Reduced folate carrier expression in acute lymphoblastic leukemia: a mechanism for ploidy but not lineage differences in methotrexate accumulation

Methotrexate (MTX) is one of the most active and widely used agents for the treatment of acute lymphoblastic leukemia (ALL). To elucidate the mechanism for higher accumulation of MTX polyglutamates (MTX-PG) in hyperdiploid ALL and lower accumulation in T-lineage ALL, expression of the reduced folate carrier (RFC) was assessed by reverse transcription-polymerase chain reaction in ALL blasts isolated from newly diagnosed patients. RFC expression exhibited a 60-fold range among 29 children, with significantly higher expression in hyperdiploid B-lineage ALL (median, 11.3) compared with nonhyperdiploid ALL (median, 2.1; P <.0006), but no significant difference between nonhyperdiploid B-lineage and T-lineage ALL. Furthermore, mRNA levels of RFC (mapped by FISH to chromosome 21) were significantly related to chromosome 21 copy number (P =.0013), with the highest expression in hyperdiploid ALL blasts with 4 copies of chromosome 21. To assess the functional significance of gene copy number, MTX-PG accumulation was compared in ALL blasts isolated from 121 patients treated with either low-dose MTX (LDMTX; n = 60) or high-dose MTX (HDMTX; n = 61). After LDMTX, MTX-PG accumulation was highest in hyperdiploid B-lineage ALL with 4 copies of chromosome 21 (P =.011), but MTX-PG accumulation was not significantly related to chromosome 21 copy number after HDMTX (P =.24). These data show higher RFC expression as a mechanism for greater MTX accumulation in hyperdiploid B-lineage ALL and indicate that lineage differences in MTX-PG accumulation are not due to lower RFC expression in T-lineage ALL.

In vivo toxicity and pharmacokinetic features of B43(Anti-CD19)-Genistein immunoconjugate

B43(anti-CD19)-Genistein immunoconjugate targets genistein, a naturally occurring protein tyrosine kinase inhibitory isoflavone to the membrane-associated anti-apoptotic CD19-LYN complexes and triggers apoptotic cell death. In this preclinical study, the toxicity profiles of B43-Genistein as well as unconjugated genistein were evaluated in mice. B43-Genistein and genistein were administered either as single bolus injections or daily injections for 10 consecutive days via the intraperitoneal route to mice. Genistein was not toxic to mice at the highest dose of 40 mg/kg and no test article-related histopathological lesions were found in any of the 64 genistein-treated mice. B43-Genistein had a significantly longer elimination half-life and slower plasma and tissue clearance than unconjugated genistein. B43-Genistein was not toxic to mice at the highest single dose of 40 mg/kg or highest cumulative dose of 100 mg/kg and no test article-related histopathological lesions were found in any of the 108 mice treated with B43-genistein. To our knowledge, this is the first preclinical toxicity and pharmacokinetic study of a tyrosine kinase inhibitor-containing immunoconjugate.

In vivo toxicity, pharmacokinetics, and anticancer activity of Genistein linked to recombinant human epidermal growth factor

Epidermal growth factor receptor (EGFR)-associated protein tyrosine kinase (PTK) complexes have vital anti-apoptotic functions in human breast cancer cells. We have shown previously that targeting the naturally occurring PTK inhibitor genistein to the EGFR-associated PTK complexes using the EGF-Genistein (Gen) conjugate triggers rapid apoptotic cell death in human breast cancer cells and abrogates their in vitro clonogenic growth. In the present study, we examined the in vivo toxicity profile, pharmacokinetics, and anticancer activity of EGF-Gen. No toxicities were observed in mice treated with EGF-Gen at dose levels as high as 40 mg/kg administered i.p. as a single dose or 140 mg/kg administered i.p. over 28 consecutive days. EGF-Gen significantly improved tumor-free survival in a severe combined immune deficiency (SCID) mouse xenograft model of human breast cancer, when it was administered 24 h after inoculation of tumor cells. At 100 microg/kg/day x 10 days (1 mg/kg total dose), which is >100-fold less than the highest tested and nontoxic cumulative dose (ie., 140 mg/kg) in mice, EGF-Gen was more effective than cyclophosphamide (50 mg/kg/day x 2 days), Adriamycin (2.5 mg/kg x 1 day), or methotrexate (0.5 mg/kg x 1 day), the most widely used standard chemotherapeutic drugs for breast cancer, and resulted in 60% long-term tumor-free survival. Furthermore, treating SCID mice with established s.c. human breast cancer xenografts of 0.5-cm diameter with EGF-Gen at this dose level resulted in disappearance of the tumors in two of five mice and >50% shrinkage in three of five mice within 10 days, whereas all of the control tumors in five PBS-treated mice as well as five mice treated with unconjugated Gen (1 mg/kg/day x 10 days) showed >200% increase in diameter during the same observation period. EGF-Gen treatment reduced the growth rate of breast cancer xenografts of 1.0-cm diameter, but unlike with tumors of 0.5-cm diameter, it failed to cause shrinkage or disappearance of these larger tumors. The level of EGF-Gen systemic exposure that was effective in SCID mice was achieved in cynomolgus monkeys without any significant side effects detectable by clinical observation, laboratory studies, or histopathological examination of multiple organs. EGF-Gen might be useful in the treatment of breast cancer as well as other EGFR-positive malignancies.

Etoposide and antimetabolite pharmacology in patients who develop secondary acute myeloid leukemia

Etoposide, an effective agent for acute lymphoblastic leukemia (ALL), can cause secondary acute myeloid leukemia (AML) in a subset of patients. Our objectives were to determine whether patients who develop secondary AML displayed altered etoposide pharmacokinetics or other pharmacologic characteristics compared to identically treated patients who did not develop AML. Children with newly diagnosed ALL were treated according to a protocol which included etoposide 300 mg/m2 given three times over 8 days during remission induction and once every 2-4 weeks during 120 weeks of continuation therapy. Characteristic 11q23 rearrangements were documented in seven of the eight patients with AML. Etoposide clearance, time that etoposide concentrations exceeded 10 microM, etoposide or etoposide catechol area-under-the-plasma-concentration vs time curve (AUC), serum albumin, and average methotrexate concentration did not differ significantly between the two groups; thiopurine methyltransferase (TPMT) activity tended to be lower in the eight children who did vs the 23 matched control children who did not develop AML (P=0.16). Further regression analyses likewise indicated that lower TPMT activity tended to be associated with shorter onset of secondary AML (P=0.11); it also tended to be lower among the eight index cases compared to the entire unmatched cohort of 105 identically treated children with ALL (P=0.10). We observed no statistically significant differences in etoposide disposition and antimetabolite pharmacology between patients who did and did not develop secondary AML.

In vivo toxicity and pharmacokinetic features of B43 (anti-CD19)-genistein immunoconjugate in nonhuman primates

B43 (anti-CD19)-genistein immunoconjugate targets genistein, a naturally occurring protein tyrosine kinase-inhibitory isoflavone to the membrane-associated antiapoptotic CD19-LYN complexes and triggers apoptotic cell death. In this preclinical study, the toxicity profiles of B43-genistein as well as unconjugated genistein were evaluated in cynomolgus monkeys. B43-genistein and genistein were administered either as single bolus injections or daily injections for 5-10 consecutive days via the i.v. route to monkeys. Neither genistein nor B43-genistein was toxic to cynomolgus monkeys, and no test article-related histopathological lesions were found in any of the two genistein-treated or five B43-genistein-treated cynomolgus monkeys. B43-genistein showed a favorable pharmacokinetics in monkeys, with a plasma half-life of 10-23 h. Plasma samples from B43-genistein-treated monkeys elicited potent and CD19 antigenspecific antileukemic activity against human CD19+ leukemia cells in vitro. To our knowledge, this is the first preclinical toxicity and pharmacokinetic study of a tyrosine kinase inhibitor-containing immunoconjugate in nonhuman primates.

In vivo biotherapy of HL-60 myeloid leukemia with a genetically engineered recombinant fusion toxin directed against the human granulocyte macrophage colony-stimulating factor receptor

Acute myeloid leukemia (AML) is the most common form of acute leukemia. Contemporary chemotherapy regimens fail to cure most patients with AML. We have genetically engineered a recombinant diphtheria toxin human granulocyte macrophage colony-stimulating factor (GMCSF) chimeric fusion protein (DTctGMCSF) that specifically targets the GMCSF receptor on fresh human AML cells and myeloid leukemia cell lines. At a nontoxic dose level, DTctGMCSF therapy was superior to the standard chemotherapeutic agents 1-beta-D-arabinofuranosylcytosine and Adriamycin, resulting in 60% long-term event-free survival of severe combined immunodeficient mice challenged with an otherwise invariably fatal cell dose of the human HL-60 myeloid leukemia. Notably, systemic exposure levels of DTctGMCSF, which were found to be therapeutic in the severe combined immunodeficient mouse xenograft model of human HL-60 myeloid leukemia, could be achieved in cynomolgus monkeys without any significant nonhematological toxicities. The recombinant DTctGMCSF fusion toxin might be useful in the treatment of AML patients whose leukemias have recurred and developed resistance to contemporary chemotherapy programs.

Differences in folylpolyglutamate synthetase and dihydrofolate reductase expression in human B-lineage versus T-lineage leukemic lymphoblasts: mechanisms for lineage differences in methotrexate polyglutamylation and cytotoxicity

Cellular accumulation of methotrexate polyglutamates (MTXPGs) is recognized as an important determinant of the cytotoxicity and selectivity of methotrexate in acute lymphoblastic leukemia (ALL). We identified a significantly lower cellular accumulation of MTXPGs in T-lineage versus B-lineage lymphoblasts in children with ALL, which is consistent with the worse prognosis of T-lineage ALL when treated with conventional antimetabolite-based therapy. Maximum MTXPG accumulation in leukemic blasts in vivo was 3-fold greater in lymphoblasts of children with B-lineage ALL (129 children) compared with those with T-lineage ALL (20 children) (p < 0.01) and was characterized by a saturable (Emax) model in both groups. The human leukemia cell lines NALM6 (B-lineage) and CCRF/CEM (T-lineage) were used to assess potential mechanisms for these lineage differences in MTX accumulation, revealing i) greater total and long-chain MTXPG accumulation in NALM6 over a wide range of methotrexate concentrations (0.2-100 microM), ii) saturation of MTXPG accumulation in both cell lines, with a higher maximum (Emax in NALM6, iii) 3-fold higher constitutive FPGS mRNA expression and enzyme activity in NALM6 cells, iv) 2-fold lower levels of DHFR mRNA and protein in NALM6 cells, and v) 4-6 fold lower extracellular MTX concentration and 2-fold lower intracellular MTXPG concentration to produce equivalent cytotoxicity (LC50) in NALM6 versus CEM. There was a significant relationship between FPGS mRNA and enzyme activity in lymphoblasts from children with newly diagnosed ALL, and blast FPGS mRNA and activity increased after methotrexate treatment. These data indicate higher FPGS and lower DHFR levels as potential mechanisms contributing to greater MTXPG accumulation and cytotoxicity in B-lineage lymphoblasts.

Enhanced proteolysis of thiopurine S-methyltransferase (TPMT) encoded by mutant alleles in humans (TPMT*3A, TPMT*2): mechanisms for the genetic polymorphism of TPMT activity

TPMT is a cytosolic enzyme that catalyzes the S-methylation of aromatic and heterocyclic sulfhydryl compounds, including medications such as mercaptopurine and thioguanine. TPMT activity exhibits autosomal codominant genetic polymorphism, and patients inheriting TPMT deficiency are at high risk of potentially fatal hematopoietic toxicity. The most prevalent mutant alleles associated with TPMT deficiency in humans have been cloned and characterized (TPMT*2 and TPMT*3A), but the mechanisms for loss of catalytic activity have not been elucidated. In the present study, we established that erythrocyte TPMT activity was significantly related to the amount of TPMT protein on Western blots of erythrocytes from patients with TPMT activities of 0.4-23 units/ml pRBC (rs = 0.99; P < 0.001). Similarly, heterologous expression of wild-type (TPMT*1) and mutant (TPMT*2 and TPMT*3A) human cDNAs in yeast and COS-1 cells demonstrated comparable levels of TPMT mRNA but significantly lower TPMT protein with the mutant cDNAs. Rates of protein synthesis were comparable for wild-type and mutant proteins expressed in yeast and with in vitro translation in rabbit reticulocyte lysates. In contrast, pulse-chase experiments revealed significantly shorter degradation half-lives for TPMT*2 and TPMT*3A ( approximately 0.25 hr) compared with wild-type TPMT*1 (18 hr). The degradation of mutant proteins was impaired by ATP depletion and in yeast with mutant proteasomes (pre-1 strain) but unaffected by the lysosomal inhibitor chloroquine. These studies establish enhanced degradation of TPMT proteins encoded by TPMT*2 and TPMT*3A as mechanisms for lower TPMT protein and catalytic activity inherited by the predominant mutant alleles at the human TPMT locus.

In vivo toxicity, pharmacokinetics, and antileukemic activity of TXU (anti-CD7)-pokeweed antiviral protein immunotoxin

We evaluated the TXU (anti-CD7)-pokeweed antiviral protein (PAP) immunotoxin in both murine and nonhuman primate models. TXU-PAP caused dose-limiting cardiac toxicity in BALB/c mice. In a SCID mouse model of invariably fatal human T-lineage acute lymphoblastic leukemia (ALL), TXU-PAP therapy resulted in a marked improvement of leukemia-free survival without any side effects. Whereas 100% of control mice treated with PBS, unconjugated TXU antibody, or B43-PAP (an immunotoxin that does not react with T-lineage ALL cells) died of disseminated human leukemia within 80 days (median survival, 37 days), 80 +/- 13% of SCID mice treated with 15 microgram of TXU-PAP (median survival, >120 days) and 100% of mice treated with 30 microgram of TXU-PAP (median survival, > 120 days) remained alive and free of leukemia for >120 days. In cynomolgus monkeys, TXU-PAP showed favorable pharmacokinetics with an elimination half-life of 8.1-8.7 h. The monkeys treated with TXU-PAP at dose levels of 0.05 mg/kg/day x 5 days and 0.10 mg/kg/day x 5 days tolerated the therapy very well, without any significant clinical compromise or side effects, and at necropsy, no gross or microscopic lesions were found. This study provides a basis for further evaluation of TXU-PAP as an investigational biotherapeutic agent in the treatment of T-lineage ALL.

Prevalence and growth characteristics of malignant stem cells in B-lineage acute lymphoblastic leukemia

We used a stroma-supported culture method to study the prevalence and growth characteristics of malignant stem cells in acute lymphoblastic leukemia (ALL). In 51 of 108 B-lineage ALL samples, bone marrow-derived stroma not only inhibited apoptosis of ALL cells but also supported their proliferation in serum-free medium. When single leukemic cells were placed in the stroma-coated wells of microtiter plates, the percentage of wells with leukemic cell growth after 2 to 5 months of culture ranged from 6% to 20% (median, 15%; 5 experiments). The immunophenotypes and genetic features of cells recovered from these cultures were identical to those noted before culture. All cells maintained their stroma dependency and self-renewal capacity. Leukemic clones derived from single cells contained approximately 10(3) to 10(6) cells after 1 month of culture; other clones became detectable only after prolonged culture. Cell growth in stroma-coated wells correlated with the number of initially seeded cells (1 or 10; r = .87). However, the observed percentages of positive wells seeded with 10 cells always exceeded values predicted from results with single-cell-initiated cultures (P < .003 by paired t-test), suggesting stimulation of leukemic cell growth by paracrine factors. In conclusion, the proportion of ALL cells with clonogenic potential may be considerably higher than previously thought.

Pharmacokinetic features, immunogenicity, and toxicity of B43(anti-CD19)-pokeweed antiviral protein immunotoxin in cynomolgus monkeys

We studied the pharmacokinetic features, immunogenicity, and toxicity of B43-pokeweed antiviral protein (PAP) immunotoxin in 13 cynomolgus monkeys. The disposition of B43-PAP in two monkeys, when administered as a single i.v. bolus dose, was characterized by a slow clearance (1-2 ml/h/kg) with a very discrete peripheral distribution. B43-PAP was retained and distributed largely in the blood as the sole compartment with no significant equilibration with the extravascular compartment. The circulating B43-PAP immunotoxin detected in monkey plasma samples by ELISA and protein immunoblotting was both immunoreactive with, and active against, human leukemic cells in vitro. In systemic immunogenicity and toxicity studies, which involved 11 cynomolgus monkeys, each monkey received a total of seven i.v. doses of B43-PAP at a specific dose level of the dose escalation schedule. B43-PAP-treated monkeys mounted a dose-dependent humoral immune response against both the mouse IgG and PAP moieties of the immunotoxin. When administered i.v. either on an every-day or every-other-day schedule, B43-PAP was very well tolerated, with no significant clinical or laboratory signs of toxicity at total dose levels ranging from 0.007 to 0.7 mg/kg. A transient episode of a mild capillary leak with a grade 2 hypoalbuminemia and 2+ proteinuria was observed at total dose levels equal to or higher than 0.35 mg/kg. At total dose levels of 3.5 and 7.0 mg/kg, B43-PAP caused dose-limiting renal toxicity due to severe renal tubular necrosis. The present study completes the preclinical evaluation of B43-PAP and provides the basis for its clinical evaluation in children with therapy-refractory B-lineage acute lymphoblastic leukemia.

Treatment of human B-cell precursor leukemia in SCID mice using a combination of the investigational biotherapeutic agent B43-PAP with cytosine arabinoside

Combined immunochemotherapy regimens using the investigational biotherapeutic agent B43(anti-CD19)-poke-weed antiviral protein (PAP) immunotoxin may offer an effective treatment for refractory B-cell precursor leukemias. The purpose of the present study was to explore and identify effective combinations of B43-PAP with standard chemotherapeutic drugs, including the anthracyclin doxorubicin, the epipodophyllotoxin etoposide, the nitrosurea carmustine, and the antimetabolite cytosine arabinoside. Here, we report that the B43-PAP plus cytosine arabinoside combination has potent antileukemic activity against human B-cell precursor leukemia in SCID mice and leads to 100% long-term event-free survival from an otherwise invariably fatal leukemia. Surprisingly, none of the other treatment protocols tested, including combinations of B43-PAP with carmustine, doxorubicin, or etoposide, proved more effective than B43-PAP alone.

Methylation of mercaptopurine, thioguanine, and their nucleotide metabolites by heterologously expressed human thiopurine S-methyltransferase

Thiopurine S-methyltransferase (TPMT), a cytosolic enzyme that exhibits genetic polymorphism, catalyzes S-methylation of mercaptopurine (MP) and thioguanine (TG), yielding S-methylated nucleobases that are inactive, whereas S-methylated nucleotides of these thiopurines are cytotoxic. A yeast-based heterologous expression system was therefore used to characterize human TPMT-catalyzed methylation of MP, TG, and their principal nucleotide metabolites [thioinosine monophosphate (TIMP) and thioguanosine monophosphate (TGMP), respectively]. MP, TG, TIMP, and TGMP were all substrates for human TPMT, exhibiting similar Michaelis-Menten kinetic parameters (Km, 10.6-27.1 microM; Vmax, 31-59 nmol/min/mg of TPMT). Consistent with these kinetic parameters, human leukemia cells (CEM) incubated for 24 hr with 10 microM MP or TG accumulated significantly higher (2.3-fold, p = 0.01) concentrations of methyl-TIMP after MP incubation than methyl-TGMP after TG incubation, due to the 2.7-fold higher concentration of TIMP after MP incubation, compared with TG nucleotides (TGN) after TG incubation. Moreover, intracellular accumulation of TGN was 2.5-fold greater after TG incubation than after MP incubation (p = 0.01). These data establish that MP, TG, and their principal nucleotide metabolites are comparable substrates for polymorphic TPMT, and they demonstrate significant differences in the accumulation of active TGN and methylated nucleotides when leukemia cells are treated with MP versus TG.

Favorable pharmacodynamic features and superior anti-leukemic activity of B43 (anti-CD19) immunotoxins containing two pokeweed antiviral protein molecules covalently linked to each monoclonal antibody molecule

Standard immunotoxin production procedures using whole IgG as the MoAb moiety yield a heterogeneous mixture of 180 kDa, 210 kDa, and 240 kDa immunotoxin species with 1 to 1, 1 to 2, and 1 to 3 MoAb to toxin ratios. This heterogeneity makes it impossible to precisely deliver a predetermined immunotoxin dose to target cells and impairs the accuracy of pharmacologic studies. In this report, we describe the preparation and characterization of B43(anti-CD19)-pokeweed antiviral protein (PAP) immunotoxins containing either one or two 30 kDa PAP toxin molecules covalently linked to each 150 kDa B43 monoclonal antibody molecule. Compared to the 180 kDa immunotoxin, the 210 kDa immunotoxin displayed greater in vitro chemical stability, resulted in higher systemic exposure levels in vivo, and was a more effective anti-leukemic agent in a SCID mouse model of human B-lineage acute lymphoblastic leukemia. Taken together, the results of this study recommend the clinical evaluation of 210 kDa B43-PAP as a potentially more effective immunotoxin against relapsed B-lineage ALL.