As an anticancer therapy, the strategy of the immunotoxins (ITs) can be described as a cancer missile. IT is composed of a monoclonal antibody, a target system, and a toxin part, the bomber. The antibody will bind the surface antigen specifically presented on the cancer cells. The toxin, which is delivered by the antibody, will enter the cytosolic part of the cancer cell to kill the cell. The toxins can be diphtheria toxin, pseudomonas exotoxin (PE), and ricin etc.. Those toxins are glycosidases. The enzymes can modify ribosomes. The modified ribosomes will not support protein synthesis. One toxin molecule can modify thousands of ribosomes in one minute, so one toxin molecule is enough to cause the apoptosis of a cancer cell(1).
ITs are designed based on the cancer specific surface antigens. Finding the cancer specific antigens can help to produce the monoclonal antibodies to target the cancer cells. To be a good cancer target, the cancer specific antigens should meet some criteria. To avoid or reduce the toxicity to normal cells, the antigen should not exist on the normal cells or exist at a very low level compared to the cancer cells. To enable toxin to enter the cancer cell, the surface antigen must internalize by antibody binding. To increase the accessibility of ITs to the cancer cells, it is better that the antigen distributes in a large surface of the cancer cell. DNA microarry and proteomic analysis are powerful tools to find the protein surface antigens on the cancer cells. Up to now, the ability to find, analyze and synthesize the non-protein antigens is quite limited. In order to circumvent the limitation of finding antigens, the whole cancer cell can be used as the antigen to immunize mouse to produce monoclonal antibodies. The cancer specific antibody can be screened from the pools of the monoclonal antibodies. The construction of human antibody phage display libraries makes the antibody screening quickly(2).
The first generation of ITs was composed of a whole monoclonal antibody and toxin molecules chemically conjugated to the antibody. The original ITs had some problems to limit their application on cancer therapy. The mouse resource monoclonal antibody can cause immune reactions to patients. The large molecular size makes the IT hard to penetrate tissues to access the tumor cells. The produce cost is high to prepare monoclonal antibodies from cell cultures. Some progress such as humanized antibody(3) and single chain antibody(4)have made the way of ITs being used as an anticancer drug smoothen. The single chain antibody is composed of a variable domain from the antibody heavy chain and a variable domain from the light chain. The two domains are either fused together by a short peptide (scFv) or cross-linked by a disulfide (dsFv). The molecular size of the truncated antibody is about 1/10 of the whole antibody. While it remains the high affinity to the antigen, it is easier to access the cancer cells. The produce cost can be reduced by expressing the recombinant small protein in E. Coli.
Mylotarg is the first FDA approved immunotoxin drug. The drug is approved in 1999 for the treatment of refractory acute myelogenous leukemia. The drug is composed of a humanized anti-CD33 antibody and a small drug calicheamicin conjugated to the antibody(5). BL22, anti-CD22 (dsFv)-PE38, is in clinical phase II trail for the treatment of chemotherapy-resistant hairy cell leukemia (6). LMB-2, anti-Tac(Fv)-PE38, is in clinical phase II trail for the treatment of CD25 positive chronic lymphocytic leukemia (7).
II. MAb8H9 and the construction of Fv-PE38 immunotoxin *
The monoclonal antibody 8H9 was produced by immunizing human neuroblastoma cells to mice. 8H9 was highly reactive human brain tumors, childhood sarcomas, and neuroblastomas. However, 8H9 was nonreactive with normal human tissues including bone marrow, colon, stomach, heart, lung, muscle, thyroid, testes, kidney, and human brain. 8H9 showed nonspecific reactive with pancreas, adrenal cortex, and liver. The nonspecific reaction may come from the Fc part of the antibody(8). The 8H9 antigen was on the external surface of tumor cell membrane.To find the 8H9 antigen, 8H9 immunoprecipitated a broad protein band centered around 90KDa from all the 8H9 positive cell lines. After N-glycanase treatment, a single protein band around 58KDa was found. The study showed that the 8H9 antigen was a protein with a heterogeneous glycosylation pattern(8). The characterization of 8H9 and its antigen presented on the surface of cancer cells suggests that 8H9 may be useful for targeted cancer therapy.
Two types of 8H9 based ITs, 8H9(scFv)-PE38 and 8H9(dsFv)-PE38( Fig. 1.), were constructed. The scFv immunotoxin is a single peptide protein. The toxin PE38, a 38-KDa truncated mutant form of Pseudomonas exotoxin A, is fused to the C-terminal of Fv portion of 8H9. The recombinant IT is expressed in E. Coli.. The protein accumulates in inclusion bodies. The refolding yield of the protein is 1.7%. The purified protein is about 62 KDa. The dsFv immunotoxin is a two subunits protein. The small subunit is the VL domain of 8H9. The large subunit is the VH domain fused with the PE38. The two subunits are separately expressed in E. Coli.. The dsFv IT is prepared by combining inclusion body protein of the two subunits. The refolding yield is 16%. Obviously, the 8H9(dsFv)-PE38 is more suitable for large-scale production.
III. In vitro cytotoxic activities of 8H9(Fv)-PE38 on malignant cell lines
Since PE can inhibits protein synthesis, the cytotoxicity activity of the 8H9(Fv)-PE38 was measured by protein synthesis inhibition. Protein synthesis can be measured by the cellular incorporation of radiolabeled amino acids such as [3H]leucine. The cytotoxicity of the 8H9(scFv)-PE38 to a variety of malignant cell lines was measured first. Three breast cancer cell lines were tested. The IC50 values are 5 ng/ml for MCF-7, 20 ng./ml for BT-474, and 35ng/ml for ZR-75-1. The IC50 values for three osteosarcoma cell lines are 30 ng/ml for U2OS, 50 ng/ml for CRL1427, and 20 ng/ml for OHS-M1. The IC50 values for three neuroblastoma cell lines are 9 ng/ml for NMB-7, 12.5 ng/ml for LAN-1, and 90 ng/ml for SK-N-BE. Two 8H9 antigen negative cell lines were used as negative controls, there was no cytotoxic effect at 1000 ng/ml on the two cell lines. The cytotoxicity of the 8H9(dsFv)-PE38 on MCF-7 cell line was also measured. The IC50 is 5 ng/ml. This indicates that 8H9(scFv)-PE38 and 8H9(dsFv)-PE38 have similar ctytotoxicity in vitro.
A control IT M1(dsFv)-PE38 was not cytotoxic to MCF-7 cell. That indicates the 8H9 Fv specific cytotoxicity to the malignant cell line. The cytotoxic activity of 8H9(scFv)-PE was competed by increasing amount of the MAb8H9, but was not by a control MAb that recognizes CD30. In addition, MAb 8H9 itself did not show cytotoxicity to the MCF-7 cell line. Therefore, the cytotoxicity of 8H9(scFv)-PE38 is mediated by specific 8H9 antigen and the PE38 toxin.
IV. Toxicology study of 8H9(dsFv)-PE38 in mice and cynomolgus monkey *
In order to obtain a reference maximum injection dose of the IT, 8H9(dsFv)-PE38 was measured for its nonspecific toxicity in mice. Groups of mice received one injection of different doses of the IT, the mortality of the mice was observed for two weeks. The LD50 of 8H9(dsFv)-PE38 is 0.78 mg/kg (95% confidential range, 0.66-0.93 mg/kg).
In general, the data collected from mouse toxicity studies are not useful in predicating human toxicities. For predicating the IT toxicity on human, two cynomologus monkeys were used for toxicology study. 8H9 showed similar reactivity to normal tissues of human and monkey (8). One monkey received three injections with a dose of 0.1mg/kg on days 1, 3, and 5. The other received three injections with a dose of 0.2 mg/kg on days 1,3, and 5. Both monkeys tolerated 8H9(dsFv)-PE38 well. The biochemical data showed a slight decrease in albumin and a boderline elevated alanine aminotransferase in the 0.2 mg/kg dose monkey (Table 4.). The observed toxicity in the two monkeys was loss of appetite.The study indicates that 8H9(dsFv)-PE38 can be administered safely to monkeys at the dose of 0.2mg/kg. Serum levels of 8H9(dsFv)-PE38 were measured in the two monkeys 10 minutes after each of the three injections. In the 0.1 mg/kg monkey, the levels were 5.0-5.4 ug/ml. In the 0.2 mg/kg monkey, the levels were 11.0-13.0 ug/ml. The IC50 value of 8H9(dsFv)-PE38 is 5 ng/ml for MCF-7 cell. The blood levels are 1000-fold higher than the IC50. The blood level of 8H9(dsFv)-PE38 is high enough to kill cancer cells.
V. In Vivo antitumor activity in SCID mice bearing human cancer cell lines *
In order to determine the antitumor activity of 8H9(Fv)-PE38 in vivo, SCID mice were used. SCID mouse has deficiency in the immune system. When cancer cells are injected into the mice, tumor can be formed in the mice. In this experiment, a human breast cancer cell line MCF-7 and an osteosarcoma cell line OHS-M1 were injected to the SCID mice to induce the tumors. The mice developed tumors of about 50 mm3 in size by day 4.
The mice bearing tumors were treated by 8H9(scFv)-PE-38 or 8H9(dsFv)-PE38 three times on days 4, 6, and 8. The doses were 0.075 mg/kg and 0.15 mg/kg for two groups. 8H9(dsFv)-PE38 showed strong tumor regression effect on MCF-7 tumor at both doses (Fig. 6C.). 8H9(dsFv)-PE38 and 8H9(scFv)-PE38 showed similar antitumor activities at the dose of 0.15 mg/kg (Fig. 6D.). 8H9(scFv)-PE38 had no effect on the OHS-M1 tumor at the dose of 0.075 mg/kg. It showed tumor regression effect at the dose of 0.15 mg/kg (Fig. 6F. ). The data indicate that OHS-M1 tumor is less sensitive to 8H9(Fv)-PE38 than the MCF-7 tumor. This can be explained by a higher 8H9(Fv)-PE38 IC50 value of OHS-M1 (MCF-7 = 5 ng/ml, OHS-M1 = 20 ng/ml). Two methods can improve the effect of 8H9(Fv)-PE38 on OHS-M1 tumor regression. One method is to increase the dose of the IT. This is limited by the toxicology of 8H9(Fv)-PE38. The other method is to decrease the IC50 of OHS-M1. The aim can be obtained by mutating the Fv portion of 8H9 to increase the affinity of the antibody to its antigen. In addition, the drug delivery method can be modified to specifically increase the drug level in the tumor regions.
VI. Summary
8H9 has selectivity to cancer cells, and it can be used for targeted cancer
therapy. The immunotoxin 8H9(dsFv)-PE38 can be produced in higher yield. The IT
showed specific cyototoxicity on cancer cells from breast cancer, osteosarcoma,
and neuroblastomas in vitro. The IT showed specific antitumor activity on
SCID mice bearing MCF-7 breast cancers or OHS-M1osteosarcomas. The IT could be
administered safely at a dose higher than that needed to cause tumor
regression.
VII. References
1. Olsnes S.. 2004. The history of ricin, abrin and related toxins. Toxicon.
44(4):361-70.
2. Smith KA, Nelson PN, Warren P, Astley SJ, Murray PG,
Greenman J. 2004. Demystified...recombinant antibodies. J
Clin Pathol. 57(9):912-7.
3. Winter G, Harris WJ. 1993. Humanized
antibodies. Trends
Pharmacol Sci. 14(5):139-43.
4. Pastan I. 1997. Targeted therapy of
cancer with recombinant immunotoxins. Biochim
Biophys Acta. 1333(2):C1-6.
5. Ludwig DL, Pereira DS, Zhu Z, Hicklin DJ,
Bohlen P. 2003. Monoclonal antibody therapeutics and apoptosis. Oncogene.
22(56):9097-106.
6. Phase II Study of BL22 Immunotoxin in Patients With
Cladribine-Resistant Hairy Cell Leukemia. NCI
Clinical Trails
7. Phase II Study of LMB-2 Immunotoxin in Patients With
CD25-Positive Chronic Lymphocytic Leukemia.
NCI Clinical Trails
8. Modak S, Kramer K, Gultekin SH, Guo HF, Cheung
NK. 2001.Monoclonal antibody 8H9 targets a novel cell surface antigen expressed
by a wide spectrum of human solid tumors.Cancer
Res. 61(10):4048-54.