Targeting MT1-MMP as an ImmunoPET-Based Strategy for Imaging Gliomas

BACKGROUND

A critical challenge in the management of Glioblastoma Multiforme (GBM) tumors is the accurate diagnosis and assessment of tumor progression in a noninvasive manner. We have identified Membrane-type 1 matrix metalloproteinase (MT1-MMP) as an attractive biomarker for GBM imaging since this protein is actively involved in tumor growth and progression, correlates with tumor grade and is closely associated with poor prognosis in GBM patients. Here, we report the development of an immunoPET tracer for effective detection of MT1-MMP in GBM models.

METHODS

An anti-human MT1-MMP monoclonal antibody (mAb), LEM2/15, was conjugated to p-isothiocyanatobenzyl-desferrioxamine (DFO-NCS) for 89Zr labeling. Biodistribution and PET imaging studies were performed in xenograft mice bearing human GBM cells (U251) expressing MT1-MMP and non-expressing breast carcinoma cells (MCF-7) as negative control. Two orthotopic brain GBM models, patient-derived neurospheres (TS543) and U251 cells, with different degrees of blood-brain barrier (BBB) disruption were also used for PET imaging experiments.

RESULTS

89Zr labeling of DFO-LEM2/15 was achieved with high yield (>90%) and specific activity (78.5 MBq/mg). Biodistribution experiments indicated that 89Zr-DFO-LEM2/15 showed excellent potential as a radiotracer for detection of MT1-MMP positive GBM tumors. PET imaging also indicated a specific and prominent 89Zr-DFO-LEM2/15 uptake in MT1-MMP+ U251 GBM tumors compared to MT1-MMP- MCF-7 breast tumors. Results obtained in orthotopic brain GBM models revealed a high dependence of a disrupted BBB for tracer penetrance into tumors. 89Zr-DFO-LEM2/15 showed much higher accumulation in TS543 tumors with a highly disrupted BBB than in U251 orthotopic model in which the BBB permeability was only partially increased. Histological analysis confirmed the specificity of the immunoconjugate in all GBM models.

CONCLUSION

A new anti MT1-MMP-mAb tracer, 89Zr-DFO-LEM2/15, was synthesized efficiently. In vivo validation showed high-specific-contrast imaging of MT1-MMP positive GBM tumors and provided strong evidence for utility of MT1-MMP-targeted immunoPET as an alternate to nonspecific imaging of GBM.

Characterization of the immune response to canine parvovirus induced by vaccination with chimaeric plant viruses

NIH mice were vaccinated subcutaneously or intranasally with chimaeric cow pea mosaic virus (CPMV) constructs expressing a 17-mer peptide sequence from canine parvovirus (CPV) as monomers or dimers on the small or large protein surface subunits. Responses to the chimaeric virus particles (CVPs) were compared with those of mice immunized with the native virus or with parvovirus peptide conjugated to keyhole limpet haemocyanin (KLH). The characteristics of the immune response to vaccination were examined by measuring serum and mucosal antibody responses in ELISA, in vitro antigen-induced spleen cell proliferation and cytokine responses. Mice made strong antibody responses to the native plant virus and peptide-specific responses to two of the four CVP constructs tested which were approximately 10-fold lower than responses to native plant virus. The immune response generated by the CVP constructs showed a marked TH1 bias, as determined by a predominantly IgG(2a) isotype peptide-specific antibody response and the release of IFN-gamma but not IL-4 or IL-5 from lymphocytes exposed to antigen in vitro. In comparison, parvovirus peptide conjugated to KLH generated an IgG(1)-biased (TH2) response. These data indicate that the presentation of peptides on viral particles could be used to bias the immune response in favor of a TH1 response.Anti-viral and anti-peptide IgA were detected in intestinal and bronchial lavage fluid of immunized mice, demonstrating that a mucosal immune response to CPV can be generated by systemic and mucosal immunization with CVP vaccines. Serum antibody from both subcutaneously-vaccinated and intranasally-vaccinated mice showed neutralizing activity against CPV in vitro.

Different architectures in the assembly of infectious bursal disease virus capsid proteins expressed in insect cells

Infectious bursal disease virus (IBDV) capsid is formed by the processing of a large polyprotein and subsequent assembly of VPX/VP2 and VP3. To learn more about the processing of the polyprotein and factors affecting the correct assembly of the viral capsid in vitro, different constructs were made using two baculovirus transfer vectors, pFastBac and pAcYM1. Surprisingly, the expression of the capsid proteins gave rise to different types of particles in each system, as observed by electron microscopy and immunofluorescence. FastBac expression led to the production of only rigid tubular structures, similar to those described as type I in viral infection. Western blot analysis revealed that these rigid tubules are formed exclusively by VPX. These tubules revealed a hexagonal arrangement of units that are trimer clustered, similar to those observed in IBDV virions. In contrast, pAcYM1 expression led to the assembly of virus-like particles (VLPs), flexible tubules, and intermediate assembly products formed by icosahedral caps elongated in tubes, suggesting an aberrant morphogenesis. Processing of VPX to VP2 seems to be a crucial requirement for the proper morphogenesis and assembly of IBDV particles. After immunoelectron microscopy, VPX/VP2 was detected on the surface of tubules and VLPs. We also demonstrated that VP3 is found only on the inner surfaces of VLPs and caps of the tubular structures. In summary, assembly of VLPs requires the internal scaffolding of VP3, which seems to induce the closing of the tubular architecture into VLPs and, thereafter, the subsequent processing of VPX to VP2.

Expression and characterization of the two outer capsid proteins of African horsesickness virus: the role of VP2 in virus neutralization

African horsesickness virus (AHSV) is a gnat-transmitted member of the Orbivirus genus of the Reoviridae family. The virus has a genome of 10 double-stranded RNA species (L1-L3, M4-M6, S7-S10). The L2 and M6 genes of AHSV serotype 4 (AHSV-4) which encode the outer capsid proteins VP2 and VP5, respectively, were inserted into recombinant baculoviruses downstream of the baculovirus polyhedrin, or p10 promoters. Recombinant baculoviruses expressing VP2, VP5, or VP2 and VP5 proteins of AHSV-4 were isolated. The expressed AHSV proteins were similar in size and antigenic properties to those of viral AHSV-4. Expressed VP2 and VP5 proteins were purified to homogeneity and utilized to differentiate sera from vaccinated and infected horses. Antigens were also used to determine whether any other AHSV serotypes are related to AHSV-4. The results indicated that AHSV-4 is distantly related to some serotypes (e.g., AHSV-2, -6, and -9) but not to others (e.g., AHSV-5 and -7). Hyperimmune monospecific antisera raised in rabbits with purified VP2 neutralized the infectivity of a virulent strain of AHSV-4 isolated from an infected horse during a recent outbreak of the disease in Spain.