WO2003006637A1

WO2003006637A1 - Leukocyte preparation

Info

Publication number: WO2003006637A1
Application number: PCT/GB2002/003151
Authority: WO
Inventors: Satish Keshav; Andrew K. Burroughs; Keith Rolles
Original assignee: University College London
Priority date: 2001-07-09
Filing date: 2002-07-09
Publication date: 2003-01-23
Also published as: GB0116752D0

Abstract

A process for preparing human hepatic dendritic cells, which comprises: (a) providing a substantially intact, human liver suitable for transplantation; (b) treating the liver by storing under conditions to promote dendritic cell availability and removing any storage liquid or perfusate; (c) perfusing the treated liver so as to produce a perfusate containing dendritic cells; and (d) isolating from the perfusate a cell fraction comprising human hepatic dendritic cells.

Description

Leukocyte Preparation

Field of the Invention

The present invention relates to a process for preparing human hepatic dendritic cells, isolated human hepatic dendritic cells obtainable from such a process and the use of such cells in medicine, for example, in the treatment of autoimmune disorders and the treatment or prevention of allograft rejection.

Background to the Invention

Dendritic cells are specialised antigen presenting cells capable of initiating and modulating the immune response. Dendritic cells are widely distributed throughout the body in both lymphoid and non-lymphoid tissue and are generally present at very low concentration. A key role for dendritic cells is to capture antigens in peripheral tissues and process them for presentation on the dendritic cell surface bound to the major histocompatibility complex (MHC) . This process involves the maturation of dendritic cells from an immature form into a mature form. Antigen presentation to T cells in this way initiates a T cell mediated immune response.

The function and potential utility of dendritic cells has become of interest to researchers. In view of the very low or trace levels of dendritic cells in the body, various attempts have been described to provide improved sources for such cells. For example, US 5994126 describes a method for producing proliferating cultures of dendritic cell precursors. Working primarily in the murine system a process is described for isolating dendritic cells from bone marrow and using a cell culture technique to proliferate the cells so as to provide a culture containing an enriched population of dendritic cells. Cultured dendritic cells were also the subject of US 5788963 in which human peripheral blood dendritic cells were used as starting material. The therapeutic aim of this work was immunotherapy for prostate cancer and the isolated dendritic cells were exposed to a prostate cancer antigen so as to activate a relevant T cell response in a subject in vivo . Accordingly, the immunogenic properties of dendritic cells from this source were exploited.

US 5871728 is directed to a method of regulating dendritic cell maturation. This disclosure aims to exploit a property of liver dendritic cells, which are thought to promote tolerogenicity . Immature mammalian dendritic cells propagated in the presence of a cytokine are administered to a host mammal in advance of transplantation. Tolerogenicity is enhanced in the host mammal in this way. This is exemplified experimentally only in mouse. Dendritic cells are prepared from whole mouse liver thereby requiring sacrifice of the donor mouse. Whilst this approach is theoretically applicable to humans, it is not practicable for ethical reasons. Any human livers which are donated as part of an organ donor program are required for use in liver transplantation where the demand for donor livers outstrips their supply. There has been a report of isolation of dendritic cells from human liver biopsy samples which have been subjected to overnight culture (Goddard et al ) (GUT, vol 46 Suppl II (2000) page A71 Goddard S. Hubscher, SG & Adams, DH Abstract W157). However, starting from such a small tissue sample does not enable isolation or purification of sufficient quantities of dendritic cells at a sufficient level of purity for them to be useful.

Jonsson et al describe in Hepatology (1997) 2_6, 1111- 1114 [1] the usefulness of liver-associated leukocytes from human liver transplant perfusate. The standard procedure for the retrieval of the liver from an organ donor is that donor viscera are perfused in si tu with a preservative and cooling solution which is usually University of Wisconsin organ preservation solution (UW) . Thereafter organs are removed surgically and further perfused ex vivo (through the portal vein in the case of the liver) until most of the blood has perfused out of the organ and perfusion fluid is minimally bloodstained. The organ is then placed in an impermeable plastic bag together with the perfusate in a volume of ice cold UW in order to surround the organ completely. The organ can then be stored and transported on ice. Jonsson et al describe a technique whereby they use the storage liquid or perfusate as a source of liver- associated leukocytes. A diverse population of cells is recovered from the liquid using filtration through gauze to remove debris and density gradient centrifugation to isolate mononuclear cells. Although Jonsson et al characterise 0.5% of the total mononuclear cell population as "dendritic cells", this characterisation is not based on a positive marker for the dendritic cells such as CD80, CD86 or MHC II. The quantity may be much lower than indicated. The quantity and purity of dendritic cells obtainable by the procedure of Jonsson et al is not sufficient for any practical therapeutic use intended for such dendritic cells. Summary of the Invention

The present invention aims to overcome the disadvantages of the prior art by providing an improved process for the production of dendritic cells and dendritic cells in greater quantity and purity than has hitherto been obtained.

Accordingly, in a first aspect there is provided a process for preparing human hepatic dendritic cells, which comprises :

(a) providing a substantially intact, optionally perfused, resected human liver suitable for transplantation;

(b) treating the liver by storing under conditions to promote dendritic cell availability and removing any storage liquid or perfusate;

(c) perfusing the treated liver so as to produce a new perfusate containing cells that are derived from the liver, including hepatic dendritic cells; and

(d) isolating from the new perfusate a cell fraction comprising human hepatic dendritic cells.

It has surprisingly been found that rather than using the storage liquid or perfusate from a treated liver but instead perfusing the treated liver so as to produce a fresh perfusate, a much higher concentration of dendritic cells in the mononuclear cell population may be obtained. The thus treated and perfused liver may then be used for transplant in the normal way without any adverse effect on the transplant recipient.

Typically the resected human liver is completely intact although it may be split or reduced in size. Once removed from the human donor, the liver is typically made suitable for transplantation by perfusion and storage in University of Wisconsin (UW) solution as described above. Preferably, the step of storing the liver under conditions to promote dendritic cell availability comprises storing for at least 6 hours, more preferably at least 12 hours, especially around 18 hours. These are typical times for storage prior to transplantation and may include transportation time to the location of the liver transplant recipient. The conditions under which the liver is stored are generally 4°C, pH 7.4 in isotonic solution (UW or similar).

In step (c) the treated liver is perfused with a perfusion solution which is usually physiological and generally comprises 4% or 4.5% human albumin or physiological saline. Most preferably the perfusion solution contains around 4% to 4.5% w/v human albumin. It is convenient to perfuse the liver by feeding perfusion solution through the portal vein. In a preferred embodiment perfusion may take place after the liver has been surgically implanted in the recipient patient's abdominal cavity.

The step of isolating a cell fraction from the perfusate may comprise isolating a cell fraction enriched with human hepatic dendritic cells or a cell fraction containing isolated human hepatic dendritic cells. Typically, mononuclear cells are separated from the perfusate to produce a cell fraction enriched with human hepatic dendritic cells and this may be achieved by density gradient centrifugation. Further steps may be employed to separate the human hepatic dendritic cells from the mononuclear cell fraction. Such steps include density gradient separation, antibody based separation using fluorescent activated cell sorting or magnetic bead separation; or centrifugal elutriation.

In a further aspect, the invention provides isolated human hepatic dendritic cells obtainable by the process described herein.

The human hepatic dendritic cells may be characterised according to the cell surface marker antigens expressed, using, for example, flow cytometry. Human hepatic dendritic cells are typically MHC class II⁺, CD3^" and CD19, and CD80⁺, CD86⁺. In addition they have a typical morphology using Giemsa staining, and are typically non- phagocytic .

Isolated human hepatic dendritic cells according to the present invention may be used in medicine. Such cells are considered to have tolerogenic properties and may be used as a tolerogen, for example in the treatment of autoimmune disorders or allograft rejection. Typical autoimmune disorders include autoimmune thyroid disease, type I diabetes mellitus, rheumatoid arthritis, inflammatory bowel disease and systemic lupus erythematosus .

In the case of autoimmune disease human hepatic dendritic cells according to the invention may be contacted with one or more autoantigens responsible for the autoimmune disease. This may be effected in vi tro under conditions to provide antigen presentation by the dendritic cells. The thus-treated dendritic cells may then be given to a patient to tolerise that patient in relation to the autoantigens.

The human hepatic dendritic cells of the present invention may also be used for treating or preventing allograft rejection in a recipient in which the allograft is a non- liver allograft from the same donor as the liver. In this therapeutic application, the cells would normally need to be donor specific and would be useful, for example, where a donor liver is going to a first recipient and a heart or other organ from the same donor is going to a second recipient. In this case, a cell fraction comprising human hepatic dendritic cells according to the present invention from the liver donor would be used to treat or prevent allograft rejection in the second recipient, who might otherwise have a reaction to antigens from the donor heart or other organ.

Brief Description of the Drawings

The invention will now be described in further detail, by way of example only, with reference to the following Examples and accompanying drawings, in which:

FIGURE 1 compares stimulatory activity of hepatic leukocytes and peripheral blood cells;

FIGURE 2 compares stimulatory activity of liver derived and peripheral blood macrophages; Figure 3 shows stimulatory activity of different cell types;

FIGURE 4 shows a giemsa-stained cytospin preparation illustrating mature dendritic cells according to the present invention;

Figure 5 shows stained cytospin preparations of liver associated leukocytes from healthy human liver; and

FIGURE 6 shows schematically apparatus configured for perfusate collection.

Detailed Description of the Invention Example

Method: Isolation of hepatic mononuclear leukocytes In the operating theatre where the liver transplant operation is to take place, the donor liver is removed from the perfusion and storage fluid. In preliminary experiments, it was found that the yield of leukocytes from the perfusion and storage fluid was extremely low, of the order of 5xl0⁶ cells, and therefore this fluid is routinely discarded, although it is the starting material for the technique of Jonsson et al . The donor organ is then surgically implanted, with the following anastomoses being fashioned: hepatic artery to hepatic artery, portal vein to portal vein, hepatic vein to suprahepatic vena cava, and infrahepatic vena cava to infrahepatic vena cava. The portal venous and suprahepatic anastomoses are cannulated, and the liver graft is perfused with 500ml of 4% or 4.5% human albumin solution, which is fed in via the portal vein. The perfusate, which is collected via a canulla in the suprahepatic vena caval anastomosis, is the starting material for the present technique of isolation of hepatic mononuclear leukocytes (Figure 6). Clinical Samples

Perfusate fluid was obtained from 25 consecutive cadeveric donors (14F:11M, mean age 47 ± 2.3 (range 13-65). The mean cold ishaemia time was llh55' (range OδhOO' -15h48' ) . The cause of death was sub arachnoid haemorrhage (n=24) and hypoxic brain damage (n=l).

Donor blood (n=14) and spleen (n=3) was collected at the time of organ retrieval and kept at room temperature and 4°C respectively until processed. In addition, peripheral blood was collected from healthy volunteers (n=29) .

Isolation of mononuclear cells

The perfusate solution was centrifuged at 500 x g for 15 mins and the pellet resuspended in 30ml Hanks Balanced Salt Solution (HBSS) (Gibco BRL, UK) supplemented with ImM EDTA (Sigma, Poole, U.K.). (5mM EDTA can be used here.) The solution was carefully layered on to a Ficoll (Histopaque®, Sigma, UK) density gradient and centrifuged at 400 x g for 30 mins. Liver associated leukocytes (LAL) at the density gradient interface were carefully aspirated and washed twice in HBSS/EDTA at 250 x g for 7 mins. The final pellet was resuspended in complete medium (CM) (RPMI 1640 (Gibco BRL, UK) supplemented with 10% human AB serum (Quest Biomedical, Knowle, UK), 25mM HEPES, 2mM L- glutamine, lOOU/ml penicillin and lOOμg/ml streptomycin (Gibco BRL, UK) . Cells were counted after Trypan Blue (Sigma, UK) exclusion and the cell concentration adjusted to 5 x lOVml. Peripheral blood and splenic lymphocytes (PBL and SL) were isolated in a similar manner. Flow Cytometry

Freshly isolated cells were analysed by flow cytometry (Coulter EPICS® XL-MCL) . The antibodies used were fluorescein isothiocyanate labelled anti-CD3, CD4, CD8, CD14, CD19, CD83 and CD86 and phycoerythrin labelled anti- CD56, CD69, HLA ABC and HLA DQ DR DP (Immunotech, UK) . 1 x 10⁵ cells were incubated with either individual antibody or a cocktail of antibodies at a dilution of 1:10 in PBS supplemented with 1% newborn calf serum (NBCS) and incubated for 20 mins at room temperature. Cells were washed, resuspended in 1% NBCS and analysed.

Mixed Lymphocyte Culture (MLC)

The allostimulatory activity of freshly isolated LAL, donor PBL, healthy PBL and SL was investigated by culturing variable numbers of adherent cells with a fixed number of responding PBL from a healthy third party. Cells were cultured in 96 well flat-bottomed plates (Triple Red Ltd, Thame, UK) in a total volume of 200μl/well. Non-adherent cells were aspirated after 1 hr incubation at 37°C in a humidified 5% C0₂ atmosphere. The adherent cells were washed once in CM and the responder cell population added (1 x 10⁵) . Mixed cultures were incubated for 5 days at 37°C in 5% C0₂. Eighteen hours before harvesting, cells were pulsed with 0.5μCi/well ³H-Thymidine (Amersham Pharmacia Biotech UK Ltd) . Each assay contained a phytohaemagglutinin (PHA) (0.5μg/ml) control as a measure of cell viability. T cell proliferation was expressed as a mean of triplicate wells . Quantitation of cell adherence

The number of adherent cells was measured using the CellTiter 96™ Aqueous assay (Promega, UK) . LAL and PBL were plated out as previously described for the MLC. After aspirating the non-adherent cells, the wells were replenished with CM and the assay reagent was added. LAL and PBL were incubated at 37°C in 5% C0₂ for 1.5 hour and 2.5 hours respectively and the absorbance read at 492nm. A standard curve was plotted and the unknown values read from the graph.

Characterization of adherent and non-adherent cells Cells in CM were seeded at 2 x 10⁶ /well in 24 well plates (Triple Red Ltd, Thame, UK) and incubated for 1 hr at 37°C in 5% C0₂. The non-adherent cells were aspirated and the adherent cells washed twice in PBS and fixed in 70% ethanol. The non-adherent cells were washed and resuspended in HBSS/EDTA. Cytospin preparations were made (Cytospin®, Shandon Life Sciences International, Runcorn, UK) and the cells were fixed in 95% ethanol.

The cell phenotype was determined by incubating cells with monoclonal antibodies to HLA Class I and II, CD4, CD8, CD19, CD56, CD68 (Dako Ltd, UK), CD80 (RND Systems, UK) and CD86 (Serotec, UK) diluted 1:100 in PBS. After 1 hr incubation at room temperature, cells were washed in PBS and biotinylated anti-mouse IgG (Vectastain Elite ABC Kit, Vector Labs, Peterborough, UK) added and incubated for 1 hr. After washing, the avidin-biotinylated peroxidase complex was added and incubated for 1 hr . After a final wash in PBS, the substrate 3, 3' -diaminobenzidine tetrahydrochloride (0.5mg/ml) in PBS containing 0.015% hydrogen peroxide was prepared and added to each well, the colour developed within 5 mins.

Isolation and propagation of dendritic cells Freshly isolated LAL and PBL were cultured in 6-well plates (Triple Red Ltd, Thame, UK) (1 x 10⁷ cells/3ml/well) based on the method of Ro ani et al [10], with some modifications. Cells were incubated for 7 days in CM supplemented with 50ng/ml each of recombinant human granulocyte macrophage colony stimulating factor (GM-CSF) and interleukin 4 (IL-4) (Autogen Bioclear, Calne, UK). The non-adherent DC-rich populations were aspirated, washed, resuspended in HBSS/EDTA and cytospins prepared. The presence and purity of DC pre and post culture was assessed by morphologic appearance (Giemsa staining) , flow cytometry and immunocytochemical analysis.

Statistical Analysis

Results are expressed as means ± sem or as medians and interquartile (IQ) ranges. The flow cytometric profiles of LAL and PBL were analysed using the Student's t-test. The MLC data was analysed using the non-parametric Mann-Whitney test .

Results

The mean number of cells isolated from perfusate fluid, donor and healthy bloods and spleen are presented in Table 1. Cell viability was > 95%. The viability of cells isolated by this technique is consistently greater than 95% as assessed by microscopy and trypan blue dye exclusion. The properties of these cells were compared with those of mononuclear cells isolated from the peripheral blood of healthy volunteers.

Flow Cytometry

Flow cytometric analysis revealed significant differences between the LAL and PBL populations (Table 2) . The most notable differences were between the CD4/CD8 ratios, natural killer (NK) (CD56) and activated cell (CD69) populations. The CD4/CD8 ratio was 1:3.5 in the LAL and 2.4:1 in the PBL population. A much higher proportion of CD56 cells was present in the LAL cell population (54% ± 7%) compared to the PBL (10% ± 3%) population. In addition, NKT cells (CD3⁺, CD56⁺) accounted for over 30% of the LAL CD3 population compared to only 6% of the PBL CD3 population. Furthermore, the activation marker CD69 was expressed by 62% ± 9% of LAL compared to only 9% ± 4% of PBL, and was predominantly associated with the CD3 and CD56 populations .

There were significant differences in flow cytometric characteristics between liver derived cells and peripheral blood cells, and the characteristics of perfusate cells mirrored those of hepatic leukocytes isolated by other groups using different techniques and of hepatic leukocytes analysed in si tu by immunohistochemistry . This supports the view that cells isolated from the perfusate are derived from the resident population of hepatic leukocytes, and are not simply blood leukocytes trapped in the hepatic sinusoids at the time of organ retrieval. The key properties distinguishing hepatic leukocytes from blood leukocytes are the greater number of CD8 cells relative to the CD4 cells, the high proportion of cells expressing CD69 (activated cells) , and the large number of cells expressing both CD3 and CD19 (NK T cells) . The liver is unique in containing a substantial number of NK T cells, which are rarely found elsewhere in the body or in the peripheral blood.

Allostimulatory capacity of hepatic leukocytes Allostimulation and proliferation of third party peripheral blood cells from healthy volunteer donors was measured in a one-way mixed lymphocyte culture. Hepatic leukocytes were consistently weaker allostimulatory cells, and at the highest concentration of stimulator cells, actually inhibited responder cell proliferation, as shown in figure 1. Results are shown as number of counts per minute (cpm) of ³H thymidine incorporated by responder cells in figure la, and as a stimulation index, which is the ratio of cpm incorporated in the presence of stimulators cells to cpm incorporated in the absence of stimulator cells in figure lb.

To compare the allostimulatory capacity of liver derived and peripheral blood macrophages, stimulator mononuclear cells were cultured in RIO for 1 hour at 37°C, to allow macrophages to adhere to the surface of the culture dish, and non-adherent cells were removed. Mixed lymphocyte cultures were set up using third party peripheral blood cells, and proliferation assessed by ³H-thymidine uptake. The results, shown in figure 2, show that hepatic macrophages are poorer allostimulators than peripheral blood macrophages, and that high numbers of hepatic macrophages actually inhibit the proliferation of responder cells .

Adherent donor LAL and healthy PBL induced cell concentration-dependent proliferation of allogeneic T cells (Figure 2a) . At low cell concentrations, both LAL and PBL demonstrated similar cell concentration dependent stimulation of responding T cells. However, at a seeding concentration of 1 x 10⁶ cells (~4 x 10⁵ adherent cells) , the potency of LAL significantly decreased compared to healthy PBL, which conversely increased in a dose dependent fashion (p=0.047). We also demonstrated that adherent SL induced cell-concentration dependent proliferation of allogeneic T cells (data not shown) . The trend of the allostimulatory activity was similar to that seen in healthy PBL samples (i.e. no reduction in proliferation was evident at 1 x 10⁶ cells) but reactivity was significantly higher than either the LAL or PBL samples.

We have shown that the allostimulatory activity of donor PBL was consistently weak compared to donor LAL, SL and healthy PBL (data not shown) . The median values for proliferative responses of donor PBL and LAL to PHA were similar, but both were much weaker than either healthy PBL or donor SL (Figure 3) .

Characterization of adherent and non-adherent cells Immunocytochemistry experiments revealed that adherent LAL and PBL were negative for CD56 but positive for Class II, CD68, CD80 and CD86 cell markers (Table 3) . Staining of the non-adherent cell populations demonstrated that cells were positive for all of the antibodies tested, but to variable degrees .

Differentiation of hepatic dendritic cells in vi tro Dendritic cells can be identified in complex cell populations by the expression of high levels of surface MHC class II molecules, and the co-stimulatory molecules CD80 and CD86. These cell populations are rare in both peripheral blood and hepatic perfusate and were detected in very few cells by flow cytometry. Nonetheless, when cells were cultured in the presence of recombinant GM-CSF and IL- 4, a standard method for promoting the differentiation of dendritic cells, mature CD86 positive cells that morphologically resemble dendritic cells were readily obtained (Figure 4).

Table 2 shows clearly that perfusate-derived cells comprise a distinct population to blood-derived cells. The key features are the reversal of the ratio of CD4⁺:CD8⁺ cells, and the high proportion of CD56⁺ and CD69⁺ cells in the perfusate-derived cells. These characteristics match those of intraheptic leukocytes studied in situ in the liver.

Characterization of dendritic cells.

Flow cytometry revealed that the number of DC present in freshly isolated LAL is comparable to the number present in PBL, based on the percentage of MHC class II and CD86 positive cells present in both cell populations (1% and <1% respectively) . However, slightly more CD83 positive cells were present in the LAL cell population. Immunocytochemical analysis of the non-adherent cells showed that a large proportion of these cells stained positive for the mature DC markers CD80 and CD86 (Table 3) .

We have also demonstrated an increase in DC populations following culture of cells for 7 days with GM-CSF and IL-4. FACS staining of LAL and PBL for CD83 showed a mean increase to 6.1 ± 2% and 3.7 ± 1% respectively; staining for CD86 showed a mean increase to 11.6 ± 3% and 8.7 ± 2% respectively (n = 5) . Immunocytochemical staining of these cells showed increased expression of the cell markers CD68 and CD86 (Figure 5a to 5f) . Furthermore, the morphological appearance changed and cells developed distinct cytoplasmic veils and processes. Fresh cells and cells cultured for 7 days without GM-CSF/IL-4 demonstrated weak expression of CD68 and CD86 although a few cells did exhibit strong antibody staining, suggesting that cells existed at various stages of maturation.

Table 1

Cell yields from donor and healthy samples

I

- 1 !

Table 2

Flow cytometric analysis of LAL and PBL.

LAL (Mean PBL (Mean ±SEM)

N=6

±SEM)

N=15

CD4 7.2 ± 1.2 27.9 ± 3.6*

CD8 25.2 ± 4.1 11.7 ± 1.8*

CD4/CD8 ratio 0.3 ± 0.03 2.6 ± 0.4*

CD14 3.7 ± 1.1 1.9 ± 0.8

CD19 4.6 ± 1.2 4.9 ± 0.9

CD3⁺ CD56 23.9 ± 4.1 47.7 ± 7*

CD3 CD56⁺ 42.3 ± 5 6.7 ± 1.5*

CD3⁺ CD56⁺ 11.5 ± 1.8 3.6 ± 1.2*

CD3⁺ CD56⁺ (% of CD3) 32.3 ± 4.1 6 ± 1.7*

CD69⁺CD3-CD14 CD19- 33.4 ± 4.8 1.6 ± 1*

CD69⁺CD3⁺CD14⁺C 28.6 ± 3.9 7.2 ± 2.8*

D19⁺

CD69 CD3⁺CD14⁺CD19⁺ 15.7 ± 2.4 53 ± 9.1*

CD83 (n=3) 1.38 ± 0.88 0.52 ± 0.35

CD86 (n=3) 0.43 ± 0.19 0.65 ± 0.58

Class II⁺CD3 CD19- 1 ± 0.1 1 ± 0.3

Class II⁺CD3⁺CD19⁺ 10.5 ± 2 6 ± 1.7

*p < 0.05

Footnote CD4 and CD8 represent helper and cytotoxic T cells respectively whilst CD14 and CD19 represent monocytes/macrophages and B cells respectively. CD3 and CD56 defined conventional T cells (CD3⁺ CD56^~) , natural killer (NK) cells (CD3^~ CD56⁺) and NKT cells (CD3⁺CD56⁺) . CD69 is an activation marker and CD69⁺CD3^"CD14^"CD19^" represents the activated NK cell population. MHC class II molecules are expressed on B cells, macrophages, monocytes, dendritic cells and activated T cells. The proportions of cells are expressed as percentages of the total number of cells. PBL were isolated from both donor and healthy blood samples, no significant difference in the proportion of different cell populations was apparent between the two.

Table 3

Characterization and comparison of LAL and PBL by immunostaining of adherent and non-adherent cell populations .

Footnote Non-adherent cells were removed after 1 hour of culture. Results represent the mean of four LAL and PBL samples respectively. mAb = monoclonal antibody. Figure Legends .

Figure la) . Median values for the stimulatory activity of irradated LAL and PBL4 against a constant third party (PBL3) (n=5) LAL refers to liver derived cells (liver associated leukocytes) , while PBL3 are peripeal blood cells from normal healthy volunteers, used as responder cells, and PBL4 are peripheral blood cells from normal healthy volunteers used as a stimulator cells, i.e. treatd n the same way as the LAL.

Figure lb) . Median stimulation indices for irradiated LAL and PBL4 against a constant third party (PBL3) n=5) . LAL refers to liver derived cells (liver associated leukocytes) , while PBL3 are peripheral blood cells from normal healthy volunteers, used as responder cells, and PBL4 are peripheral blood cells from normal healthy volunteers used as stimulator cells, i.e. treated in the same way as the LAL.

Figure 2. Allostimulatory activity of liver > associated leukocytes versus peripheral blood lymphocytes. Median allostimulatory activities of adherent LAL (n=15) and PBL

(n=14) against a constant third party. The actual number of adherent LAL was 3.7xl0⁵ ± 4xl0⁴ (1 x 10⁶) , 2xl0⁵ ± 2xl0⁴ (5 x 10⁵), 1.3xl0⁵± lxlO⁴ (2.5 x 10⁵) , and 6.2xl0⁴ ± 2xl0⁴

(1.25 x 10⁵) ; and adherent PBL was 4.3xl0⁵ ± 4xl0⁴ (1 x 10⁶) , 2.5xl0⁵ ± 4xl0⁴ (5 x 10⁵) , l.lxl0⁵± 3xl0⁴ (2.5 x 10⁵) and 5.4xl0⁴ ± 2xl0⁴ (1.25 x 10⁵) . Adherent LAL and PBL were separately quantified in four different samples. Figure 3. Stimulatory activity of different cell types. Median values for the stimulatory activity of adherent LAL (n=15) , donor PBL (n=14), healthy PBL (n=29) and splenic lymphocytes (n=3) against PHA.

Figure 4. Cytospin preparations of liver associated leukocytes . Giemsa stained cytospin preparation: Four larger cells with pale nuclei and extensive pale cytoplasm and prominent cytoplasmic processes show typical features of mature dendritic cells. The majority of cells shown are lymphocytes, which are smaller and round with dense nuclear and cytoplasmic staining.

Figure 5. Cytospin preparations of liver associated leukocytes isolated from healthy human liver. Freshly isolated LAL were stained for CD68 (a) and CD86 (b) cell markers (magnification x40) . Many of the CD68 stained cells had irregular shaped nuclei and perinuclear staining. CD86 demonstrated a more diffuse pattern of cytoplasmic staining. No staining was evident on cells with irregular nuclei. After 7 days in culture, the morphology of the cells changed. Cells cultured without GM-CSF and IL-4 demonstrated weak or no staining for CD68 (c) and CD86 (e) (magnification x20) . However, cells cultured with GM-CSF and IL-4 had developed the classical dendritic cell morphology, consisting of a large, irregular nucleus, and extensive cytoplasm with many folds and processes. In addition, a number of these cells were found clustered with small round cells that were presumed to be lymphocytes. Strong intracellular and surface staining was demonstrated for CD68 (d) and CD86 (f) respectively (magnification x40) . Summary

A new technique has been developed to isolate functional mononuclear cells from normal human liver in quantities that make feasible further in vivo manipulation and investigation. The cells isolated include unique lymphocyte populations, macrophages and dendritic cells that are not available by other techniques. These cells will be useful to study the immunological and cell biological properties of human hepatic leukocytes, and may also be useful therapeutically, as they have the potential to be cultured and manipulated in vitro, and administered to patients. Potential uses include regulation and control of the immune system in allograft rejection and autoimmune disease.

Discussion

The invention provides a robust and reproducible method, which allows us to readily obtain important sub-populations of cells, such as DC and KC, as well as the mainly lymphocytic population previously described (1) . There is increasing evidence to suggest that DC are central to many of the mechanisms that have been proposed to explain the unique tolerogenic properties of liver allografts. These include T cell anergy and T cell suppression [11,12], activation-induced apoptosis [13,14], induction of regulatory cells [13,15], veto function [16] and immune deviation toward a Th2 response [8,9,17]. However, a more recent study has shown that DC progenitors can also act as immunogenic antigen-presenting cells in the liver [18]. The balance between immunity and tolerance is dependent on the amount and nature of the antigen, its route of entry into the body, and whether it manifests itself intracellularly or extracellularly [19]. The composition of liver leukocyte populations has already been determined by in si tu immunohistochemistry and flow cytometry, however there have been few functional studies, mainly due to difficulties in isolating sufficient numbers of cells. We have confirmed that the cells we have isolated are identical to hepatic leukocytes isolated by other techniques and distinct from peripheral blood cells [1,20]. By studying the earliest responses of donor associated hepatic leukocytes with third party cells, we tried to recapitulate in vi tro some of the immunological events occurring immediately after engraftment. While this limited interaction does not reproduce all features of immunological engagement, it provides important and unique data on the earliest cellular interactions.

DC are now considered to play a leading role in the induction of transplantation tolerance, however KC have also been considered in this context, and it is unknown how the two cell populations may be inter-linked. However, recent studies have shown that KC may play a role in the trafficking of DC from the blood to the liver. Immunohistochemical studies revealed that immature phagocytic DC translocated from the blood to celiac lymph via hepatic sinusoids, down-regulated their phagocytic ability as they migrated. Interestingly, selective binding of DC to KC was revealed [21] . More recently, Uwatoku et al demonstrated that DC-KC binding through N- acetylgalactosamine-specific C-type lectin like receptors is crucial for DC recruitment to the liver [22]. Cellular manipulation with cytokines has made it possible to generate pure DC in sizeable numbers. In vitro, DC can be generated from human CD34⁺ bone marrow, cord blood, and peripheral blood progenitor cells after culture with different cytokine combinations including GM-CSF, stem cell factor, and either IL-4 or TNFD. Monocytes cultured in the presence of IL-4 acquire a macrophage like DC morphology where cells increase in size and develop extensive processes. IL-4 appears to suppress overgrowth by macrophages and controls differentiation of monocytes into DC. GM-CSF alone induces the formation and survival of aggregates of DC progenitors [10]. In the present study we have been able to generate large numbers of DC from human liver leukocytes using a combination of GM-CSF and IL-4. We have shown that mature DC may have been present in freshly isolated cell samples as increased CD86 expression was observed in immunocytochemistry experiments, despite very weak expression on FACS staining (<1%). We have shown varying degrees of CD86 expression pre and post culture with GM-CSF and IL-4. Generally, immature DC differentiate locally to acquire their ability to endocytose antigens [19]. These DC capture foreign antigens, process them with high efficiency and thus generate ligands for the antigen- specific T cell receptors. Immature DC synthesize high levels of MHC class II antigens but only few or no adhesion and co-stimulatory molecules. The shape of immature DC is dendritic in the strict sense of the word. Migratory or veiled DC leave peripheral tissues carrying on their surfaces immunogenic MHC/peptide complexes to extravascular sites or to the T cell areas of draining lymphatic organs. During migration DC change their phenotype as they mature for high antigen presenting capacity. Final maturation may be induced by interaction with T cells.

There is now compelling evidence for considerable heterogeneity among DC with respect to their origin and phenotype, and the nature of the responses they elicit [23] . Whereas unfractionated murine DC from the spleen preferentially mount Thl like responses, those associated with the liver and mucosal sites such as the respiratory tract and Peyer's Patches [8,9,17] polarise T cells toward Th2 cytokine profiles. It is unknown whether a distinct subset of DC exists, dedicated to the induction and maintenance of peripheral tolerance. There are indications however that DC function is less predetermined than once thought and is subject to environmental factors; tolerogenicity may relate to the maturation state of the individual DC. There are various lines of evidence to suggest that the protolerogenic properties of immature DC may be specifically harnessed for the establishment of transplantation tolerance.

We have demonstrated significantly poor allostimulatory capacity in adherent LAL populations compared to PBL. We are currently investigating what factors may be responsible for this effect. It is feasible that at high cell concentrations, KC, DC or both may induce the production of various inhibitors of T cell proliferation or T cell apoptosis. However, recent evidence implicates immune deviation toward a Th2 response as the most likely candidate for this reduced reactivity in liver-derived cells [8,9]. In addition, we showed that irradiated (i.e. total cells) LAL (n=5) and PBL (n=5) induced cell- concentration dependent proliferation of allogeneic T cells (data not shown) , the optimum response occurred at a stimulator cell concentration of 5 x 10⁵. LAL were notably less potent than PBL, but did not exhibit the same contrasting reactivity observed in the adherent MLC experiments .

Splenic lymphocytes isolated from tissue that had been preserved in Marshall's solution at 4°C, demonstrated very potent allostimulatory activity in MLC (data not shown) . Thus, exposure of donor liver to reduced temperature and preservative solution was not responsible for the reduced allostimulatory activity demonstrated by LAL in vi tro . In contrast to Jonsson' s study [1], we showed that the allostimulatory reactivity of donor PBL was consistently weak. However, their study did not include comparative data with healthy PBL. We demonstrated that donor PBL and LAL were similar in their response to stimulation by PHA, but both were weaker than either healthy PBL or donor spleen. The decreased proliferative capacity of LAL may have been partly due to the low percentage of CD3 cells in that population compared to PBL (35 ± 6% vs. 51 ± 8% (Table 2).

No morphological difference was evident between freshly isolated LAL and PBL, although CD80 and CD86 staining was weaker in the latter. Immunocytochemical staining with anti-CD68 demonstrated that adherent cells consisted mainly of macrophages/monocytes, but may also have included DC. In our MLC experiments, non-adherent cells were aspirated after 1 hr of culture to remove T cells, B cells and NK cells. However, as DC are loosely adherent at this early stage of culture [10] , DC progenitors are likely to have remained in this antigen-presenting adherent cell population, despite washing of the cells prior to addition of T cell responders. After 1 hour in culture adherent LAL and PBL were strongly CD68 positive but weakly CD80 and CD86 positive. In contrast, non-adherent LAL were weakly positive for CD68 but strongly positive for CD80 and CD86, suggesting that mature DC were present in the non-adherent cell fractions. Importantly, immature DC progenitors present in the loosely adherent cell population may have contributed towards the reduced allostimulatory reactivity observed in the LAL population. Overall, PBL demonstrated weak CD80 and CD86 expression and did not exhibit reduced allostimulatory activity, suggesting that LAL must exhibit some unique characteristics.

In conclusion, we have demonstrated that consistent with other reports hepatic leukocytes demonstrate reduced allostimulatory activity in MLC. These findings support the increasingly popular view that liver leukocytes, particularly DC, play an important role in the tolerogenicity of liver allografts.

References

1. Jonsson JR, Hogan PG, Balderson GA, et al . Human liver transplant perfusate: an abundant source of donor liver- associated leukocytes. Hepatology 1997; 26: 1111-1114.

2. Calne RY, Sells RA, Pena JR, et al . Induction of immunological tolerance by porcine liver allografts. Nature 1969; 223: 472-6.

3. Kamada N, Brons G, Davies HS . Fully allogeneic liver grafting in rats induces a state of systemic nonreactivity to donor transplantation antigens. Transplantation 1980; 29: 429-31

4. Qian S, Demetris A, Murase N, et al . Murine liver allograft transplantation: tolerance and donor cell chimerism. Hepatology 1994; 19: 916-24.

5. Kamada N, Wight DG. Antigen-specific immunosuppression induced by liver transplantation in the rat. Transplantation 1984; 38: 217-21.

6. Thompson AW, Drakes ML, Zahorchak AF et al . Hepatic dendritic cells: immunobiology and role in liver transplantation. J Leuk Biol 1999; 66: 322-330.

7. Bell D, Young JW, Banchereau J. Dendritic cells. Adv Immunol 1999; 72: 255-324.

8. Khanna A, Morelli AE, Zhong C, et al . Effects of liver- derived dendritic cell progenitors on Thl- and Th2-like cytokine responses in vitro and in vivo. J Immunol. 2000; 164 (3) : 1346-54.

9. Iwasaki, A & Kelsall BL . Freshly isolated Peyer's patch, but not spleen, dendritic cells produce interleukin 10 and induce the differentiation of T helper type 2 cells. J Exp Med 1999; 190 (2) : 229-39.

10. Romani N, Gruner S, Brang D, et al . Proliferating Dendritic Cell Progenitors in Human Blood. J Exp Med. 1994, 180: 83-93

11. Grohmann U, Bianchi R, Ayroldi E, et al . A tumor- associated and self antigen peptide presented by dendritic cells may induce T cell anergy in vivo, but IL-12 can prevent or revert the anergic state. J Immunol 1997; 158: 3593-3602.

12. Chen W, Sayegh MH, Khoury SJ. Mechanisms of acquired thymic tolerance in vivo: intrathymic injection of antigen induces apoptosis of thymocytes and peripheral T cell anergy. J Immunol 1998; 160: 150

13. Wells AD, Li XC, Li Y, et al . Requirement for T cell apoptosis in the induction of peripheral transplantation tolerance. Nat Med 1999; 5: 1303-07.

14. Li Y, Li XC, Zheng XX, et al . Blocking bothe signal 1 and signal 2 of T cell activation prevents apoptosis of alloreactive T cells and induction of peripheral allograft tolerance. Nat Med 1999; 5: 1298-1302. 15. Clare-Salzler MJ, Brooks J, Chai A, et al . . Prevention of diabetes in nonobese diabetic mice by dendritic cell transfer. J Clin Invest 1992; 90: 741-748.

16. Thomas JM, Carver FM, Kasten-Jolly, et al . Further studies of veto activity in rhesus monkey bone marrow in relation to allograft tolerance and chimerism. Transplantation 1994; 57: 101-115.

17. Stumbles PA, Thomas JA, Pimm CL, et al . Resting respiratory tract dendritic cells preferentially stimulate T helper cell type 2 (Th2) responses and require obligatory signals for induction of Thl immunity. J Exp Med 1998; 188: 2019-2031.

18. Abe M, Akbar SMF, Horiike N, et al . Induction of cytokine production and proliferation of memory lymphocytes by murine liver dendritic cell progenitors: role of these progenitors as immunogenic resident antigen-presenting cells. J Hepatol 2001; 34: 61-67.

19. Banchereau J and Steinman RM. Dendritic cells and the control of immunity. Nature 1998; 392: 245-252.

20. Norris S, Collins C, Doherty DG, et al . Resident human hepatic lymphocytes are phenotypically different from circulating lymphocytes. J Hepatol. 1998; 28: 84-90.

21. Matsuno K, Ezaki T, Kudo S, et al . A life stage of particle-laden rat dendritic cells in vivo: their terminal division, active phagocytosis, and translocation from the liver to the draining lymph. J Exp Med 1996; 183: 1865- 1878.

22. Uwatoku R, Suematsu M, Ezaki T, et al . Kupffer cell- mediated recruitment of rat dendritic cells to the liver: roles of N-acetylgalactosamine-specific sugar receptor. Gastroenterology 2001; 121: 1460-1472.

23. Reid SC, Penna G, Adorini L. The control of T cell responses by dendritic cell subsets. Curr Opin Immunol 2000, 12: 114-121.

Claims

CLAIMS :

1. A process for preparing human hepatic dendritic cells, which comprises:

(a) providing a substantially intact, human liver suitable for transplantation;

(c) perfusing the treated liver so as to produce a perfusate containing dendritic cells; and

(d) isolating from the perfusate a cell fraction comprising human hepatic dendritic cells.

2. A process according to claim 1, wherein the step of storing the liver under conditions to promote dendritic cell availability comprises storing for at least 6 hours.

3. A process according to claim 2, wherein the liver is stored for at least 12 hours.

4. A process according to any one of claims 1 to 3, wherein the step (c) of perfusing the treated liver comprises perfusing with a perfusion solution containing around 4% to 4.5% human albumin.

5. A process according to any preceding claim, wherein step (d) comprises a step of separating mononuclear cells from the perfusate to produce a cell fraction enriched with human hepatic dendritic cells.

6. A process according to claim 5, wherein the mononuclear cells are separated by density gradient centrifugation.

7. A process according to any preceding claim, which further comprises further density gradient separation, antibody based separation^' using fluorescent activated cell sorting or magnetic bead separation; or centrifugal elutriation.

8. A process according to any preceding claim, wherein the human hepatic dendritic cells are MHC class II⁺, CD3^~ , CD19⁺, CD80⁺ and CD86⁺.

9. Isolated human hepatic dendritic cells obtainable by:

(a) providing a substantially intact, optionally perfused, resected human liver;

10. Isolated human hepatic dendritic cells according to claim 9, obtainable by a process according to any one of claims 2 to 8.

11. Isolated human hepatic dendritic cells according to claim 9 or claim 10, for use in medicine.

12. Use of isolated human hepatic dendritic cells according to claim 9 or claim 10, for the preparation of a composition for treating an autoimmune disorder.

13. Use of isolated human hepatic dendritic cells according to claim 9 or claim 10, for the preparation of a composition for treating or preventing allograft rejection in a recipient, wherein the allograft is a non-liver allograft from the same donor as the liver.