG in culture supernatants Anti-Rh(D) activity in the supernatants was quantified against British national standards by

Auto Analyser. The mean of at least two determinations was calculated. The quantitative estimation of IgG was performed by ELISA (modification of the method of Wakefield et al. in Clin. Chi . Acta. (1982) 12 , 303-310) with at least eight determinations for each supernatant. Coating antibody (affinity purified goat anti-human IgG (Sigma)) was used at 1/200 in 0.05M carbonate buffer pH 9.6. Supernatants and standard (purified human IgG (Sigma) ) were diluted in RPMI 1640 + 10% FCS. Peroxidase-conjugated goat anti-human IgG (Sigma) was diluted 1/500 in PBS + 0.05% Tween 20 and the substrate was TMB (3,3 ^s , 5,5'-tβtra ethyl benzidine) .

Not all the cell lines established under different experimental conditions showed stable antibody production. However, all the cell lines

which were subsequently cloned had maintained high titres (over 1/33,000 by microtitre) of anti-D for over 6 months. All clones from these cell lines were positive for anti-D and maintained their titres throughout the duration of continuous culture (Al, A2 and A3 - over 2 years; Bl, B2, B3 and Bll - over 8 months) . The doubling time was 3-7 days. The cells grew well in suspension culture with no loss of antibody production.

(f) Immunoglobulin class and subclass determinatio

An immunodot assay (McDougal et al. (1983) 63, 281-290) was used to determine the reaction of the monoclonal anti-Rh(D) antibodies absorbed to nitrocellulose with anti-IgG, anti-IgM, anti- kappa and anti-lambda antiserum (Serotec} ; positive reactions were detected with peroxidase-conjugated anti-sheep IgG (Serotec) followed by colour development with 4-chloro-l-naphthole The IgG subclass was evaluated by agglutination of anti-D coated RBCs by monoclonal anti-subclass antibodies (Unipath) .

The presence of single discrete bands for antibody derived from the selected cell lines after SDS-polyacrylamide gel electrophoresis was evidence of monoclonal!ty.

(g) SDS-Page and Western Blotting.

Iscove's supernatants (serum free) were electrophoresed under reducing conditions on 15% polyacrylamide gels (Laemmli, Nature (1970) 227, 680-685) . The separated proteins were then electrophoretically transferred to nitrocellulose membranes (Brunette, Annals Biochem. (1981) 112, 195-203) , which were probed with anti-IgG antiserum (Serotec) and detected as above.

(h) Protein A absorption

2 ml volumes of supernatants were run twice down a 25 mm (1ml) column of Protein A Sepharose C1-4B (Sigma) and absorption of anti-Rh(D) assessed by titration.

(i) Gm allotyping

RBCs were coated with the monoclonal anti- D antibodies and agglutination assessed using panels of Gm allotyping reagents (Birmingham or Amsterdam) . The Gm allotypes of the monoclonal antibodies of clones Al, A2, A3, Bl, B2, B3, Bll, B12 and . B13 together with other characteristics are given in Table V below.

Table V

Identification of immunoglobulin class, subclass, and Gm allotype

Monoclonal Al A2 A3 Bl B2 B3 Bll B12 antibody:

_r (kDa) : 55.5 55 55 54 54 54 55

Protein A absorption ÷÷+ +++ 4,4.4. +++ +++ +++ +++ + ^• ++

Gm allotype: Gl Gl Gl Gl m(l,17) m(l,17) m(l,17) m(3) Km-1 Km=l m- Km-1

* no.t determined

( ii ) Serology

Anti-D titration was performed by adding 50 pi of 0.1% suspension of bromelain-treated OR-^R _j (CDe/cDE) erythrocytes to 50 pi of supernatant, neat and serially diluted two-fold, in V-well microplates After incubation at 37°C for 60 mins, and centrifugation at 600 rpm for 3 mins, they were read macroscopically by tilting the plate at 70° and allowing negative reactions to trail. The degree of agglutination was graded in conventional manner and the titres given as the highest dilution giving complete agglutinati (see Table Ilia) . Anti-D titres were also determined by the IAG test using rabbit anti-human IgG and 3% R,r, R,R,, R,r or R r cells in low ionic strength saline (LISS) . The anti-D activity of the monoclonals against a panel of D variant red cells was also assessed using saline, albumin, papain and IAG (in LISS) tests. In a separate series of tests, the reactivity of the monoclonal anti-Ds of Al, A2, A3, Bl, B3 and Bll with D ^U red cells (15 sets of R, ^u r cells and 10 sets of ₂ ^U r cells, each set being taken from a different individual) was assessed by IAG (in LISS) using untreated supernatants. Using the same procedure, culture supernatant of Bl was further compared with culture supernatant of B2 against 1 set of R^r cells and 3 sets of Rur cells. The supernatants were tested on a Technicon

Autogrouper 16C at dilutions ranging between 1:5 to 1:10,000.

Tests against cells of "normal" RhD positive or RhD negative phenotypes by albumin, papain and indirect antiglobulin techniσues revealed that all the IgGl monoclonal antibodies showed specificity for the D antigen (see Table VI below) . None of

the monoclonal anti-Ds were reactive by saline. When tested by IAG against "partial" D positive cells, the antibodies agglutinated D , D and D ^Tar red cells, but not D ^VI or D ^B (see Table VII below). As shown in Table VIII, the supernatants of Al, A2, A3, Bl, B3 and Bll were all found to show stronger reactivity by IAG tests with 25 D cells (i.e. weak D) than polyclonal antiserum (the routine reagent employed by the South Western ^{^} Regional Transfusion Service in the U.K) . Using the same IAG test procedure, culture supernatant of Bl was found to exhibit similar or identical reactivity to culture supernatant of B2 against 4 further sets of D ^U cells (see Table IX) . All the IgGl monoclonal antibodies reacted more strongly with R-^ (CDe/cde) or ^R _χ (CDe/CDe) cells than with either D ^u or D cells, though there were some differences in titres between the supernatants. When tested separately on a Technicon Autogrouper 16C, supernatants of A2, Bl and Bll could be used diluted to 1:1,000 for use as a routine anti-D and at 1:10 dilution to distinguish D from D negative cells.

(Grade: 0 to 5]

(Grade: 0 to 5)

Table IX

Monoclonal Bl B2 antibody (titres by IAG)

RBCs ^r

^R o ^r

^R «o ^r

(iii) Lymphocyte ADCC assay

Equal volumes (50 μl) of target cells (chromium-51 labelled R-^ RBC suspension), effector cells (K lymphocytes) and anti-D culture supernatant were incubated overnight at 37 ^Θ C in microplates after mild centrifugation, and the chromium-51 release measured (Urbaniak (1979) Br. J. Haematol. 42, 303-314). The effector:target cell ratio was 15:1. The culture supernatants of clones Bl, B2, B3 _r Bll, B12 and B13 containing a monoclonal anti- Rh(D) of the Glm(3) allotype, unlike the other culture supernatants with antibody of the Glm(1,17) allotype, were found to be highly active in mediating lysis of sensitised RBCs in the presence of K cells (see Table X below) <_

Table X

* not determined

(iv) U937 Monocyte rosetting and phagocytosis assay lOOμl packed 0R,R ₂ RBCs were sensitised with

500ul anti-Rh(D) (previously adjusted to 1 ug/ml) at 37° for 60 minutes, washed and resuspended at

1 x 10 cells/ml in RPMI. U937 cells were taken

5 in the log phase of growth and cultured for two days either in the presence or absence of interferon-/ g (Amersham) at 50 U/ml. 45 x 10 RBCs were then added to a pellet of 1.5 x 10 U937 cells and mixed, giving a ratio of 30:1. For the rosetting assay,

10 the cells were incubated at room temperature for

5 minutes, spun at 600 rpm for 3 minutes and examined in a hae ocytometer after a further 5-20 minutes. Phagocytosis was assessed immediately after incubating the cells at 37° for 3 hours. Results were expressed

15 as the percentage of monocytes with one or more adherent or phagocytosed RBCs.

A comparision. of the results obtained wth culture supernatants of clones Al, A2, A3, Bl, B3 and Bll and a conventional polyclonal anti-Rh(D)

20 serum is given in Figures 1 and 2.

(v) Macrophage binding assay

RBCs (R ₂ r) (1 volume) were sensitised with monoclonal anti-Rh(D) (2 volumes of untreated culture supernatant) and incubated with monocyte-derived cultured macrophages. The macrophages were stimulated with 500 U/ml of recombinant immune interferon (Biogen, Geneva) during the 48 hours prior to their use in the assay. The binding of RBCs to macrophages was assesed microscopically and expressed as the macrophage binding index (= number of red cells attached to or ingested by 100 macrophages) .

Table XI below shows the results obtained with culture supernatants of clones Al, A2, A3, Bl, B3 and Bll. The ability of these supernatants 5 to bring about RBC-macrophage interaction was very poor compared to that of the polyclonal anti-Rh(D) serum (anti-Rh(D) - 43 U/ml) , which served as a positive control.

Table XI

Source of polyclonal anti-Rh(D) Al A2 A3 Bl B3 Bll anti- Rh(D) serum.

Macrophage 20 28 24 5 20 17 416-

500 binding index

EXAMPLE 2

Solution for Rh(D)-Phenotyping of RBCs

In general, it is preferred for the above purpose to use a blend of an anti-Rh(D) monoclonal antibody according to the invention (e.g. Bll)

with a further IgG monoclonal antibody having anti- D activity, for example the monoclonal antibody of the cell line B7 (ECACC 86091603) of our copending International patent application No. of even date herewith, the contents of which are incorporated by reference into the present specification.

Solution for Manual Use

The final blend is 1:1:1 Bll:B7: diluent,

Diluent 100 ml 30% Bovine Serum Alubumin

2.42g KH ₂ P0 ₄ 2.77g Na ₂ HP0 ₄ .2H ₂ 0 4.50g NaCl 0.2 ml Tween 20

1.0Og a ₃ To 1.0 litre with distilled water: pH 6.8

This blend can be used in all manual tests for

D and D ^u typing, e.g. microtitre, microplate, IAG.

A blend of 1:1:2 Bll:B7: diluent may also be used.

Blend for Machine Use

This blend may be used, for example, in a Technicon autogrouper 16C

Pre-blend reagents B11:B7 1:1

For D-Phenotyping (D-positive v. D-negative) the solution comprises 1:1000 blend: diluent, and for D determinations (D v. D-negative) the solution comprises 1:5 blend:diluen .

ilαent 2% Bovine Serum Albumin in 1.3% physiological saline containing 13.5% methylcellulose.

The monoclonal antibody B7 may be similarly blended with other antibodies of the present invention, e.g. the monoclonal antibodies of the deposited cell lines A3 (ECACC 86091605) and B2 (ECACC 86091604) to provide an anti-Rh(D) reagent with broad specificity for the various D-variant antigens, including D T, and exhibiting activity against D ^u cells by an TAG test.