WO2012045913A1

WO2012045913A1 - Method for isolating cells and a cell population thereof

Info

Publication number: WO2012045913A1
Application number: PCT/FI2011/050862
Authority: WO
Inventors: Jukka Partanen; Jari Natunen; Annika Kotovuori; Suvi Natunen; Leena Valmu
Original assignee: Suomen Punainen Risti, Veripalvelu; Glykos Finland Oy
Priority date: 2010-10-06
Filing date: 2011-10-06
Publication date: 2012-04-12
Also published as: FI20106031A0

Abstract

The present invention is directed to a method for the isolation of a cell population from a sample of human umbilical cord blood comprising steps of contacting said sample with an antibody binding to core-2 type O-glycan, with a sialyl Lewis x (s Lex) and Neu5Acα2- 3Galβ1-3(Neu5Acα2-3Galβ1-4Glc NAcβ1-6)Gal NAcα epitopes, such as CHO-131, and isolating the cells bound to said antibody. The invention also provides compositions comprising human cells isolated from umbilical cord blood.

Description

METHOD FOR ISOLATING CELLS AND A CELL POPULATION THEREOF

The present invention relates to a composition comprising an isolated cell population that can be used for cellular therapy. The cell population having a characteristic content of cells can be isolated and defined by an antibody having binding specificity to core-2 type O-glycan with a sialyl Lewis x (sLex) and also to a non-fucosylated disialylated core -2 type O-glycan. One example of an antibody with such specificities is CHO-13. The isolated cell population has a low number of T_hei_per, T_cytotoXi_C and B lymphocytes known to be potentially harmful in haematopoietic stem cell transplantation. On the other hand, the cell population contains not only the original haematopoietic stem cell population of the cord blood unit but also essentially all natural killer (NK) cells of the unit. The present invention also relates to the method for isolating said cell population by the use of said antibody.

BACKGROUND OF THE INVENTION

Haematopoietic stem cells (HSC) transplantation is routinely used to as a treatment for malignant blood diseases, in particular leukaemia. Worldwide more than 15 000 allogenic, that is, the donor is not the same individual as the recipient, and more than 30 000 autologous transplantations are carried out (J Kersey. In Atkins et al. (eds) Clinical bone marrow and blood stem cell transplantation, pp 1- 10. Cambridge Univ Press, 2004). The source of haematopoietic stem cells has traditionally been bone marrow but recently cord blood and peripheral blood-derived HSCs are more common. Typically, the malignant bone marrow including the HSC of the patient is destroyed by using chemotherapy and irradiation to kill malignant cells. After this preconditioning the bone marrow is replaced by a healthy HSC graft.

However, it is known that the current types of stem cell grafts are not optimal: the

transplantation usually leads to severe graft-versus-host disease, the patient suffers from serious infections and sometimes the graft does not 'seed', that is, it does not produce novel hematopoietic cells. On the other hand, it is established (SR Solomon and AJ Barrett. In Atkins et al. (eds) Clinical bone marrow and blood stem cell transplantation, pp 277- 290. Cambridge Univ Press, 2004), that a pure HSC preparation, such as CD34 or CD133 positive cells alone, is not a functional option for successful transplantation, although one might assume that in theory use of pure stem cells should solve many problems. A fine balance including other non-HSC cells are needed for a transplant or graft to survive, for the graft cells to kill tumours, that is, the graft-versus-leukaemia effect, and finally for the stem cells of the graft to be able to produce novel haematological cells. Hence, in the clinical practise of today, the graft can be even a non-processed sample of bone marrow or cord blood unit. Alternatively, some cell populations, in particular CD4 positive T-helper lymphocytes and CD8 positive T-cytotoxic lymphocytes can be partially depleted to decrease the probability or magnitude of graft- ersus-host disease, the major drawback of HSC transplantation. About 20-50% of the patients develop a severe, even life-threatening graft-versus-host-disease, depending on the protocols used.

Umbilical cord blood is a valuable source of HSC, although the small size of a cord blood unit is often a limiting factor in these therapies. The cell population given to patients in cord blood transplantation is most often unspecified fraction of mononuclear cells, although possibility to enrich HSC using either anti-CD34 and/or anti-CD133 antibodies exists. The purified cord blood HSC population has not, however, proven as efficient as it was expected. Hence, some supporting cell population can be assumed to be beneficial in the therapy process. The nature of this supporting cell type is not known. It would be of great value to be able to identify an effective cell population or composition. Cell populations that can be assumed (Morris and Hill. Br J Haematol 2007; 137: 3 - 19) to be helpful in HSC transplantation, and hence, their presence in the graft desirable, are natural killer (NK) cells and regulatory T lymphocytes (Treg). NK cells have been shown to be potentially beneficial in HSC transplantation (Leung 2011. British Journal of Haematology, 155: 14-29), particularly in partially HLA-identical HSC transplantation. NK and NK-like cells may kill those effector cells of the recipient that are essential to graft- versus-host disease (GVHD), decreasing the risk for relapse, and combat infections and cancer. Also, NK cells may improve immune reconstitution of the donor (Symons and Fuchs. Bone Marrow

Transplant 2008; 42: 365-377). Furthermore, NK cells can be used for immunotherapy after HSC transplantation. Treg cells are central inhibitors of the immune response and they induce immunological tolerance. GVHD is mediated by the immune system: the cells or tissues of the host are recognised as immunologically foreign by the cells in the transplant and the immune system is targeted to kill these cells, leading to severe destruction of tissues. Hence, adding or increasing the number of NK and/or NK-like cells in the graft or after the transplantation may reduce the risks related to the transplantation. Cell populations can be identified by their surface antigens. A useful tool for detecting and enriching specific population is an antibody specific for the desired antigen. The monoclonal antibody CH0131 (R&D) is known to recognize the core-2 type O-glycan with a sialyl Lewis x (sLex) epitope (Walcheck 2002), but it appers to have a dual specificity as it also binds to the non-fucosylated structure Neu5Aca2-3Gai i-3(Neu5Aca2-3Gaipi-4GlcNAcpi- 6)GalNAca as demonstrated by the glycan microarray analysis carried out by the Consortium for Functional Glycomics (www.functionalglycomics.org) and described in patent application PCT/FI2010/050338. The epitope is known to be expressed by a subset of T-cells, where it acts as a high affinity P-selectin ligand (Ni 2006), and by colorectal adenocarcinoma (St Hill 2010). Due to cell-type and species specificity of glycan structures or epitopes, no predictions on the expression of a particular glycan epitope on a certain cell type can be done beforehand or based on literature only.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 illustrates relative proportions of cell populations within typical umbilical cord blood derived mononuclear cells.

Figure 2 illustrates the percentage of cells stained by the two antibodies CHO-131 and CS- LEX1, within each cell population derived from cord blood.

DETAILED DESCRIPTION OF THE INVENTION

The present invention is based on the finding that by using a single monoclonal antibody, here CHO-131, with a dual specificity to core-2 type O-glycan with a sialyl Lewis x (sLex) and non-fucosylated Neu5Aca2-3Galpl-3(Neu5Aca2-3Gal l-4GlcNAcpl-6)GalNAca epitopes, such as CHO- 131, a novel composition or population of human cells were detected and could be isolated from cord blood samples. In other words, the antibody or binder has binding specificity to at least one, preferably two, sialylated core-2 type O-glycan according to the Formula {Neu5Aca2-3 }_nGaipi-3[Neu5Aca2-3Gaipi-4(Fucal-3)_pGlcNAcpi- 6]GalNAca wherein n and p are 0 or 1 and n indicates presence or absence of terminal Neu5Ac and p indicates presence or absence of fucose branch, [ ] indicates a branch in the structure, optionally with two structures, optionally at least with a sialyl Lewis x (sLex) Neu5Aca2-3Galpl-3[Neu5Aca2-3Galpl-4(Fucal-3)GlcNAcpl-6]GalNAca and

Neu5Aca2-3Gai i-3(Neu5Aca2-3Gal l-4GlcNAc l-6)GalNAca epitopes. The advantage of the invention is that by using only a single anti-glycan antibody a particular, mixed cell population useful for hematopoietic stem cell transplantation can be obtained. The population so isolated still contained the haematopoietic stem cell population of the original cord blood, but surprisingly, the population additionally contained essentially all (80% - 90%) human natural killer (N ) cells, that is cells positive for CD56 antigen, and NK-like CD8+CD94 positive cells of the original cord blood unit. The number of B lymphocytes (CD20 positive cells), T helpers (CD4 positive cells), and cytotoxic T lymphocytes (CD8 positive cells), all potentially harmful in HSC transplantation, were clearly decreased in number down to 5% to 40% of that found in the original sample. Hence, the cell population obtained by antibody CHO-131 has a unique composition of cellular subpopulations, which can be more useful and effective as a cell therapy graft than the current transplants. In one embodiment of the invention, the cell population obtained by using a single antibody can be characterized by containing:

(i) CD34 positive (HSC) cells: at least 60%, preferably over 70% of the CD34 positive cells in the original cord blood sample;

(ii) CD56 positive NK cells: at least 60%, preferably over 70% of the CD56 positive cells in the original cord blood sample;

(iii) CD8+CD94 positive NK-like cells: at least 60%, preferably over 70% of the CD8+CD94 positive cells in the original cord blood sample.

In one embodiment of the invention, the cell population obtained by the antibody or binder comprises:

15-25% CD3 positive cells,

40-55% CD14 positive cells,

less than 0.5% CD20 positive cells,

25-33% CD45 positive cells, and

2-5% CD34 positive cells. In another embodiment of the invention, the cell population obtained can be characterized by containing:

(iii) CD8+CD94 positive NK-like cells: at least 60%, preferably over 70% of the CD8+CD94 positive cells in the original cord blood sample;

(iv) CD4 positive T helper cells: less than 50%, preferably less than 40% of the CD4 positive cells in the original cord blood sample;

(v) CD20 positive B lymphocytes: less than 30%, preferably less than 20% of the CD20 positive cells in the original cord blood sample;

(vi) CD8 cytotoxic T lymphocytes: less than 50%, preferably less than 40% of the CD8 positive cells in the original cord blood sample.

In another embodiment, the cell population obtained contains

(i) at least 60%, preferably over 70%, optionally over 80 % or 85 %, or 90 %, of the CD34 positive cells in the original cord blood sample; and

(ii) at least 60%, preferably over 70%, optionally over 80 % or 85 %, of the CD56 positive cells in the original cord blood sample; and

(iii) at least 60%, preferably over 70%, optionally over 80 % or 85 %, or 90 %, of the CD8 and CD94 positive cells; and

(iv) CD4 positive T helper cells: less than 50%, preferably less than 40% of the CD4 positive cells; and

(v) CD20 positive B lymphocytes: less than 30%, preferably less than 20%, optionally less than 15 % or 10 %, of the CD20 positive cells; and

(vi) CD8 cytotoxic T lymphocytes: less than 50%, preferably less than 40% of the CD8 positive cells

found in the original cord blood sample.

In a preferred embodiment of the invention, the cell population is depleted from the remaining T lymphocytes or a harmful subpopulation of T lymphocytes by using anti-T lymphocyte or anti CD3, anti CD8 or anti CD4 antibodies or by other means well known in the art. This step, often called T cell depletion, can be done either before or after the step using CHO-131 antibody. The cells bound by the anti-T lymphocyte antibodies are removed from the graft or are destroyed. There are a number of different ways to carry out the T cell depletion as described by Champlin et al 2000 (Blood 95(12) :3996-4003), antibodies commonly used in the art are summarized in Table 4. In an embodiment of the invention, the graft is treated with humanized anti-CD52 (Campath-1) antibody or antibodies against T cell antigen receptor.

In another embodiment, potentially harmful B lymphocytes are removed by antibodies against B cell specific antigens such as CD20. In still another embodiment, monocytes are removed by using anti-CD14 antibody. The rationale in the cell depletion is further removal of harmful effector T and/or B lymphocytes, in particular cytotoxic T cells, antibody-producing B cells and Thelper cells or monocytes; hence, a product even more enriched particularly with CD34 positive HSC and CD56 positive natural killer (NK) cells can be obtained. The cell populations so removed are known to mediate graft-versus-host disease and their removal reduce the severity and incidence of graft-versus-host disease, as discussed e.g. by Champlin et al. (Blood 95(12) :3996-4003). In addition, the removal of specific cell populations enables further tailoring of the therapeutic product.

The specific composition of the product depends on the depleting antibody used. For example, in one embodiment anti CD3 antibody is used after CH0131 enrichment so that practically all (80% - 99%) T cell populations or T cell antigen receptor expressing cell populations of the original sample can be removed, and the product obtained contains essentially all (80% - 99%) CD34 positive HSC, CD56 positive NK cell and CD 14 positive monocyte populations of the original sample. Their proportions based on simulations depicted in Table 3, in the product are: 3% - 5% or even up to 9% are CD34 positive HSC, 25% - 35% or even 40% are CD56 positive NK cells, 50% - 60% are CD14 positive cells, and less than 0.5% or even less than 0.2% are CD20 positive B cells. Similar simulation can be readily done for other antibodies based on data of Table 3. In another embodiment, the depleting antibody is selected from the list: anti CD3, anti CD8, anti CD20, anti CD14. There are various manufacturers and sources known for possible depleting antibodies (Champlin et al. 2000).

The invention is directed to methods for isolating cell populations by an antibody or binder having binding specificity to at least one (optionally to two structures of) sialylated core -2 type O-glycan according to the Formula {Neu5Aca2-3}_nGaipi-3[Neu5Aca2-3Gaipi- 4(Fucal-3)_pGlcNAcpi-6]GalNAcoc wherein n and p are 0 or 1 and n indicates presence or absence of terminal Neu5Ac and p indicates presence or absence of fucose branch, [ ] indicates a branch in the structure. In an embodiment, the antibody or binder binds at least to one of the following glycans a core-2 sialyl Lewis x (sLex) glycan with a sialylated Ga^3- branch Neu5Aca2-3Gaipi-3[Neu5Ac 2-3Galpl-4(Fucal-3)GlcNAc i-6]GalNAc or a non- sialylated Gal 3-branch Gai i-3[Neu5Aca2-3Gaipi-4(Fucal-3)GlcNAc i-6]GalNAca, and the non-fucosylated core 2 epitope Neu5Aca2-3Gai i-3(Neu5Aca2-3Gaipi- 4GlcNAc i-6)GalNAca epitopes, and in another embodiment to all three of these glycans. In an embodiment the binder binds at least to a sialyl Lewis x (sLex) Neu5Aca2-3Gaipi- 3[Neu5Aca2-3Gaipi-4(Fucal-3)GlcNAcpi-6]GalNAca and Neu5Aca2-3Gai i- 3(Neu5Aca2-3Gai i-4GlcNAc i-6)GalNAca epitopes. In an embodiment, the glycans are O-linked glycans linked to a serine or threonine residue(s), on cell surface proteins. In another embodiment, the specific binding includes the three sialyl core two O-glycans and especially SLex glycan with sialylated Galp3-branch and also non-reducing end terminal sialyl-Lewis x and Neu5Aca3Gaip4(Fuccc3)GlcNAcP- (fucose is branch), and rare

Neu5Aca3(SE6)Gal(P4[Fuca3]nGlcNAcP)_m-epitopes, wherein n is 0 or 1 and m is 0 or 1, independently and [Fuca3] is a fucose branch and ( ) indicates a linear sequence, as defined in WO2010122232 fully incorporated herein by reference. The present data with a sLex-control antibody shows that the sialylated core-2 type glycan specificity is essential for the production of the preferred cell population. In an embodiment, the binding is specific and binding is 10%, 30 %, 50 %, or 75 % lower non-core two glycan and/or other target glycans when measured by solid phase binding assay e.g. by a glycan array as defined in WO2010122232 fully incorporated herein by reference. The invention is especially directed to the use of an antibody with a specificity similar to CH0131 antibody with regard to the target glycans described above.

Accordingly, the present invention is directed to a method for the isolation of a cell population from a sample of human umbilical cord blood comprising steps of contacting said sample with an antibody binding to core-2 type O-glycan with a sialyl Lewis x (sLex) and Neu5Aca2-3Gal l-3(Neu5Aca2-3Gai i-4GlcNAcpi-6)GalNAca epitopes, such as CHO- 131, and isolating the cells bound to the said antibody. In a preferred embodiment, the isolated cells have therapeutic use. For example, the cells can be used either directly or after further processing as a graft for hematopoietic stem cell transplantation. The cells can also be used as a cellular support therapy after hematopoietic stem cell transplantation to augment the curative process. Another use in cell therapy is prevention of graft-versus-host-disease when the cells are given before transplantation or inhibition of on-going graft-versus-host disease after stem cell transplantation. The original graft and possible supportive cell therapy product can be from the same individual or from a third party.

The present invention is further directed to a composition comprising human cells isolated from umbilical cord blood, wherein all said cells are detected by antibodies binding to the core-2 type O-glycan with a sialyl Lewis x (sLex) and Neu5Aca2-3Gaipi-3(Neu5Aca2-

epitopes. Consequently, these cells or this cell population can be isolated by the use of an antibody binding to these epitopes, one such antibody is CHO-131. Preferably, the present invention is also directed to composition, wherein said cells are bound to said antibody. As disclosed in Table 3, the proportion of CD34+ cells in the composition of the present invention can be about 3%. Consequently, the present invention is directed to a composition comprising at least 3 %, preferably at least 4, 5, 7 or 9 %, CD34+ cells. Accordingly, in one specific embodiment of the invention, the composition contains (i) at least 3% CD34 positive cells and (ii) about 29% CD56 positive cells. Preferably, the aim of this invention is to provide a composition that contains at least

(i) CD34 positive cells; and (ii) CD56 positive cells; and (iii) CD8+CD94 positive cells, since these are supposed to be advantageous in the prevention of graft-versus-host-disease.

Further, the invention also relates to the use of an antibody binding to core-2 type O-glycan with a sialyl Lewis x (sLex) and Neu5Accc2-3Gal l-3(Neu5Aca2-3Gal l-4GlcNAcP 1- 6)GalNAcoc epitopes, such as CHO-131, to isolate the preferred cell population of the invention. The isolation or enrichment can be done using methods known by a person skilled in the art. For example, the antibody can be linked to magnetic beads, with which the starting cell material or cord blood sample is incubated followed by washing with a buffer. The cells detected by the antibody can be collected using magnetic device. Alternatively, the cells can be enriched using cell sorting apparatus (for example, the FACSAria® device) wherein cells labelled by a fluorescent-linked antibody can be "sorted" or separated from all other cells in the sample. Instead of using an antibody other molecules, "binders", binding to particular glycan structures can be applied. Examples of such are Fab or Fab2 fragments of an antibody, derivatives of enzymes or combinations thereof, such as glycosyltransferases including a3- sialyltransferase or 3-fucosyltransferase or glycosidases such as sialidase or fucosidases that are specific for glycan structures but rendered inactive, or lectins.

The invention is directed to the use of an amount of the binder such as an antibody, which can saturate the glycan binding sites of the binder such as an antibody on all target cells. In an example, the cord blood samples were incubated with 1.5 - 2 μg of the antibodies per 200 000 cord blood mononuclear cells. This corresponds to 7.5 - 10 micrograms of antibody, e.g. IgM or equivalent antibody, per million cells or per about 100 000-900 000 target cells, or 400 000 -700 000 target cells or about 500 000- 700 000 target cells or about 600 000 target cells. The targeted cells account for about one tenth fraction of the cells in a sample. In case the sample is already depleted from some cell populations, the number or fraction of the target cells may accordingly vary. The invention is directed in one embodiment to the use of a large amount of antibodies: at least 0.1 micrograms per million cord blood mononuclear cells (or 0.1 million target cells) in a sample. It is obvious that the amount of antibody can vary. In other embodiments the amount may be at least 0.5 micrograms, at least 1.0 microgram, at least 2.5 microgram, at least 5.0 microgram, at least 6.0 microgram, at least 7.5 microgram, at least 10 microgram, at least 12.5 microgram, at least 15 microgram, or at least 20 microgram per million sample cells or 0.1 million target cells. In one embodiment of the invention a high excess of antibody is used, an amount by far more than necessary for a stem cell population preparation, and a cell population comprising e.g. NK cell population is thus efficiently provided. It is realized that the high excess of antibody may reveal a group of less strongly labelled cells in competition of labelling of subpopulations. It is further realized that very high amounts of antibodies may not be useful or increase the effect. In an embodiment the amount of antibody is between 0.1- 100 micrograms, between 0.5-50 micrograms or between 1- 20 micrograms or between 2.5 and 15 micrograms per million sample cells or 0.6 million target cells. It is realized that IgM has ten antigen binding sites and it binds oligovalently to cell surface. Equivalently to IgM, also other antibodies with suitable affinity may be used, such as monovalent antibody fragments such as single chain antibodies or divalent antibodies such as IgG or equivalents or multivalent antibodies such as oligomeric nanobodies or oligomerically conjugated FAb or FAb2 fragments. When calculating corresponding amount of any antibody type the molecular weight of antibody and the number binding sites are considered. In one embodiment, the composition is mononuclear cell population comprising at least about 30 % of the cord blood mononuclear cells, optionally comprising 40-80 % of the mononuclear cells of the original sample, or preferably comprising 50-70 % of the mononuclear cells or 55-65 % to about 60 % of the mononuclear cells of the original sample.

Cell populations that can be assumed (Morris and Hill. Br J Haematol 2007; 137: 3 - 19) to be helpful in HSC transplantation, and hence, their presence in the graft desirable, are natural killer (NK) cells and regulatory T lymphocytes (Treg). NK cells have been shown to be potentially beneficial in HSC transplantation (Leung 2011. British Journal of Haematology, 155: 14-29). NK and NK-like cells may kill those effector cells of the recipient that are essential to graft-versus-host disease (GVHD), decreasing the risk for relapse, and combat infections and cancer. Also, NK cells may improve immune reconstitution of the donor (Symons and Fuchs. Bone Marrow Transplant 2008; 42: 365-377). Furthermore, NK cells can be used for immunotherapy after HSC transplantation. Hence, adding or increasing the number of NK and/or NK-like cells in the graft or after the transplantation may reduce the risks related to the transplantation. In one embodiment of the invention, the novel cell population enriched with HSC and potentially useful NK cells is used for treatment of complications related to hematopoietic stem cell transplantation. In another embodiment, the complications are at least one from the list: graft versus host disease, relapse, infections, cancers. In a further embodiment, the cell population can be used to imporove immune reconstitution after stem cell tranplantation. There are various ways to administer the cells to patients known in art, depending many clinical factors. The glycan and monosaccharide structures, their linkages and formulas used here are well known in the art and are described in e.g. Varki et al (Essentials of Glycobiology, 2^nd ed, Cold Spring Harbor Laboratory Press, New York, 2009). Glycolipid and carbohydrate

nomenclature is according to recommendations by the IUPAC-IUB Commission on

Biochemical Nomenclature (Carbohydrate Res. 1998, 312, 167; Carbohydrate Res. 1997, 297, 1; Eur. J. Biochem. 1998, 257, 29). Galactose (Gal), N-acetylglucosamine (GlcNAc), mannose (Man), fucose (Fuc) and N-acetylneuraminic acid (Neu5Ac) are in pyranose form and D-form except fucose which is in L-form, N-acetyl group of GlcNAc is on carbon 2. References

Ni Z, Campbell JJ, Niehans G, Walcheck B (2006) J Immunol. 177(7):4742-8.

St Hill CA, Farooqui M, Mitcheltree G, Gulbahce HE, Jessurun J, Cao Q, Walcheck B. (2009) BMC Cancer 9:79

Walcheck B, Leppanen A, Cummings RD, Knibbs RN, Stoolman LM, Alexander SR, Mattila PE, McEver RP (2002) Blood 99(11):4063-9.

EXAMPLES

Materials and methods common to all examples

Mononuclear cells from umbilical cord blood

Human term umbilical cord blood (UCB) units were collected after delivery with informed consent of the mothers and the UCB was processed within 24 hours of the collection as described earlier (Jaatinen T, Hemmoranta H, Hautaniemi S, Niemi J, Nicorici D, Laine J, Yli-Harja O and Partanen J. Stem Cells 2006; 24:631-641). The mononuclear cells (MNCs) were isolated from each UCB unit diluting the UCB 1:4 with phosphate-buffered saline (PBS) containing 2mM EDTA followed by Ficoll-Paque Plus (Amersham Biosciences, Uppsala, Sweden) density gradient centrifugation (400 g / 40 min). The mononuclear cell fragment was collected from the gradient and washed three times with PBS - 2mM EDTA.

Antibodies

The antibodies used in the examples are listed in Table 1.

Fluorescence activated cell sorting (FACS) analysis

UCB mononuclear cells were washed with buffer (PBS - 0,3 % BSA - 2mM EDTA) before antibody labelling. To avoid oligosaccharide contamination, ultra pure BSA (Sigma- Aldrich, Germany) was used. Primary antibodies (VPU037 or VPU020) were incubated (0^g / 100 μΐ cell suspension / 200 000 cells) for 30 minutes on ice and washed once with buffer. Cells were then labelled with a secondary antibody, Alexa Fluor 488 conjugated goat anti-mouse (1:2000) for 30 minutes on ice in the dark. After washing with the buffer, tertiary antibodies (1 or 4 μΐ of each cell marker/isotype control) were added and incubated for 30 minutes on ice in the dark. After washing the cells were fixed with stabilizing fixative (BD biosciences). As a negative control cells were incubated without primary antibody and otherwise treated similar to labelled cells. Cells were analysed by FACSAria and FACSDiva™ Version 5.0.2. software (Becton Dickinson).

EXAMPLE 1. Characterization of different cell populations within mononuclear cells of umbilical cord blood unit In this example, the relative amount of different cell populations within cord blood derived mononuclear cells were calculated. The cell populations were stained with marker antibodies and the amount of cells expressing these marker epitopes were then calculated.

Here mononuclear cells were isolated from umbilical cord blood unit by Ficoll centrifugation. Cells were then stained by different cell surface marker antibodies as listed in Table 2 using only tertiary antibodies in Table 1. Thereafter, the cell populations were analyzed using flow cytometry by FACSaria instrument (Figure 1).

We analyzed the fraction of HSC to be less than two percent of the entire mononuclear cell population, depending on cell surface marker used. The CD34+ cell population was 1.4% of the overall mononuclear cell fraction, and CD133+ cell population was only 1.2% of the same population. These calculations were representative of three replicate analyses when CD34+ cells were in question, whereas CD 133+ cells were only analyzed once.

EXAMPLE 2. Cord blood derived cell populations expressing core-2 type O-glycan, with a sialyl Lewis x (sLex) epitope recognized by antibody CHO-131 and sLex epitope recognized by antibody CSLEX1.

The cell populations depicted in Example 1 were analyzed using flow cytometry for the cell surface expression of sLex epitope using both CH0131 antibody, that recognizes core-2 type O-glycan, with a sLex epitope , and CSLEX1 antibody, which recognizes the sLex epitope as such. The percentage of CH0131 epitope positive cells within each cell population is visualized in Figure 2. As much as 90% of CD34 positive HSCs were detected by CH0131 antibody and approximately 80% of CD133 positive HSCs do the same. Also essentially all NK cells and NK-like T cells (80-90%) could be detected with the antibody. Furthermore nearly all monocytes are positive in this cell surface expression. A much smaller proportion of the most abundant T cell types (helper and cytotoxic T cells) (30-40%) as well as B cells (10%) expressed this epitope. The result was based on three biological replicates, despite CD 133 positive cells, where only one cord blood unit was analyzed.

A clear difference in the results could be seen when CSLEX1 antibody was used

(Figure 2). In case of T cells, B cells and NK type of cells the expression of sLex epitope recognized by CSLEX1 antibody was markedly lower as compared to the expression of the epitope recognized by antibody CH0131. This indicates that the CHOI 31 antibody recognized also other epitopes than the core-2 type O-glycan with a sLex epitope.

EXAMPLE 3. Simulation of purified cell population generated by CHO-131 antibody In this example the cell population detected by antibody CH0131 was simulated.

The antibody investigated in Example 2 can be used to purify cell populations expressing the target epitopes on their surface. Here, the antibody can either be coupled to solid carrier, for example magnetic bead, or sorting by flow cytometry can be utilized. The advantage of using glycan binders in these types of isolation protocols is the possibility to elute the cells from the antibody and its carrier using epitope-like glycan structures, here disialylated structures.

The characteristic cell population derived from umbilical cord blood mononuclear cells using antibody CH0131 (Table 3) was simulated using the data obtained in the above mentioned staining experiments. The cell populations stained by tertiary marker antibodies (Table 1) were gated and the number of cells was estimated (Table 3, left column).

Within these gated cell populations, staining by CH0131 antibody was investigated.

According to this data, the simulated number of cells was derived (Table 3, right column).

The cell population isolated using CH0131 antibody was relatively large in the above simulation: about 60% of the original mononuclear cell fraction. The enrichment of monocytes, from 28.2% to 47.6% and NK cells, from 14.8% to 28.9%, within the population was highly significant; the enrichment of HSC within the population varied from 2.2% to

3.4%. EXAMPLE 4. Isolation of a cell population from UCB-MNC by antibody CH0131 using fluorescence activated cell sorter (FACS) or magnetic beads UCB mononuclear cells recognized by antibody CHO-131 can be isolated by cell sorting using FACS flow cytometry. UCB mononuclear cells are filtered using a 30 pm cell filter and 1 -5 x 108 cells are suspended in cold buffer (standard Phosphate Buffered Saline (PBS), 0,3 % Bovine Serum Albumin, 2mM EDTA) at 108 cells/ml. Appropriate amount of CH0131 antibody, typically approximately 2-10 pg of antibody per 108 cells, the optimal amount has to be titrated, is added. After a 30-minute incubation on ice the cells are washed once with the buffer and labelled with a secondary antibody, Alexa Fluor 488 conjugated goat anti-mouse (1:2000) for 30 minutes on ice in the dark. The cells are washed twice. After filtering (30 pm cell filter) the cells are suspended in cold buffer and the volume is adjusted to approximately 3 x 107 cells/ml. The cell sorting can be performed by FACSAria (Becton Dickinson) using standard protocols.

UCB mononuclear cells recognized by antibody CHO-131 (mouse IgM) can be isolated using anti-mouse IgM MicroBeads (Miltenyi Biotec) according to the manufacturer's instructions. Briefly, UCB mononuclear cells (up to 10⁷ cells) are suspended in buffer (PBS - 0,3 % BS A - 2mM EDTA), filtered and labelled with an appropriate amount of CHO 131 antibody for 30 minutes on ice. After washing the cells are magnetically labelled with 20 μΐ of anti-mouse IgM MicroBeads for 15 minutes (2-8 °C). After a wash step the cell suspension is loaded onto a MACS Column which is placed in the magnetic field of a MACS Separator. The magnetically labelled CH0131 positive cells (the cell population of interest) are retained within the column. After washing the unbound cells the magnetically retained cells can be eluted by removing the column from the magnetic field.

Table 1. Antibodies used in the examples

Table 2. Cell surface markers stained in the definition of different cell populations

Table 3. Simulated cell populations enriched from umbilical cord blood mononuclear cells using antibody CHO-131

Cell population Cell number/ % CHO-131positive %

100 000 MNC cell number /

100 000 MNC

CD3+ (T cells) 30690 51.3 7049 19.8

CD 14+ (monocytes) 16894 28.2 16921 47.6

CD20+ (B cells) 2020 3.4 115 0.3

CD56+ (NK cells) 8835 14.8 10286 28.9

CD34+ (HSC) 1320 2.2 1192 3.4

Summary 59759 100 35563 100 Table 4 Different ways for cell depletion and their frequency, based on Champlin et al 2000.

Narrow-specificity Antibodies targeting T10B9 (cc/P T-cell receptor), CD3 + CD7, antibodies CD3 + CD6 + CD8, CD4 ± CD8, CD4 + CD5 + CD8, CD5 + CD8,

N = 450 (52%) CD6 ± CD7, CD6 + CD8, CD6 + CD7 + CD8, CD 8 ± CD7, CD14, CD20 Broad- specificity Antibodies targeting CD2 + CD3, CD2 + CD5, CD2 + CD7, CD2 + CD8, antibodies N - 73 CD2 + CD4 + CD8, CD2 + CD5 + CD7, CD2 + CD5 + CD8,

(8%) CD2 + CD3 + CD4 + CD8, CD2 + CD3 + CD4 + CD5 + CD6 + CD8 +

CD28; ATG incubation

Campath Antibodies targeting CD52 including Campath 1G (IgG-2b), Campath

1 N = 131 (15 ) lM (IgM)

Elutriation N = 75 Elutriation; density-gradient centrifugation (DGR)

(9%)

Lectins/SRBC Lectins + CD5 + CD8; lectins ± sheep red blood cell; rosetting (SRBC); N = 141 (16%) SRBC ± DGR

Claims

1. Method for the isolation of a cell population from a sample of human umbilical cord blood comprising steps of contacting said sample with an antibody or binder having binding specificity to at least one sialylated core-2 type O-glycan according to the Formula

{Neu5Aca2-3 }_nGai i-3[Neu5Ac 2-3Gaipi-4(Fucal-3)_pGlcNAcpi-6]GalNAc wherein n and p are 0 or 1 and n indicates presence or absence of terminal Neu5Ac and p indicates presence or absence of fucose branch, [ ] indicates a branch in the structure, and isolating the cells bound to said antibody or binder.

2. The method according to claim 1 , wherein said antibody or binder have binding specificity to core-2 type O-glycan with a sialyl Lewis x (sLex) Neu5Aca2-3Galpl-3[Neu5Aca2- 3Gai i-4(Fucal-3)GlcNAcpi-6]GalNAca and Neu5Acoc2-3Gai i-3(Neu5Aco 2-3Gal l- 4GlcNAc i-6)GalNAc epitopes.

3. The method according to claim 1 or 2 comprising a further step of contacting the cells with a depleting antibody and removing the cells detected by said depleting antibody.

4. The method according to claim 3, wherein the depleting antibody is at least one selected from Table 4.

5. The method according to claim 3 or 4, wherein the depleting antibody is at least one selected from the list consisting of: anti CD3, anti CD20, and anti CD 14.

6. The method according to claim 1, wherein said antibody is CHO-131.

7. The method according to claim 1, wherein the cells are isolated for cellular therapy.

8. The method according to any one of claims 1-7, wherein the product cell population is mononuclear cell population comprising at least about 30 % of the sample cord blood mononuclear cells, optionally comprising 40-80 % or 50-70 % of the mononuclear cells of the original sample.

9. The method according to any one of claims 1-7, wherein the amount of the antibody is saturating with regard to the glycans on all cells of the sample, optionally the amount of the antibody is between 0.1 -100 micrograms per 1 million mononuclear cells in the sample

10. A composition comprising human cells isolated from umbilical cord blood, wherein all said cells are detected or detectable by antibody or binder binding to the sialylated core-2 type O-glycan.

11. The composition according to claim 10, wherein said cells are isolated by the use of an antibody or binder binding to the sialylated core-2 type O-glycan.

12. The composition according to claim 10 or 11, wherein said antibody or binder bind to the core-2 type O-glycan with a sialyl Lewis x (sLex) and Neu5Aca2-3Gaipi-3(Neu5Aca2- 3Galβl-4GlcNAcβl-6)GalNAca epitopes.

13. The composition according any one of claims 10- 12, wherein said cells are isolated by contacting the cells with a depleting antibody and removing the cells detected by said depleting antibody.

14. The composition according to claim 13, wherein the depleting antibody is at least one selected from Table 4.

15. The composition according to claim 13, wherein the depleting antibody is at least one selected from the group consisting of: anti CD3, anti CD20, and anti CD 14.

16. The composition according to claim 15, in which the depleting antibody is anti CD3 antibody.

17. The composition according to any one of claims 10-16, wherein 3% - 5% or even up to 9% of the cells in the composition are CD34 positive cells, 25% - 35% or even 40% are CD56 positive cells, 50% - 60% are CD 14 positive cells, and less than 0.5% or even less than 0.2% are CD20 positive cells.

18. The composition according to any one of claims 10-16, wherein said antibody is CHO- 131.

19. The composition according to any one of claims 10-16, wherein said cells are bound to said antibody or binder.

20. The composition according to any one of claims 10-16, wherein said composition contains at least

(i) CD34 positive cells;

(ii) CD56 positive cells; and

(iii) CD 8 and CD94 positive cells.

21. The composition according to claim 20, wherein said composition contains

(i) at least 60%, preferably over 70% of the CD34 positive cells in the original cord blood sample; and

(ii) at least 60%, preferably over 70% of the CD56 positive cells in the original cord blood sample; and

(iii) at least 60%, preferably over 70% of the CD8 and CD94 positive cells in the original cord blood sample.

22. The composition according to claim 21 , wherein said composition contains

(iii) at least 60%, preferably over 70%-, optionally over 80 % or 85 %, or 90 %, of the CD8 and CD94 positive cells; and

found in the original cord blood sample.

23. The composition according to claim 20, wherein said composition contains at least 3, 4, 5, 7, or 9% of CD34 positive cells, of the total mononuclear cell amount of said composition.

24. The composition according to claim 23, wherein said composition contains

(i) at least 3% of the CD34 positive cells;

(ii) about 25-35 %, or about 27-31 % or about 29% of the CD56 positive cells;

of the total mononuclear cell amount of said composition.

25. The composition according to claim 20, wherein said composition contains at least 5%, 5.5%, or 6 %, or optionally at least 6.5%, 7%, or 7.5 % cells positive for both CD8 and CD94.

26. The composition according to claim 20, wherein the cell population comprises less than 3, 2, or 1 % of CD20 positive B-lymphocytes, optionally less than 0.7%, 0.5%, or 0.3 % of CD20 positive B-lymphocytes.

27. The composition according to claim 20, wherein the cell population comprises less than 9, 8, or 6 % of CD8 positive T-lymphocytes, optionally less than 5. %, 4.5%, or 4 % of the CD8 positive T-lymphocytes.

28. The composition of any one of claims 10- 16, wherein the cell population is mononuclear cell population comprising at least about 30 % of the cord blood mononuclear cells, optionally comprising 40-80 % of the mononuclear cells of the original sample.

29. The composition of any one of claims 10- 16, wherein the cell population is effectively decreased with regard to B-cells and part of T-cells of cord blood, and optionally increased with regard to NK cells.

30. The composition of any one of claims 10- 16, wherein the cell fraction bound to said antibody or binder having specificity for the sialylated core two glycan is separated from the sample and obtaining a binding negative cell fraction, which is a highly enriched B-cell and a T-cell subpopulation comprising cell fraction, and, optionally further testing the fraction with regard to presence of B-cells and/or T-cell subpopulations, and/or CD34+ or CD 133+ hematopoietic stem cells and/or the sialylated core-2 type O-glycans of the invention.

31. The composition of any one of claims 10-16,wherein the cell population is mononuclear cell population comprising at least about 30 % of the cord blood mononuclear cells, optionally comprising 40-80 % of the mononuclear cells of the original sample. 50-70 % of the mononuclear cells or 55-65 % about 60 % of the mononuclear cells

32. The composition according to any one of claims 10-31 for use in the treatment of complications related to hematopoietic stem cell transplantation.

33. The composition according to claim 32, wherein said complications are selected from the group consisting of: inhibition of graft versus host disease, decreasing the risk for relapse of leukaemia, infections, cancer, and improvement of immune reconstitution.

34. The composition according to claim 32, wherein the treatment is obtained by the composition containing 30% - 40% CD56 positive cells.

35. The composition according to claim 32, wherein the treatment is obtained by the composition in which additionally the number of cells positive for both CD8 and CD94 is increased to 5% - 7% of the total cell number.

36. The composition according to claim 32, wherein the composition is depleted from cells detected by at least one antibody selected from group consisting of: anti CD3 antibody, anti CD 8 antibody, anti CD20 antibody, and anti CD 14 antibody.

37. The composition of any one of claims 10-36 produced by the method according to any one of claims 1-9