Technical Field This invention relates to therapeutics, transplantation and immunology. More specifically, it relates to a newly discovered class of chemically modified oligonucleotides with increased physiological stability that are capable of interacting with transplantation antigen nucleotide sequences, and methods for making cells that are more easily transplanted into a recipient host using these oligonucleotides to reduce the level of transplantation antigens expressed on the cell surface of transplanted cells.

Background of the Invention

Among gene products that relate to transplantation antigens are the products of the Human Leukocyte Antigen (HLA) complex, located on the short arm of chromosome 6. The HLA antigens are divided into two classes depending on their structure. The genetic loci denoted HLA-A -B, and -C code for the HLA Class I antigens, and HLA-DP, -DQ and -DR code for the HLA Class II antigens. HLA Class II molecules are composed of two non- covalently linked glycoproteins, the α chain and the highly polymorphic β chain. Each chain contains one extracellular domain, a transmembrane segment and a cytoplasmic tail. The structure of the α and β chains and their genes have been elucidated. All known Class II

genes are similar in structure and encoded by exons 1 - 4, with exon 5 coding for an untranslated region. The DP, DQ and DR loci all consist of multiple genes. A total of twelve class II genes have been identified. In some haplotypes, some class II genes do not code for a functional peptide and are classified as pseudogenes. Regulation of HLA class II antigen expression by binding anti-gene oligonucleotides to the structural region of the gene has not been reported in the literature. Regulation of HLA class II antigen expression occurs in part through a series of promoter regions such as the J, , X (including X-*_ and X2) ■ and Y boxes, and the gamma interferon response element. The X (including ^■ X-*_ and X ₂ ) and Y boxes are known to be required in the transcriptional regulation of all class II promoters.

Ono, S.J. et al., Proc. Natl. Acad. Sci. (USA) (1991) 88: ^• 4304-4308.

Transcription of HLA-DRα Class II is activated by RF-X (Regulatory Factor-X) which binds to the X-box region (-110 to -95) of the DRot promoter. RF-X and its binding site, the X-box are unique and have a high specificity for each other. The DNA binding domain of RF-X consists of 91 amino acids with a basic stretch and shares no notable homology with other known DNA binding motifs (Reith et al . (1990), Genes Dev. , Vol. 4(9), pp. 1528-40) . The X-box sequence is an atypical promoter site, being neither palindromic nor dyad symmetric. Additionally no other sequence (using the program Eugene) shows exact homology with the X-box. The X-box is conserved in humans. No other known cloned transcription factors bind to this entire region of the X-box in the same manner.

HLA antigens are implicated in the survival of cell grafts or transplants in host organisms. Although there is acceptable graft survival in the first year for

nearly all types of transplants, by five and ten years after transplantation only 40-50% of all grafts are still functioning. This low rate is due to the relentless attack of the immune system on the graft. In addition, death rates of 1-5% are recorded even at the best transplant centers. Drugs are commonly used to control immune responses and prevent graft rejection, and death is often an indirect result of this drug administration. The drugs used to control immune responses usually cause a non-specific depression of the immune system. A patient with a depressed immune system is far more susceptible to develop life-threatening infections and a variety of neoplasia. The low rate of long term success, and serious risks of infection and cancer are the two main challenges now facing the entire field of tissue and organ transplantation.

It has been suggested that graft rejection can be prevented or reduced by reducing the levels of exposed HLA antigens on the surface of transplant cells. Faustman, D. et al. , Science (1991) 252;1700-1702, observed that xenograft survival was increased by masking HLA class I surface antigens with F(ab')2 antibody fragments to HLA class I or tissue specific epitopes. One way to reduce the level of cell surface transplantation antigens is to downregulate the expression of the transplantation antigen genes.

Generally, eucaryotic gene expression may be regulated at any of the steps from DNA transcription to RNA translation to protein; and it is generally agreed that the expression of most genes is regulated primarily at the level of transcription. In order for transcription to occur, transcription factors must bind distinct regulatory sites or promoters on the gene. Once bound, transcription factors may interact with RNA polymerase or other factors to activate or repress

transcription. Some transcription factors are constitutively expressed in specific cells while others may be transiently activated in response to various physiological signals (such as cAMP, γ-IFN, etc.). Thus in a given cell transcription of particular genes depends on which transcription factors are present in that cell type and/or whether the signals to activate the transcription factors are present.

Agents such as actinomycin (an intercalator) have been used to block transcription in a nonspecific manner. A variety of approaches to sequence-specific gene modulation include use of antisense oligonucleotides and antigene oligonucleotides (triple helix formers) . These are limited in general or in particular instances. Antisense oligonucleotides block gene expression by targeting mRNA while triple helix forming oligos target double-stranded DNA. Inaccessibility of the target mRNA due to RNA secondary structure can limit the usefulness of antisense methods; triple helix approaches are limited by poor nuclear access, chromatin structure (bypassing histones) , and the need for targeting a homopurine-homopyrimidine stretch. In addition, degradation of oligonucleotides by exonucleases can limit the effectiveness of both antisense and triple helix methods. Chemical modification of oligos can improve nuclease resistance but can also result in increased toxicity, reduced binding affinity, and lower activity (Cook 1991, Crooke 1991)..

Anti-gene code molecules are short RNA or DNA transcripts that are "antisense" (i.e., complementary to a DNA or RNA strand in a Watson-Crick pairing manner) to a portion of the normal mRNA and are not translated. Regulation of expression of genes by anti-gene code RNA, one of the natural modes of gene regulation, was first recognized in prokaryotes. Green, P.M. et al., Ann. Re .

Biochem (1986) 5_5:569. Natural anti-gene codes and artificial anti-gene codes have been used in prokaryotes to downregulate prokaryotic proteins. Simmons, R.W. et al., Cell (1983) 34*683; Mizuno, T. et al. , Proc. Natl. Acad. Sci. (1984) 11:1966; Okamoto, K. et al. , Proc.

Natl. Acad. Sci. (1986) £2-5000; Pestka, S. et al. , Proc. Natl. Acad. Sci. (1984) .81:7525; Coleman, J. et al. , Cell (1984) 37:429; Farnham, P.J. et al. , Proc. Natl. Acad. Sci. (1985) 82:3978; Kindy, M.S. et al. , Mol. Cell. Biol. (1987) 7:2857.

Artificial anti-gene codes have also been synthesized and used to regulate eukaryotic gene expression. Microinjection or transfection of thymidine kinase (TK) anti-gene codes has been shown to inhibit expression of the TK protein. Izant, J.G. et al., Cell (1984) 3_6:1007; Kim, S.K. et al. , Cell (1985) 4.2.129. Additionally, short anti-gene codes to the 5' untranslated region of the thymidine kinase gene successfully downregulates protein expression. Izant, J.G. et al., Science (1985) 229:345. Other examples of the regulation of eukaryotic gene expression by anti-gene codes are the pp66 c-src gene (by transfected full length anti-gene codes) , and the c-fos gene (by an anti-gene code spanning the 5' untranslated region of the first exon) . Amini, S. et al. , Mol. Cell. Biol. (1986) 6.:2305; Holt, J.T. et al., Proc. Natl. Acad. Sci. (1986) j32:4794. These anti-gene codes have been introduced under constitutive or heterologous inducible promoters.

Synthetic oligomers have also been used to downregulate the expression of c-myc in promyelocytic leukemia cells, and T-lymphocytes. Wickstrom, E.L. et al., Proc. Natl. Acad. Sci. (1988) J3_5:1028; Heikkila, R. et al., Nature (1987) 328:445. C-myc anti-gene code oligonucleotides have been shown to inhibit proliferation in normal hematopoietic cells. Gewirtz, A.M. et al. ,

Science (1988) 242:1303. An anti-gene code to a CD8 fragment downregulated the expression of CD8 molecules on the surface of human cytotoxic T-cells. Hambor, J.E. et al., Proc. Natl. Acad. Sci. (1988) 15:4010. Lotteau et al., J. Exp. Med. (1989) 169:351 used episomal vectors to introduce DR A coding sequences in B-lymphoblastoid cell lines and downregulated the expression of DR A-DQ B mixed isotype heterodimers, but did not observe any changes in the levels of isotype matched DR A-DQ B heterodimers. Anti-gene code oligonucleotides may act to prevent transcription by inhibiting RNA polymerase, by binding to mRNA and preventing ribosomal translation, by decreasing the stability of mRNA through enhancement of mRNA degradation by RNase H, or by preventing or inhibiting the processing to mature mRNA. Maher, L.J. et al., Science (1989) 245.:725; Moser, H.E. et al. , Science (1987) 238:645: Melton, D.A. et al. , Proc. Natl. Acad. Sci. (1985) H:144; Gewirtz, A.M. et al. , Science (1989) 245:180; Walder, R.Y. et al. , Proc. Natl. Acad. Sci. (1988) 15_:5011. Absolute homology between the target and the antisense sequences is preferred but not required for the inhibition. Holt, J.T., supra.

Anti-gene code oligonucleotides may also form a triplex DNA structure with the intact duplex gene. Moffat, A.S., Science (1991) 152:1374-1375. This technique of making anti-gene code oligonucleotides involves the formation of a triplex structure according to certain binding rules. When this triplex structure is formed in the promoter region of a gene, it has been shown to disrupt transcription of that gene. Orson, F.M. et al., Nuc. Acids Res. (1991) 19:3435-3441.

Summarv of the Invention

The present invention is drawn to a newly discovered class of chemically modified oligonucleotides with increased physiological stability that are also capable of interacting with transplantation antigen nucleotide sequences. In specific, these modified oligonucleotides comprise a 3' -CH ₂ CH(OH)CH ₂ NH ₂ group that dramatically increases the physiological stability of the oligonucleotides of the invention. The present invention is also drawn to methods for making cells that are more easily transplanted into a recipient host using these modified oligonucleotides to reduce the level of transplantation antigens expressed on the cell surface of target cells. Oligonucleotides capable of binding to a transplantation antigen nucleotide sequence are described ^" in WO 93/14769, which also describes methods for employing oligonucleotides that are capable of binding in some fashion to a nucleotide sequence relating to a transplantation antigen, and preventing expression of that antigen. Cells treated with these oligonucleotides will express significantly less of the targeted antigen, and when transplanted will be more easily tolerated by the recipient host. Although these oligonucleotides were designed to be capable of binding to a transplantation antigen nucleotide sequence, it was contemplated that their ultimate mode of action may be different.

The present invention contemplates the use of the chemically modified oligonucleotides in the development of a "universal donor cell" reduced in one or more transplantation antigen, preferably one or more MHC antigen. The absence of certain MHC antigens on the surface of donor cells, tissues or organs comprising these cells will cause them not to be recognized as foreign and not to elicit a rejection response. By the

selective introduction of anti-gene codes into a cell it is possible to block the expression of targeted MHC genes, thereby rendering a graft "invisible" to the immune system. Thus, the problem of rejection is eliminated without nonspecific suppression of the immune system, and the immune system remains active to defend against infection and neoplasia. By implementing the methods of the present invention, even xenogeneic donor cells, tissues and organs can be rendered "invisible" to the immune system. Thus, the target cells or organs of the present invention may be derived from non-human mammalian species, made into transplantation antigen- depleted cells or universal donor organs, respectively, by a method of the invention, and subsequently transplanted into a human individual.

The class I and class II genes of the major histocompatibility complex as well as ICAM-1 are regulated by interferon-γ ("IFN-7") in a variety of cell types. We have now identified a family of modified oligonucleotides whose members are capable of inhibiting IFN-γ mediated MHC-I induction and/or ICAM-1 induction in a variety of cell types. The inhibition of cell surface expression of MHC-I and of ICAM-1, at least by some of the oligonucleotides in the family, is selective for IFN- γ mediated induction; that is to say, these members do not inhibit MHC-I induction or ICAM-1 induction by comparable doses of either IFN-α. or IFN-/3. Particularly, we have identified several members of this oligonucleotide family, which selectively inhibit IFN-γ mediated cell surface enhancement of the MHC Class I ("MHC-I") and ICAM-1 proteins in the K562 cell line. The present invention embodies methods for making a transplantation antigen-depleted cell from a target cell, the method comprising: (a) obtaining the target cell; and

(b) exposing the target cell to a oligonucleotide capable of binding to a transplantation antigen nucleotide sequence, said- oligonucleotide being present in an amount sufficient to make the target cell a transplantation antigen-depleted cell; said oligonucleotide being selected from the group consisting of

I 5' GGG GTT GGT TGT GTT GGG TGT TGT GT

AGG GTT CGG GGC GCC ATG ACG GC -RNH ₂ , ' CAG CCT TGA GGA TTC CCC AAC TCC G

V 5' GCC ACG GAG CGA GAC ATC TCC G —RNH ₂ , where R= —CH ₂ CH(OH)CH ₂ — , and 3' -RNH ₂ -containing fragments thereof that retain binding activity. The present invention also embodies transplantation antigen-depleted cells prepared by a method of the invention.

The present invention also embodies oligonucleotides for use in the preparation of a composition for treating target cells to make them transplantation antigen-depleted, wherein the oligonucleotide is capable of binding to a portion of the transplantation antigen nucleotide sequence; said oligonucleotide being selected from the group consisting of

I 5' GGG GTT GGT TGT GTT GGG TGT TGT GT

AGG GTT CGG GGC GCC ATG ACG GC -RNH ₂ ,

CAG CCT TGA GGA TTC CCC AAC TCC G

V 5' GCC ACG GAG CGA GAC ATC TCC G —RNH ₂ , where R= —CH ₂ CH(OH)CH ₂ — , and 3' -RNH ₂ -containing fragments thereof that retain binding activity. The present invention further embodies universal donor organs prepared by the method comprising:

(a) obtaining a target organ from an individual; and (b) exposing the target organ to an oligonucleotide capable of binding to a transplantation antigen nucleotide sequence, said oligonucleotide being preset in an amount sufficient to make the target organ a universal donor organ,* said oligonucleotide being selected from the group consisting of

I 5' GGG GTT GGT TGT GTT GGG TGT TGT GT

' AGG GTT CGG GGC GCC ATG ACG GC —RNH ₂ ,

CAG CCT TGA GGA TTC CCC AAC TCC G

V 5' GCC ACG GAG CGA GAC ATC TCC G -RNH ₂ , where R= —CH ₂ CH(OH)CH ₂ — , and 3' -RNH ₂ -containing fragments thereof that retain binding activity.

The present invention further embodies methods for modulating the expression of a polypeptide, wherein the expression is regulated by IFNγ, said method comprising, exposing a cell to an oligonucleotide capable of binding to a transplantation antigen nucleotide sequence, said oligonucleotide being present in an amount sufficient to disrupt the IFNγ-regulated expression of said polypeptide in said cell; said oligonucleotide being selected from the group consisting of 1 5' GGG GTT GGT TGT GTT GGG TGT TGT GT

-RNH _2i

III 5' AGG GTT CGG GGC GCC ATG ACG GC —RNH ₂ ,

IV 5' CAG CCT TGA GGA TTC CCC AAC TCC G

— NH ₂ , and V 5' GCC ACG GAG CGA GAC ATC TCC G -RNH ₂ . where R= —CH ₂ CH(OH)CH ₂ — , and 3'-RNH ₂ -containing fragments thereof that retain binding activity.

Brief Description of the Drawings

Figs. 1.1 - 1.3 are graphs showing data from flow cytometry expressed as a percentage of the level obtained with IFN-γ in the absence of oligonucleotide. K562 cells were treated with either —■— oligonucleotide I or —O— oligonucleotide II and stained for cell surface MHC-I expression 24, 48 and 72 hours later (n = 3; error bars= ±1 SD) . Oligonucleotide I inhibition of MHC-I induction by IFN-γ is dose-dependent. Oligonucleotide II treatment does not inhibit the effects of IFN-γ. Control cells were not treated with IFN-γ or oligonucleotide.

Figs. 2.1 - 2.3 are graphs showing data from flow cytometry expressed as a percentage of the level obtained with IFN-γ in the absence of oligonucleotide.

K562 cells were treated with either —■— oligonucleotide I or —O— oligonucleotide II and stained for cell surface 0 ₂ microglobulin expression 24, 48 and 72 hours later (n = 3; error bars: ±1 SD) . Oligonucleotide I inhibition of MHC-I induction by IFN-γ is dose-dependent. Oligonucleotide II treatment does not inhibit the effects of IFN-γ. Control cells were not treated with IFN-γ or oligonucleotide.

Figs. 3.1 - 3.3 are graphs showing data from flow cytometry expressed as a percentage of the level obtained with IFN-γ in the absence of oligonucleotide. K562 cells were treated with either —■— oligonucleotide I or —O— oligonucleotide II and stained for cell surface ICAM-1 expression 24, 48 and 72 hours later (n = 3; error bars: ±1 SD) . Oligonucleotide I inhibition of MHC-I induction by IFN-γ is dose-dependent. Oligonucleotide II treatment does not inhibit the effects of IFN-γ. Control cells were not treated with IFN-γ or oligo.

Figs. 4.1 - 4.3 are histograms showing cell surface MHC-I levels 24, 48, and 72 hours after K562 cells (n = 3; error bars: ±1 SD) were treated with 25 μM oligo I and either IFN-γ (800 U/ml) , IFN-α (6400 U/ml) or IFN-/3 (6400 U/ml) . Levels are expressed as a percentage of the appropriate interferon induced value for (A) Control cells (B) Cells treated only with the indicated interferon, and (C) Cells treated with the indicated interferon plus 25 μM oligonucleotide I. Oligonucleotide I is selective for IFN-γ and does not inhibit MHC-I induced by either IFN-α or IFN-0 under these conditions.

Figs. 5.1 - 5.3 are histograms showing cell surface β ₂ Microglobulin levels 24, 48, and 72 hours after K562 cells (n = 3; error bars: +1 SD) were treated with 25 μM oligo I and either IFN-γ (800 U/ml) , IFN-α

(6400 U/ml) or IFN-/3 (6400 U/ml) . Levels are expressed as a percentage of the appropriate interferon induced value for (A) Control cells (B) Cells treated only with the indicated interferon, and (C) Cells treated with the indicated interferon plus 25 μM oligonucleotide I.

Oligonucleotide I is selective for IFN-γ under these conditions.

Figs. 6.1 - 6.3 show cell surface ICAM-1 levels 24, 48, and 72 hours after K562 cells were treated with 25 μM oligonucleotide I and either IFN-γ (800 U/ml) or TNF-α (800 U/ml) . Levels are expressed as a percentage of the appropriate induced value for (A) Control cells (B) Cells treated with the indicated cytokine, and (C) Cells treated with the indicated cytokine and 25 μM oligonucleotide I. Oligonucleotide I is selective for IFN-γ under these conditions.

Fig. 7 are histograms showing data from staining for: (i) -MHC-I, (ii) β ₂ microglobulin, (iii) ICAM-1, (iv) MHC-II DR, (v) transferrin receptor; Raji cells (Top row), HeLa S3 cells (Middle row) and HUT78

cells (Bottom row) were either mock treated, treated with IFN-γ (800 U/ml) or treated with IFN-γ (800 U/ml) and 25 μM oligo I (n = 2; error bars: ±1 SD) . Cells were stained 48 hours later and the results are expressed as a percentage of the appropriate induced value.

Figs. 8.1 - 8.3 are histograms showing data from flow cytometry expressed as a percentage of the level obtained with IFN-γ i the absence of oligonucleotide. K562 cells were treated with either oligonucleotide I, III, IV, V or VI and stained for cell surface MHC-I, 0 ₂ microglobulin and ICAM-1 expression 48 hours later (n = 3; error bars: ±1 SD) . The following activity relationship summarizes Figs. 4.1 - 4.3: Oligo I « Oligo III > Oligo IV *=» Oligo V > Oligo VI « No treatment.

Detailed Description of the Invention

The present invention provides modified oligonucleotides and methods of their use in the making of transplantable cells for medical treatment. These transplantation antigen-depleted cells give rise to improved graft survival rates in the recipient or require lower levels of immunosuppressant drug administration in the recipient. These cells may also be useful in treating patients with autoimmune diseases.

In one aspect, this invention provides a method for making a transplantation antigen-depleted cell from a target cell comprising obtaining the target cell, and then exposing the target cell to a modified oligo- nucleotide capable of binding to a transplantation antigen nucleotide sequence, wherein the oligonucleotide is presented or produced locally in an amount sufficient to make the target cell a transplantation antigen- depleted cell. The modified oligonucleotide is capable of binding to the nucleotide sequence according to

Watson-Crick or triplex binding rules (which includes Hoogsteen-like bonds) . In preferred embodiments the transplantation antigen is an MHC class I or II antigen. In another aspect of this invention, a transplantation antigen-depleted cell is provided, prepared by obtaining a target cell, and exposing the target cell to a modified oligonucleotide of the invention capable of binding to a transplantation antigen nucleotide sequence, wherein the oligonucleotide is presented or produced locally in an amount sufficient to make the target cell a transplantation antigen-depleted cell.

In yet another aspect of this invention, an oligonucleotide capable of binding to a double-stranded transplantation antigen nucleotide sequence is provided.

In still another aspect of this invention, a universal donor organ is provided, prepared by obtaining a target organ from an individual, and exposing the target organ to a modified oligonucleotide of the invention capable of binding to a transplantation antigen nucleotide sequence, wherein the oligonucleotide is presented or produced locally in an amount sufficient to make the target organ a universal donor organ.

In a further aspect of this invention, a method of treating an individual with an autoimmune disease characterized by dysfunctional expression of a transplantation antigen is provided, comprising administering to that individual a modified oligonucleotide of the invention capable of binding to a portion of the transplantation antigen nucleotide sequence, in an amount sufficient to inhibit expression of the transplantation antigen.

The practice of the present invention encom¬ passes conventional techniques of chemistry, molecular biology, biochemistry, protein chemistry, and recombinant

DNA technology, which are within the skill of the art. Such techniques are explained fully in the literature. See, e.g.. Oligonucleotide Synthesis (M.J. Gait ed. 1984) ; Nucleic Acid Hybridization (B.D. Hames & S.J. Higgins, eds., 1984); Sambrook, Fritsch & Maniatis,

Molecular Cloning: A Laboratory Manual. Second Edition (1989); PCR Technology (H.A. Erlich ed. , Stockton Press); R.K. Scope, Protein Purification Principles and Practice (Springer-Verlag) ; and the series Methods in Enzymology (S. Colowick and N. Kaplan eds., Academic Press, Inc.). All patents, patent applications and publications mentioned herein, whether supra or infra, are hereby incorporated by reference in their entirety.

Definitions:

As used herein, the term "transplantation antigen" is used to refer to antigenic molecules that are expressed on the cell surface of transplanted cells, either at the time of transplantation, or at some point following transplantation. Generally these antigenic molecules are proteins and glycoproteins. The primary transplantation antigens are products of the major histocompatibility complex (MHC) , located on chromosome 6 in humans. The human MHC complex is also called the human leukocyte antigen (HLA) complex. MHC antigens are divided into MHC class I antigens (in humans, this class includes HLA-A, -B, and -C antigens) and MHC class II antigens (in humans, this class includes HLA-DP, -DQ, and -DR antigens) . Transplantation antigens also include cell surface molecules other than MHC class I and II antigens. These antigens include the following: (1) the ABO antigens involved in blood cell recognition; (2) cell adhesion molecules such as ICAM, which is involved in leukocyte cell-cell recognition; and (3) |S ₂ - microglobulin, a polypeptide associated with the 44 kd

heavy chain polypeptide that comprises the HLA-I antigens but is not encoded by the MHC complex.

As used herein, the term "transplantation antigen nucleotide sequence" refers to nucleotide sequences associated with genes encoding transplantation antigens. Nucleotide sequences associated with genes include the region of the gene encoding the structural product, including intron and exon regions, and regions upstream of the structural gene associated with transcription, transcription initiation (including transcription factor binding sites) , translation initiation, operator and promoter regions, ribosome binding regions, as well as regions downstream of the structural gene, including termination sites. Nucleotide sequences associated with genes also include sequences found on any form of messenger RNA (mRNA) derived from the gene, including the pre-mRNA, spliced mRNA, and polyadenylated mRNA.

As used herein, the term "transplantation antigen-depleted cell" refers to cells that are in some way depleted in the expression of at least one transplantation antigen. This depletion may be manifested by a reduced amount of antigen present on the cell surface at all times. Preferably, at least about 50%, more preferably about 80%, and even more preferably about 90% of the antigen is eliminated at the cell surface. Most preferably, this depletion results in essentially total absence of the antigen at the cell surface. Certain transplantation antigens are not always constitutively expressed on the cell surface. These antigens have their expression increased at some point shortly after transplant. In these cases, the depletion is manifested by a reduced amount of antigen or complete

lack of antigen at the cell surface at the post- transplant point of normal increased expression.

A transplantation antigen-depleted cell will have at least one of two properties: (1) the cell will survive in the transplant recipient for time periods significantly longer than normal cells; or (2) the cell will survive in the transplant recipient for time periods commensurate to normal or untreated cells, but will require lower doses of immunosuppressive agents to the transplant recipient.

As used herein, "oligomers" or "oligo¬ nucleotides" include RNA, DNA, or RNA/DNA hybrid sequences of more than one nucleotide in either single chain or duplex form. "Nucleic acids", as used herein, refers to RNA, DNA, or RNA/DNA hybrid sequences of any length in single-stranded or duplex form.

As used herein, the term "binding" refers to an interaction or complexation between an oligonucleotide and a target transplantation antigen nucleotide sequence, mediated through hydrogen bonding or other molecular forces. As used herein, the term "binding" more specifically refers to two types of internucleotide binding mediated through base-base hydrogen bonding. The first type of binding is "Watson-Crick-type" binding interactions in which adenine-thymine (or adenine-uracil) and guanine-cytosine base-pairs are formed through hydrogen bonding between the bases. An example of this type of binding is the binding traditionally associated with the DNA double helix. The second type of binding is "triplex binding" which follows a set of still-developing binding rules. In general, triplex binding refers to any type of base- base hydrogen bonding of a third oligonucleotide strand with a duplex DNA (or DNA-RNA hybrid) that is already paired in a Watson-Crick manner. Triplex binding is more

fully described in PCT Application No. WO 90/15884 (published 27 December 1990) . In one set of triplex binding rules, the third strand is designed to match each A on a purine-rich strand of the duplex with a T, and each G on the purine-rich strand of the duplex with a C, and the third strand runs parallel to the purine-rich (or matched) strand. In another set of triplex rules, the third strand is designed to match each A on a purine-rich strand of the duplex with a T, and each G on the purine-rich strand of the duplex with a G, and the third strand runs antiparallel to the purine-rich (or matched) strand. Other types of triplex binding rules are described in PCT Application No. WO 90/15884. One or more types of triplex binding may occur for a given oligonucleotide.

As used herein, Hoogsteen-like bonds refers to hydrogen bonding between bases.

As used herein the terms "modified oligonucleotides", "chemically modified oligonucleotides", and "oligonucleotides of the invention" refer to oligonucleotides that are : (1) capable of binding to a transplantation antigen nucleotide sequence; and (2) comprise a 3'- CH ₂ CH(OH)CH ₂ NH ₂ group. As used herein, the term "3' -RNH ₂ -containing fragments" refers to fragments of the oligonucleotides I, III, IV, and V or chemically modified oligonucleotides thereof, which comprise a 3' -CH ₂ CH(OH)CH ₂ NH ₂ group. These fragments need not comprise the original 3' nucleotides of the oligonucleotides I, III, IV, and V, but rather "3'-RNH ₂ -containing fragments" may contain any segment of the oligonucleotides I, III, IV, and V, so long as the 3'-terminal nucleotide of the fragment is linked to a 3' -CH ₂ CH(0H)CH ₂ NH ₂ group.

The MHC gene locus encodes a family of heterodimeric proteins that are involved in antigen presentation to T lymphocytes. The protein ICAM-1 (CD54) belongs to the immunoglobulin superfamily, binds LFA-1 (CDlla/CD18) and mediates lymphocyte-target cell adhesion. These proteins and processes are central to the initiation of T cell-mediated responses and initiation of graft rejection. The MHC genes and ICAM-1 are up-regulated by interferon-γ ("IFN-γ") , a powerful immunodulatory cytokine that is produced by lymphocytes and natural killer cells. IFN-γ is known to induce or enhance expression of MHC Class I (MHC-I) , MHC Class II and ICAM-1 proteins in a variety of cell types by transcriptional enhancement. It also acts at the cellular level in the immune system, primes macrophages and stimulates natural killer cells. These and other immunomodulatory effects of IFN-γ increase the immunogenicity of tissues and enhance both antigen specific and nonspecific cell-mediated host response. However, these responses may be undesirable in those clinical situations where they promote graft rejection and inflammation. Agents capable of blocking one or more physiological effects of IFN-γ can therefore have useful therapeutic applications, particularly in transplantation.

In the Examples Section we have identified and evaluated' oligonucleotides that inhibit the effects of IFN-γ in a variety of cell lines at micromolar concentrations, with emphasis on the inhibitory effects on the IFN-γ mediated enhancement of two markers—MHC-I and ICAM-1.

In particular, we have shown that the inhibition by (I) :

5' GGG GTT GGT TGT GTT GGG TGT TGT GT— NH ₂

is dose-dependent, with an ED _5Q 24 hours after dosing of approximately 4 μM for 800 U/ml interferon-γ. The reverse complement

5' AC ACA ACA CCC AAC ACA ACC AAC CCC—RNH ₂ was inactive. Three other oligonucleotides with little sequence similarity to I also have activity. Constitutive ICAM-1 expression in K562 cells was not inhibited at the 25 μM level. The inhibition is selective in that I inhibits induction of cell surface MHC-I by interferon-γ, but does not inhibit induction by 6400 U/ml of either interferon-α or interferon-jS. ICAM-1 induction by 800 U/ml TNF-α was also not inhibited.

The present invention encompasses methods ^' for modulating the expression of a polypeptide, wherein the expression is regulated by IFN-γ. These methods comprise exposing a cell to an oligonucleotide of the invention in an amount sufficient to disrupt the IFNγ- regulated expression of said polypeptide in said cell. Thus, the oligonucleotides of the invention can be used to inhibit the IFN-γ-induced expression of any protein, wherein the protein's expression is subject to induction by IFN-γ. For example, the oligonucleotides of the invention were shown to disrupt the IFN-γ-induced expression of the MHC-I and ICAM-I transplantation antigens in Examples 1, 2, and 3, below. Further, these methods include using the oligonucleotides of the invention to inhibit the IFN-γ-stimulated down-regulation of any protein's expression, wherein the protein's expression is subject to down-regulated by IFN-γ. For example, oligonucleotide I was shown to disrupt the IFN- γ-stimulated down-regulation of the transferrin receptor in Example 4.

The mechanisms by which the oligonucleotides of the invention interfere with or inhibit the production of one or more transplantation antigens is not clearly

established, and is not a necessary part of the invention. A number of mechanisms for the mode of action of the modified oligonucleotides are proposed herein, without limiting the oligonucleotides of the invention to any mechanistic theory. Hereafter we examine several proposed mechanisms and indicate how our data support or do not support the mechanism as a candidate.

Oligonucleotides are known to exert therapeutic effects on cells via several known mechanisms and are currently being investigated as therapeutic agents in a variety of strategies. These include the mRNA directed antisense, antigene directed triple helix, DNA binding protein directed DNA decoys, and aptamer strategies. Oligonucleotide treatment sometimes produces effects through unknown mechanisms and questions relating both to side effects and to "appropriate" controls have been discussed in the antisense literature. Oligonucleotides and other nucleic acids based approaches have also been directed against the MHC by several workers in order to study the effect of decreased MHC expression on immune function. Lichtenstein et al. (1992) found decreased cell surface MHC-I in a variety of human cell lines treated with an antisense oligo targeted to the AUG region of β ₂ microglobulin and used the results to support the hypothesis that Her2/neu oncogene expressing cells resist NK and LAK cytotoxicity through mechanisms other than enhanced MHC-I expression. Kanbe et al . (1992) effected H-2K antigen down regulation using antisense oligonucleotides and examined the effect of H-2K loss on metastatic potential. Both Lichtenstein et al . and Kanbe et al . implied that an antisense mechanism was operative and responsible for the down regulation but the mechanism of action was not investigated. Neither studied the effect of these oligos on the IFN-γ enhanced component of MHC-I. Chiang et al .

(1991) studied the effects of two phosphorothioate oligos, one complementary to the AUG initiation codon and the other targeted to the 3' untranslated region of ICAM-1 in human umbilical vein endothelial cells and the A549 lung carcinoma cell line induced with either IL-1 β , TNF-α or IFN-γ. The oligo directed to 3' untranslated region was found to inhibit ICAM-1 mRNA levels while the AUG codon targeted oligo did not and Chiang et al . suggested that the two oligos acted via distinct mechanisms. Chiang et al . also found that IFN-γ induced ICAM-1 was more sensitive to phosphorothioate oligo-mediated down regulation than either IL-1 β or TNF-α induced ICAM-1. This is may be due in part because the phosphorthioates used exhibit weak IFN-γ inhibitory activity in addition to the other mechanisms that result in inhibition of IL-1 b or TNF-α induced ICAM-1.

Oligo I is apparently a representative of a family of oligonucleotides having specific and selective activities. Since a variety of cell surface markers are impacted by I we suggest that an early step or steps in the regulation of genes by IFN-γ is probably modulated. This is supported by the surprising similarities in the dose response curves for all three markers in K562 cells. We suggest further that the modulation of separate and late MHC-I and ICAM-1 unique steps by I would likely have resulted in dose responses curves differing significantly from each other, unless the changes were orchestrated coordinately on all the pathways involved. An aptamer-like mechanism, in which oligos bind IFN-γ, IFN-γ receptor or other essential accessory proteins in the

IFN-γ induction process is plausible and not inconsistent with our data. IFN-γ has a pi of 8.6-8.7 and is positively charged at physiological pH. Further, the carboxyl terminus of IFN-γ is essential for activity and has several positively charged amino acids that could

possibly interact with negatively charged oligos via charge-charge interactions. Such a pathway can plausibly inhibit the effects of IFN-γ provided the structural requirements needed for efficient charge-charge interaction account for the differences in activity between II and I. The differences in activity may however be due to differences in half-lives between apparently active and inactive oligos caused by differences in nuclease susceptibility. Another possibility is that active oligonucleotides of the invention, through some undetermined mechanism, accelerate the rate at which IFN-γ is committed to a futile degradative pathway or compartment. Yet another possibility is that the internalization of IFN-γ is essential for activity and that active oligos block this internalization process. All the speculative mechanisms outlined so far act outside the cell or on the cell membrane. Oligonucleotides are taken up, albeit inefficiently and heterogeneously by cells and could enter cells and either bind to DNA binding proteins or alter binding of these proteins to DNA thereby inhibiting induction by IFN-γ. However: i) the same interferon response element (IRE) is necessary for responsiveness to IFN-α, IFN-0 and IFN-γ

(Sugita et al . (1987); Korber et al. (1987)); and ii) the IFN-γ and IFN-α transcriptional activation pathways for guanylate binding proteins are known to overlap (Reid et al . (1989); Shuai et al . (1992); Schindler et al . (1992) ) , and any proposed mechanism must account plausibly for the low inhibition of IFN-α and IFN-/3 by oligo I.

The low toxicity of oligonucleotides suggests useful applications for these molecules in: i) the pharmacological inhibition of IFN-γ in experimental

situations; and ii) as attractive lead candidates for the inhibition of IFN-γ in therapeutic situations such as transplantation, inflammation and autoimmunity.

According to this invention, the modified oligonucleotides described above are used in a method of treatment to make a transplantation antigen-depleted cell from a normal target cell. The cells are created by incubation of the cell with one or more of the above- described oligonucleotides under standard conditions for uptake of nucleic acids, including electroporation or lipofection.

Alternatively, the chemically modified oligonucleotides can be derivatized or co-administered for targeted delivery to the nucleus. The cell nucleus is the likely preferred site for action of the triplex- forming oligonucleotides of this invention, due to the location therein of the cellular transcription and replication machinery. Also, improved stability for the oligonucleotides of the invention is expected in the nucleus due to: (1) lower levels of DNases and RNases; (2) higher oligonucleotide concentrations due to lower • total volume; (3) higher concentrations of key enzymes such as RNase H implicated in the mechanism of action of these oligonucleotides. The cytoplasm, however, is the likely preferred site for action of the traditional antisense oligonucleotides of this invention.

A primary path for nuclear transport is the nuclear pore. Targeted delivery can thus be accomplished by derivatizing the oligonucleotides by attaching a nuclear localization signal. One well-characterized nuclear localization signal is the heptapeptide PKKKRKV (pro-lys-lys-lys-arg-lys-val) .

Any transplantable cell type is a potential target cell for this invention. Preferably, the target cell is selected from corneal endothelial cells, thyroid

cells, parathyroid cells, brain cells, adrenal gland cells, bone marrow cells, pancreatic islet cells, hepatic cells, lymphoid cells, fibroblasts, epithelial cells, chondrocytes, endocrine cells, renal cells, cardiac muscle cells, and hair follicle cells. Most preferably, the target cell is selected from corneal endothelial cells, thyroid cells, parathyroid cells, brain cells, adrenal gland cells, bone marrow cells, pancreatic islet cells and hepatic cells. Furthermore, this invention is applicable to the field of solid organ transplants. Organs are normally perfused ex vivo prior to transplantation. By adding an amount of the above-described oligonucleotides to the perfusion medium, transplantation antigen-depleted cells can be created from perfusion-accessible cells in the organ to create a transplantation antigen-depleted organ useful in solid organ transplants.

Finally, local administration of the anti-gene oligonucleotides directly into the transplanted organ during the first day after the transplant is within the scope of this invention. Also, sustained releases of the of these drugs are also contemplated.

The oligonucleotides of the invention are useful in creating the transplantation antigen-depleted cells of the present invention. These cells are then directly transplanted to an individual. This technique can be used for any individual with an immune system, including humans.

The chemically modified oligonucleotides of this invention are also useful in treating autoimmune diseases characterized by dysfunctional or aberrant expression of a transplantation antigen. In such a case, the modified oligonucleotides described herein may be administered in an amount sufficient to inhibit expression of the transplantation antigen.

It may be commented that the mechanism by which the modified oligonucleotides of the invention interfere with or inhibit the production of one or more transplantation antigens is not always established, and is not a part of the invention. The oligonucleotides of the invention are characterized by their capability to bind to a specific target nucleotide sequence regardless of the mechanisms of binding or the mechanism of the effect thereof. Described below are examples of the present invention which are provided for illustrative purposes, and not to limit the scope of the present invention. In light of the disclosure, numerous embodiments within the scope of the claims will be apparent to those of ordinary skill in the art.

Examples

Cell Lines and-Culture

All cell lines were obtained from the Cell Culture Facility, University of California and were maintained at 37 °C with 5 % carbon dioxide in RPMI 1640 containing 10 % heat inactivated (56 °C for 30 minutes) fetal bovine serum (FBS) (GIBCO-BRL, Grand Island, NY) , 100 U/ml penicillin and 10 μg/ml streptomycin. The human myelogenous leukemia derived K562 cell was selected as the primary model system because it has a very low background level of constitutive MHC-I that can be induced with IFN-γ. The Burkitts lymphoma derived Raji cell line, the uterine carcinoma derived HeLa S3 cell line, and the T cell lymphoma derived HUT78 were studied for comparison.

Interferons and Tumor Necrosis Factor α

Recombinant human IFN-γ (Collaborative Research Incorporated, Bedford, MA) was diluted in

Dulbecco's PBS (D-PBS) to give 50 U/mL, stored at -70 °C in aliquots, and thawed prior to use. Recombinant human IFN-α (Chemicon, Temecula, CA) and IFN-3 (Accurate Chemical Corporation, Westbury, NY) were diluted in D-PBS to 100 U/μL and 1000 U/mL respectively and stored at -20 °C. Aliquots of recombinant TNF-α (Genzyme Corporation, Cambridge, MA) were stored at -70 °C.

Oligonucleotide Synthesis Oligonucleotides were synthesized using standard phosphoramidite protocols and purified by high performance liquid chromatography (Keystone Laboratories, Menlo Park, CA) . Oligonucleotide were stored at -20 °C as a 1250 μM stock solution in twice autoclaved diethyl pyrocarbonate treated deionized water. The following

oligonucleotides were used in our identification of the oligonucleotide family of the invention:

I . 5' GGG GTT GGT TGT GTT GGG TGT TGT GT

—RNH.

II 5' AC ACA ACA CCC AAC ACA ACC AAC CCC

*-RNH ₂ ;

III 5' AGG GTT CGG GGC GCC ATG ACG GC —RNH ₂

IV 5' CAG CCT TGA GGA TTC CCC AAC TCC G

--RNH ₂ ; V 5' GCC ACG GAG CGA GAC ATC TCC G -RNH ₂ ;

VI 5' CAT CTT CTG CCA TTC TGA AGC CGG -RNH ₂ ; where R= —CH ₂ CH(OH)CH ₂ — (Glen Research, Sterling, VA) .

Antibodies and Flow Cytometry Fluorescein isothiocyanate conjugated mouse monoclonal antibodies to human MHC-I heavy chain, β ₂ microglobulin and a mouse IgG ₂ b control were purchased from Olympus, Lake Success, NY. The MHC-I heavy chain antibody recognizes an SDS stable determinant on the MHC-I heavy chain associated with β ₂ microglobulin.

Additionally, fluorescein isothiocyanate conjugated mouse monoclonals to human ICAM-1 and an IgG-^ control were purchased from AMAC (Westbrook, ME) . Fluorescein isothiocyanate .conjugated mouse monoclonals to human MHC-II DR and transferrin receptor were purchased from Beeton Dickinson (San Jose, CA) .

Example 1: Oligonucleotide I Selectively Downregulates IFN-γ Induced Expression of MHC-I and ICAM-I: The treatment of cells with the oligonucleotides of the invention was performed as follows: samples in triplicate were set up in Falcon 2051 tubes and made up of 0.5 x 10 ⁶ K562 cells and 0.25 ml of either: a) RPMI 1640, b) RPMI 1640.and 800 Units/ml IFN-γ, or c) RPMI 1640, 800 Units/ml IFN-γ and

oligonucleotide. For the selectivity experiments IFN-α and IFN-/3 were used at 6400 U/ml in place of IFN-γ. After 1 hour at 37 °C, RPMI 1640 containing heat inactivated (65 °C for 30 minutes) FBS was added to give a final FBS concentration of 10 % in a' volume of 1 ml. Samples were drawn from each tube at predetermined times and stained for flow cytometry using fluorescein isothiocyanate conjugated mouse monoclonal antibodies to human MHC-I heavy chain, β ₂ microglobulin, ICAM-1 and a mouse IgG ₂ b control.

Approximately 7.5 x 10 ⁴ cells were used with each antibody. The cells were washed with 1 mL and resuspended in 50 mL of D-PBS containing 2 % FBS and 0.1 % sodium azide. After the addition of 1 mg of antibody, the mixture was vortexed, incubated for 30 minutes on ice in the dark and then washed twice with D-PBS containing 0.1 % sodium azide to remove unbound antibody. The cell pellet was resuspended in 100 mL of D-PBS containing 0.1 % sodium azide and analyzed on a FACScan (Becton Dickinson, San Jose, CA) flow cytometer calibrated daily with QuickCal beads (Flow Cytometry Standards Corporation, Research Triangle, NC) . Propidium iodide (50 mL of 50 mg/ml) was added prior to cytometry to exclude dead cells from analysis. The calibration curve used to determine the number of specific antibody binding sites on each sample was obtained by concomitantly staining 50 mL of Quantitative Simply Cellular (Flow Cytometry Standards Corporation, Research Triangle, NC) beads with each antibody. The data were analyzed using Lysys II software (Becton Dickinson, San Jose, CA) on a Hewlett Packard computer.

A dose dependent down regulation of the IFN-γ induced expression of MHC-I and ICAM-I transplantation antigens was observed: we obtained a dose response curves for the effect of I on the IFN-γ mediated

enhance ent of MHC-I and ICAM-1 in K562 cells. Fig. 1 summarizes data obtained with an antibody that recognizes a SDS stable determinant on the MHC-I heavy chain, and shows that I inhibits IFN-γ mediated induction of MHC-I heavy chain proteins on the cell surface at 24, 48 and 72 hours. The inhibition is dose-dependent, and interpolating from Fig. la, we find an ED ₅₀ of approximately 4 μM for 800 U/ml IFN-γ at 24 hours. The human MHC-I protein is heterodimeric and consists of a 45 kD polymorphic heavy chain closely associated with β ₂ microglobulin, a nonpolymorphic, 12 kD, light chain.

Fig. 2 shows that I concomitantly inhibits the expression of β ₂ microglobulin.

Fig. 3 shows that I also inhibits induction of ICAM-1. The data are normalized and expressed as a percentage of the protein expression observed with IFN treatment so as to allow a direct comparison of the results obtained with the various markers. From such a comparison, it is evident that the dose response relationships at any given time for all three markers are surprisingly similar in K562 cells. The results in Figs. 1-3 show that I inhibits a step in the cascade of IFN-γ mediated enhancement of cell surface MHC-I and ICAM-1 and the similarity in the dose response relationships for all three markers suggests that an early step or steps in the induction of proteins by IFN-γ is probably altered by I. Constitutive ICAM-1 is unaffected because K562 cells treated with 25 μM of either I or II in the absence of IFN-γ did not decrease ICAM-1 expression (data not shown) . Oligo II does not exhibit dose dependent activity against IFN-γ enhanced MHC-I or ICAM-1 at 24 and 72 hours. However, an unusual dose dependent up regulation of IFN-γ enhanced MHC-I and ICAM-1 is seen with II at 48 hours.

Example 2: A Family of Oligonucleotides Downregulates IFN-γ Induced Expression of MHC-I and ICAM-I:

In a separate set of experiments, oligos I, III, IV, V and VI were compared using the same assay as in Example 1. K562 cells were treated with 25 μM of each oligo and analyzed for MHC-I, β ₂ microglobulin and ICAM-1 expression at 48 hours. These results are shown in Fig. 4 and are broadly similar for all three markers. The following activity relationship summarizes Fig. 4: Oligo I - Oligo III > Oligo IV « Oligo V > Oligo VI - No treatment.

Taken together, the data suggest that a family or class of oligos exists that can inhibit MHC-I induction by IFN-γ in K562 cells and that I, III, IV and V are members of this family. Alignment of I-VI using a dynamic programming algorithm did not reveal significant sequence similarity. The only structural characteristics separating the active oligos from the inactive oligos is the presence of guanosine bases in active oligos (e.g. I and III) and the presence of cytosine bases (absence of guanosine) in the inactive oligos (e.g. II and VI) . We have been unable to identify structure-activity relationships based on an examination of the oligos for the presence of hairpins and inverted repeats which may result in the formation of secondary structures or facilitate the binding of DNA binding proteins. Searches of the primate database in GenBank with oligos I-VI as probes did not turn up targets different from those against which the oligos were originally directed.

Example 3: Oligonucleotides Demonstrate Inhibition Specific to IFN-Ύ Induction of MHC-I and ICAM-I:

MHC-I can also be induced by the cytokines IFN-α, IFN-0 (Lindahl e al. (1976)) and TNF-α (Collins et al . (1986)). Fig. 5b shows that, at 48 hours, oligo I

selectively inhibits IFN-γ mediated MHC-I but does not inhibit MHC-I induction by 6400 U/ml of either IFN-α or IFN-jS. The flow cytometry results with the anti-/? ₂ microglobulin monoclonal antibody confirmed those obtained with the anti-MHC-I heavy chain antibody (data not shown) . IFN-α and IFN-3 do not induce cell surface ICAM-1 in K562 cells. At 24 hours (Fig. 5a), cells treated with IFN-/3 and 25 μM of oligo I appear to express slightly lower levels of MHC-I heavy chain, but recover almost completely by 48 hours. The IFN doses were selected to give comparable enhancements of cell surface MHC-I. The results show that 25 μM oligo I inhibits IFN-γ induction of MHC-I, but has no significant effect on induction by IFN-α and IFN-/3 under these conditions. We also examined the effect of I on the TNF-α mediated enhancement of ICAM-1 (Rothlein et al. (1988). We compared the effect of 25 μM oligo I on ICAM-1 induction by 800 U/ml TNF-α and IFN-γ. The results are shown in Fig. 6 and demonstrate that I does not inhibit induction of ICAM-1 by TNF-α in this model.

Taken together, the foregoing data suggest that the mechanism(s) leading to blockage of cell surface MHC-I and ICAM-1 in K562 cells by I may be specific for IFN-γ.

Example 4: Oligonucleotides of the Invention Have Activity in other Cell Lines:

We extended the studies of the activity of I to Burkitts lymphoma-derived Raji cells, to uterine carcinoma derived HeLa S3 cells, and to T cell lymphoma derived HUT78 cells. These cell lines are reasonable in vitro models for B cells, epithelial and mature T cells respectively. Cells were treated with 25 μM oligo I according to the protocol used in the K562 cell experiments and stained with monoclonals for MHC-I heavy

chain, β ₂ microglobulin, MHC Class II DR and transferrin receptor protein expression at 48 hours.

Raji cells constitutively express high levels of MHC-I, MHC Class II DR and are not induced significantly by IFN-γ. As shown in Figure 7, I does not appear to down regulate any of the markers examined. The poor induction of these markers by IFN-γ, coupled with relatively high standard deviation for the measures for MHC-I and β ₂ microglobulin expression did not allow an absolute determination of oligo I activity in this cell line. It is clear however, that I does not significantly impact the constitutive levels of any of the markers examined. The relative stability of I in the presence of these metabolically active cells was not determined. _f The second panel in Fig. 7 shows that MHC-I levels are enhanced significantly in HeLa S3 by IFN-γ, and that I is only slightly effective in inhibiting this induction. This low level of inhibition is a departure from the experience with K562 cells but is not due to low relative stability of I in HeLa S3 cells because ICAM-1

(unlike MHC-I) induction is inhibited by I. The observed differences in the activity of I between HeLa S3 cells and K562 cells could therefore be attributed to differences in*IFN-γ or oligo I sensitivities of the MHC-I or to other factors in the two cell lines. The

HeLa S3 MHC Class II DR data in Fig. 7 suggest that I is effective in inhibiting DR induction in HeLa S3, even though only a modest induction was observed at 48 hours. Unlike the other markers examined, the transferrin receptor is down regulated by IFN-γ in the HeLa S3 cell line and 25 μM oligo I appears to effectively block this down regulation. Thus, I is capable of inhibiting both the up regulation and the down regulation of IFN-γ regulated genes.

The IFN-γ mediated MHC-I and β ₂ microglobulin induction in HUT78 cells was small under the conditions of the experiment and the effects of I on these markers is not clear. The induction of ICAM-1 by IFN-γ is however, inhibited. The results with the transferrin receptor in HUT78 are unique in that I appears to impact the basal levels of the transferrin receptor. This activity may be advantageous for future therapeutic applications of this family of oligos because concomitant local inhibition of host T cell activation is usually desirable in these applications.

In summary, the ability of I to inhibit IFN-γ induction of specific proteins is not restricted to one cell line, suggesting that I and similarly acting oligos intervene directly in the cascade of events initiated by IFN-γ. Finally, the low activity against MHC-I in HeLa S3 cells suggests that the enhancement of the same cell surface marker in different cell types may have different sensitivities/susceptibilities to inhibition by this drug class.