Evaluation of 225Ac for vascular targeted radioimmunotherapy of lung tumors.

Several alpha particle emitting radioisotopes have been studied for use in radioimmunotherapy. Ac-225 has the potential advantages of a relatively long half life of 10 days, and a yield of 4 alpha emissions in its decay chain with a total energy release of approximately 28 MeV. A new, 12 coordination site chelating ligand, HEHA, has been chemically modified for coupling to targeting proteins without loss of chelating ability. HEHA was coupled with MAb 201B which binds to thrombomodulin and accumulates efficiently in murine lung. Ac-225 was bound to the HEHA-MAb 201B conjugate and injected into BALB/c mice bearing lung tumor colonies of EMT-6 mammary carcinoma. Biodistribution data at 1 and 4 h postinjection indicated that, as expected, 225Ac was delivered to lung efficiently (> 300% ID/g). The 225Ac was slowly released from the lung with an initial t1/2 = 49 h, and the released 225Ac accumulated in the liver. Injection of free HEHA was only partially successful in scavenging free 225Ac. In addition to the slow release of 225Ac from the chelate, data indicated that decay daughters of 225Ac were also released from the lung. Immediately after organ harvest, the level of 213Bi, the third alpha-decay daughter, was found to be deficient in the lungs and to be in excess in the kidney, relative to equilibrium values. Injected doses of 225Ac MAb 201B of 1.0 microCi, delivering a minimum calculated absorbed dose of about 6 Gy to the lungs, was effective in killing lung tumors, but also proved acutely radiotoxic. Animals treated with 1.0 microCi or more of the 225Ac radioconjugate died of a wasting syndrome within days with a dose dependent relationship. We conclude that the potential for 225Ac as a radioimmunotherapeutic agent is compromised not only by the slow release of 225Ac from the HEHA chelator, but most importantly by the radiotoxicity associated with decay daughter radioisotopes released from the target organ.

Labeling and distribution of linear peptides identified using in vivo phage display selection for tumors.

To develop targeting molecules to be used for vascular targeting of short half-lived alpha-emitters for radioimmunotherapy, linear peptide phage display libraries were selected in vivo for binding to IC-12 rat tracheal tumors growing in severe combined immune deficient mice. After three rounds of selection, 15 phage clones were analyzed for DNA sequence, and the deduced translation products of cDNA inserts were compared. Three consensus sequences were chosen from three separate experimental selection series and peptides of these sequences with added -gly-gly-tyr were obtained. Peptides were radiolabeled on tyrosine with (125)I and the biodistribution in tumor-bearing mice was determined. The radioiodinated peptides were stable in vitro and when injected in tumor-bearing mice approximately 3.0 %ID/g accumulated in the tumor; however, much of the (125)I was found in the gastrointestinal tract and thyroid, indicative of dehalogenation of the labeled peptide. Radiolabeling peptide 2 with N-succinimidyl-3-(125)I-iodobenzoate resulted in faster excretion, which in turn resulted in lower levels in tumor and other organs, especially thyroid and gastrointestinal tract. Peptide 2 was derivatized with the bifunctional isothiocyanates of cyclohexyl-B diethylenetriaminepentaacetic acid (DTPA) or CHX-A" DTPA by direct conjugation or with a hydroxylamine derivative of 1B4M-DTPA (2-(p-[O-(carboxamylmethyl)hydroxylamine]benzyl)-6-methyl-diethylenetriamine-N,N,N',N",N"-pentaacetic acid ) coupled at the N-terminus. The primary molecular species in the conjugated products were shown by mass spectrometry to have one DTPA per peptide. Peptide chelate conjugates were radiolabeled with (213)Bi and the products tested for biodistribution in tumor-bearing mice. The data show that chelation of (213)Bi to peptides was accomplished by both the direct method of DTPA attachment and by the method using the linker at the N-terminus. Only small amounts of peptide accumulated at tumor sites. We conclude that phage display is a powerful tool to select peptides with restricted binding specificity; however, the peptides isolated to date do not bind with high retention to tumor sites in vivo.

Combination vascular targeted and tumor targeted radioimmunotherapy.

Rat MAb 201B, which binds to murine thrombomodulin, can deliver up to 50% of the injected dose of attached radioisotopes to the lung vascular endothelium. We have shown previously that intravenous injection of about 30 microCi of 213Bi-MAb 201B, which delivers about 15 Gy of alpha irradiation to the lung, is capable of eradicating small lung colonies (500-1000 cells) of the mammary tumor line, EMT-6. Larger tumors (> 5000 cells) were not completely cured by this vascular targeted radioimmunotherapy (VT-RAIT) approach. We reasoned that VT-RAIT might make the lung vessels serving the tumor cells more permeable, allowing MAb targeted to the tumor cells to extravasate more readily and mediate more efficient standard radioimmunotherapy (RAIT). Distribution experiments with the tumor targeted MAb 13A (RAIT MAb), following VT-RAIT, did not demonstrate a large increase in tumor uptake; however, microautoradiography did indicate that MAb 13A was distributed more evenly throughout the tumor when administered after VT-RAIT. Therapy experiments on lung tumors of approximately 5000 cells each, combining 213Bi-MAb 201B (VT-RAIT) with 213Bi-MAb 13A (RAIT) 24 hours later, resulted in a better outcome (3 cured/10 at risk) than for control groups: RAIT only (0/10), VT-RAIT only (1/10), or no therapy (0/10). RAIT therapy delivered 48 hours after VT-RAIT had no apparent benefit. 213Bi-MAb 201B VT-RAIT followed by 90Y-MAb 13A Fab' RAIT showed only a slight improvement in tumor cures (2/10) over that in control groups: (0/9), (0/10), (0/10), respectively. These results suggest that optimal timing, dosage, and choice of MAb for RAIT should enhance the double MAb therapy approach significantly.

CD44 expression on murine tissues.

CD44 is a cell surface glycoprotein found on lymphoid and epithelial cells. Its primary function on lymphocytes and macrophages is to mediate interaction with endothelium, while its function on epithelial cells is not known. The protein has many different forms, generated by alternative mRNA splicing and by post-translational modification, which may mediate different functions. During previous work on murine lung tumor cells, mAb 133-13A was isolated and shown to recognize a surface glycoprotein, P100, of 90-100 x 10(3) M(r). Amino acid sequence analysis of purified P100 indicates that it is CD44. Since few data exist to indicate which forms of CD44 are present in different normal tissues, mAb 133-13A was used to analyze CD44 expression in mouse tissue. Quantitative data on the distribution of CD44(P100) in mice show that spleen, thymus, liver, intestine, uterus and choroid of the eye are major sites of expression. In addition, epithelia of adrenals, esophagus and trachea are CD44(P100) positive. Previous work on human cell lines has implicated a high molecular mass (130-160 x 10(3) M(r)) form of the glycoprotein as the form expressed in epithelial cells and carcinomas. Isolation of CD44 proteins from lymphoid tissues in the mouse indicate that, as in human lymphoid tissue, the low molecular mass form (80-90 x 10(3) M(r)) is predominately expressed. These data show that both small (approximately 81 x 10(3) M(r)) and large forms of the glycoprotein are expressed in basal epithelia of esophagus and trachea and in salivary gland, while only the small form is expressed in epithelium of the adrenal cortex and in the murine lung and mammary carcinomas studied. While these data cannot distinguish between specific splice variants, they show that the large forms of CD44 are minor components in normal tissue and seem to be found only in basal epithelium. The CD44 of low M(r) found in epithelial tissues is probably associated with lymphoid cell types in the tissues.

The beta 4 subunit of the integrin family is displayed on a restricted subset of endothelium in mice.

More than 15 subunits of the integrin family of cell surface adhesion molecules have been identified. The alpha 6 beta 4 integrin has recently been identified as a component of hemidesmosomes of stratified squamous epithelium. The monoclonal antibody (mAb) 346-11A binds to the beta 4 subunit in mice. Sequence analysis of a cDNA clone coding for this epitope localizes reaction of the mAb to a portion about half way through the extracellular domain at the beginning of the cysteine-rich region. Sites of beta 4 expression in mice were detected by autoradiographic analysis of tissues collected from mice 24 or 48 h after intravenous injection of 125I-labeled mAb 346-11A. This non-quantitative technique emphasizes detection of antigen exposed in the vascular space. These data show that in addition to the epithelia of several organs, the endothelia of intermediate vessels throughout the body are sites of beta 4 expression. In particular, endothelium of larger vessels but not capillaries in lung, vessels in thymus, spleen and Peyer's patches, and portal vessels but not the central veins of the liver, are positive. The expression of the beta 4 integrin in blood vessels may indicate a specialized function for beta 4 at these sites that is distinct from its role in hemidesmosome-mediated attachment.

Second generation monoclonal antibodies to the human integrin alpha 6 beta 4.

Second generation monoclonal antibodies to the alpha 6 beta 4 subunits of human integrins have been prepared. MAbs 450-9D, 10D, and 11A1 react at different sites on the beta 4 molecule and MAbs 450-30A1 and 33D react at the same site on the alpha 6 subunit. Double determinant (two-site) radioimmunoassays using combinations of these MAbs have been developed. Two assays for beta 4 distinguish between the whole beta 4 molecule and the beta 4 molecule truncated from the C-terminus (form c) while another assay measures the presence of alpha 6 subunits. Data from the two-site assays support the following conclusions: (1) Colon tumors and normal colon mucosa express large amounts of alpha 6 beta 4 although only form c of the beta 4 was detected; (2) There is no evidence for alpha 6 beta 1 expression in colon; however, some of this complex may be present in certain lung tumors. The extracellular domains of alpha 6 and beta 4 can associate with each other even if the cytoplasmic domain of the beta 4 subunit is not present. MAbs to specific domains of the beta 4 molecule may be useful in analyses of forms a and c in normal and malignant tissue. The fact that only the largest beta 4 molecule "a" retains the phosphorylation site may have functional significance.

Enhancement of lung tumor colony formation by treatment of mice with monoclonal antibodies to pulmonary capillary endothelial cells.

Two rat monoclonal antibodies, 34A and 201B, have been isolated and shown to bind preferentially to capillary endothelial cells in the lung. Administration of these antibodies to mice increases the number of lung colonies derived from i.v. injection of tumor cells. The antibodies increase lung colonization in C57BL/6 mice following i.v. injection of B16-F10 melanoma cells and in BALB/c mice following injection of line 1 lung carcinoma cells. Neither 34A nor 201B monoclonal antibody binds to B16 melanoma or line 1 carcinoma and so must exert its effect by interaction with endothelial cells. Antibodies injected i.v., s.c., or i.p. are active from 1 h to 1 wk if injected before cell injection. The effect is optimal when 0.1 ml of ascites fluid containing 120 micrograms of antigen binding capacity of both MoAbs 34A and 201B is injected. Significant damage to endothelial cells could not be documented by histopathological examination at the light microscope level or by protein leakage into the air space as measured by lung lavage. However, electron micrographs taken 3 h after monoclonal antibody injection show minor damage to endothelial cell membranes throughout the lung with some areas of mild edema. The increased colonization may be mediated by this subtle damage to endothelial cells, or antibody interactions with endothelial cells may trigger secondary reactions such as altered expression of growth factors.