Blockade of SIRPα-CD47 axis by anti-SIRPα antibody enhances anti-tumor activity of DXd antibody-drug conjugates.

Signal regulatory protein alpha (SIRPα) is an immune inhibitory receptor on myeloid cells including macrophages and dendritic cells, which binds to CD47, a ubiquitous self-associated molecule. SIRPα-CD47 interaction is exploited by cancer cells to suppress anti-tumor activity of myeloid cells, therefore emerging as a novel immune checkpoint for cancer immunotherapy. In blood cancer, several SIRPα-CD47 blockers have shown encouraging monotherapy activity. However, the anti-tumor activity of SIRPα-CD47 blockers in solid tumors seems limited, suggesting the need for combination therapies to fully exploit the myeloid immune checkpoint in solid tumors. Here we tested whether combination of SIRPα-CD47 blocker with antibody-drug conjugate bearing a topoisomerase I inhibitor DXd (DXd-ADC) would enhance anti-tumor activity in solid tumors. To this end, DS-1103a, a newly developed anti-human SIRPα antibody (Ab), was assessed for the potential combination benefit with datopotamab deruxtecan (Dato-DXd) and trastuzumab deruxtecan (T-DXd), DXd-ADCs targeting human trophoblast cell-surface antigen 2 and human epidermal growth factor receptor 2, respectively. DS-1103a inhibited SIRPα-CD47 interaction and enhanced antibody-dependent cellular phagocytosis of Dato-DXd and T-DXd against human cancer cells. In a whole cancer cell vaccination model, vaccination with DXd-treated cancer cells led to activation of tumor-specific T cells when combined with an anti-mouse SIRPα (anti-mSIRPα) Ab, implying the benefit of combining DXd-ADCs with anti-SIRPα Ab on anti-tumor immunity. Furthermore, in syngeneic mouse models, both Dato-DXd and T-DXd combination with anti-mSIRPα Ab showed stronger anti-tumor activity over the monotherapies. Taken together, this study provides a preclinical rationale of novel therapies for solid tumors combining SIRPα-CD47 blockers with DXd-ADCs.

Preclinical Development of U3-1784, a Novel FGFR4 Antibody Against Cancer, and Avoidance of Its On-target Toxicity.

The FGFR4/FGF19 signaling axis is overactivated in 20% of liver tumors and currently represents a promising targetable signaling mechanism in this cancer type. However, blocking FGFR4 or FGF19 has proven challenging due to its physiological role in suppressing bile acid synthesis which leads to increased toxic bile acid plasma levels upon FGFR4 inhibition. An FGFR4-targeting antibody, U3-1784, was generated in order to investigate its suitability as a cancer treatment without major side effects.U3-1784 is a high-affinity fully human antibody that was obtained by phage display technology and specifically binds to FGFR4. The antibody inhibits cell signaling by competing with various FGFs for their FGFR4 binding site thereby inhibiting receptor activation and downstream signaling via FRS2 and Erk. The inhibitory effect on tumor growth was investigated in 10 different liver cancer models in vivo The antibody specifically slowed tumor growth of models overexpressing FGF19 by up to 90% whereas tumor growth of models not expressing FGF19 was unaffected. In cynomolgus monkeys, intravenous injection of U3-1784 caused elevated serum bile acid and liver enzyme levels indicating potential liver damage. These effects could be completely prevented by the concomitant oral treatment with the bile acid sequestrant colestyramine, which binds and eliminates bile acids in the gut. These results offer a new biomarker-driven treatment modality in liver cancer without toxicity and they suggest a general strategy for avoiding adverse events with FGFR4 inhibitors.

POSH promotes cell survival in Drosophila and in human RASF cells.

In Drosophila, Eiger, a tumor necrosis factor α (TNFα) superfamily ligand, induces cell death by activating the c-Jun N-terminal kinase (JNK) pathway. Here, we report that overexpression of Plenty of SH3s (POSH) suppresses Eiger-induced cell death and produces highly deformed tissues. These results imply that high levels of POSH protect tissues from cell death. In humans, rheumatoid arthritis synovial fibroblasts (RASF) are generally resistant to apoptosis. We show that POSH is expressed at relatively high levels in RASF, and its reduction by RNAi sensitizes these cells to Fas-mediated apoptosis. Thus, we demonstrate that POSH promotes cell survival in Drosophila and in human RASF.

Fine-tuning of ENA gapmers as antisense oligonucleotides for sequence-specific inhibition.

For gene validation and the development of oligonucleotide agents, 2'-O,4'-C-ethylene-bridged nucleic acid (ENA) antisense gapmers are widely available. An in vitro Escherichia coli RNase H reaction analysis using ENA gapmers and an RNA oligonucleotide with mouse peptidylarginine deiminase 4 (PADI4) gene sequences revealed that the RNA oligonucleotide was specifically cleaved in the only reported case of the use of an ENA gapmer with an antisense sequence. On the other hand, duplexes of the full-length transcripts of PADI4 mRNA and ENA gapmers with a wide DNA window were cleaved not only at the target site, but also at nontarget sites by RNase H derived from partial base-pairing between the transcript and the ENA gapmer. When the DNA window region of the ENA gapmer was shortened to 5 or 6 nucleotides, the nontarget cleavage was effectively diminished. Moreover, the specific inhibition of PADI4 mRNA expression was observed in the cotransfection of PADI4 cDNA and ENA gapmers containing a short DNA region into NIH3T3 cells. These results demonstrated that ENA gapmers with a short DNA region improved the sequence-specificity of mRNA downregulation. These optimized ENA gapmers could reduce the "off-target" effect and be applicable to gene validation and oligonucleotide therapeutics.

Design of ENA gapmers as fine-tuning antisense oligonucleotides with sequence-specific inhibitory activity on mouse PADI4 mRNA expression.

In order to evaluate the function of the mouse peptidylarginine deiminase 4 (PADI4) gene, gapmer-designed ENA antisense oligonucleotides were utilized. Antisense ENA gapmers were found to inhibit the mRNA expression of the PADI4 gene, but ENA gapmers with sense sequences as controls only partially inhibited this mRNA expression. An in vitro E. coli RNase H reaction analysis using transcripts of the PADI4 gene and their ENA gapmers revealed that the mixtures of the transcripts and ENA gapmers with sense sequences were cleaved at non-target sites by RNase H derived from partial base-pairing between the transcript and the ENA gapmer. In an in vitro E. coli RNase H reaction analysis, when the DNA region of the ENA gapmer was shortened to 5 or 6 nucleotides, the non-target cleavage disappeared and the specific inhibition of PADI4 mRNA expression was observed. These results demonstrated that ENA gapmers with a short DNA region improved the sequence-specificity of mRNA down-regulation. As well, they suggest that we should be alert to the utility of antisense oligonucleotides with a wide DNA region in terms of sequence specificity, and additionally, that optimized ENA gapmers could be useful for application to gene validation.

Ethnic differences in allele frequency of autoimmune-disease-associated SNPs.

Several multiple, large-scale, genetic studies on autoimmune-disease-associated SNPs have been reported recently: peptidylarginine deiminase type 4 (PADI4) in rheumatoid arthritis (RA); solute carrier family 22 members 4 and 5 (SLC22A4 and 5) in RA and Crohn's disease (CD); programmed cell death 1 (PDCD1) in systemic lupus erythematosus (SLE), type 1 diabetes mellitus (T1D), and RA; and protein tyrosine phosphatase nonreceptor type 22 (PTPN22) in T1D, RA, and SLE. Because these reports on association were not always evaluated in multiple ethnic groups and because ethnic difference in allele frequency of the variants has been also reported, we investigated allele frequencies of nine SNPs in four autoimmune-disease-associated loci in Caucasian, African-descent, and Japanese populations. Although SNPs in PADI4 had similar allele frequency among three groups [maximal difference 11%; (P >0.05)], the other three loci revealed statistically significant allele frequency differences (maximal difference 39% (P <0.00001), 13% (P <0.00001), and 8% (P <0.00001) in SLC22A4, PDCD1, and PTPN22, respectively). Of note, three SNPs in the three loci that had allele frequency more than 8% in the Caucasian population were either not polymorphic at all or extremely rare in the Japanese population. Our data suggest that ethnic variations of polymorphisms should be evaluated in detail, and differences should be incorporated into investigations of susceptibility variants for common diseases.

Comparison of enzymatic properties between hPADI2 and hPADI4.

In the sera of rheumatoid arthritis (RA) patients, autoantibodies directed to citrullinated proteins are found with high specificity for RA. Peptidylarginine deiminases (PADIs) are enzymes responsible for protein citrullination. Among many isoforms of PADIs, only PADI4 has been identified as an RA-susceptibility gene. To understand the mechanisms of the initiation and progression of RA, we compared the properties of two PADIs, human PADI2 and human PADI4, which are present in the synovial tissues of RA patients. We confirmed their precise distribution in the RA synovium and compared the stability, Ca2+ dependency, optimal pH range, and substrate specificity. Small but significant differences were found in the above-mentioned properties between hPADI2 and hPADI4. Using LC/MS/MS analysis, we identified the sequences in human fibrinogen indicating that hPADI2 and hPADI4 citrullinate in different manners. Our results indicate that hPADI2 and hPADI4 have different roles under physiological and pathological conditions. Further studies are needed for the better understanding of the role of hPADIs in the initiation and progression of RA.

Functional haplotypes of PADI4, encoding citrullinating enzyme peptidylarginine deiminase 4, are associated with rheumatoid arthritis.

Individuals with rheumatoid arthritis frequently have autoantibodies to citrullinated peptides, suggesting the involvement of the peptidylarginine deiminases citrullinating enzymes (encoded by PADI genes) in rheumatoid arthritis. Previous linkage studies have shown that a susceptibility locus for rheumatoid arthritis includes four PADI genes but did not establish which PADI gene confers susceptibility to rheumatoid arthritis. We used a case-control linkage disequilibrium study to show that PADI type 4 is a susceptibility locus for rheumatoid arthritis (P = 0.000008). PADI4 was expressed in hematological and rheumatoid arthritis synovial tissues. We also identified a haplotype of PADI4 associated with susceptibility to rheumatoid arthritis that affected stability of transcripts and was associated with levels of antibody to citrullinated peptide in sera from individuals with rheumatoid arthritis. Our results imply that the PADI4 haplotype associated with susceptibility to rheumatoid arthritis increases production of citrullinated peptides acting as autoantigens, resulting in heightened risk of developing the disease.

Jun dimerization protein 2 (JDP2), a member of the AP-1 family of transcription factor, mediates osteoclast differentiation induced by RANKL.

Osteoclasts are multinucleated cells that resorb bones, and are derived from hematopoietic cells of the monocyte/macrophage lineage. The receptor activator of NF-kappaB ligand (RANKL, also called ODF/TRANCE/OPGL) stimulates both osteoclast differentiation from osteoclast progenitors and activation of mature osteoclasts. To identify genes responsible for osteoclast differentiation, we used a molecular indexing technique. Here, we report a clone of one of these genes whose transcription is induced by soluble RANKL (sRANKL) in both the RAW264.7 cells of the mouse macrophage cell line and the mouse primary bone marrow cells. The predicted protein was found to be a mouse homologue of Jun dimerization protein 2 (JDP2), a member of the AP-1 family of transcription factors, containing a basic region-leucine zipper motif. Transient transfection experiments revealed that overexpression of JDP2 leads to activation of both tartrate-resistant acid phosphatase (TRAP) and cathepsin K gene promoters in RAW264.7 cells. Infection of mouse primary bone marrow cells with retroviruses expressing JDP2-facilitated sRANKL-mediated formation of TRAP-positive multinuclear osteoclasts. Importantly, antisense oligonucleotide to JDP2 strongly suppressed sRANKL-induced osteoclast formation of RAW264.7 cells. Our findings suggest that JDP2 may play an important role in the RANK-mediated signal transduction system, especially in osteoclast differentiation.

Effects on RNA interference in gene expression (RNAi) in cultured mammalian cells of mismatches and the introduction of chemical modifications at the 3'-ends of siRNAs.

The highly specific posttranscriptional silencing of gene expression induced by double-stranded RNA (dsRNA) is known as RNA interference (RNAi) and has been demonstrated in plants, nematodes, Drosophila, and protozoa, as well as in mammalian cells. The suppression of expression of specific genes by chemically synthesized 21-nucleotide (21-nt) RNA duplexes has been achieved in various lines of mammalian cells, and this technique might prove to be a valuable tool in efforts to analyze biologic functions of genes in mammalian cells. In order to investigate the utility of potential modifications that can be introduced into small interfering RNAs (siRNAs) and also to study their functional anatomy, we synthesized different types of siRNA targeted to mRNA of Jun dimerization protein 2 (JDP2). Our detailed analysis demonstrated that siRNAs with only one mismatch, relative to the target, on the antisense strand had reduced RNAi effect, whereas the corresponding mutation on the sense strand did not interfere with the RNAi. Moreover, one 2-hydroxyethylphosphate (hp) substitution at the 3'-end of the antisense strand but not of the sense strand also prevented RNAi, whereas a related modification at the 3'-end of either strand, using 2'-O,4'-C-ethylene thymidine (eT), which is a component of ethylene-bridge nucleic acids (ENA), completely abolished RNAi. These results support the hypothesis that the two strands have different functions in RNAi in cultured mammalian cells and indicate that their chemical modification of siRNAs at the 3'-end of the sense strand exclusively is possible, without loss of RNAi activity, depending on the type of modification. Because modification at the 3'-end of the antisense strand by hp or eT abolished the RNAi effect, it appears possible that the 3'-end is recognized by the RNA-induced silencing complex (RISC).

Humanization of the mouse anti-Fas antibody HFE7A and crystal structure of the humanized HFE7A Fab fragment.

Binding of Fas ligand to Fas induces apoptosis. The Fas-Fas ligand system plays important roles in many biological processes, including the elimination of autoreactive lymphoid cells. We have previously obtained the mouse anti-Fas antibody HFE7A (m-HFE7A), which specifically induces apoptosis in inflammatory cells. In order to apply m-HFE7A for human therapy, we performed antibody humanization of m-HFE7A by grafting the mouse complementarity-determining regions (CDRs) to a human antibody. Five versions of humanized HFE7A (h-HFE7A) demonstrated the same antigen-binding affinity and same competition-binding activity against Fas as the chimeric HFE7A. Furthermore, these h-HFE7As induced the same degree of apoptosis in WR19L12a cells that express human Fas on their surface as chimeric HFE7A does. To further probe the structural basis for antibody humanization, we determined the three-dimensional structure of the h-HFE7A antigen-binding fragment (Fab) by X-ray crystallography and compared it with the crystal structure of the parent m-HFE7A Fab previously determined. The main-chain conformation in each h-HFE7A CDR is almost identical to that in m-HFE7A with root mean square (rms) deviations of 0.14-0.77 A. However, a significant segmental shift was observed in the CDR-L1 loop. Together with the high temperature factors of the CDR-L1 residues, both the loops are flexible, suggesting that the CDR-L1 loop would undergo conformational change upon binding to the antigen. Our results indicate that the humanization of m-HFE7A succeeded in maintaining the main-chain conformation as well as the flexibility of the CDR loop.